# CENTRAL NERVOUS SYSTEM METASTASES IN LUNG CANCER PATIENTS: FROM PREVENTION TO DIAGNOSIS AND TREATMENT

EDITED BY : Lizza E. L. Hendriks, Deepa S. Subramaniam and Anne-Marie C. Dingemans PUBLISHED IN : Frontiers in Oncology

#### Frontiers Copyright Statement

© Copyright 2007-2019 Frontiers Media SA. All rights reserved. All content included on this site, such as text, graphics, logos, button icons, images, video/audio clips, downloads, data compilations and software, is the property of or is licensed to Frontiers Media SA ("Frontiers") or its licensees and/or subcontractors. The copyright in the text of individual articles is the property of their respective authors, subject to a license granted to Frontiers.

The compilation of articles constituting this e-book, wherever published, as well as the compilation of all other content on this site, is the exclusive property of Frontiers. For the conditions for downloading and copying of e-books from Frontiers' website, please see the Terms for Website Use. If purchasing Frontiers e-books from other websites or sources, the conditions of the website concerned apply.

Images and graphics not forming part of user-contributed materials may not be downloaded or copied without permission.

Individual articles may be downloaded and reproduced in accordance with the principles of the CC-BY licence subject to any copyright or other notices. They may not be re-sold as an e-book.

As author or other contributor you grant a CC-BY licence to others to reproduce your articles, including any graphics and third-party materials supplied by you, in accordance with the Conditions for Website Use and subject to any copyright notices which you include in connection with your articles and materials.

All copyright, and all rights therein, are protected by national and international copyright laws.

The above represents a summary only. For the full conditions see the Conditions for Authors and the Conditions for Website Use. ISSN 1664-8714 ISBN 978-2-88945-707-6 DOI 10.3389/978-2-88945-707-6

#### About Frontiers

Frontiers is more than just an open-access publisher of scholarly articles: it is a pioneering approach to the world of academia, radically improving the way scholarly research is managed. The grand vision of Frontiers is a world where all people have an equal opportunity to seek, share and generate knowledge. Frontiers provides immediate and permanent online open access to all its publications, but this alone is not enough to realize our grand goals.

#### Frontiers Journal Series

The Frontiers Journal Series is a multi-tier and interdisciplinary set of open-access, online journals, promising a paradigm shift from the current review, selection and dissemination processes in academic publishing. All Frontiers journals are driven by researchers for researchers; therefore, they constitute a service to the scholarly community. At the same time, the Frontiers Journal Series operates on a revolutionary invention, the tiered publishing system, initially addressing specific communities of scholars, and gradually climbing up to broader public understanding, thus serving the interests of the lay society, too.

#### Dedication to Quality

Each Frontiers article is a landmark of the highest quality, thanks to genuinely collaborative interactions between authors and review editors, who include some of the world's best academicians. Research must be certified by peers before entering a stream of knowledge that may eventually reach the public - and shape society; therefore, Frontiers only applies the most rigorous and unbiased reviews.

Frontiers revolutionizes research publishing by freely delivering the most outstanding research, evaluated with no bias from both the academic and social point of view. By applying the most advanced information technologies, Frontiers is catapulting scholarly publishing into a new generation.

#### What are Frontiers Research Topics?

Frontiers Research Topics are very popular trademarks of the Frontiers Journals Series: they are collections of at least ten articles, all centered on a particular subject. With their unique mix of varied contributions from Original Research to Review Articles, Frontiers Research Topics unify the most influential researchers, the latest key findings and historical advances in a hot research area! Find out more on how to host your own Frontiers Research Topic or contribute to one as an author by contacting the Frontiers Editorial Office: researchtopics@frontiersin.org

# CENTRAL NERVOUS SYSTEM METASTASES IN LUNG CANCER PATIENTS: FROM PREVENTION TO DIAGNOSIS AND TREATMENT

Topic Editors:

Lizza E. L. Hendriks, Maastricht University Medical Center+, Netherlands Deepa S. Subramaniam, Georgetown University, United States Anne-Marie C. Dingemans, Maastricht University Medical Center+, Netherlands

Image: Crevis/Shutterstock.com

Approximately 40% of lung cancer patients will develop central nervous system (CNS) metastases during the course of their disease. Most of these are brain metastases, but up to 10% will develop leptomeningeal metastases. Known risk factors for CNS metastases development are small cell lung cancer (SCLC), adenocarcinoma histology, epidermal growth factor receptor (EGFR) mutant or anaplastic lymphoma kinase (ALK) rearranged lung cancer, advanced nodal status, tumor stage and younger age. CNS metastases can have a negative impact on quality of life (QoL) and overall survival (OS). The proportion of lung cancer patients diagnosed with CNS metastases has increased over the years due to increased use of brain imaging as part of initial cancer staging, advances in imaging techniques and better systemic disease control. Post contrast gadolinium enhanced magnetic resonance imaging (gd-MRI) is preferred, however when this is contra-indicated a contrast enhanced computed tomography (CE-CT) is mentioned as an alternative option. When CNS metastases are diagnosed, local treatment options consist of radiotherapy (stereotactic or whole brain) and surgery. Local treatment can be complicated by symptomatic radiation necrosis for which no high level evidence based treatment exists. Moreover, differential diagnosis with metastasis progression is difficult. Systemic treatment options have expanded over the last years. Until recently, chemotherapy was the only treatment option with a poor penetration in the CNS. Angiogenesis inhibitors are promising in the treatment of primary CNS tumors as well as radiation necrosis but clinical trials of anti-angiogenic agents in NSCLC have largely excluded patients with CNS metastases. Furthermore, research has also focused on methods to prevent development of CNS disease, for example with prophylactic cranial irradiation. Recently, checkpoint inhibitors have become available for NSCLC patients, and tyrosine kinase inhibitors (TKIs) have improved prognosis significantly in those with a druggable driver mutation. Newer TKIs are often designed to have better CNS penetration compared to first-generation TKIs. Despite advances in treatment options CNS metastases remain a problem in lung cancer and cause morbidity and mortality.

This Research Topic provides an extensive resource of articles describing advances in CNS metastases management in lung cancer patients, from prevention to diagnosis and treatment.

Citation: Hendriks, L. E. L., Subramaniam, D. S., Dingemans, A-M. C., eds. (2019). Central Nervous System Metastases in Lung Cancer Patients: From Prevention to Diagnosis and Treatment. Lausanne: Frontiers Media. doi: 10.3389/978-2-88945-707-6

# Table of Contents

*05 Editorial: Central Nervous System Metastases in Lung Cancer Patients: From Prevention to Diagnosis and Treatment*

Lizza E. L. Hendriks, Deepa S. Subramaniam and Anne-Marie C. Dingemans

#### PREDICTION AND PREVENTION OF CNS METASTASES DEVELOPMENT AND DIAGNOSTIC PITFALLS


Willem J. A. Witlox, Bram L. T. Ramaekers, Jaap D. Zindler, Daniëlle B. P. Eekers, Judith G. M. van Loon, Lizza E. L. Hendriks, Anne-Marie C. Dingemans and Dirk K. M. De Ruysscher

*27 Invasive Aspergillosis Mimicking Metastatic Lung Cancer* Michiel J. E. G. W. Vanfleteren, Anne-Marie C. Dingemans, Veerle F. Surmont, Karim Y. Vermaelen, Alida A. Postma, Astrid M. L. Oude Lashof, Cordula C. M. Pitz and Lizza E. L. Hendriks

#### LOCAL TREATMENT OF BRAIN METASTASES

*34 Stereotactic Radiosurgery in the Management of Patients With Brain Metastases of Non-Small Cell Lung Cancer: Indications, Decision Tools and Future Directions*

Dianne Hartgerink, Britt van der Heijden, Dirk De Ruysscher, Alida Postma, Linda Ackermans, Ann Hoeben, Monique Anten, Philippe Lambin, Karin Terhaag, Arthur Jochems, Andre Dekker, Janna Schoenmaekers, Lizza Hendriks and Jaap Zindler

*46 Brain Radiation Necrosis: Current Management With a Focus on Non-small Cell Lung Cancer Patients* Gokoulakrichenane Loganadane, Frédéric Dhermain, Guillaume Louvel, Paul Kauv, Eric Deutsch, Cécile Le Péchoux and Antonin Levy

#### SYSTEMIC TREATMENT OF CNS METASTASES

	- William J. Kelly, Neil J. Shah and Deepa S. Subramaniam

Ed Schuuring and Anthonie J. van der Wekken

*84 Management of Leptomeningeal Metastases in Non-oncogene Addicted Non-small Cell Lung Cancer*

Ana Turkaj, Anna M. Morelli, Tiziana Vavalà and Silvia Novello

# Editorial: Central Nervous System Metastases in Lung Cancer Patients: From Prevention to Diagnosis and Treatment

Lizza E. L. Hendriks <sup>1</sup> \*, Deepa S. Subramaniam<sup>2</sup> and Anne-Marie C. Dingemans <sup>1</sup>

*<sup>1</sup> Department of Pulmonary Diseases, GROW School for Oncology and Developmental Biology, Maastricht University, Maastricht, Netherlands, <sup>2</sup> Division of Hematology-Oncology, Georgetown University, Washington, DC, United States*

Keywords: brain metastases from lung cancer, leptomeningeal metastases, treatment—contemporary views, diagnosis, driver mutations, cranial irradiation

**Editorial on the Research Topic**

**Central Nervous System Metastases in Lung Cancer Patients: From Prevention to Diagnosis and Treatment**

#### INTRODUCTION

Edited and reviewed by: *Stephen V. Liu, Georgetown University, United States*

> \*Correspondence: *Lizza E. L. Hendriks lizza.hendriks@mumc.nl*

#### Specialty section:

*This article was submitted to Thoracic Oncology, a section of the journal Frontiers in Oncology*

Received: *30 September 2018* Accepted: *19 October 2018* Published: *06 November 2018*

#### Citation:

*Hendriks LEL, Subramaniam DS and Dingemans A-MC (2018) Editorial: Central Nervous System Metastases in Lung Cancer Patients: From Prevention to Diagnosis and Treatment. Front. Oncol. 8:511. doi: 10.3389/fonc.2018.00511* Approximately 40% of lung cancer patients will develop central nervous system (CNS) metastases during the course of their disease (1). Most of these are brain metastases (BM), but 3–9% will develop leptomeningeal metastases (2, 3). The proportion of lung cancer patients diagnosed with CNS metastases has increased over the years due to increased use of brain imaging as part of initial cancer staging, advances in imaging techniques and better systemic disease control (4–6). CNS metastases can have a negative impact on quality of life (QoL) and overall survival (OS) (7– 9). As such, prevention of CNS metastases development, as well as optimal treatment of already established CNS metastases is important.

Contributors in this Research Topic of Frontiers in Oncology, section Thoracic Oncology describe the advances in CNS metastases management in lung cancer patients, from prediction and prevention, to diagnosis and treatment.

#### PREDICTION AND PREVENTION OF CNS METASTASES DEVELOPMENT AND DIAGNOSTIC PITFALLS

Known risk factors for CNS metastases development are small cell lung cancer, adenocarcinoma histology, advanced nodal status, tumor stage and younger age (10–13). Patients with a driver mutation have a high risk of CNS metastases (14). This seems mainly due to their long survival, combined with the poor blood-brain-barrier penetration of the older generation tyrosine kinase inhibitors (TKIs) (Pedrosa et al.). However, these factors alone cannot predict accurately enough which patients will develop CNS metastases and better prediction models are needed. In the review of Pedrosa et al. the process of BM development and the evidence for the clinical and molecular factors associated with increased risk of BM diagnosis in lung cancer is summarized. In addition, they provide an excellent overview of new promising strategies to identify patients at high risk for BM development,

including signatures derived from circulating tumor DNA measurements, single-nucleotide polymorphisms, copy number alterations, microRNAs and long non-coding RNAs (Pedrosa et al.). As it is known that among the non-small cell lung cancer (NSCLC) patients treated with curative intent, stage III patients have the highest risk of brain metastases development [30% (9, 15, 16)], preventive treatments such as prophylactic cranial irradiation (PCI) have been studied especially in this patient population (15, 16). Most studies have been published before the year 2000, but afterwards three new trials incorporating adequate baseline brain imaging, have reported their results. All studies showed a reduction of BM incidence after PCI compared to no PCI (Witlox et al.). Other important factors to consider before administering PCI are toxicity (acute as well as long-term), QoL and OS. Witlox et al. provide an up-to-date systematic review and meta-analysis of all published studies in this field. They provide detailed data on the effects of PCI on BM reduction, toxicity, QoL and OS and discuss areas for future research.

Although brain imaging techniques have improved over the years facilitating the diagnosis of BM, differential diagnosis can be challenging, especially in a patient with a medical history of cancer, as is discussed in the case report by Vanfleteren et al. It is stressed that even in an immunocompetent patient, a diagnosis of cerebral aspergillosis cannot be excluded. A short review of existing literature on this topic is also provided.

#### LOCAL TREATMENT OF BRAIN METASTASES

When BM are diagnosed, local treatment options consist of radiotherapy [stereotactic radiosurgery (SRS) or whole brain radiotherapy (WBRT)] and surgery (17, 18). SRS is more and more often used in the treatment of BM. In the past, SRS was mainly used for up to four BM, but a recent trial showed that SRS is feasible for a higher number of BM (19). More than fifty percent of BM patients treated with SRS will experience an intracranial relapse (19, 20), and especially in this palliative setting patient participation in the decision making around available treatment options [e.g., SRS, WBRT, systemic treatment (with/without concurrent cranial irradiation), best supportive care] is important. Hartgerink et al. discuss the current evidence of SRS for NSCLC BM, and the incorporation of decision aid tools in the future directions for NSCLC BM treatment. Furthermore, local treatment can be complicated by symptomatic radiation necrosis for which no high-level evidence-based treatment exists, although bevacizumab is a promising treatment option (21–23). Differential diagnosis of radiation necrosis and BM progression can be difficult

#### REFERENCES


(21, 22). Loganadane et al. provide a very nice overview of the pathobiology, epidemiology, predictive factors, diagnosis and emerging treatment of radiation necrosis, with a specific focus on NSCLC.

#### SYSTEMIC TREATMENT OF CNS METASTASES

Systemic treatment options for BM have expanded over the last years (24, 25). Until recently, chemotherapy was the only treatment option with a poor penetration in the CNS (26). Angiogenesis inhibitors are promising in the treatment of NSCLC BM (23, 27, 28), but clinical trials of anti-angiogenic agents in NSCLC have largely excluded BM patients (29). Furthermore, TKIs have improved prognosis significantly in those with a druggable driver mutation (25, 30, 31). Newer TKIs are often designed to have better CNS penetration compared to first-generation TKIs (24, 30, 31). In the review of Kelly et al. the management of CNS metastases in EGFR mutated NSCLC patients is discussed, including the role of newer generation EGFR-TKIs, immunotherapy, and EGFR-TKIs combined with cranial irradiation or angiogenesis inhibition. Remon and Besse provide a broader overview of incidence and treatment of BM in oncogene addicted NSCLC patients, including rare driver mutations such as ALK, ROS1, RET, BRAF, and NTRK. Relevant research questions such as optimal sequence of treatment (upfront cranial irradiation or upfront TKI, sequence of TKI) are also discussed. In the case report of Meedendorp et al. treatment of a BM patient with acquired resistance to EGFR-TKI is discussed, stressing the importance of a molecular tumor board for decision making. Last, leptomeningeal metastases remain challenging to treat, especially in non-oncogene addicted NSCLC patients. Turkaj et al. provide an up-to-date review of possible treatment options for these patients, including systemic as well as intrathecal chemotherapy as well as radiotherapy options.

#### CONCLUSION

With the articles in this Research Topic, we hope to provide a review of present and future treatment options for lung cancer CNS metastases, including evidence for predictive markers, preventive treatments, and local as well as systemic treatment options for already diagnosed CNS metastases.

## AUTHOR CONTRIBUTIONS

All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.

treatment era. Cancer Treat Rev. (2017) 53:128–37. doi: 10.1016/j.ctrv.2016. 12.006

3. Kuiper JL, Hendriks LE, van der Wekken AJ, de Langen AJ, Bahce I, Thunnissen E, et al. Treatment and survival of patients with EGFRmutated non-small cell lung cancer and leptomeningeal metastasis: a retrospective cohort analysis. Lung Cancer (2015) 89:255–61. doi: 10.1016/j.lungcan.2015.05.023


(JLGK0901): a multi-institutional prospective observational study. Lancet Oncol. (2014) 15:387–95. doi: 10.1016/S1470-2045(14)70061-0


**Conflict of Interest Statement:** LH: none related to current manuscript, outside of current manuscript: research funding Roche, Boehringer Ingelheim (both institution), advisory board: Boehringer, BMS, (both institution), travel reimbursement: Roche, BMS (self). DS: speaker's bureau Astra Zeneca, Takeda; advisory board Astra Zeneca, Takeda, BMS; institutional research funding as PI: BMS, Genentech, Incyte, Lilly Oncology, Astra Zeneca, Novocure. A-MD: none related to current manuscript, outside of current manuscript: advisory board BMS, MSD, Roche, Eli Lilly, Takeda, Pfizer, Boehringer Ingelheim (all institution).

Copyright © 2018 Hendriks, Subramaniam and Dingemans. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Potential Molecular Signatures Predictive of Lung Cancer Brain Metastasis

*Rute M. S. M. Pedrosa1 , Dana A. M. Mustafa1 , Joachim G. J. V. Aerts2 and Johan M. Kros1 \**

*1Department of Pathology, Erasmus Medical Center, Rotterdam, Netherlands, 2Department of Lung Diseases, Erasmus Medical Center, Rotterdam, Netherlands*

Brain metastases are the most common tumors of the central nervous system (CNS). Incidence rates vary according to primary tumor origin, whereas the majority of the cerebral metastases arise from primary tumors in the lung (40–50%). Brain metastases from lung cancer can occur concurrently or within months after lung cancer diagnosis. Survival rates after lung cancer brain metastasis diagnosis remain poor, to an utmost of 10 months. Therefore, prevention of brain metastasis is a critical concern in order to improve survival among cancer patients. Although several studies have been made in order to disclose the genetic and molecular mechanisms associated with CNS metastasis, the precise mechanisms that govern the CNS metastasis from lung cancer are yet to be clarified. The ability to forecast, which patients have a higher risk of brain metastasis occurrence, would aid cancer management approaches to diminish or prevent the development of brain metastasis and improve the clinical outcome for such patients. In this work, we revise genetic and molecular targets suitable for prediction of lung cancer CNS disease.

Keywords: brain metastasis, lung cancer, molecular mechanisms, genetic alterations, chemotherapy

#### INTRODUCTION

Brain metastases are the most common tumors of the central nervous system (CNS). Metastatic brain lesions outnumber primary brain tumors with a 10-fold (1) with incidence rates varying according to the primary tumor origin. The majority of the cerebral metastases arise from primary tumors in the lung (40–50%) and it is estimated that 50% of the patients with small-cell lung cancer (SCLC) or non-small-cell lung cancer (NSCLC) will develop brain metastasis (2, 3). In contrast to cerebral metastases from other primary cancers, where generally a metastatic latency period takes place, brain metastasis from lung cancers often occur months after, or even concurrently, with the diagnosis of the primary tumor (4). Metastatic brain lesions carry a clinical burden of morbidity and mortality, as well as significant neurological deficits, cognitive impairment, and emotional difficulties (5). Despite treatment, lung cancer brain metastases are usually fatal for 90% of patients within two years after the initial diagnosis, with a median survival of 7–10 months five years after diagnosis (2). Previous efforts to characterize patients that are at high risk of developing brain metastasis have been fairly disappointing.

Currently, only clinical and pathologic variables are used to predict the risk of brain metastasis in patients with lung cancer. However, data on predictive parameters are diverse and not clinically usable (**Table 1**). Identifying patients at highest risk of developing brain metastases on the basis

#### *Edited by:*

*Anne-Marie Dingemans, Academic Hospital Maastricht, Netherlands*

#### *Reviewed by:*

*K. Shilo, The Ohio State University, United States Timothy F. Burns, University of Pittsburgh Cancer Institute, United States*

#### *\*Correspondence:*

*Johan M. Kros j.m.kros@eramusmc.nl*

#### *Specialty section:*

*This article was submitted to Thoracic Oncology, a section of the journal Frontiers in Oncology*

*Received: 07 March 2018 Accepted: 25 April 2018 Published: 11 May 2018*

#### *Citation:*

*Pedrosa RMSM, Mustafa DAM, Aerts JGJV and Kros JM (2018) Potential Molecular Signatures Predictive of Lung Cancer Brain Metastasis. Front. Oncol. 8:159. doi: 10.3389/fonc.2018.00159*

**8**


Table 1 | Conflicting clinical and pathological risk factors associated with the development of brain metastases.

*ND, not determined; "*−*", no predictive value (p* > *0.05); "*+*", predictive value, no significance. \*Significant for univariate analysis only.*

of standard clinical and pathological factors, such as status of primary tumor, tumor histology, nodal involvement, and patient age, may not be reliable due to small hazard ratios and unknown prognostic factors (6). Recently, Hung et al. (7) demonstrated in a study on 182 lung adenocarcinomas with distant metastases that the micropapillary histology subtype was significantly associated with brain metastasis (*p* = 0.01). However, a more robust method to identify which patients are at risk of developing brain metastasis is urgently needed.

Molecular classification by correlating distinct molecular markers with oncogenic mechanisms has been practiced to improve risk stratification of the TNM staging system (21). The potential of molecular biologic distinction would direct appropriate therapy, thereby improving patient outcome. Among earlystage (I/II) NSCLC patients, the 5-year overall survival (OS) rate is only 45.1% (22). Many clinical trials have confirmed that postoperative adjuvant therapy can prolong the survival of NSCLC patients. In a recent meta-analysis of 3,923 patients, Chen et al. (23) demonstrated the efficacy of postoperative chemotherapy – both cisplatin based (*p* < 0.0001) and single tegafur-uracil (UFT, *p* = 0.002), in stages I–II, IA, and IB NSCLC, and no significant benefit was found in stage IA patients (*p* = 0.43). In addition, cisplatin was shown to be better than single UFT chemotherapy in OS (*p* = 0.0005 and *p* = 0.81, respectively) (23). More trials should be conducted in order to confirm the efficacy of diseasefree survival therapies in future clinical practice.

In order to predict the rise of cerebral metastasis of lung cancer, we would need a measurable biomarker that correlates well with brain seeding of the lung cancer cells. Molecular markers may be classified into subgroups based on their mechanism of action in the metastatic cascade to the brain (6). The optimal marker to disclose concealed (brain) metastatic disease would be displayed in primary tumors while not detectable in the serum of control subjects (24). The capacity to identify metastatic disease based on proto-oncogenes such as Kirsten rat sarcoma viral oncogene homolog (KRAS) and tumor suppressor p53 (TP53), present in only half of the lung cancer patients (47 and 50%, respectively) (25), demands a more broaden and deepened spectrum of the investigation of primary lung cancers, the molecular interactions with other cells, and the tumor microenvironment.

The process of metastasis is a selective and refined event called organotropism whereby, apart from an overall tendency to spread and invade, primary tumors show predilection for particular distinct organs (26). Cancers that metastasize to brain need to take a number of anatomic, physiologic, and molecular hurdles. The first requirement is intravasation into the blood stream, dependent upon a reversible epithelial-to-mesenchymal transition (EMT). The epithelial cell traits, such as cell polarity and E-cadherin-mediated cell adhesion, are suppressed and replaced by mesenchymal cell characteristics. The cells become motile, invasive, and resistant to apoptosis (27). Through the EMT process, tumor cells acquire stem cell-like features such as self-renewal, differentiation and ability to seed, justifying the term "tumor-initiating cells" (27). EMT molecular regulation is accomplished through an intricate network arranged by different genes and molecule inducers of EMT (28–30). AXL, a receptor tyrosine kinase belonging to the TAM family, and its ligand GAS6, growth arrest-specific gene 6, have been reported to downregulate several oncogenic signaling pathways (31), through activation of MAPK/ERK and PI3K/AKT signaling pathways (32, 33). Recently, AXL-GAS6 signal axis has been reported to have a potential key role in NSCLC tumor progression and may be suitable as a prognostic biomarker for identifying high-risk NSCLC brain metastasis patients (34). Tumor cell growth in the brain microenvironment is the result of genetic predisposition and cellular adaptation mechanisms and is largely dependent on cross-talk between tumor and brain-resident cells.

Genomic instability and mutations are just two of the characteristics of cells associated with the transition from a preneoplastic lesion to an invasive tumor state and consequent progression to metastatic disease. During tumorigenesis, a sequence of genetic modifications such as gene deletions, copy number alterations (CNAs), and chromosomal rearrangements occur. This review focuses on the use of molecular characteristics that are predictive of tumor progression and development of metastatic NSCLC brain metastasis in particular.

#### GENETIC ALTERATIONS

Due to the recent discoveries of targetable genetic alterations in the treatment of NSCLC, patients have been stratified according to genetic variations in the primary tumor, including epidermal growth factor receptor (EGFR), KRAS, and anaplastic lymphoma kinase (ALK) (35). A summary of all genetic alterations that will be addressed in this review are presented in **Figure 1**. There are considerable differences in reported incidence and time to development of brain metastases for these genetic alterations.

#### EPIDERMAL GROWTH FACTOR RECEPTOR

In the Caucasian population, EGFR-activating mutations are present in 10–15% of adenocarcinomas and in less than 5% of squamous cell carcinomas (36). Roughly 90% of all known EGFR mutations reside in exon 19 (in-frame deletions) and in exon 21 (L858R, point mutation) (37, 38). The prevalence of activating EGFR mutations appears to be dependent on gender, smoking status and ethnicity. In patients from East-Asia, EGFR mutation is reportedly up to five times higher than in Caucasian patients (39–41). The relation between EGFR status, brain metastasis and survival are complex and not fully understood. It has been shown that lung cancer patients suffering from tumors with particular EGFR mutations survive longer, probably due to effective treatment. However, data also suggest that brain metastases arise more frequently in patients with primary lung tumors bearing EGFR mutations (42, 43) and the development of brain metastases is relatively frequent during treatment. There are discordance rates of EGFR mutational status between primary tumors and their CNS metastases that vary from 0 to 32% (44–50). In a series of 55 NSCLC primary tumors with matched cerebral metastases, EGFR was found to be more

frequently amplified in the metastatic adenocarcinomas than corresponding primary tumors, with 30 and 10%, respectively (50). Discrepancies regarding the response of the brain metastases may well be due to the timing of administering adjuvant chemotherapy for the primary tumors relative to the occurrence of the brain metastases. The choice of the agents is currently based on the molecular characteristics of the primary, not the metastatic, tumors. Paradoxically, prolonging survival times due to successful response of the primary tumors would create more time for brain metastases to develop as late complication (51, 52). Similar to EGFR, the KRAS status may also be discordant between primary and metastatic tissues (44) and a KRAS mutation in a small subset of tumor cells may confer resistance to EGFR tyrosine kinase inhibitors (TKIs) therapy.

#### KRAS

Epidermal growth factor receptor and KRAS mutations are generally mutually exclusive (53, 54), but cases of EGFR and KRAS co-mutations have been identified (55–57). Roughly 15–30% of NSCLCs harbor activating mutations in codons 12 and 13 of the KRAS gene (58). KRAS mutations are associated with advanced tumor progression and clinical aggressiveness (59), forming a persistent risk of lung adenocarcinoma and implying to be an early event in the tumorigenesis process (53). The correlation of the presence of KRAS mutations with a smoking history (60) suggests that KRAS mutations are a sequel of the actions of carcinogens of tobacco products (53). However, in a cohort of 482 lung adenocarcinomas, it was demonstrated that KRAS mutations do occur in patients with lung adenocarcinomas without a smoking history (61), but the mutations are different. Significantly more transition mutations (G>A) are being found in non-smokers than the transversion mutations (G>T or G>C, *p* < 0.0001) that occur in former- or current smokers (61). This observation supports the idea that the distinct transition profile – replacement of a purine for a purine or a pyrimidine for a pyrimidine (62) – of never smokers is very unlikely to be caused by passive tobacco vulnerability. No specific KRAS targeting treatment has so far shown efficacy. There is little available data on the KRAS mutational status in primary lung cancers as compared to that in their brain metastases (44, 57). In a relatively small series, Munfus-McCray et al. found 23.5% of brain metastatic lung adenocarcinomas with KRAS mutation exclusively in patients with a smoking history (*p* < 0.01) (59).

## ANAPLASTIC LYMPHOMA KINASE

Anaplastic lymphoma kinase rearrangements occur in 2–7% of all NSCLC, with predominance in non- or light smokers, younger age, and adenocarcinomas (63, 64). Fusion between EML4 (echinoderm microtubule-associated protein-like 4) and ALK yields at least 15 molecular variants with different biological behaviors and affected signaling pathways and consequences for therapy choice (65). ALK testing is particularly recommended for non-squamous lung cancers in the absence of EGFR mutation, of patients with non- or light smoking history (66). The recommended method for testing the presence of ALK translocation is fluorescent *in situ* hybridization (FISH) and immunohistochemistry (IHC) as confirmation (67, 68). In a large Western cohort, functional ALK rearrangements appeared to be mutually exclusive with EGFR and KRAS mutations (69). Although ALK translocations seem to be similar in primary tumors and their brain metastases, ALK amplifications are found more frequently in CNS metastasis with discordance rates of only 12.5% (70). Similar to EGFR, ALK rearrangements are predictive of response to TKIs, but the development of brain metastasis in patients with ALK translocations receiving ALK directed TKI is a major clinical problem (71, 72). Recently, second-generation TKI alectinib has shown to delay the development of brain metastases compared to first-generation TKI and also demonstrated promising efficacy in the CNS for crizotinib-resistant ALK-positive NSCLC patients (73, 74). Similar to EGFR and KRAS mutations and ALK rearrangements, several other molecules such as liver kinase B1 (LKB1, also known as STK11), proto-oncogene tyrosine-protein kinase ROS1, and C-MET that encodes the hepatocyte growth factor receptor were found to be implicated in the development of lung cancer (75–79). However, the connection of these molecules with the development of brain metastases is still under investigation and not yet implicated in clinical decision making. KRAS aberrations have a synergistic effect with LKB1 inactivation on lung cancer development and distant metastasis formation (80, 81). In a cohort of 154 NSCLC patients, Zhao et al. demonstrated that a lower LKB1 copy number (CN), along with KRAS mutation, were significantly associated with a higher number of brain metastasis. Moreover, the odds ratio of brain metastasis was ~20 times higher in patients with one decrease in LKB1 CN values (82). LKB1 is observed to be inactivated in ~30% of all NSCLCs (83).

## OTHER MUTATIONS

Although several potential targets may not regularly be expressed in a high number of lung cancer brain metastasis, their potential use for personalized treatment of selected lung cancer patients harboring actionable mutations should not be discarded. In a cohort of 874 brain metastases samples, of which 295 NSCLC, Capper et al. showed that, although a total of 51/874 samples harbored a BRAF V600E mutation, only 1/295 NSCLC brain metastases (~0.3%) was BRAF mutant (84). Despite this low frequency of BRAF mutations in lung brain metastasis, regression of both visceral and brain metastases by BRAF inhibitor vemurafenib was reported in a patient with a BRAF V600E-mutated NSCLC (85). While 3% of primary lung cancers harbor ROS1 alterations, only 1/99 adenocarcinomas bore ROS1 translocations and 1/11 squamous cell carcinomas showed ROS1 amplifications (86).

Activating mutations in EGFR are associated with sensitivity to TKI therapy, but nearly 30% of EGFR positive patients show primary resistance to EGFR inhibitor therapy (87). While C-MET amplification is one of the factors commonly associated with disease progression (88), Benedettini et al., in a first cohort of 23 NSCLC samples of patients harboring an EGFR activating mutation, showed that both C-MET phosphorylation and expression were significantly associated with shorter time to progression, correlating with *de novo* resistance to EGFR TKI. In a second cohort of 40 patients, englobing 18 primary NSCLC from patients who later developed brain metastases and 22 NSCLC from patients that did not develop brain metastases, Benedettini et al. demonstrated that both C-MET expression and phosphorylation, but not C-MET amplification, were significantly higher in the tumors from patients who developed brain metastasis. In 18 matched brain metastasis, amplification was demonstrated (89). In addition, in a cohort of 196 NSCLC brain metastasis samples, Presseur et al. found C-MET gene amplification and overexpression in 21.6 and 44.4%, respectively, confirming that C-MET is commonly activated in brain metastasis manifestation (90). Furthermore, a significant correlation between C-MET and ALK amplification status was observed (*p* = 0.039). In another study, these authors demonstrated that fibroblast growth factor receptor 1 (FGFR1) amplification in brain metastases of adenocarcinomas – but not squamous cell carcinomas, is fivefold more frequent than reported for primary tumors (~3%). Similar to C-MET, a positive correlation of ALK and FGFR1 amplification status in brain metastasis was reported as significant (*p*< 0.001) (91). In a recent study, Keap1, Nrf2, and P300, key genes of the Keap1–Nrf2–ARE survival pathway, were found to be mutated in brain metastatic tissue of progressive NSCLC patients (92). Moreover, mutations in Keap1-Nrf2-ARE pathway were found in circulating tumor cells (CTCs), suggesting a role in the ability of CTCs to bear the rough environment in blood-circulation and attain distant organs (92).

## CIRCULATING TUMOR DNA (ctDNA)

An adequate characterization of somatic genetic modifications in human cancers is critical for an optimal diagnosis and subsequent therapy. In brain metastatic tissue, as for all other brain malignancies, repeated biopsies are not a feasible approach to portray the tumor clonal diversity. Several studies have shown that cell-free ctDNA in the plasma could serve to characterize and monitor tumors (93–95). Nevertheless, ctDNA analysis of plasma from patients with brain malignancies has disclosed very low levels of tumor DNA (96). Recently, ctDNA analysis from cerebrospinal fluid (CSF) has been shown promising for brain cancer patients (97–99) and brain metastatic cancer patients (100, 101). CSF is in direct contact with the brain and, therefore, with tumor cells of brain cancer patients. In a comparative study of ctDNA derived from plasma and from CSF of patients with primary or metastatic brain tumors, De Mattos-Arruda et al. showed ctDNA levels of brain malignancies to be more abundantly present in the CSF than in the plasma (100). Moreover, ctDNA from CSF appeared to recapitulate the brain metastasis-specific mutations – private mutations, absent in extracranial tumors of a patient with Her2-positive metastatic breast cancer (100). In addition, the CSF ctDNA proficiency to monitor responses to systemic therapy and brain tumor progression (98, 100), i.e., the capacity of the CSF ctDNA to recapitulate the modulation of mutant allele frequency over time in the brain tumor burden, suggests that genomic CSF analysis may be useful not only in facilitating diagnosis of tumor in the CNS or as guidance to second-line agents choice, but also in pinpointing pathways' intimate related with cancer spread to the CNS and predictive of brain metastases (98).

## SINGLE-NUCLEOTIDE POLYMORPHISMS (SNPs) ASSOCIATED WITH BRAIN METASTASES

Studying SNPs in signaling pathways that regulate cell proliferation and migration and assessing the relationship between multiple SNPs can be used to estimate the risk of brain metastasis. The PI3K–PTEN–AKT–mTOR pathway, important in the control of cell growth, tumorigenesis, and cell invasion, has been shown to be abnormally activated in several cancer types, including NSCLC (102, 103). In a study of genetic variations in the PI3K– AKT–mTOR pathway to predict brain metastasis in NSCLC patients, Quianxia et al. identified three SNPs that appeared to be exclusively associated with higher risk of brain metastasis: the GT/GG (*p* = 0.006) and CT/TT (*p* = 0.002) genotypes of AKT1, variant alleles rs2498804 and rs2494732, respectively, and AG/AA (*p* = 0.010) genotype of PIK3CA, variant allele rs2699887 (103). Furthermore, patients carrying at least one variant allele in PIK3CA had roughly twice the risk of brain metastasis as those without those variants (103). Multiple mechanisms of PI3K activation may be responsible for activation of the PI3K pathway (104), and increased PI3K activity would result in increased metastases. In concordance, Paik et al. reported that patients with aberrant PI3K squamous lung carcinomas (*n* = 9) had worse survival (median OS: 8.6 vs 19.1 months, *p*< 0.001), higher metastatic burden (>3 organs, 18 vs 3%, *p* = 0.025), and higher incidence of brain metastases (27 vs 0%, *p* < 0.001) (105). Similar to PIK3K–AKT–mTOR pathway, it was hypothesized that common genetic variants in the TGF-β pathway would be associated with the risk of brain metastasis (106). TGF-β pathway has been demonstrated to suppress early-stage tumor development and to stimulate tumor cell growth and invasiveness at later stages of tumorigenesis (107). Quianxia et al. found the GG genotype of SMAD6: rs12913975 (*p* = 0.014) and the TT genotype of INHBC: rs4750259 (*p* = 0.024) to be associated with risk of brain metastasis in a cohort of 161 blood samples from NSCLC patients. Furthermore, de combination of both genetic variants was shown to be higher for prediction of brain metastasis (*p* = 0.001) (106).

## CNAs ASSOCIATED WITH BRAIN METASTASIS

Activation or inhibition of a gene occurs through a variety of mechanisms such as, for example, activating mutations and deletions. Gene deletion can be evaluated by CNA. Animal models have given clear evidence that LKB1 haploinsufficiency stimulates KRAS driven lung cancer in mice (81), and a single copy inactivation of LKB1 can considerably ease brain recurrence (82). Although EGFR CN status is still controversial and some of the available data do not support EGFR CN as a prognostic factor (108, 109), Bonanno et al. have shown, despite the less predictive accuracy of FISH analysis compared to EGFR mutation analysis, that patients with EGFR-FISH-positive tumors have better outcomes (median OS: 177 vs 57 weeks, *p* = 0.048) (110). Considering that primary lung adenocarcinomas with early development of brain metastasis would contain more CNAs predictive of metastatic potential, Lee et al. compared the CN changes of four lung adenocarcinomas with coexistent brain metastasis with 8 lung adenocarcinomas with metachronous brain metastasis (111). Amplification in 5q35.1-2 and 17q23.3- 24.1 was detected in 100% and that in 10q23.31 and 17q24.1 was detected in 75% of the cases with synchronous brain metastasis. On the other hand, and in a less frequent ratio, only 5q35.1-2 and 17q24.1 amplification status was found in 12.5% of the metachronous brain metastasis. Moreover, gained regions specific for early (simultaneous) brain metastasis were found to contain ACTA2, FAS, RGS9, and ICAM2 as putative metastasis promoting genes, the latter being most significant (*p* = 0.002) (111). In the same line, another study compared CNAs of primary NSCLC tumor and matched brain metastasis from one single patient (112). Brain metastatic tissue exhibited a higher degree of genetic heterogeneity when compared with the primary tumor with common regions of gain including 7p, 7q, and 19q and common regions of loss including 20q13 (112). In a stage IV SQCLCs study, four brain metastases and matched archived FFPE primary cancers were shown to have complete loss of PTEN by IHC and whole exome sequencing (105). In an early-stage NSCLC report, 30 (24%) of the total of 125 specimens analyzed for PTEN-IHC showed a lack of staining (113). Although genetic alterations of the PTEN gene are unusual in NSCLC, loss of PTEN protein is not a unique event in early-stage NSCLC and Soria et al. demonstrated that besides being a reversible event, PTEN loss may be partially explained by promotor methylation, in addition to point mutations and homozygous deletions (113).

#### MICRORNAs (miRNAs) ASSOCIATED WITH BRAIN METASTASIS

Recently, molecular studies have stressed the role of miRNAs which are small non-coding endogenous RNAs containing 18–24 nucleotides that regulate gene expression at the posttranscriptional level thereby acting as negative regulators of mRNA translation and/or stability (114). miRNAs appear to regulate several hundred genes and could serve as a better classifier than gene expression profiling (115). miRNAs are known to play a crucial role in normal development, proliferation, differentiation, and apoptosis, and dysregulation of miRNAs has been linked to various pathological conditions, including cancer (116). The role of miRNAs in the development of brain metastases has been recently explored (117, 118).

Several studies have addressed the miRNA expression as biomarkers to predict the occurrence of brain metastases in lung cancer. miRNA-328 appeared to be significantly overexpressed in both primary tumor samples and cerebral metastases of patients with NSCLC, when compared with NSCLC patients without brain metastasis. Moreover, miRNA-328 overexpression has been found to promote migration and subsequent brain metastasis formation of NSCLC cells through PRKCA deregulation (119). PRKCA mediates the expression of urokinase plasminogen activator, leading to the migration of the tumor cells (120). Similar to miRNA-328, miRNA-378 has also been demonstrated as a potential biomarker to assist clinicians in stratifying patients for highrisk of brain metastasis, because miRNA-378 was also found to be overexpressed in NSCLC primary tumor samples and matched brain metastasis of NSCLC patients (121). Also, miRNA-378 promotes cell migration, invasion, tumor growth, and angiogenesis, *in vitro* and *in vivo* (121). Recently, Remon et al. have identified miRNA-197 and miRNA-184 as two significantly overexpressed miRNAs in EGFR-mutant patients with brain metastases, when compared with EGFR-mutant patients with no brain metastasis (122). However, because of lack of patients with EGFR wild-type (EGFRwt) tumors without BM, no comparison between patients with EGFRwt tumors, with and without BM, could be made. Therefore, the effects of these miRNAs, irrespective of the EGFR status, need further scrutiny.

MicroRNAs' expression status varies according to their targeted genes. Zhao et al. have reported the significant upregulation of miRNA-1471 and miRNA-9 and down-regulation of miRNA-214 and miRNA-145 in 11 brain metastatic lung cancer samples, when compared with 40 primary lung adenocarcinomas (*p* < 0.001 for all four miRNAs) (123). The up-regulation of miRNA-145 in primary lung adenocarcinomas was shown to suppress proliferation of tumor cells (123), consistent with other reports that show inhibition of cell proliferation in human lung adenocarcinomas through miRNA-145 targeting c-Myc, EGFR and NUDT1 (124, 125). Subramani et al. have shown the miRNA-768-3p to be underexpressed in several brain metastases, compared to matched primary tumors (126). miRNA-768-3p was found to be underexpressed in *in vitro* lung cancer cells after co-culture with astrocytes, driving to increased KRAS protein and downstream effectors ERK1/2 and BRAF, thereby boosting tumor cell viability and promoting metastasis. From various studies, it appears that miRNAs regulate the growth of metastases either by under- or overexpression, within the tumor tissue or in the tumor environment. The brain microenvironment negatively regulates miRNA-768-3p to enhance KRAS expression that promotes the propagation of lung cancer brain metastasis (126). miRNA-146 was shown to be significantly up-regulated in NSCLC tissue when compared to healthy adjacent lung tissue (*p* < 0.05) (127). In another study, miRNA-146a expression in primary NSCLC was correlated with advanced clinical TNM stages and distant metastasis (*p* < 0.05). The patients with a high miRNA-146a expression showed longer progression-freesurvival times than those with a low expression of miRNA-146a (25.6 and 4.8 weeks, respectively, *p* < 0.05) (128). In the same line with these findings, in a xenograft model, Hwang et al. showed high expression of miRNA-146a in parental cells, while diminished expression in the brain-seeking cells. Moreover, miRNA-146a overexpression in the brain-seeking cancer cells suppressed their metastatic potential, which was correlated to the up-regulation of β-catenin and down-regulation of heterogeneous nuclear ribonucleoprotein C1/C2 (129). Taken together, these findings suggest that miRNA-146a serve as a valid clinical biomarker for prediction of brain metastasis in lung cancer patients. However, validation of miRNA-146a expression levels in a large cohort of human matched primary and brain metastatic lung tumors is essential to confirm this finding. Similar to miRNA-146a, overexpression of miRNA-95-3p suppresses brain metastasis of lung adenocarcinoma through down-regulation of cyclin D1 (130). miRNA-95-3p is decreased in brain metastases of lung cancers as compared to the primary tumors and higher cyclin D1 expression correlates with poorer prognoses (130). In a recent study, Chen et al. reported miRNA-375 deregulation to be associated with NSCLC brain metastasis (131). miRNA-375 is another miRNA documented to be down-regulated in primary tumors of NSCLC patients with brain metastasis. miRNA-375 expression was significantly decreased in matched brain metastatic NSCLC tissues (*p* < 0.05) and significantly correlates with total number of brain metastasis (*p* < 0.001). In addition, VEGF and MMP9 – which roles have been extensively studied in the development of brain metastasis – were over-expressed in downregulated miRNA-375 tumors (131).

MicroRNAs are linked with several molecular pathways. Several studies have correlated the overexpression of ADAM9 in NSCLC patients with brain metastases (4, 132). ADAM9 has been demonstrated to enhance the ability of tissue plasminogen activator to cleave and stimulate the function of CUB domain containing protein 1 (CDCP1) – promigratory protein, to promote brain metastasis (4). Recently, Chiu et al. reported that ADAM9 down-regulates miRNA-1 via EGFR signaling pathways activation, enhancing CDCP1 expression to promote lung cancer progression (133). miRNA-1 expression was shown to be downregulated in primary lung tumors but increased in ADAM9 knockdown lung cancer cells. Moreover, miRNA-1 negatively correlates with CDCP1 expression and with migration ability of lung cancer cells (133). Another study has identified miRNA-21 as a target of signal transducers and activators of transcription 3 (STAT3) pathway activity in lung-derived brain metastasis initiating cells (134). STAT3 is admitted as a central regulator in the metastatic process (135), and STAT3-knockdown has been demonstrated to reduce expression of known downstream targets of miRNA-21, while STAT3 and miRNA-21 act as cooperative regulators of stemness, migration and tumor initiation in lungderived brain metastasis (134). miRNAs appear very promising as diagnostics, prognostics and therapeutics to improve cancer patient outcome; however, the clinical use of miRNA therapeutics to treat brain metastases has yet to be achieved. Advances in pre-clinical and translational studies to identify miRNAs that change after growth in the brain microenvironment have been made, but validation of large cohorts from patient tumor samples is required.

#### LONG NON-CODING RNAs (lncRNAs) ASSOCIATED WITH BRAIN METASTASIS

Long non-coding RNAs have been recently identified as effective players in tumorigenesis. lncRNAs represent a class of nonprotein coding transcripts longer than 200 nucleotides (136) that covers a broad spectrum of physiological and pathological functions by implementing different modes of action (137). Similar to miRNAs that regulate several hundred genes, lncRNAs are involved in the regulation of multiple miRNAs, impacting the expression of thousands of genes (136). Besides performing a single function, some lncRNAs act at multiple functional levels in different types of cells. Metastasis-associated lung adenocarcinoma transcript 1 (MALAT1), localized in nuclear speckles and highly conserved among mammals, regulates alternative splicing (138) and gene expression through additional splicingindependent mechanisms in lung cancer metastasis (139). In a recent study, Shen et al. have shown lncRNA-MALAT1 levels to be significantly higher in primary NSCLC from patients who developed brain metastasis when compared with primary NSCLC from patients without brain metastasis (*p* < 0.001) (140). Additional *in vitro* functional studies showed overexpression of vimentin in a highly invasive subline of brain metastasis lung cancer cells overexpressing MALAT1, while overexpression of E-cadherin was observed when MALAT1 was silenced, indicating that MALAT1 overexpression promotes lung cancer brain metastasis by inducing EMT (140). Accordingly, RNAi-mediated suppression of MALAT1-RNA, negatively influenced migration and clonogenic growth in established human NSCLC cell lines. Forced expression of MALAT1 in mouse NIH 3T3 fibroblasts significantly increased migration (141). Concordantly, long noncoding MALAT1 expression was found to enhance cell motility through transcriptional and post-transcriptional regulation of motility related gene expression (142), displaying the strongest association with genes involved in cancer, like cellular growth, movement, proliferation, signaling and immune regulation genes (141). MALAT1 and thymosin β4 expression levels were identified as prognostic parameters for patient survival in stage I NSCLC that are at high risk to develop metastasis (*p* = 0.04 and *p* = 0.01, respectively) (143). Tumorigenesis and metastases may be driven by tumor suppressive and oncogenic pathways deregulation through aberrant expression of cancer metastasisassociated lncRNA (144). In a recent *in vitro* study, the lncRNA brain cytoplasmatic RNA 1 (BCYRN1) was found up-regulated and targeted by c-MYC in human NSCLC cell lines (145). c-MYC is a commonly inhibited oncogene and becomes activated in onco-

genic pathways, and correlates with metastasis of NSCLC (146). Besides demonstrating that IncRNA BCYRN1 is essential in the c-MYC-regulated cell migration and invasion, BCYRN1 positively correlates with the expression levels of MMP9 and

## REFERENCES


MMP13 (145). MMP9 and MMP13, two members of the matrixin subfamily of the metzincin superfamily of Zn-dependent metalloproteinases (147), are extracellular matrix degrading proteins proven to induce migration and invasion of tumor cells (147, 148), thereby regulating cancer cell metastasis (149).

#### CONCLUDING REMARKS

Lung adenocarcinoma establishes distant clinical detectable metastasis within months of initial diagnosis (26, 150). This short abeyance indicates that metastatic ability would arise from early oncogenic events that stimulate primary tumor growth rather than late-arising, scarce genomic alterations specific for metastasis (151). Thus, monitoring persistent chromosomal changes in the primary NSCLC alongside with prospective multicenter studies of patient-matched primary and CNS metastatic lesions could help identify targetable approaches for brain metastasisspecific signatures.

## AUTHOR CONTRIBUTIONS

RP, DM, JA, and JK wrote and revised the article.

## FUNDING

This paper was carried out in the framework of a project on cerebral metastases of common cancers by the Dutch cancer society (KWF).

irradiation be reconsidered? *Cancer* (2001) 91:2394–400. doi:10.1002/1097- 0142(20010615)91:12<2394::AID-CNCR1273>3.0.CO;2-6


and local control in the setting of brain metastases treated with radiation? *Neuro Oncol* (2015) 17:1022–8. doi:10.1093/neuonc/nov043


of the K-ras gene in patients with primary adenocarcinoma of the lung. *Cancer* (2001) 92:1525–30. doi:10.1002/1097-0142(20010915)92:6<1525:: AID-CNCR1478>3.0.CO;2-H


non-small-cell lung cancer. *J Thorac Oncol* (2013) 8:e112–3. doi:10.1097/ JTO.0000000000000038


growth factor receptor (EGFR) gene copy number and protein expression in non-small cell lung cancer. *Lung Cancer* (2006) 54:387–98. doi:10.1016/j. lungcan.2006.08.015


non-small cell lung cancer and induces migration and tumor growth. *J Thorac Oncol* (2011) 6:1984–92. doi:10.1097/JTO.0b013e3182307eac


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Copyright © 2018 Pedrosa, Mustafa, Aerts and Kros. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# The Prevention of Brain Metastases in Non-Small Cell Lung Cancer by Prophylactic Cranial irradiation

*Willem J. A. Witlox1 \*, Bram L. T. Ramaekers1 , Jaap D. Zindler <sup>2</sup> , Daniëlle B. P. Eekers2 , Judith G. M. van Loon2 , Lizza E. L. Hendriks3 , Anne-Marie C. Dingemans3 and Dirk K. M. De Ruysscher 2,4*

*1Department of Clinical Epidemiology and Medical Technology Assessment, Maastricht University Medical Centre, Maastricht, Netherlands, 2Department of Radiation Oncology (Maastro Clinic), Maastricht University Medical Centre, GROW School of Oncology and Developmental Biology, Maastricht, Netherlands, 3Department of Pulmonology, Maastricht University Medical Centre, GROW School of Oncology and Developmental Biology, Maastricht, Netherlands, 4Radiation Oncology, KU Leuven, Leuven, Belgium*

#### *Edited by:*

*Yanis Boumber, Fox Chase Cancer Center, United States*

#### *Reviewed by:*

*Meng Xu Welliver, The Ohio State University James Cancer Center, United States Desiree Hao, University of Calgary, Canada*

> *\*Correspondence: Willem J. A. Witlox willem.witlox@mumc.nl*

#### *Specialty section:*

*This article was submitted to Thoracic Oncology, a section of the journal Frontiers in Oncology*

*Received: 30 January 2018 Accepted: 18 June 2018 Published: 26 July 2018*

#### *Citation:*

*Witlox WJA, Ramaekers BLT, Zindler JD, Eekers DBP, van Loon JGM, Hendriks LEL, Dingemans A-MC and De Ruysscher DKM (2018) The Prevention of Brain Metastases in Non-Small Cell Lung Cancer by Prophylactic Cranial Irradiation. Front. Oncol. 8:241. doi: 10.3389/fonc.2018.00241*

Background: Non-small cell lung cancer (NSCLC) patients frequently develop brain metastases (BM), even though the initial imaging with brain CT or MRI was negative. Stage III patients have the highest risk to develop BM, with an incidence of approximately 30%. BM can lead to neurocognitive disorders, loss of quality of life (QoL), and they are the most important factors influencing patient's overall survival (OS). Although a radical local treatment of BM may be possible with primary radiosurgery or after resection, the prognosis often remains poor. Preventing the development of BM through prophylactic cranial irradiation (PCI) may improve the outcome of these patients.

Methods: Data from published randomized trials comparing PCI with non-PCI were sought using electronic database (PubMed) searching, hand searching, and by contacting experts. Trials were included if they considered a randomized comparison of PCI and non-PCI, enrolled NSCLC patients, excluded patients with recurrent or metastatic disease, and reported results on BM occurrence. Each randomized controlled trial (RCT) was assessed for methodological quality using the Cochrane collaboration's tool for the assessment of risk of bias. Study estimates were pooled using a fixed effects sample-weighted meta-analysis approach to calculate an overall estimate and 95% confidence interval (CI). Results on PCI-related toxicity, QoL, and OS were only reported descriptively.

Results: Seven RCTs were included in the meta-analysis. In total, 1,462 patients were analyzed, including 717 patients who received PCI and 745 patients who did not. The risk of developing BM was significantly decreased through PCI (13% reduction, RR 0.33; 95% CI 0.22–0.45). PCI-related toxicity and QoL data were limited. Acute toxicity mostly included fatigue, skin-related toxicity, and nausea or vomiting. Late toxicities such as headache, dyspnea, lethargy, and low grade cognitive impairments were also reported in some of the included RCTs. Results on OS were inconclusive.

Conclusion: The risk of developing BM was reduced in patients who received PCI compared to patients who did not. To implement PCI as the standard treatment for patients

**20**

with NSCLC, the impact of PCI-related toxicity on QoL should be further investigated, as well as long-term OS. A future individual patient data meta-analysis could produce definitive answers to this clinical question.

Keywords: non-small cell lung cancer, prophylactic cranial irradiation, brain metastases, toxicity, survival, quality of life

#### INTRODUCTION

Non-small cell lung cancer (NSCLC) is the most important cause of death due to cancer worldwide, and accounts for about 85% of all lung cancers. At present, more than 50% of all patients are diagnosed with adenocarcinoma, less than 10% are diagnosed with large cell cancer and the rest with squamous cell carcinoma. One-third of NSCLC present with locally advanced (stage III) disease, 20% with stage I–II, and the rest have metastases (stage IV) at diagnosis (1).

Non-small cell lung cancer patients frequently develop brain metastases (BM), even though the initial staging with brain CT or MRI was negative. The more advanced the disease stage is, the more frequent BM occur. They are also more frequent in adenocarcinoma than in squamous cell cancer (1). Stage III patients have a BM incidence of approximately 30% (2). With longer overall survival (OS) and better imaging techniques, this percentage might increase. For example, in drive-mutated patients (e.g., EGFR and ALK) with a survival beyond 5 years, this percentage increases to more than 50% (3). BM can lead to neurocognitive disorders, loss of quality of life (QoL), and they are the most important factors influencing patients' OS (2). Although a radical local treatment of BM may be possible with radiosurgery or resection, the prognosis often remains poor. In order to improve QoL as well as OS, there is an unmet need to prevent the occurrence of BM (4).

Prophylactic cranial irradiation (PCI) was shown to significantly improve OS (5.4% improvement of 3-year OS) in localized small cell lung cancer with complete remission or stable disease after multimodality treatment, as a result of decreasing BM incidence by about 50% (5). Also in patients with NSCLC, several randomized controlled trials (RCTs) studied the value of PCI in the prevention of BM (6–14). However, PCI might deteriorate QoL as a result of neurocognitive decline associated with cranial irradiation. Recently, a randomized phase III trial conducted by the NVALT/DLCRG (14) showed that PCI reduced the incidence of symptomatic BM [7.0% in PCI vs 27.2% in no PCI, hazard ratio 0.25; 95% confidence interval (95% CI) 0.11–0.58]. Therefore, it is time to update the previously published literature and revisit the role of PCI in the prevention of BM in NSCLC patients. Here, we report on the results of a meta-analysis assessing the impact of PCI on the reduction of BM in primary stage I–III NSCLC patients, with PCI-related toxicity, QoL, and OS as secondary endpoints.

#### METHODS

#### Data Collection

Data from published RCTs comparing PCI with non-PCI were sought using electronic database searching between 1980 and December 1, 2017 (PubMed), hand searching (reference checking of individual studies and review articles), and by contacting experts in the field. The following keywords were used as search terms: "Carcinoma, Non-Small-Cell Lung," "NSCLC," "Cranial Irradiation," "Cranial Neoplasms/radiotherapy," "Brain Metastasis," "Overall Survival," and "RCT." Details of the search strategy and corresponding flow chart can be found in Appendix S1 in Supplementary Material.

#### Selection Criteria

Trials were included if they considered a randomized comparison of PCI and non-PCI, enrolled NSCLC patients, excluded patients with recurrent or metastatic disease, and reported results on BM occurrence.

#### Quality Assessment

Two investigators (Willem J. A. Witlox and Bram L. T. Ramaekers) independently assessed each RCT for methodological quality using the Cochrane collaboration's tool for the assessment of risk of bias (15). This tool consists of seven items, including random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other sources of bias. Each item was scored "low risk," "unclear risk," or "high risk" of bias (Appendix S2 in Supplementary Material).

#### Statistical Analysis

Data of the primary endpoint (BM occurrence) was analyzed using Stata/SE 14.2 (16). Relative risk (RR) and accompanying 95% CI of the individual studies were calculated based on the number of events and group totals. Subsequently, the estimates were pooled using a fixed effects sample-weighted meta-analysis approach to calculate an overall estimate and 95% CI. Heterogeneity of the studies was tested using chi-square- and *I* 2 -tests (15). Publication bias and small study effects were assessed by visual inspection of funnel plots and performing Egger's test, respectively (17, 18). If Egger's test is significant, a sensitivity analysis will be performed excluding small studies (weight < 10%).

Results on PCI-related toxicity, QoL, and OS will only be reported descriptively.

#### RESULTS

#### Literature Search and Quality Assessment of Publications

The electronic literature search yielded 360 unique publications. Another two publications were identified through hand searching and contacting experts. After screening of titles and/or abstracts, 354 trials were excluded. One RCT (12) was excluded after reading the full text, because local treatment was different between both arms. Methodological quality of the remaining seven RCTs was checked and most of the items were at "low risk" of bias (**Figure 1**). Three studies (6–8) did only perform a brain scan when indicated by a change in neurological status of the patients, without adequately defining how neurological status was assessed. Therefore, the reviewers judged these studies to be at high risk of introducing bias in assessing the outcome. The reviewers suspected possible selection bias in the study of Cox et al. (6), because randomized patients were excluded from evaluation, which is not in accordance with the intention to treat principle. Although blinding of participants and personnel was not performed in any of the included studies, based on the nature of the intervention, this item was judged by the reviewers as low risk of introducing bias (Appendix S2 in Supplementary Material).

#### Characteristics of the Included Trials

Characteristics of the seven included trials are listed in **Table 1** below. There were little differences between the selected patient groups of the trials. Four studies (9, 10, 13, 14) included stage III

Figure 1 | Methodological quality assessment of the included randomized controlled trials evaluating the effect of prophylactic cranial irradiation on brain metastases in non-small cell lung cancer.



*a Number of eligible patients.*

*bAll inoperable patients; stage not clear.*

*PCI, prophylactic cranial irradiation; RT, radiotherapy; Chemo; chemotherapy; BM, brain metastases; CNS, central nervous system; OS, overall survival; DFS, disease-free survival; RCTs, randomized controlled trials.*

NSCLC patients only, two studies (7, 8) included stage I, II, and III patients, and in one study (6) staging was unclear. For treatment of the primary tumor, two trials (6, 14) treated their patients with radiotherapy alone, one trial (11) used chemo-radiotherapy, and the four remaining trials used either combinations of chemoradiotherapy and radiotherapy alone (7, 9, 10), chemotherapy and surgery (13), or radiotherapy and surgery (8). Brain imaging was mainly done by a radionuclide scan in two studies (6, 7). One study (8) used CT scans, three more recent studies (11, 13, 14) used MRI and in one study (9, 10) the technology of brain imaging was unclear. Dosing of cranial irradiation ranged from 20 to 37.5 Gy (10 fractions of 2 Gy to 15 fractions of 2.5 Gy).

#### Incidence of BM After PCI

Taken all RCTs together, in total, 1,462 patients were analyzed, including 717 patients who received PCI and 745 patients who did not. The BM incidence in the PCI arm ranged from 0.9 to 12.3%, and from 11.0 to 30.7% in the non-PCI arm (**Table 2**). The overall effect estimate of the impact of PCI on the occurrence of BM is presented in **Figure 2**. The risk of developing BM was significantly decreased in the PCI arm compared to no PCI (13% reduction, RR 0.33; 95% CI 0.22–0.45). Heterogeneity across the studies was low (*I* <sup>2</sup> = 0%; *p* = 0.468). Furthermore, Egger's test indicated that smaller studies showed larger effect sizes (*p* = 0.048), which is also reflected in the asymmetric funnel plot (**Figure 3**). Nevertheless, results of the sensitivity analysis excluding small studies (7, 9, 10) (weight < 10%) were similar (RR 0.38; 95% CI 0.24–0.51).

## PCI-Related Toxicity, QoL, and OS

Few trials reported on PCI-related toxicity and QoL with most details in the study of Gore et al. (11) (**Table 3**). Acute toxicity mostly included

Table 2 | Data on BM events [(a)symptomatic BM occurrence] and incidence of the included RCTs evaluating PCI.


*PCI, prophylactic cranial irradiation; BM, brain metastasis; RCTs, randomized controlled trials.*

interval; RCTs, randomized controlled trials; PCI, prophylactic cranial irradiation; BM, brain metastases; NSCLC, non-small cell lung cancer.

#### Table 3 | RCTs evaluating PCI-related toxicity and QoL in NSCLC.

controlled trials; PCI, prophylactic cranial irradiation; BM, brain metastases; NSCLC, non-small cell lung cancer.


*PCI, prophylactic cranial irradiation; QoL, quality of life; MMSE, Mini Mental Status Exam; ADLS, Activities of Daily Living Scale; HVLT, Hopkins Verbal Learning Test; CTC-AE, Common Terminology Criteria Adverse Events; RTOG/EORTC-LRMSS, RTOG/ERTOC late radiation morbidity scoring schema; EORTC QLQ-C30, EORTC quality of life questionnaire-C30; EORTC QLQ-BN20, EORTC quality of life questionnaire-BN20; FACT-L questionnaire, Functional Assessment of Cancer Therapy-Lung questionnaire; NSCLC, non-small cell lung cancer; RCTs, randomized controlled trials.*

fatigue, skin-related toxicity, and nausea or vomiting. Late toxicities such as headache, dyspnea, and lethargy were also reported in some of the included RCTs. Low grade (1 and 2) memory impairments and cognitive disturbances were only reported in the study of De Ruysscher et al. (14). Results reporting on QoL were limited, and no significant differences were observed between both arms.

#### Table 4 | RCTs evaluating PCI-related OS in NSCLC.


*OS, overall survival; PCI, prophylactic cranial irradiation; NSCLC, non-small cell lung cancer; RCTs, randomized controlled trials.*

In addition to BM occurrence, all included trials, except for the trial of Cox et al. (6), also reported on PCI-related OS in NSCLC (**Table 4**). Nearly all studies only report on short-term survival, and most trials did not use contemporary staging and systemic therapy. Taking these shortcomings in mind, no statistically significant OS difference was found between the PCI arm and no PCI arm, except for the study of Miller et al. (9, 10), which showed a significant OS benefit in favor of the no PCI arm (*p* = 0.004).

## DISCUSSION

This study aimed to review published literature and revisit the role of PCI in the prevention of BM in NSCLC. The analysis included seven RCTs, involving 1,462 NSCLC patients in total. The current meta-analysis shows that the risk of developing BM was reduced in patients who received PCI compared to patients who did not (RR 0.33; 95% CI 0.22–0.45).

Previously published results from reviews that investigated the role of PCI in the prevention of BM in NSCLC are in line with our findings. The most recently published meta-analyses of Sun et al. (4) and Xie et al. (19), evaluating the impact of PCI on BM occurrence in NSCLC, showed highly significant results in favor of PCI. However, these reviews could not evaluate the recent RCTs of Li et al. (13) and De Ruysscher et al. (14). Including the two most recent RCTs does not only add to the sample size of our meta-analysis, but the proper methodological quality and the use of more advanced brain imaging methods also add value to the conclusiveness of our results. Furthermore, unlike our study, the RCT of Pottgen et al. (12) was included in these meta-analyses. Local treatment was different between the two arms in this study (primary curative resection followed by postoperative thoracic radiation therapy vs chemotherapy and concurrent chemoradiotherapy followed by thoracic surgery). Therefore, we judged that it was not possible to assess the impact on prevention of BM attributable to PCI.

Although this is the largest, most recent review incorporating all available evidence from RCTs on the impact of PCI on the prevention of BM in NSCLC, there are also limitations. Egger's test and visual inspection of the funnel plot indicated the presence of publication bias. However, sensitivity analysis showed that excluding smaller studies from the meta-analysis did not much alter the results. Furthermore, assessment of the methodological quality of the included studies showed that for some items risk of bias was high. Nevertheless, most of these high bias risk items could be found in the studies (7, 9, 10) that were excluded in the sensitivity analysis, and results remained largely similar to the original results.

The meta-analysis showed that the BM incidence was lower in patients who received PCI, but few trials also reported that PCI could cause toxicity resulting in a decline in QoL. Most occurring acute toxicities were fatigue, skin-related toxicity, and nausea or vomiting. Toxicities occurring on longer term were headache, dyspnea, and lethargy. In addition, the study of De Ruysscher et al. (14) also reported low-grade memory and cognitive functioning impairments. Therefore, the indications of PCI should be considered in the light of its potential (neuro)toxicity. QoL data were limited and not significantly different between the groups, no short-term OS benefit was shown, and the influence of PCI on long-term OS should be further investigated. It is necessary to further study the role of PCI in relation to neurocognitive decline and thus deterioration of QoL, and whether PCI could improve patients' long-term OS. In the era of more effective targeted therapy and immunotherapy, when extracranial disease is better controlled and patients are living longer, there may be increased importance of PCI. Moreover, hippocampal sparing techniques and medications such as memantine could be interesting future areas of research as alternatives to reduce toxicity and thus loss of QoL. Other areas of future research might include the role of MRI surveillance in combination with radical local treatment such as stereotactic radiosurgery or whole-brain radiotherapy. However, studies of the EORTC and RTOG showed that cure remains elusive in the overwhelming majority of these patients (20, 21).

#### CONCLUSION

The risk of developing BM was reduced in patients who received PCI compared to observation. To implement PCI as the standard treatment for patients with NSCLC, the impact of PCI on toxicity and QoL should be further investigated, as well as the impact on long-term OS. A future individual patient data meta-analysis with updated long-term OS could potentially produce definitive answers to these clinical questions.

#### AUTHOR CONTRIBUTIONS

WW wrote the initial draft of this review, with edits and revisions from all other authors, and especially BR and DR.

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at https://www.frontiersin.org/articles/10.3389/fonc.2018.00241/ full#supplementary-material.

## REFERENCES


cerebral metastases after adjuvant chemotherapy. *Ann Oncol* (2015) 26:504–9. doi:10.1093/annonc/mdu567


**Conflict of Interest Statement:** LH: Consulting or Advisory Role: Boehringer Ingelheim and BMS. Fees for lectures: MSD, Astra Zeneca, Roche. Research grant: Roche. A-MD: Consulting or Advisory Role: Genentech (Inst), MSD Oncology (Inst), AstraZeneca (Inst), Pfizer, Eli Lilly (Inst), Boehringer Ingelheim (Inst), Bristol-Myers Squibb, Clovis Oncology (Inst). DR: Consulting or Advisory Role: Bristol-Myers Squibb (Inst), Genentech (Inst), Merck Serono (Inst), AstraZeneca (Inst), Celgene (Inst). Research Funding: Bristol-Myers Squibb (Inst).

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Copyright © 2018 Witlox, Ramaekers, Zindler, Eekers, van Loon, Hendriks, Dingemans and De Ruysscher. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Invasive aspergillosis Mimicking Metastatic Lung Cancer

*Michiel J. E. G. W. Vanfleteren1,2,3\*, Anne-Marie C. Dingemans <sup>2</sup> , Veerle F. Surmont <sup>1</sup> , Karim Y. Vermaelen1 , Alida A. Postma4 , Astrid M. L. Oude Lashof <sup>5</sup> , Cordula C. M. Pitz <sup>6</sup> and Lizza E. L. Hendriks <sup>2</sup>*

*1Department of Respiratory Medicine, University Hospital Ghent, Ghent, Belgium, 2Department of Respiratory Medicine, Maastricht University Medical Centre (MUMC*+*), Maastricht, Netherlands, 3Department of Respiratory Medicine, Sint-Jozefskliniek Izegem, Izegem, Belgium, 4Department of Radiology and Nuclear Medicine, Maastricht University Medical Centre (MUMC* +*), Maastricht, Netherlands, 5Department of Medical Microbiology, Section Infectious Diseases, Maastricht University Medical Centre (MUMC* +*), Maastricht, Netherlands, 6Department of Respiratory Medicine, Laurentius Hospital Roermond, Roermond, Netherlands*

#### *Edited by:*

*Marco Lucchi, Università degli Studi di Pisa, Italy*

#### *Reviewed by:*

*Christopher William Seder, Rush University Medical Center, United States Janaki Deepak, University of Maryland, Baltimore, United States*

#### *\*Correspondence:*

*Michiel J. E. G. W. Vanfleteren michielvanfleteren@msn.com*

#### *Specialty section:*

*This article was submitted to Thoracic Oncology, a section of the journal Frontiers in Oncology*

*Received: 18 February 2018 Accepted: 11 May 2018 Published: 05 June 2018*

#### *Citation:*

*Vanfleteren MJEGW, Dingemans A-MC, Surmont VF, Vermaelen KY, Postma AA, Oude Lashof AML, Pitz CCM and Hendriks LEL (2018) Invasive Aspergillosis Mimicking Metastatic Lung Cancer. Front. Oncol. 8:188. doi: 10.3389/fonc.2018.00188*

In a patient with a medical history of cancer, the most probable diagnosis of an 18FDGavid pulmonary mass combined with intracranial abnormalities on brain imaging is metastasized cancer. However, sometimes a differential diagnosis with an infectious cause such as aspergillosis can be very challenging as both cancer and infection are sometimes difficult to distinguish. Pulmonary aspergillosis can present as an infectious pseudotumour with clinical and imaging characteristics mimicking lung cancer. Even in the presence of cerebral lesions, radiological appearance of abscesses can look like brain metastasis. These similarities can cause significant diagnostic difficulties with a subsequent therapeutic delay and a potential adverse outcome. Awareness of this infectious disease that can mimic lung cancer, even in an immunocompetent patient, is important. We report a case of a 65-year-old woman with pulmonary aspergillosis disseminated to the brain mimicking metastatic lung cancer.

Keywords: lung cancer, lung neoplasms, brain metastasis, brain neoplasms, brain abscess, aspergillosis, differential diagnosis

#### CASE DESCRIPTION

A 65-year-old woman, never-smoker, was referred for a second opinion in January 2014 because of an abnormal computed tomography (CT) of the chest with a mass in the right lower lobe. Extensive evaluation in the referring hospital had not revealed a diagnosis. A clear overview of the medical disease history is demonstrated in a timeline (**Figure 1**). Her medical history consisted of a rightsided mastectomy for breast cancer in 2006, with no adjuvant treatment indicated. On the staging 18-fluordeoxyglucose positron emission tomography-computed tomography (18FDG-PET-CT) for the breast cancer in 2006 an asymptomatic, 30-mm diameter, lobulated 18FDG-negative solitary mass was seen in the right lower lobe. Bronchoscopic sampling for cytology and microbiological cultures showed neither proof of malignancy nor infection, and follow-up was chosen. Serial follow-up chest CTs up to December 2011 (total follow-up of 5 years) showed no change and follow-up was ended.

In August 2013, she was seen in the referring hospital because of a productive cough, dyspnea on exertion, tiredness, and weight loss since the last 3 months. There was no fever nor night sweats. 18FDG-PET-CT revealed an intense 18FDG-avid mass in the right lower lobe, just next to the old pulmonary lesion, with intense hilar lymphadenopathy (**Figure 2**). Differential diagnosis consisted of malignancy (metastatic breast cancer or primary lung cancer) or infection. Bronchoscopic

**27**

magnetic resonance imaging.

FIGURE 2 | Continued

FIGURE 2 | Evolution of thoracic lesions. Top: Follow-up chest computed tomography (CT) in 2011 showing a right-sided lobulated pulmonary mass at the right lower lobe (3.0-cm diameter). Middle: CT (left) and fusion 18-fluordeoxyglucose positron emission tomography-computed tomography (18FDG-PET-CT) (right) in August 2013 shows an increase at the medial side of the mass and right hilar lymphadenopathy, with intense 18-fluordeoxyglucose (18FDG) uptake. Bottom: CT (left) and fusion 18FDG-PET-CT (right) in January 2014 showing further growth of the 18FDG-avid mass in the right lower lobe with hilar invasion and a mild 18FDG-avid subcarinal lymph node.

#### TABLE 1 | Diagnostic test results.


*NA, not available.*

sampling, CT-guided biopsy, and endobronchial ultrasound with transbronchial needle aspiration (EBUS-TBNA) revealed no malignancy and cultures were negative (**Table 1**). Transthoracic biopsy of the mass showed fibrosis and a chronic inflammation with histiocytic reaction. A video-assisted thoracoscopy with partial wedge resection of the pulmonary nodule was performed to obtain a definitive diagnosis and to rule out or confirm malignancy. Pathologic analysis of the resection specimen showed fibrosis with bronchiectasis, focal inflammation, and bronchiolisation of the alveoli, but no malignancy or microorganisms (**Table 1**). With a diagnosis suggestive of cryptogenic organizing pneumonia, prednisone 0.5 mg/kg/day was initiated, although, due to steroid side-effects, limited in dose and duration (prednisone 0.5 mg/kg/day for 3 weeks followed by gradual tapering of the dose, total duration of prednisone treatment was 2 months). Despite the steroids, her complaints worsened and the patient was referred to our tertiary center in January 2014. A new 18FDG-PET-CT showed progression of the 18FDG-avid lesion with extension into the mediastinum and lymph node station 4R, with also mass effect on the right pulmonary artery and invasion of the left superior pulmonary vein (**Figure 2**). There was no evidence of extrathoracic lesions. Because of the invasive growth, malignancy was again in the differential diagnosis. In our hospital, renewed analysis was performed with bronchoscopy and mediastinoscopy, both without evidence of malignancy or infection (**Table 1**). Cytology of the bronchial aspirate showed sporadic hyphae (most probable *Aspergillus*), but without growth on culture and these were considered contamination or colonization. In March 2014, a multidisciplinary decision was made for a treatment with a higher dose of prednisone 1 mg/kg/day in combination with macrolide antibiotic treatment for 3 months (with slow tapering of the steroids) under suspicion of cryptogenic organizing pneumonia, but without clinical or radiologic response. Additional diagnostics were considered for the growing part of the lesion, but a CT-guided biopsy and surgical sampling were both not possible because of the risk of a massive bleeding. A follow-up 18FDG-PET-CT was scheduled to evaluate whether in the follow-up lesions would become better accessible for further diagnostic work-up. During the pulmonary work-up, in March 2014, the patient developed new complaints of progressive muscle weakness and sensibility loss of the right upper arm. She was hospitalized in the referring hospital. Additional brain imaging with magnetic resonance imaging (MRI) revealed multiple brain lesions in the cortex and watershed region, in the left corpus callosum, the left thalamus and partially in the right semioval center, which were considered brain metastases by the referring hospital (**Figure 3**). The MRI was revised by an experienced neuro-radiologist in our hospital who withheld a differential diagnosis of ischemia and metastasis. The brain lesions were not accessible for a biopsy because of the location and small size. Because of a poor clinical condition combined with a differential diagnosis of ischemia, no brain radiation was initiated. Patient was discharged after 1½-week hospitalization in the referring hospital. The neuro-oncology multidisciplinary team advised follow-up MRI 3 months later. On this MRI (May 2014), two lesions had enlarged significantly with marked perilesional edema but other lesions showed shrinkage (**Figure 4**). The radiologic appearance with restrictive diffusion of these lesions on diffusion-weighted images was suggestive for (atypical) cerebral abscesses rather than metastases. The 18FDG-PET-CT showed further growth of the mass in the right lower lobe but without distant lesions. At the same time, the patient developed multiple ill-defined skin lesions. Because of her worse clinical condition, she was hospitalized in our tertiary hospital and cerebral and skin biopsies were performed: both showed marked inflammation and fungal hyphae with dichotomous branching, suggestive for *Aspergillus*, there was no evidence for malignancy (**Table 1**; **Figures 5** and **6**). Cultures of the wound fluid after skin biopsy also revealed *Aspergillus* (**Table 1**). The definitive diagnosis of proven invasive aspergillosis with pulmonary, mediastinal, cerebral, and skin involvement was made. Treatment with voriconazole was initiated with monitoring of serum and cerebral spinal fluid (CSF) voriconazole levels. Because of progressive somnolence caused by hydrocephalus, repeated CSF drainage was necessary. Eventually, five neurosurgical procedures were needed for effective control of the infection and adequate drainage of the CSF, with in the end placement of an internal ventriculo-peritoneal drain. During treatment with voriconazole, there was a slow clinical recovery. Additional immunological analysis did not reveal an immunity disorder; there were normal titers of total IgG, IgM, IgG, and IgA, there were no complement abnormalities, screening for antinuclear antibodies and antineutrophil cytoplasmic antibodies was negative. Only the use of prednisone could be identified as immunosuppressant factor which aided the further dissemination of this opportunistic infection. MRI of the brain performed after 6 months of treatment showed marked improvement without evidence of hydrocephalus, the chest CT also improved. Patient was discharged in December 2014, 7 months after admission. She rehabilitated and made a near complete recovery.

## DISCUSSION

We report our experience of the diagnostic dilemma in this patient with disseminated aspergillosis mimicking metastatic cancer. The

FIGURE 3 | Brain magnestic resonance imaging in March 2014. T1-weighted image after gadolinium of the brain shows a small right frontal enhancing cerebral lesion.

FIGURE 4 | Brain magnetic resonance imaging in May 2014. There is an increase in size of the right frontal lesion with surrounding perilesional edema. T2-weighted image (left) demonstrates a hypo-intense rim with ring-enhancement after gadolinium (contrast-enhanced T1-weighted middle). At diffusion imaging (right panels) there is restricted diffusion in a part of the central area.

stain on skin biopsy with fungal hyphae stained purple. Two fungal hyphae with dichotomous branching (diagnostic of *Aspergillus*) are depicted (arrows).

18FDG-avid pulmonary lesions were highly suspicious for malignancy, especially in a patient with a history of breast cancer. Even in the presence of brain lesions, this suspicion remained high, as these brain lesions were first thought to be of metastatic origin. Eventually skin and cerebral biopsies and wound cultures did reveal the definitive diagnosis of disseminated invasive aspergillosis.

As a tertiary center, patients are frequently referred to our center with presumed lung cancer. In a retrospective analysis of a tertiary US hospital, the majority of such patients were proven to have a neoplastic process, only 1.3% had an infection (1). Pulmonary aspergillosis mimicking lung malignancy remains rare and only sporadic case reports are available in literature (1–7). In addition, symptoms (such as malaise, weight loss, cough, and hemoptysis) are non-specific and are overlapping those of a pulmonary neoplasm. Moreover, pulmonary aspergillosis can present as an infectious pseudotumour with radiological

FIGURE 6 | Cerebral biopsy with presence of fungal hyphae. Hematoxylin and eosin stain on cerebral biopsy showing nectrotic tissue with moderate numbers of septate fungal hyphae with parallel walls. Two fungal hyphae with dichotomous branching (diagnostic of *Aspergillus*) are depicted (arrows).

appearance and features similar and indistinguishable from lung cancer. When clinical and radiological features are suspected for malignancy, it is of utmost importance to strive for a definitive histopathological diagnosis. Many minimal invasive techniques such as bronchoscopy, CT-guided biopsy or EBUS-TBNA are available today to obtain this histopathological diagnosis. If not amenable or feasible or histopathological diagnosis cannot be obtained, a surgical approach might be necessary. The differential diagnosis can be very challenging and perseverance for diagnostic accuracy can be a hard and exhaustive exercise, which is demonstrated in our case. Despite several attempts with noninvasive and invasive procedures to obtain a histopathological diagnosis, there was no clear evidence for malignancy or infection. Maybe in retrospect, the hyphae in the bronchial aspirate could have raised the suspicion for invasive aspergillosis, although this is very rare in an immunocompetent host such as our patient. In retrospective studies with immunocompromised patients with a diagnosis of invasive aspergillosis, cytological examination of bronchial washings had a sensitivity of 64.0%, specificity of 99.1%, and positive predictive value of 88.9% (8), However, predictive values depend upon the prevalence of disease in the population tested and our patient did not have an impaired immunity. Moreover, various species of *Aspergillus* spp*.* can colonize the airways, especially in patients with a chronic pulmonary disease, without any pathogenic consequences (as was thought to be the case in our patient), but they are also capable of causing several and severe types of disease as has been described in patients with bronchiectasis (9). Pathological features in the surgical specimen suggestive for cryptogenic organizing pneumonia made this case more complex. Furthermore, the suspicion of malignancy remained high, with further growth of her thoracic disease and development of brain lesions, suspicious for brain metastasis. Indeed, cerebral abscesses caused by *Aspergillus* spp. can also mimic cerebral metastasis. Contrast-enhanced CT and MRI are the modalities of choice for imaging of the brain when brain metastasis is suspected, with MRI more appropriate in characterizing lesions (10–13). Although many MR features have been described, the differentiation between abscesses and necrotic brain tumors cannot be made in many cases with conventional MR imaging since its signal appearance can be similar to that of a cystic or necrotic tumor on conventional series (14, 15). A combination with diffusion-weighted MRI has been shown to be useful in the diagnosis of acute cerebral ischemia, malignancy, abscesses, cysts, and various forms of white matter disorders (16). In our case, the dissociated response, the hypo-intense rim at T2-weighted imaging and the diffusion-weighted MRI aided toward an infectious diagnosis. Previous reports in literature of a cerebral *Aspergillus* abscess mimicking a solid tumor are sparse; we could only identify two case resports (15, 17).

Coexistence of infectious pseudotumours and solid tumors at initial diagnosis have previously been reported (18), but are rather rare, especially in an immunocompetent patient. However,

#### REFERENCES


an endobronchial aspergilloma is thought to be able to infect endobronchial cancer lesions (19, 20). In general, most cases of coexisting infectious pseudotumours and lung cancer are rather a consequence of treatments with corticosteroids and/or chemotherapy.

## CONCLUSION

Pulmonary aspergillosis, even in the presence of cerebral abscesses, can present as an infectious pseudotumour with clinical and imaging characteristics resembling lung malignancy. These clinical and radiological similarities can cause significant diagnostic difficulties, with a subsequent therapeutic delay and a potential adverse outcome. A definitive histopathological diagnosis should always be strived for when malignancy is suspected, but awareness that this infectious disease can mimic lung cancer even in immunocompetent patients is of great diagnostic and prognostic importance.

## ETHICS STATEMENT

Patient gave written informed consent in accordance with the Declaration of Helsinki.

## AUTHOR CONTRIBUTIONS

Conception and design and drafting of the manuscript: MV, AD, and LH. Drafting the manuscript for important intellectual content: all authors. All co-authors critically revised the article and gave final approval of this version to be published.

#### ACKNOWLEDGMENTS

The authors wish to thank C. J. Kootstra, department of Pathology, Maastricht University Medical Centre (MUMC+), Maastricht, The Netherlands, for the selection and preparation of artwork.


metastatic lung cancer: a case report and literature review. *Respir Investig* (2016) 54(3):211–5. doi:10.1016/j.resinv.2015.12.005

20. Zhou XM, Li P, Zhao L, Shi WJ. Lung carcinosarcoma masked by tracheobronchial aspergillosis. *Intern Med* (2015) 54(15):1905–7. doi:10.2169/ internalmedicine.54.3945

**Conflict of Interest Statement:** MV has nothing to disclose. AD: none related to current manuscript. Outside of the submitted work: advisory boards: Roche, Pfizer, Eli Lilly Boehringer Ingelheim, MSD, Astra Zeneca, BMS; fees for lectures: BMS, Roche, Ely Lilly. VS: none related to current manuscript. KV: none related to current manuscript. AP: none related to current manuscript. Outside of the submitted work: fees for lectures: Bayer. AL: none related to current manuscript. CP has nothing to disclose. LH: none related to current manuscript. Outside of the submitted work: advisory boards: Boehringer Ingelheim and BMS; fees for lectures: MSD, Astra Zeneca, Roche, research grant: Roche.

*Copyright © 2018 Vanfleteren, Dingemans, Surmont, Vermaelen, Postma, Oude Lashof, Pitz and Hendriks. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Stereotactic Radiosurgery in the Management of Patients With Brain Metastases of Non-Small Cell Lung Cancer: Indications, Decision Tools and Future Directions

*Dianne Hartgerink1 , Britt van der Heijden1 , Dirk De Ruysscher1,2, Alida Postma3 , Linda Ackermans <sup>4</sup> , Ann Hoeben5 , Monique Anten6 , Philippe Lambin1 , Karin Terhaag1 , Arthur Jochems <sup>1</sup> , Andre Dekker 1,2, Janna Schoenmaekers <sup>7</sup> , Lizza Hendriks7 and Jaap Zindler1,2\**

*1Department of Radiation Oncology (MAASTRO Clinic), GROW School for Oncology and Developmental Biology, Maastricht University Medical Centre, Maastricht, Netherlands, 2Proton Therapy Department South-East Netherlands (ZON-PTC), Maastricht, Netherlands, 3Department of Radiology, Maastricht University Medical Centre, Maastricht, Netherlands, 4Department of Neurosurgery, Maastricht University Medical Centre, Maastricht, Netherlands, 5Department of Medical Oncology, Maastricht University Medical Centre, Maastricht, Netherlands, 6Department of Neurology, Maastricht University Medical Centre, Maastricht, Netherlands, 7Department of Pulmonary Diseases, Maastricht University Medical Centre, Maastricht, Netherlands*

Brain metastases (BM) frequently occur in non-small cell lung cancer (NSCLC) patients. Most patients with BM have a limited life expectancy, measured in months. Selected patients may experience a very long progression-free survival, for example, patients with a targetable driver mutation. Traditionally, whole-brain radiotherapy (WBRT) has been the cornerstone of the treatment, but its indication is a matter of debate. A randomized trial has shown that for patients with a poor prognosis, WBRT does not add quality of life (QoL) nor survival over the best supportive care. In recent decades, stereotactic radiosurgery (SRS) has become an attractive non-invasive treatment for patients with BM. Only the BM is irradiated to an ablative dose, sparing healthy brain tissue. Intracranial recurrence rates decrease when WBRT is administered following SRS or resection but does not improve overall survival and comes at the expense of neurocognitive function and QoL. The downside of SRS compared with WBRT is a risk of radionecrosis (RN) and a higher risk of developing new BM during follow-up. Currently, SRS is an established treatment for patients with a maximum of four BM. Several promising strategies are currently being investigated to further improve the indication and outcome of SRS for patients with BM: the effectivity and safety of SRS in patients with more than four BM, combining SRS with systemic therapy such as targeted agents or immunotherapy, shared decision-making with SRS as a treatment option, and individualized isotoxic dose prescription to mitigate the risk of RN and further enhance local control probability of SRS. This review discusses the current indications of SRS and future directions of treatment for patients with BM of NSCLC with focus on the value of SRS.

Keywords: brain metastases, non-small cell lung cancer, stereotactic radiosurgery, isotoxic dose prescription, shared decision

#### *Edited by:*

*Yanis Boumber, Fox Chase Cancer Center, United States*

#### *Reviewed by:*

*Gregory Gan, University of New Mexico, United States Michael T. Milano, University of Rochester, United States Stephanie Elizabeth Weiss, Fox Chase Cancer Center, United States Rachel E. Sanborn, Providence Cancer Center, United States*

#### *\*Correspondence:*

*Jaap Zindler jaap.zindler@maastro.nl*

#### *Specialty section:*

*This article was submitted to Thoracic Oncology, a section of the journal Frontiers in Oncology*

*Received: 30 January 2018 Accepted: 24 April 2018 Published: 09 May 2018*

#### *Citation:*

*Hartgerink D, van der Heijden B, De Ruysscher D, Postma A, Ackermans L, Hoeben A, Anten M, Lambin P, Terhaag K, Jochems A, Dekker A, Schoenmaekers J, Hendriks L and Zindler J (2018) Stereotactic Radiosurgery in the Management of Patients With Brain Metastases of Non-Small Cell Lung Cancer: Indications, Decision Tools and Future Directions. Front. Oncol. 8:154. doi: 10.3389/fonc.2018.00154*

## INTRODUCTION

Brain metastases (BM) are the most frequent intracranial malignancies and originate mainly from lung cancer (1). In patients with driver mutations of non-small cell lung cancer (NSCLC), systemic therapies have become more effective in patients with metastatic disease, resulting in longer overall survival (OS). Due to the screening for BM, the longer OS, and the often poor drug penetration through the blood–brain barrier (BBB), more and more patients are diagnosed with BM. BM may cause neurologic symptoms, a decrease in quality of life (QoL), and are often associated with poor OS (2).

Overall, patients with BM are treated with the intention to maintain QoL during their remaining lifespan. Traditionally, the treatment consists predominately of radiotherapy, mainly wholebrain radiotherapy (WBRT), but in selected patients, surgery, systemic therapy, or a combination of treatment modalities is used. Depending on the prognostic subgroups, the OS after WBRT in patients with BM of NSCLC without systemic treatment remains poor with an estimated survival of weeks or months (3–5). For instance, for patients treated with WBRT and optimal supportive care in the QUARTZ trial, the median survival was 8.5 weeks, and there was no OS benefit (4). Physicians are increasingly reluctant in the use of WBRT, as the results of the QUARTZ trial did not show a benefit of WBRT in NSCLC patients over steroids alone in patients with an intermediate or unfavorable prognosis [recursive partitioning analysis (RPA 2–3)] (4).

It is increasingly important to accurately estimate the prognosis after all treatment options, to support decision-making of both patients and physicians. OS of patients with BM of NSCLC ranges from several weeks to several years depending on relevant prognostic factors, such as performance status, age, control of extracranial disease, number of BM, and the presence of driver mutations. Gaspar et al. published a report in 1997 on a prognostic index for patients with BM, the RPA based on patients mainly treated with WBRT. The RPA was externally validated, and the favorable prognostic RPA score had a median survival of 7.1 months. The unfavorable RPA score had a survival of only 2.3 months (6, 7). The weakness of the RPA score is that the majority of the patients are classified into the intermediate prognostic class, and for clinical decision-making the favorable and unfavorable prognostic classes are the most important ones Also, the RPA score was developed in the pre-immunotherapy era, and cancers other than NSCLC were included into this score which limits its utility in patients with BM of NSCLC. Therefore, the Graded Prognostic Assessment (GPA) was developed from a database of almost 2,000 patients with BM, validated and refined with diagnosis-specific (DS) indices based on a second retrospective analysis of 4,259 patients with BM (8–10). For patients with BM of NSCLC and an unfavorable DS GPA score, the median survival time is 3 vs 15 months for patients with a favorable DS-GPA score (11). Recently, a refined Lung-molGPA score was developed and validated specifically for patients with BM of NSCLC. molGPA integrates molecular features such as epidermal growth factor receptor (EGFR) and anaplastic lymphoma kinase (ALK) alterations (12, 13). The overall median survival for the cohort was 12 months, and those patients with a Lung-molGPA score of 3.5–4.0 had a median survival of almost 4 years. It is for these reasons that lung-moGPA is the most useful prognostic tool for clinical practice in the era of personalized medicine and targeted agents.

In the last decades, a local alternative treatment became widely available, stereotactic radiosurgery (SRS). SRS has the advantage of achieving higher local tumor control while sparing healthy brain tissue, which results in less severe side effects such as neurocognitive damage and hair loss (14). SRS is currently a wellestablished treatment modality for patients with a maximum of four BM. SRS is also a more complex and costly treatment which may not be available in every radiotherapy department. Other disadvantages of SRS compared with WBRT are a higher risk of new BM during follow-up (e.g., distant brain recurrences), and an increased risk of radionecrosis (RN) depending on the volume of healthy brain tissue which is irradiated to a relatively high dose, tumor biology factors, and the location of the tumor. RN is focal damage of the nearby brain tissue caused by a high dose of radiation. The result of RN can be temporary or permanent neurologic symptoms. These symptoms can be treated with steroids, but steroids have several side effects such as obesity, sleeping disorders, hyperglycemia, and muscle weakness (15–17). The risk of RN is mainly correlated with the SRS dose in the brain and the size of the lesion, for example, if the volume of brain tissue which receives ≥12 Gy is more than 10 cm3 , the risk of RN increases to above 10% and can go as high as 25%. However, other factors may contribute. The risk of distant brain recurrences is mainly correlated with the number of SRS treated BM and varies between 40 and 90% if SRS is applied as a single treatment modality (14, 18). From all patients treated with SRS as a single treatment modality for a maximum of three BM, 25% of patients will receive WBRT, because a significant proportion of patients die from extra cranial disease progression. It should be stated that almost all long-term survivors will undergo WBRT at some time point (14, 19, 20).

For small BM, SRS is equally effective as surgical resection (21). SRS is typically delivered in one to five fractions of multiple photon beams, but more recently even multiple, i.e., more than four, BM can be treated with a single fraction (22). For patients with more than four BM, the current Dutch guidelines advises WBRT as standard of care, but trials are ongoing and already conducted to investigate the value of SRS in patients with more than three BM (23–26).

The aim of this review is to discuss the current evidence of SRS for BM of NSCLC, potential improvements in patient information with focus on shared decision-making (SDM), and promising future treatment strategies to improve the clinical outcome.

## CURRENT EVIDENCE FOR SRS AS A SINGLE TREATMENT MODALITY

As mentioned previously, the indication of WBRT is currently a matter of debate: according to the QUARTZ trial, there was no beneficial effect of WBRT over steroids, in NSCLC patients, most with an RPA class 2–3, with respect to QoL or survival (4). There are several remarks with respect to clinical application of the results: patients were only included if the patient or the multidisciplinary team had doubts if WBRT should be applied, favorable prognostic patients (RPA class I) did have a survival benefit of WBRT, and the effect on symptom control such as seizures or headache were not described in detail. The latter is relevant, because the reason for applying WBRT is sometimes symptom control, prevention of neurologic symptom progression, or prevention of dying due to a neurologic cause. The availability of drugs used to target different areas for these patients is constantly increasing which may influence their outcome. Taking into account the limitations of the study which is conducted in a poor prognostic population, physicians should be more reluctant to apply WBRT (RPA class 3).

As an alternative to SRS, patients with small asymptomatic BM of NSCLC can be treated with systemic therapy (chemotherapy, targeted therapy, or immunotherapy in the second line). Targeted therapy can be considered if there is a driver mutation (27, 28). Cranial radiation can be considered if a low response rate is to be expected from systemic treatment or if neurological symptoms are to be expected in the event of disease progression.

Nowadays, the preference for either SRS or WBRT depends on the size and number of BM, but only if patients are in a good physical condition (Karnofsky performance status 70 or more). Patients with a maximum of four BM are usually treated with SRS, and patients with more than four BM are treated with WBRT. The size and location of the lesions are decisive factors: a very large BM with a diameter of 6 cm is inappropriate for treatment with SRS or a large brainstem metastasis, whereas five minor lesions located in the cerebral hemispheres are technically less challenging for treatment with SRS. In patients with a poor performance status (Karnofsky performance status of less than 70) are usually treated with supportive care. Patients treated with SRS despite a poor physical condition still have a very poor prognosis with a median survival of around 3 months (2).

The treatment of patients with asymptomatic BM from nonsquamous NSCLC depends on molecular diagnostics, but primary systemic treatment with deferral of radiotherapy is a treatment option that could be considered (29, 30). For NSCLC, Lim et al. randomized 105 patients with 1–4 asymptomatic BM to receive SRS followed by chemotherapy or upfront chemotherapy alone (31). The trial closed early due to slow recruitment and was therefore underpowered. SRS followed by chemotherapy did not improve the OS compared with upfront chemotherapy (14.6 vs 15.3 months, *p* = 0.418). The time to central nervous system (CNS) progression was not significant different between the two arms [9.4 months (SRS) vs 6.6 months (upfront chemotherapy), *p* = 0.248].

Also for patients with asymptomatic BM SRS is an attractive first line of treatment. Approximately 25% of the patients with an EGFR-mutated NSCLC have BM at first presentation. For patients with an EGFR targetable mutated NSCLC, an alternative first line of treatment consists of an EGFR inhibitor like erlotinib, gefitinib, osimertinib, or afatinib with a response rate of approximately 60–70% (30, 32). Only a small percentage passes through the BBB and the penetration is different between the treatment options whereas osimertinib has a greater penetration. However, the response probability is equal to the extra-cerebral response probability (33, 34). Approximately 24% of ALK translocated NSCLC patients have BM at presentation (35). In case of an ALK translocation the treatment consists of alectinib, ceritinib, or crizotinib, with an expected response rate of approximately 50–60%, whereas alectinib has a superior CNS activity compared with crizotinib (36). If cerebral progression occurs during treatment with an EGFR or ALK inhibitor, SRS is considered, or a second line of systemic therapy. A recent study provides evidence for upfront SRS in patients with asymptomatic BM. Patients who are tyrosine kinase inhibitors (TKI) naïve and have an EGFR mutation had a better survival than patients who were treated with primary systemic therapy (37). This study was limited by its retrospective nature. Prospective randomized studies are needed to investigate if upfront SRS without TKI, directly followed by TKI or even concurrent, is beneficial over primary systemic treatment without SRS in TKI naïve patients.

#### SRS for a Maximum of Four BM

The definition of limited BM traditionally consisted of patients presenting with a single BM, often treated with surgery. This definition has evolved to encompass patients presenting with up to three metastasis for treatment with SRS (38–41). The management of patients presenting with a limited number of BM and a good performance status has developed from WBRT alone to a more aggressive approach consisting of WBRT in combination with SRS (41–43). The necessity of WBRT was evaluated *via* clinical trials, which compared SRS alone to SRS with adjuvant WBRT (44–48).

Aoyama et al. reported the first randomized control trial comparing SRS alone with SRS plus WBRT, randomizing 132 patients with 1–4 BM from histologically confirmed systemic cancer, mainly NSCLC (67%) (46). The primary endpoint was cranial recurrence. Although the 1-year local control rate was higher in the SRS plus WBRT group (88.7 vs 72.5%, *p* = 0.002), there was no OS difference between the two study arms (trial was underpowered for the secondary endpoint survival). There was also no advantage with respect to cognition based on Mini-Mental Status Exam (MMSE) for patients receiving SRS plus WBRT. It should be taken into account that the MMSE is a crude measurement for neurocognition compared with a standardized test such as the Hopkins Verbal Learning Test (HVLT) or other neurocognitive tests proposed by the European Organization for Research and Treatment of Cancer (EORTC), for example (49–51). A secondary analysis of the JROSG 99-1 randomized clinical trial investigated the feasibility of SRS alone for patients with different prognoses defined by the DS-GPA (24). Significantly better OS was observed in the DS-GPA favorable group (scores 2.5–4.0) in WBRT + SRS vs SRS alone. However, it is questionable if the side effects of adding WBRT to SRS, justify a potential limited survival benefit in favorable prognostic patients.

Chang et al. also randomized patients into SRS alone or SRS plus WBRT treatment arms, but they took a different approach by evaluating patients' neurocognition using the HVLT-revised (HVLT-R) as the primary endpoint (47). Patients presenting with 1–3 BM from different primary cancers, mainly NSCLC (55%), were randomized, 30 patients to SRS alone and 28 to SRS plus WBRT. The trial was prematurely stopped because at the interim analysis, patients in the SRS plus WBRT arm were more likely to have a decline in neurocognitive function 4 months posttreatment. They found an unexpected survival advantage (secondary endpoint) in the SRS alone arm, with an OS of 15.2 vs 5.7 months in the in the SRS alone, and SRS plus WBRT arms, respectively. The reasons for higher survival rates in the SRS alone arm were unclear. It may be explained by a higher rate of salvage cranial treatment. Moreover, chemotherapy was administered to more patients with a longer duration in the SRS group compared with the SRS plus WBRT group (52). The authors concluded that the management for patients presenting with one to three BM with SRS alone is the optimal treatment.

The EORTC evaluated SRS alone vs SRS plus WBRT with the primary endpoint of functional outcomes, using the World Health Organization (WHO) performance status scale, in patients with one to three BM from mainly NSCLC (53%) (44). They concluded that WBRT did not improve the duration of functional independence (WHO ≤ 2, SRS alone 10.0 months vs SRS plus WBRT 9.5 months). SRS plus WBRT reduced the incidence of radiological endpoints, such as distant brain failure and 2-year local control failure rate (radiosurgery group: 31–19%, *p* = 0.040). Despite the secondary outcomes, the QoL was worse in several domains for patients who received WBRT (31, 44). A secondary analysis of the EORTC 22952-26001 trial investigated the impact of WBRT on patients with a favorable GPA prognostic score. The primary endpoint was OS (45). There was no significant survival benefit for NSCLC patients with a favorable GPA score of WBRT + SRS over SRS alone. There was also no survival benefit for patients with controlled extracranial disease. This secondary analysis supports the practice of treatment with SRS alone for patient with limited BM.

Recently, Brown et al. reported the results of the North Coast Cancer Treatment Group (NCCTG) phase III study in patients with one to three BM, mainly from lung cancer, treated with SRS alone or SRS plus WBRT (53). Regarding the primary endpoint of neurocognitive function, the trial is comparable with Chang et al., except that Brown et al. randomized 208 patients (47). Cognitive progression, defined as a decline of >1 SD from baseline on ≥1 of the cognitive tests 3 months post-SRS, was higher in the SRS plus WBRT arm compared with SRS alone (91.7 vs 63.5%, *p* < 0.001), and cognitive deterioration was more frequent in long-term survivors (living 12 months or more) receiving WBRT plus SRS compared with SRS alone (94.4 vs 60%). The 1-year intracranial disease control was 50.5% in the SRS alone arm and almost 85% in the SRS plus WBRT arm. The secondary survival analysis showed a median OS of 10.4 months for SRS alone vs 7.4 months for SRS plus WBRT. The authors concluded that SRS alone is preferred for patients presenting with one to three BM, supporting the results of Chang et al.

A secondary analysis of the NCCTG randomized control trial from Brown et al. was performed by Churilla et al. to determine whether WBRT is associated with improved OS among NSCLC patients with favorable prognoses (DS-GPA ≥ 2.0 or ≥2.5) at diagnosis (53, 54). They used two separate cut-points of DS-GPA, ≥2.0 vs <2.0 and ≥2.5 vs <2.5 in a study population consisting of 126 NSCLC patients with 1–3 BM. For patients with DS-GPA ≥ 2.0 treated with SRS alone, the median survival was 17.9 vs 11.3 months in the SRS plus WBRT arm (*p* = 0.63), and 6.6 vs 3.7 months for patients with DS-GPA < 2.0 (*p* = 0.85). They observed no significant differences in survival analysis in favorable-prognosis NSCLC patients treated with SRS, with or without WBRT, which further supports the approach of SRS alone. The above trials, summarized in **Table 1**, demonstrate that adjuvant WBRT results in reduced QoL and neurocognitive function without improvement of OS.


*KPS, Karnofsky performance status; WBRT, whole-brain radiotherapy; SRS, stereotactic radiosurgery; WHO, World Health Organization; HVLT-R, Hopkins Verbal Learning Test revised; OS, overall survival. NSCLC, non-small cell lung cancer. MMSE, Mini-Mental State Examination; DS-GPA; diagnosis-specific Graded Prognostic Assessment.*

The current available evidence supports the use of SRS as a single treatment modality in patients with a maximum of three BM. This is supported by the American Society for Therapeutic Radiation Oncology (ASTRO).

#### SRS for More Than Four BM

For patients presenting with >4 metastases, traditionally WBRT has been the standard of care. In patients with >4 metastases, the application of SRS is controversial, because the currently discussed randomized trials were done in patients with a limited number of BM. The additional palliative value of SRS over WBRT in patients with four or more BM remains to be determined. An international practice survey reported in 2010 showed the difference in consensus on SRS. In a responder-population consisting of SRS-specialists, 83% would consider patients >4 brain metastasis as potential candidate for SRS. By contrast, only 7% of a responder-population consisting of radiation oncologists agreed for SRS in patients with >4 BM (55). Physicians that support SRS in patients with >4 metastases indicate that in the past there where technical issues concerning the long treatment times and safety of the radiation doses (56–59).

The multi-institutional prospective study from Yamamoto et al. was the first evaluating SRS alone for four and more BM (22). The trial population consisted of favorable prognostic patients with low volume BM, three-quarters originate from primary lung cancer, the majority had an RPA 2 and KPS ≥ 80 (largest tumor <10 mL in volume, <3 cm in longest diameter; total cumulative volume ≤15 mL). This study included 1,194 patients with 1–10 metastases and where treated with SRS, split in to three cohorts: 208 patients with 5–10 metastases, 531 patients with 2–4 metastases, and 455 patients with a single metastasis. The intention was to determine non-inferiority in the cohort with 5–10 BM compared with 2–4 BM with OS as the primary endpoint. The OS did not differ between patients with 5–10 BM or 2–4 BM (HR 0.97, *p* = 0.78). As well as local control rates, distant brain relapses were not significantly different between both cohorts. This suggests that SRS without WBRT in patients with 5–10 BM is non-inferior to patients with 2–4 BM. SRS may be an alternative for WBRT as SRS is minimally invasive and has fewer side effects.

A second case-matched, retrospective cohort trial from Yamamoto et al. compared treatment results for patients with 10 or more BM vs 2–9 metastases (60). The primary endpoint was OS, whereas the secondary endpoints consisted of neurological death and deterioration, local recurrence and repeat SRS, and major complications of SRS. The median survival time between the two arms did not differ significantly, 6.8 months for patients with 2–9 BM vs 6.0 months for patients ≥10 BM (*p* = 0.10). Considering the incidence of neurological deterioration (defined as any brain disease-caused neurological worsening), there was no difference between the groups, including radiation-related complications. They concluded that treatment results after SRS were not inferior for patients with 10 or more BM than for patients with 2–9 metastases.

Serizawa et al. conducted a small retrospective study to compare the effectiveness of SRS (*N* = 62) with WBRT (*N* = 34) for multiple cranial metastases from non-small-cell lung cancer (61). They included patients with 1–10 BM with a life expectancy of more than 2 months, lesions >3 cm were surgically removed. The OS time and the neurological survival in the SRS arm were significantly longer. The risk of neurologically impaired QoL was also lower in the SRS arm.

The results of these studies support the hypothesis that SRS is a viable treatment option in patients with four or more BM. The main question if SRS is beneficial over WBRT with respect to QoL, survival, and maintenance of neurocognitive function, needs to be answered in randomized trials (25).

There are no published randomized trials evaluating the role of SRS in patients with ≥4 BM. In the Netherlands a randomized phase III trial (NCT02353000, https://clinicaltrials.gov/show/ NCT02353000) is enrolling patients with 4–10 BM, KPS ≥ 70, and any primary solid tumor including NSCLC. The standard treatment WBRT is compared with SRS for all lesions, with the primary endpoint being of QoL at 3 months after radiotherapy (25). Another randomized phase III trial (NCT01592968) at the MD Anderson Cancer Center is randomizing patients with 4–15 BM to SRS alone vs WBRT alone. The primary endpoints are cognitive function and local tumor control at 4 months. If these trials are successful, SRS will replace WBRT as the standard treatment for patients with more than four BM. Several other trials are currently being initiated to evaluate the role of SRS in patients with multiple BM, such as Heidelberg (NCT0329778) and Boston (NTC03075072), among others.

Another treatment option for patients with multiple BM, especially in patients with asymptomatic BM of a driver mutation, is to only treat the larger and worrisome BM with SRS (**Figure 1**). This strategy is of specific interest in subgroups of patients who may survive over several years due in part to several lines of targeted agents and to postpone the radiotherapy (either SRS or WBRT) until progression of intracranial disease.

## IMPROVING INDICATION OF CRANIAL IRRADIATION AS A TREATMENT OPTION FOR PATIENTS WITH BM DURING SDM

Especially in the palliative setting of BM, it will be increasingly important to better inform the patient about the available treatment options, such as SRS, to individualize the multimodality treatment of patients with BM of NSCLC.

Shared decision-making is based on the principle of the person's autonomy and to improve the participation of patients in making decision concerning their personal health and treatment. It can be divided into four parts: the health-care professional needs to inform the patient that a decision needs to be made and that the opinion of the patient also important is, the explanation from the health-care professional to the patient about the different options inclusive of advantages and disadvantages of the different treatment options, a discussion between the health-care professional and the patient about their preferences, and finally to make a decision about the chosen treatment (62).

Traditionally, physicians had a more paternalistic approach with respect to treatment choices based on national guidelines. However, for patients with BM with incurable cancer, it is of

specific relevance to consciously make choices between treatment options taking into account all advantages and disadvantages. With the availability of an increasing number of treatment options, SDM is challenging and complex, and work is in progress to make SDM tools available for the patient and caretakers.

## Decision Support Systems (DSSs) Including SRS as a Treatment Option

A DSS is a direct aid in clinical decision-making, in which characteristics of individual patients are used to generate patientspecific assessments or recommendations that are then presented to clinicians for consideration (63).

In radiotherapy oncology, the DSSs use advanced and innovative information technology, combined with available medical data to achieve the highest possible accuracy in the prediction of everything from the response of the tumor, to the treatment response and toxicity of normal tissue (64). The base of SDM is individualized prognostic models for outcome prediction per patient. Individualized prognostic models use machine learning algorithms to learn from patients treated in the past to make predictions for patients we currently see in the clinic. Traditionally, outcome was calculated per group of patients, such as the RPA (65).

Recently, individualized prognostic models became available for patients treated with SRS for a limited number of BM of NSCLC (66). These two models can predict individualized chances of both early death (<3 months), and long-term survival (>12 months), after SRS for BM of NSCLC, with an area under the curve (AUC) for early death of 0.79 with a 95% confidence interval (CI) of 0.72–0.86, and an AUC for long-term survival of 0.77 with a 95% CI of 0.70–0.84. The nomograms were more accurate in the prediction of early death and long-term survival than commonly used prognostic scores after SRS for a limited number of brain metastasis of NSCLC. Examples of commonly used prognostic scores are the RPA, the Golden Grading System, DS-GPA, and Score Index for Radiosurgery in brain metastasis (SIR) (66). Other than the increased accuracy of prediction, these nomograms are also easy to use in routine clinical practice and are available at www.predictcancer.org (66). Other tumor-specific individualized models are necessary for more relevant clinical endpoints, such as the occurrence of distant brain recurrences, local control probability, and the risk of RN or cognitive toxicity. The current individualized prognostic models are relatively simple and may become more accurate in outcome prediction if more patient-, treatment-, and tumor characteristics are added into these models. With respect to the latter genomic and radiological factors with biomarkers, radiomics, and deep learning are of specific interest (64, 67–70). Model performance is in part, dependent on the volume of patient data used on which to base these models (71). Unfortunately, sharing of data between hospitals is hampered by ethical, political, administrative, and legal barriers (72). Efforts have been made to learn from multiple hospitals while avoiding the hurdles associated with sharing data. The EuroCAT project is an example of one such effort, in which the distributed learning approach is championed (see the animation at https://www.youtube.com/watch?v=nQpqMIuHyOk) (73). Distributed learning is defined as learning from data stored at the hospital without the data leaving the hospital. As only model coefficients are transmitted, patient privacy is preserved while the data can still be used for teaching models. Models for survival and treatment-related side effects have been learned using this approach (64, 73–75).

## Patient Decision Aids Including SRS as a Treatment Option

Patient decision aids are based on prognostic models which can be individually adapted to the characteristics of the patient and their disease. The model consists of two key steps: information exchange and deliberation between the health-care professional and the patient (76). The goal of these tools is to achieve an optimal individualized treatment strategy. The patient benefits of SDM have been proven with level 1 evidence on more than 30,000 patients, and yet clinical decision-making remains complex; patients must not only weigh several treatment options in terms of benefits and harms but also absorb a large amount of technical information while dealing with the emotional burden of their disease (25, 77). Lack of awareness about treatment options can lead patients to choose treatments that are more expensive or not aligned with their values. To improve patients' QoL, health-care quality, and to reduce unnecessary procedures, it is crucial to empower patients with solid decision support. Evidence shows that patients who use decision aids are better informed about their treatment options, and experience less decisional conflict, and less anxiety about their care. Despite the evidence, patient decision aids are not routinely applied in practice (77). The shared decision tools are focused on the patient, which means that the level of these tools is also adjusted to the average patient. Medically complicated terms will be avoided as much as possible (78). An example of a potential patient decision aid is a tool which will be designed for SRS treatment in NSCLC patients consists of several headings: who is the main health-care professional during the treatment, what is the path during the diagnosis to brain metastasis, what makes the treatment eligible for the patient, what does the treatment consist of, what are the advantages of the treatment, what are the disadvantage of the treatment including possible adverse events, what is the influence of the treatment on the life of the patient, and what does the follow-up consist of. Ideally, every treatment option will be incorporated within this decision aid.

When the patient has run through this decision tool, the patient and the health-care professional will discuss this tool, and the patient has the opportunity to ask questions about the content and the information in this tool. After this discussion, the patient and the health-care professional will determine, together, if the SRS treatment is the best treatment option for this patient. The goal of this SDM is to obtain an optimal individualized treatment strategy by making use of the shared decision tool in which there is a deliberation between the health-care professional and the patient (78).

#### FUTURE DIRECTIONS OF SRS FOR BM OF NSCLC TO FURTHER IMPROVE OUTCOME

## Combining Systemic Therapy With SRS for BM of NSCLC

To further improve the clinical outcome of SRS in patients with BM of NSCLC, it can be considered to combine targeted therapies with either SRS or WBRT. The hypothesis is that systemic therapies have a superior penetration through the BBB after radiotherapy leading to a better response in the brain of the systemic agent. Considering multiple reports on the efficacy of targeted therapy on BM, it is interesting to investigate if the combination of SRS with systemic treatment improves efficacy above systemic treatment only (79, 80).

A few trials combined cranial radiation with targeted therapies in patients with BM from primary NSCLC. Lee et al. randomized 80 NSCLC patients with KPS ≥ 70 and multiple BM to erlotinib or placebo arms, both concurrent with WBRT (81). Patients continued with treatment of erlotinib or placebo until progression of disease. The neurological progression-free survival (nPFS) and median OS are not significantly different with an equal nPFS in both treatment arms of 1.6 (*p* = 0.84), and a median OS of 2.9 and 3.4 months in the placebo and erlotinib arms, respectively (*p* = 0.83). More grade 3/4 toxicity was found in the erlotinib arm compared with the placebo arm for fatigue and rash and there was no reported difference in the QoL. They concluded that there was no improvement of nPFS or OS for erlotinib concurrent to WBRT for patients with EGFR wild-type NSCLC and multiple BM.

A phase III trial from Sperduto et al. randomized 126 NSCLC patients with 1–3 BM into WBRT plus SRS vs WBRT plus SRS plus erlotinib or temozolomide (82). Erlotinib or temozolomide could be continued up to 6 months after radiation. The trial closed early because of slow accrual. There was no statistically difference between the groups concerning OS and time to CNS progression. The performance status was inferior in the group with erlotinib or temozolomide compared with the group treated with WBRT plus SRS. They found more grade 3–5 toxicity in the patients treated with targeted therapy concurrent with radiation. The trial did not support the concurrent use of systemic treatment with WBRT plus SRS. The question remains if SRS only combined with erlotinib or temozolomide provides benefit over either SRS or systemic treatment only, in these patients.

Another phase II trial from Welsh et al. enrolled patients with BM from NSCLC who received erlotinib concurrently with WBRT, followed by maintenance erlotinib (83). The overall radiologic response rate was 86% and there was no increase in neurotoxicity and no grade ≥4 toxicity. The median survival was 9.3 months for EGFR wild-type patients and 19.1 months for patients with an EGFR mutation. The addition of erlotinib to cranial radiation was well tolerated, but there was no control arm.

Magnuson et al. preformed a multi-institutional analysis with 351 EGFR-mutant NSCLC patients who developed BM with no prior treatment with EGFR-TKI (37). Patients received EGFR-TKI followed by SRS or WBRT at intracranial progression (*N* = 131), WBRT followed by EGFR-TKI (*N* = 120), or SRS followed by EGFR-TKI (*N* = 100). The analysis demonstrated that the use of upfront EGFR-TKI, and deferral or radiotherapy is associated with inferior OS. SRS followed by EGFR-TKI resulted in the longest OS (46 vs 25 months for the EGFR-TKI arm). The high effective doses of SRS may ablate BM, whereas systemic therapy simultaneously controls the extracranial diseases and possibly intracranial micro metastatic disease. A prospective trial comparing SRS followed by EGFR-TKI vs EGFR-TKI only is urgently needed.

Hendriks et al. published a systematic review to address the toxicity of combining cranial radiotherapy (SRS, WBRT, or SRS + WBRT) with TKI such as erlotinib and gefitinib, focusing on neurological toxicity (84). Fifteen articles were included, with only one phase III randomized trial. The authors concluded that there are arguments that TKI can safely be combined with WBRT, lacking high-level evidence. Grade 3 or higher toxicity may increase combining TKI with SRS and WBRT. The systematic review emphasizes the need for further investigation. Two retrospective studies analyzed patients with concurrent crizotinib, an ALK inhibitor, and SRS for cranial and extracranial metastatic NSCLC patients. They concluded that SRS can be used safely in patients receiving crizotinib. When SRS was administered to the patient, crizotinib could be longer admitted to the patient, leading to a longer 2-year OS (72% duration of crizotinib >12 months vs 12% when duration of crizotinib <12 months) (85, 86). For ALK-rearranged patients, minimal data are available, further high quality studies evaluating the use of ALK inhibitors concurrent with SRS are needed.

#### Combining SRS With Immunotherapy

A potential radiobiological advantage of SRS is an enhanced antitumor immune response after radiation of the tumor as mediator of systemic effects, better known as the abscopal effect (87, 88). Radiotherapy changes the tumor environment resulting in the release of tumor antigens and therefore increases the antitumor effect of immunotherapy. For example, IL2 is a cytokine, which plays a role in the activation of the immune response against tumor cells. L19 targets the extra domain B of fibronectin, which is a marker for tumor neoangiogenesis and is, among others, overexpressed in NSCLC. L19 can significantly increase the immune response. Zegers et al. provide evidence for an increased therapeutic potential due to the combination of radiation therapy with L19-IL2 (89).

Reynders et al. published an overview of preclinical and clinical data about the abscopal effect, resulting in 37 articles. They found a median time to abscopal effect of 5 months, and an ongoing median time of abscopal response of 13 months before disease progression (88). The included data point toward a synergy between immune treatment and radiotherapy, but further research is needed before the abscopal effect can become relevant in clinical practice.

The abscopal effect is well discussed in the literature concerning patients with BM of metastatic melanoma. Schoenfeld et al. reviewed 16 patients with BM of melanoma treated with ipilimumab and cranial radiation to evaluate the abscopal effect. They found a superior OS for patients who received SRS before ipilimumab compared with patients receiving ipilimumab before SRS (26 vs 6 months, *p* < 0.001) (90). Also, Knisely et al. found an increased survival rate for patients with BM of melanoma treated with SRS in addition to ipilimumab, with a median survival of 21.3 months in the SRS plus ipilimumab group vs 4.9 months in the ipilimumab group and a 2-year survival of 47.2 vs 19.7% (91). Both trials suggesting a role for radiation therapy in enhancing immunotherapy.

This phenomenon was also demonstrated in a case report regarding a patient with metastatic lung adenocarcinoma treated with ipilimumab concurrent with radiation (92). A complete response was received at the initial site and all metastatic sites. Posttreatment, an increase in tumor-infiltrating cytotoxic lymphocytes and tumor regression was seen. They concluded that this approach in NSCLC may establish clinical trials for patients with advanced metastatic disease in the future. Combining immunotherapy with SRS in patients with BM of NSCLC to induce an abscopal effect and improve outcome is a strategy which is currently being investigated in clinical trials, NTC02086721, for example.

## Individualized Isotoxic Dose Prescription (IDP) With SRS for Avoidance of RN and Improvement of Local Control

A disadvantage of SRS is the risk of RN with current SRS dose prescription. The SRS dose is prescribed based on the size of the planning target volume (PTV) of the BM and ranges between 15 and 27 Gy in one or three fractions. In larger BM, the dose is hypofractionated and lowered to avoid the risk of RN. Despite this strategy, the risk of RN is still increased in BM with a diameter of more than 2 cm as the volume of healthy brain tissue which is irradiated to a high dose remains relatively high (93).

A relatively new possible strategy to mitigate the risk of RN and to increase the local control probability is *IDP* (94). The idea of IDP is to prescribe the dose based on the normal tissue tolerance level instead of the size of the PTV so that the risk of complications is always minimized or even avoided. The dose in the PTV is escalated until the highest dose which is technically achievable. If the local control probability is unsatisfactory for the patient, the number of fractions can be increased to compensate. In the literature fractionated stereotactic radiotherapy (FSRT), approaches have been described to improve the local control in large BM (95, 96). These studies use a fixed prescribed dose, for example, 25 Gy in five fractions. FSRT has the increased risk of observing RN with increasing size of the BM due to the fixed prescribed dose, while IDP has the advantage that the tolerance level of the healthy brain tissue is always respected. IDP will increase the local control probability of patients with a diameter of less than 2 cm compared with current SRS dose prescription and decrease the risk of RN in BM with a diameter of more than 2 cm. IDP, therefore has the potential to increase the therapeutic ratio, e.g., ratio local control/RN, for all sizes of BM (94). IDP is expected to yield the best results as the margins used are minimized or even avoided, with an optimal beam arrangement (non-coplanar vs coplanar SRS beams). Predictive studies for IDP have already been published, such as an *in silico* study for NSCLC patients (97). Individualized IDP, compared with conventionally prescribed fractionated radiotherapy, enabled a therapeutic gain in almost 80% of the patients. In a predictive modeling study, a 25% increase in the estimated tumor control probability was expected with IDP for patients with NSCLC (98). Nowadays IDP is not yet in active clinical use, clinical studies are needed to validate the results achieved with this *in silico* study.

#### CONCLUSION

In recent years, the management of lung cancer has changed dramatically. At present, patients having NSCLC with driver mutations are treated with multiple lines of systemic therapy leading to an increasing importance of the management of BM.

## REFERENCES


The indication of WBRT is a matter of debate because of its side effects and relatively poor outcome in terms of QoL and survival after treatment. SRS is an emerging strategy for patients with BM of NSCLC and the standard treatment for patients with a maximum of three BM. SRS is also a promising treatment option for patients with four or more BM and randomized trials are ongoing to determine its value. Promising future strategies include combining SRS with systemic treatments, for example, upfront TKIs to improve survival by destruction of the BBB and better penetration of SRS. The combination of SRS with immunotherapy is promising to induce an abscopal effect. Finally, a promising strategy is the potential improvement of outcome of SRS in large BM by individualized IDP. With this strategy, the risk of RN is minimized or even avoided with a simultaneous improvement of the local control probability.

## AUTHOR CONTRIBUTIONS

DH and BH have written the manuscript with input from all the authors.

## ACKNOWLEDGMENTS

The authors would like to thank Mark Podesta, PhD, MAASTRO Clinic for the language correction.

the graded prognostic assessment. *J Neurosurg* (2008) 109:87–9. doi:10.3171/ JNS/2008/109/12/S14


**Conflict of Interest Statement:** MAASTRO Clinic has a research agreement with Varian Medical Systems, Palo Alto, CA, USA. DR: Advisory board of Merck/Pfizer, Roche/Genentech, Bristol Myers Squibb, Celgene.

The reviewer SW and handling Editor declared their shared affiliation.

*Copyright © 2018 Hartgerink, van der Heijden, De Ruysscher, Postma, Ackermans, Hoeben, Anten, Lambin, Terhaag, Jochems, Dekker, Schoenmaekers, Hendriks and Zindler. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Brain Radiation Necrosis: Current Management With a Focus on Non-small Cell Lung Cancer Patients

Gokoulakrichenane Loganadane<sup>1</sup> , Frédéric Dhermain<sup>2</sup> , Guillaume Louvel <sup>2</sup> , Paul Kauv <sup>3</sup> , Eric Deutsch2,4,5, Cécile Le Péchoux <sup>2</sup> and Antonin Levy 2,4,5 \*

<sup>1</sup> Department of Radiation Oncology, AP-HP, CHU Henri Mondor, University of Paris-Est, Créteil, France, <sup>2</sup> Department of Radiation Oncology, Gustave Roussy, Université Paris-Saclay, Villejuif, France, <sup>3</sup> Department of Neuroradiology, AP-HP, CHU Henri Mondor, University of Paris-Est, Créteil, France, <sup>4</sup> INSERM U1030, Molecular Radiotherapy, Gustave Roussy, Université Paris-Saclay, Villejuif, France, <sup>5</sup> Université Paris-Sud, Université Paris-Saclay, Le Kremlin-Bicêtre, France

As the prognosis of metastatic non-small cell lung cancer (NSCLC) patients is constantly

#### Edited by:

Deepa S. Subramaniam, MedStar Georgetown University Hospital, United States

#### Reviewed by:

Meng Xu Welliver, The Ohio State University, United States David Ball, Peter MacCallum Cancer Centre, Australia

\*Correspondence: Antonin Levy antonin.levy@gustaveroussy.fr

Specialty section: This article was submitted to Thoracic Oncology, a section of the journal Frontiers in Oncology

Received: 24 January 2018 Accepted: 02 August 2018 Published: 05 September 2018

#### Citation:

Loganadane G, Dhermain F, Louvel G, Kauv P, Deutsch E, Le Péchoux C and Levy A (2018) Brain Radiation Necrosis: Current Management With a Focus on Non-small Cell Lung Cancer Patients. Front. Oncol. 8:336. doi: 10.3389/fonc.2018.00336

improving with advances in systemic therapies (immune checkpoint blockers and new generation of targeted molecular compounds), more attention should be paid to the diagnosis and management of treatments-related long-term secondary effects. Brain metastases (BM) occur frequently in the natural history of NSCLC and stereotactic radiation therapy (SRT) is one of the main efficient local non-invasive therapeutic methods. However, SRT may have some disabling side effects. Brain radiation necrosis (RN) represents one of the main limiting toxicities, generally occurring from 6 months to several years after treatment. The diagnosis of RN itself may be quite challenging, as conventional imaging is frequently not able to differentiate RN from BM recurrence. Retrospective studies have suggested increased incidence rates of RN in NSCLC patients with oncogenic driver mutations [epidermal growth factor receptor (EGFR) mutated or anaplastic lymphoma kinase (ALK) positive] or receiving tyrosine kinase inhibitors. The risk of immune checkpoint inhibitors in contributing to RN remains controversial. Treatment modalities for RN have not been prospectively compared. Those include surveillance, corticosteroids, bevacizumab and local interventions (minimally invasive laser interstitial thermal ablation or surgery). The aim of this review is to describe and discuss possible RN management options in the light of the newly available literature, with a particular focus on NSCLC patients.

Keywords: complication, stereotactic radiotherapy, radiosurgery, vascular endothelial growth factor (VEGF), lung cancer, immunotherapy

## INTRODUCTION

Due to its incidence and specific brain tropism, non-small cell lung cancer (NSCLC) represents the most common source of brain metastases (BM) (1). Given advances in systemic treatments with prolonged overall survival and better imaging [brain magnetic resonance imaging (MRI)] detection, BM incidence rate is increasing. The prognosis of BM NSCLC patients with targetable mutations has improved (2, 3), and recently available immune checkpoint blockers (ICI) provide promising prolonged outcome in non-mutated patients (4, 5). Altogether, up to 22% of NSCLC patients may have BM at the time of initial diagnosis, and BM will develop in approximately half of

**46**

patients during their disease (6, 7). The BM rate may then be even higher in molecularly selected groups, such as epidermal growth factor receptor (EGFR) mutated or anaplastic lymphoma kinase (ALK) positive NSCLC patients (8).

The main focal treatment options for BM include surgery, stereotactic radiosurgery (SRT), and whole brain radiation therapy (WBRT). In the past decade, SRT has become the most frequently delivered focal treatment in patients with good prognosis criteria, and a limited number (<4) of BM (9–11). Frameless SRT delivers "ablative" dose, in a single or multiple course, as a definitive or postoperative treatment (12, 13). Focal high dose irradiation, as compared with neurosurgery, has the ability to treat inoperable sites, several lesions, and has the advantage to be less invasive. WBRT alone or in combination with SRT has been challenged in randomized trial, and its role is now limited to selected patients with multiple BMs ineligible for SRT (12, 14, 15). SRT is now often favored over WBRT due to a lower rate of adverse neurocognitive side effects. It has also been suggested that SRT without WBRT was feasible as the initial treatment for patients with 5–10 BMs (16). Local control at 1 year is generally high (88% in recent series), and SRT is generally considered as a cost-effective treatment (12, 17).

However, rare but potentially debilitating secondary late effects (3 months to several years post-irradiation) have been described after SRT. The most common delayed complication SRT is brain radiation necrosis (RN). RN may be particularly challenging in terms of diagnosis and treatment. Few studies have highlighted that RN may be more frequent in NSCLC patients harboring an oncogenic driver mutation. Within this review we aimed to describe and discuss the current knowledge regarding RN, with a special attention to NSCLC patients.

## PATHOBIOLOGY

The physiopathology of radiation necrosis is still elusive and several hypotheses have been proposed. Implicated mechanisms in delayed RN include vascular injury, immune-mediated mechanisms and direct neural effects.

The vasculature damages are characterized by an increased permeability and a disruption of the blood brain barrier (BBB). High dose focal radiotherapy induces an endothelial cell loss through acid sphingomyelinase-dependent apoptosis (18) leading to vasogenic edema and ischemia. Tissue ischemia and vasogenic edema induce hypoxia, leading to reactive oxygen species production, affecting many cellular functions, and produce an increase of the hypoxia inducible factor (HIF-1α). HIF-1α subsequently upregulates the vascular endothelial growth factor (VEGF) secreted by astrocytes and endothelial cells. Immunohistochemistry of surgical samples of RN showed increased levels of VEGF in reactive astrocytes surrounding the core of necrotic tissue. VEGF exacerbates edema by increase of vascular permeability (19). These data indicate a crucial role of VEGF in the development and progression of RN and its inhibition could decrease the vascular permeability and therefore edema. Following these observation, anti-VEGF therapy has been one of the most tested compounds in the preclinical setting, and the sole pharmacological agent translating to clinical efficacy in the treatment of RN (cf. below, chapter on VEGF inhibition) (20, 21).

The immune system and peri-necrotic inflammation are also implicated in RN formation. Local infiltration of immune cells likely aggravates RN. VEGF induces the expression of adhesion proteins such as ICAM-1 on endothelial cells, and trigger pro-inflammatory cytokines [e.g.,: interleukin (IL)-1α, IL-6 and tumor necrosis alpha (TNF)-α] in animal models (18). Yoritsune et al. have also shown in human RN specimens, that astrocyte cells expressing the chemokine CXCL12 might attract CXCR4 expressing immune cells into the perinecrotic area, which in turn aggravates the local hypoxia (18, 22). The introduction of ICIs has significantly modified the therapeutic landscape of advanced NSCLC. As those agents are immunostimulatory, they could potentially exacerbate a preexisting inflammatory reaction in the context of RN.

Radiation induces white matter necrosis and oligodendrocytes demyelination. In the periphery of this necrotic zone, astrocytes, microglial cells and oligodendrocytes produce factors promoting cytokine release. A decrease of oligodendrocytes with incomplete neural stem cells or neuroblasts repopulation has been described (23, 24). Remyelinisation after human embryonic stem cell-derived oligodendrocyte progenitors transplantation is subsequently also assessed in preclinical models (25). Following these observations, many other agents than anti-VEGF have been tested in the experimental setting, but without reported favorable clinical effects (18).

## CLINICAL SPECIFICITIES OF BRAIN RADIONECROSIS

The diagnosis of RN may be challenging. The main issue is to distinguish between RN and local recurrence (LR). When analyzing epidemiology or predictive factors of RN, one should keep in mind the possible subsequent bias related to diagnosis difficulties, as described below.

#### Epidemiology and Predictive Factors

Reported clinical rate of RN is approximately 10%, with or without prior surgery (**Table 1**) (32, 35, 37). However, the rate of asymptomatic radiographical RN is higher: up to 25–30% in some series (29, 31). The cumulative incidence of RN is increasing over time after SRT. As an example, in a series from the Memorial Sloan Kettering, the actuarial incidence of RN was 5.2% at 6 months, 17.2% at 12 months, and 34% at 24 months (31). In another Japanese series, 16 patients with MRI contrast enhancement >18 months following SRT were identified. With a median follow-up of 48.2 months, 12 adverse radiation events (suspected radiological or pathological confirmed RN) occurred in a median follow-up of 33.2 months (38).

Predictive risk factors associated with the development of RN cited in the literature link to BM, and treatment characteristics. Main accepted ones are a larger BM size, reirradiation, and higher total delivered radiotherapy dose (39, 40). Others criteria including BM features (location and deepness), radiotherapy


TABLE

1


Literature

overview.

+Postop

hypofractionated

 stereotactic

 radiation therapy only.

parameters (high dose per fraction, volume of irradiated normal brain parenchyma [generally total volume of irradiated brain at a dose 12Gy or more]), and the use concurrent systemic therapy (including ICI) have been evocated but not systematically described (29, 41–45). In any case, fractionation (i.e., to increase the number of radiotherapy fractions), or the use of formulas for optimal individual SRT dose based on BM volume is encouraged to prevent RN (46).

Some authors advocated that RN occurrence might be more frequent in NSCLC patients (**Table 1**). In a NSCLC cohort of 836 patients (2,276 lesions), Miller et al. showed that lung adenocarcinoma histology (1-year incidence of 5.9% vs. 3.1–3.9% for other histologies), and ALK (HR 6.36, p < 0.001), but not EGFR lesions had increased rates of RN. The 1-year cumulative incidences of RN among EGFR+, ALK+, and ALK/EGFR wild-type lesions were 7.6, 17.3, and 3.7%, respectively. EGFR or ALK inhibitors, as compared to conventional treatments, were not associated with the occurrence of RN (32). Another series included 699/1,650 (42%) NSCLC patients who underwent SRS, with or without WBRT. Patients also received systemic treatments, including targeted therapies. NSCLC patients who received concurrent EGFR tyrosine kinase inhibitors (TKIs) had an increased of 12-month cumulative incidence of RN (15.6 vs. 6%, p = 0.04) as compared to other patients. This was more specifically observed in patients that received SRT+WBRT (p = 0.02) as compared with those receiving SRT without WBRT (p = 0.45) (34). It should anyway be emphasized that BM NSCLC patients with an oncogenic driver mutation generally receive more intensive local treatment, partly explaining the excess risk of toxicity (47).

The risk of ICI in contributing to RN is controversial. Prospective data is lacking and most retrospective series included melanoma patients (48). A retrospective SRT series reported a higher incidence of symptomatic RN for patients who received ICI as compared to those who did not. Among 480 patients with BM (289 [61%] of 480 NSCLC) who had been treated with SRT, 115 (24%) received an anti-PD1 (nivolumab or pembrolizumab) or an anti-cytotoxic T-lymphocyte associated protein 4 (ipilimumab). Patients treated with ICI had a significantly higher rate of symptomatic RN after adjustment for tumor type (HR: 2.6; p = 0.004). The risk of neurotoxicity was however highest for melanoma patients treated with ipilimumab (36). Other retrospective studies focusing on the outcome of patients with NSCLC with BM who received both cranial RT and an ICI did not report RN increase (49–52). However, it should be emphasized that pseudoprogression, observed with ICI may be difficult to be distinguished from RN or brain progression (53).

#### Challenges in RN Diagnosis

Radiographic changes (grade I, approximately 50%) from the symptomatic RN (grade II–IV) should be distinguished. In the latter case, an intervention may be required whereas a simple surveillance is sufficient for the former case. The symptoms depend on the location of the lesion, but manifest usually with focal neurologic signs and symptoms related to cerebral edema.

The main difficulty is to distinguish between RN and LR. Histology is the gold standard for a confirmed diagnostic. A recent series of BM patients who had a brain biopsy for RN or LR suspicion on MRI included 11/34 (31%) lung cancers patients. Most biopsies (24/35; 69%) showed RN only, and time from SRT to biopsy was significantly longer (>9 months) in the RN group (p = 0.004) as LR seemed to occur earlier than RN (54). On the other hand, brain biopsies are invasive and may not be accessible for all patients. Histopathologic interpretation of brain specimens could also be challenging due to heterogeneity of the lesion mixing irradiated residual tumor cells of indeterminate viability with RN that can be missed by the sampling, and some authors suggested that excision of the lesion only is able to determine its true histological nature (55).

More often, non-invasive (clinical and radiographic) criteria are used, but the distinction between the RN and tumor can be particularly challenging. In most cases, conventional MRI shows a contrast-enhancing mass lesion with central necrosis and reactive edema contiguous to the site of the initial BM. "T1/T2 mismatch" (i.e., larger mass lesion seen in T2 sequence as compares with the T1 contrast-enhanced residual lesion) may favor RN (56). Dynamic (perfusionand diffusion-weighted) MRI (**Figure 1**), and spectro-MRI (or magnetic resonance spectroscopy: MRS) have extensively been assessed to differentiate RN from LR. Dynamic susceptibility contrast-enhanced (DSCE) MR perfusion decreased parameters such as relative cerebral blood volume (rCBV) and relative peak height (rPH) or percentage of signal-intensity recovery (PSR) increase correlate with RN (57). On diffusion-weighted MR, decreased signal on diffusion-weighted imaging (DWI) and increased apparent diffusion coefficient (ADC) maps values reflect tumor control (58). MRS is an analytical technique that can be used to complement MRI in the characterization of tissue. Low lipid peak or high choline-tocreatine ratio and high choline-to-N-acetylaspartate (NAA) ratio on MR spectroscopy suggest tumor recurrence (59). Regarding positron emission tomography (PET), lower uptake with various radiotracers [fluorodeoxyglucose (FDG), methionine, fluorodihydroxyphénylalanine (Fdopa), fluoroéthyl-L-tyrosine, fluorocholine or thallium chloride-201 single-photon emission computed tomography (SPECT)] suggests necrosis. FDG has been the most commonly studied radiotracers but specificity is low, and the use of the couple dynamic MRI/PET is encouraged (60–64). Altogether, these imaging studies underline the difficulties to diagnose RN. Finally, the beneficial effect of steroids has also been incorporated to the diagnosis strategy, as depicted in the existing proposed algorithm to diagnose and treat RN (37, 64).

## TREATMENT OPTIONS OF RADIATION NECROSIS

RN can generally be managed conservatively without intervention. In symptomatic patients, moderate dose of glucocorticoids may produce prompt symptomatic improvement by reducing cerebral edema. Corticosteroids can then be gradually tapered. If not sufficient, RN management consists of VEGF inhibitors or laser interstitial thermal therapy (LITT).

FIGURE 1 | A 66-year-old man with history of brain metastasis of non-mutated NSCLC and treated by surgical resection and postop SRT. (A) Axial T2w FLAIR sequence showed a hyperintense signal appeared around the treated region 13 months after SRT. (B) T1w contrast sequence showed an inhomogeneous ring enhancement within the treated region. (C) DWI showed a low signal within the enhanced margin, with a high ADC (not shown). (D) Dynamic susceptibility contrast-enhanced perfusion weighted imaging showed a low hyperperfusion with a relative cerebral blood volume of 1.5, suggesting absence of tumor recurrence. Surgical resection confirmed the diagnosis of cerebral RN.

NSCLC, non-small cell lung cancer; T2w, T2 weighted; T1w, T1 weighted; DWI, diffusion weighted imaging; ADC, apparent diffusion coefficient; SRT, stereotactic radiotherapy; RN, radionecrosis.

Ultimately, surgery may be required in patients who are resistant to other treatments, and/or to obtain a definitive diagnosis if a LR is suspected. Alternative approaches have been reported in some cases (therapeutic anticoagulation, antiplatelet therapy, and hyperbaric oxygen therapy), but may not be currently recommended.

#### VEGF Inhibition

As previously described, VEGF plays a critical role in the RN pathogenesis. Bevacizumab is the most commonly used anti-VEGF monoclonal antibody, and was prospectively evaluated in only one small prospective trial in the context of RN. Fourteen patients were randomized 1:1 to receive four cycles of intravenous (IV) bevacizumab at a dose of 7.5 mg/kg every 3 weeks vs. IV saline placebo. The primary endpoint was the change in edema volume on MRI (T2 FLAIR images) from baseline to the first evaluation at 6 weeks. Of note, there were no BM patients included but only prior irradiated primary central nervous system or head and neck tumors. Crossover was permitted, and the sample size was estimated to 16 patients. The 7 patients in the bevacizumab arm had a decreased volume of FLAIR edema with clinical amelioration whereas placebo arm patients demonstrated an increase in the volume of T2 weighted FLAIR edema (−59 vs. +14%, respectively; p = 0.01). Similarly, in patients receiving bevacizumab, a median decrease in the T1 weighted gadolinium enhancement (−63 vs. +17%; p = 0.006), and of the endothelial transfer constant (K-trans; a measure of capillary permeability in DCE MRI; −99 vs. +49%; p = 0.02) were reported. Six of 11 patients receiving bevacizumab had adverse events, with 3 serious adverse events: one aspiration pneumonitis, one pulmonary embolism secondary to deep vein thrombosis and one superior sagittal sinus thrombosis (65). Other retrospective series also reported a clinical benefit of bevacizumab, including reduction in steroid requirement (66–68).

Those promising results should nevertheless be tempered. One should not forget that bevacizumab has a certain activity on BM in NSCLC patients, especially when we know that LR and RN can be associated in a significant proportion of cases (69). Development of RN was also observed among 24/271 (9%) patients receiving SRT with concurrent bevacizumab (31). Worsening of symptoms may occur, and RN recurrences after bevacizumab withdrawal have been described (70). In a series including a majority (11/14; 79%) of BM from primary lung cancer, clinical improvement was seen in 13/14 cases (92.9%), but the 10/13 responsive patients (76.9%) exhibited a recurrence of brain necrosis after bevacizumab discontinuation (71). Bevacizumab is a promising treatment option for RN, but needs to be validated in larger prospective studies.

#### Invasive Interventions

LITT is a stereotactic-guided minimally invasive ablative technique that generates high temperature, resulting in tissue coagulation necrosis, angiogenesis eradication, and cellular apoptosis. The use of LITT-guided MRI allows to control accurately the delivery, and to spare the surrounding healthy tissues. LITT has been used in several situations in neurology, including RN. Most of the available data come from small retrospective studies. Rao et al. reported the results of MRIguided LITT for 12/15 (80%) NSCLC patients with suspected RN or LR after SRT for BM. On average, the lesion size measured 3.7 cm. Authors were able to perform 3.3 ablations per treatment, in a total ablation time of 7.5 min. The local control was high (76%) at a median follow-up of 6 months, with two patients experiencing recurrence at 6 and 18 weeks after the procedure (72). The largest series, from the University of Arizona, consisted of 25 patients with suspected RN, occurring after treatment for 18 primary brain tumors and 7 BM. Progression free and overall survival rates in patients with BM were 11.4 and 55.9 months, respectively. The quality of life analysis showed an improvement on mental health and vitality at 12 months (73). One of the advantages of this technique is the possibility to perform a biopsy prior to treatment to confirm the diagnosis of RN. Moreover, LITT is a reasonable option in case of LR, considering the efficacy and secondary effects of other treatment modalities (i.e., reirradiation).

Surgery allows pathological confirmation, and the rapid relief from mass effect and brain edema. In a series of 15 patients with RN, the surgery improved the neurological symptoms in 14 cases. Pure RN was histologically determined for 50% of operated patients. In the algorithm proposed by the authors, patients with significant increased edema volume with mass effect, or becoming symptomatic despite steroids trial should undergo surgery (37). Another surgical series for patients with RN reported that 9 had a steroid dose reduction, 4 improved their performance status score (4 stable and 3 deterioration), and neurologic deficits were ameliorated in 4 (4 stable). Nonetheless, 2 worsened their neurologic deficit and one patient developed a new neurologic deficit after surgery. This study highlights the potential morbidity of surgical resections of RN, and suggests reserving surgery for symptomatic patients in whom medical treatment has failed (74).

#### PERSPECTIVES AND CONCLUDING REMARKS

Newer generation TKIs will possibly modify the therapeutic sequences in advanced mutated NSCLC patients. In retrospective studies, the deferral of radiation therapy (SRT or WBRT) was usually associated with inferior survival rates in oncogenic driver mutation patients (75, 76). However newer generation TKI such as first-line alectinib (ALK+ patients) and osimertinib (EGFR mutated patients) provided superior intracranial control

#### REFERENCES


compared to standard of care (2, 3). This, with the increased use of ICI, may then possibly lead to a decreased use of SRT, and subsequently change the RN rate occurrences in NSCLC patients. Moreover, NSCLC mutated patients have potentially an increased incidence of RN due to tumor biology or the use of concurrent TKI, but this remains to be confirmed.

An ongoing randomized phase II trial (BeSt Trial;Alliance A221208; NCT02490878) from the MD Anderson is investigating whether the addition of bevacizumab (10 mg/kg IV at day one and 15 for four cycles) to standard corticosteroid therapy could result in greater improvement of RN symptoms (primary endpoint: patient-reported outcome of RN up to 8 weeks). One hundred thirty patients should be included and eligibility criteria encompass perfusion-imaging parameters of RN susceptibility (high PSR and low rCBV). Another multicenter prospective French trial (CV-METANEC;NCT02636634) has recently been completed. It compared PET-FET (1-Fluoro-Ethyl-Tyrosine) and magnetic resonance spectroscopy to histological results in patients receiving brain biopsy for active persistent and increased brain lesion 4 months after SRT. The results of such studies should help to differentiate RN from LR after SRT, and help to guide clinicians to select an appropriate treatment for patients.

#### AUTHOR CONTRIBUTIONS

GL and AL designed and outlined structure and contents of the review. All authors contributed to the literature analysis, interpretation, and writing of the review.


therapy in non-small cell lung cancer. J Thorac Oncol. (2018) 13:553–61. doi: 10.1016/j.jtho.2018.01.012


radiation necrosis. Int J Radiat Oncol Biol Phys. (2013) 87:449–57. doi: 10.1016/j.ijrobp.2013.05.015


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Loganadane, Dhermain, Louvel, Kauv, Deutsch, Le Péchoux and Levy. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Management of Brain Metastases in epidermal Growth Factor Receptor Mutant Non-Small-Cell Lung Cancer

*William J. Kelly, Neil J. Shah and Deepa S. Subramaniam\**

*Division of Hematology-Oncology, Georgetown University, Washington, DC, United States*

Lung cancer remains a leading cause of mortality with 1.69 million deaths worldwide. Activating mutations in epidermal growth factor receptor (EGFR), predominantly exon 19 deletions and exon 21 L858R mutations, are known oncogenic drivers identified in 20–40% of non-small-cell lung cancers (NSCLC). 70% of EGFR-mutant NSCLC patients develop brain metastases (BM), compared to 38% in EGFR wild-type patients. Firstgeneration tyrosine kinase inhibitors (TKIs), such as erlotinib and gefitinib have proven to be superior to chemotherapy in the front-line treatment of EGFR-mutant NSCLC, as has afatinib, a second-generation TKI. The most common acquired resistance mechanism is the development of a gatekeeper mutation in exon 20 T790M. Osimertinib has emerged as a third-generation EGFR TKI with proven activity in the front-line setting as well as in patients with a T790M acquired resistance mutation with remarkable CNS activity. As long-term survival outcomes in EGFR-mutant NSCLC continue to improve, the burden of BM becomes a greater challenge. Here, we review the literature related to the management of BM in EGFR-mutant NSCLC including the role of the three generations of EGFR TKIs, immunotherapy, and brain radiation.

Keywords: epidermal growth factor receptor, non-small-cell lung cancer, brain metastases, targeted therapy, osimertinib

## OVERVIEW OF BRAIN METASTASES IN EPIDERMAL GROWTH FACTOR RECEPTOR (EGFR) MUTANT NON-SMALL-CELL LUNG CANCER (NSCLC)

#### Epidemiology and Molecular Alterations in EGFR Mutant NSCLC

Lung cancer remains a leading cause of mortality with 1.69 million deaths worldwide (1). An estimated 234,030 new cases will occur in the United States in 2018 with a median age at diagnosis of 70 and 64% predominance for males (2). Approximately 84% of these lung cancers are non-small cell lung cancers (NSCLC) (3). NSCLC has traditionally been classified by histology (adenocarcinoma, squamous, and large cell) but the classification paradigm has evolved to incorporate molecular subtypes that guide treatment decision making.

Epidermal growth factor receptor (EGFR) is a transmembrane tyrosine kinase receptor which activates Jak, PI3K, ROS, and RAS pathways leading to cell survival (4, 5). The most common activating mutations are exon 19 deletions or point mutations in exon 21 *via* Leu858Arg (L858R) (6, 7). Reports of the prevalence of EGFR mutations in NSCLC ranges from 46.7% in the East Asian population as reported by Liu et al. (8) to 38.4% (range 36.5–40.3%) in China and 14.1% (range 12.7–15.5%) in Europe seen in Zhang et al. (9) and 22% in African Americans enrolled in the Lung Cancer Mutation Consortium (10). The landmark BR21 trial demonstrated the survival advantage in chemo-refractory NSCLC with the use erlotinib, a first-generation EGFR inhibitor (11).

#### *Edited by:*

*Miguel Angel Villalona, Baptist Health South Florida, United States*

#### *Reviewed by:*

*Joel W. Neal, Stanford University, United States Rabab Mohamed Gaafar, Cairo University, Egypt Gregory Alan Otterson, The Ohio State University, United States*

#### *\*Correspondence:*

*Deepa S. Subramaniam dss26@gunet.georgetown.edu*

#### *Specialty section:*

*This article was submitted to Thoracic Oncology, a section of the journal Frontiers in Oncology*

*Received: 01 February 2018 Accepted: 22 May 2018 Published: 03 July 2018*

#### *Citation:*

*Kelly WJ, Shah NJ and Subramaniam DS (2018) Management of Brain Metastases in Epidermal Growth Factor Receptor Mutant Non-Small-Cell Lung Cancer. Front. Oncol. 8:208. doi: 10.3389/fonc.2018.00208*

**54**

Subsequently, three additional drugs (gefitinib, afatinib, and osimertinib) have now been approved to treat newly diagnosed EGFR-mutated advanced NSCLC. Among NSCLC patients who progress on first- or second-generation EGFR TKI therapy, most do so through a unique gatekeeper mutation, viz. the exon 20 point mutation Thr790Met (T790M) in the ATP-binding site of EGFR (12). Incidence of the T790 gatekeeper mutation has been reported to be between 49 and 63% (13, 14). The methionine side chain acts as a "gatekeeper" residue causing steric hindrance thus decreasing hydrophilicity and preventing tyrosine kinase binding (15). The T790M mutation also increases ATP affinity (16). Other rare mechanisms of TKI resistance include MET amplifications or mutations, HER2 amplifications, and rarely BRAF mutations (12). Additionally, transformation to small cell histology is another possible mechanism of EGFR TKI resistance (13).

## Prevalence of Brain Metastases (BM) in EGFR-Mutant NSCLC

Among NSCLC patients, those with BM have an increased frequency of EGFR mutations than those without brain metastasis and conversely, among EGFR mutant NSCLC patients the incidence of BM (70%) greatly surpasses the incidence of BM in wild-type (wt) EGFR NSCLC patients (38%) (17). Approximately, one-third of EGFR-mutant NSCLC patients develop central nervous system (CNS) progression during the course of their illness (18). Among Asian populations, the prevalence of EGFR mutations in NSCLC BM ranges from 39 to 63% (19, 20). Among North American and European populations this ranges from 2 to 40% (21, 22). At initial diagnosis, EGFR mutation discordance estimates between primary and BM range are minimal (23). Prevalence of T790M mutations in CNS lesions among EGFR mutant NSCLC patients with TKI failure is much lower than anticipated at around 17% (24). This may reflect a pharmacokinetic failure of the first-generation EGFR TKIs to penetrate the brain and thus induced acquire resistance *via* the gatekeeper T790M mutation. Case reports have detailed patients on gefitinib and erlotinib, first-generation TKIs with modest brain penetrance, who have developed T790M-mediated resistance at primary tumor locations but not in the brain metastasis (25, 26). CNS progression appears to be higher in those with L858R point mutations (18). Interestingly, a retrospective radiologic analysis of 57 NSCLC patients suggested that exon 19 deleted patients may have more of a miliary pattern of BM (27). **Table 1** summarizes the prospective trials of three generations of EGFR tyrosine inhibitors in EGFR-mutant NSCLC with BM.

## FIRST-GENERATION TYROSINE KINASE INHIBITORS (TKIs)

#### Erlotinib

Erlotinib is a first-generation (EGFR) tyrosine kinase inhibitor (TKI) (28). The drug reduces EGFR autophosphorylation in intact tumor cells at a median inhibitory concentration of 20 nM although this ranges from 5 (nM) and 6 (nM) in exon 19 deletion and L858R cell models respectively to >2,000 (nM) in T790M models (29, 30). High-performance liquid chromatography studies have shown that erlotinib penetrates the cerebrospinal fluid (CSF) at a rate of between 2 (67 nM) and 4% (31, 32). Radiolabeled 11C-erlotinib injected to one NSCLC patient was shown to accumulate in brain metastasis (33). Additionally, the average concentrations of erlotinib in CSF appear to be higher in those with partial responses (PRs) (35 ng/ml) compared to those who have progressive disease (16 ng/ml) (32).

Several studies have evaluated the effect of erlotinib in NSCLC patients with BM. Deng et al. reported on six unselected NSCLC patients with BM treated with erlotinib and noted that four of the six harbored an EGFR mutation in the tumor; two PRs and two stable diseases (SD) were noted in EGFR mutant patients (32).

Table 1 | Prospective studies in epidermal growth factor receptor (EGFR) mutant non-small-cell lung cancers (NSCLC) patients with brain metastases (BM).


Porta et al. retrospectively reviewed 69 NSCLC with BM patients treated with erlotinib (34). 17 patients had EGFR mutations, 82% of whom had an objective response rate (ORR) to erlotinib including eight complete responses (CRs) as well as a median time to progression of 11.7 months compared to 5.8 in EGFR wt patients and an overall survival of 12.9 months versus 3.1, respectively (34). Moreover, no patients without EGFR mutations had an objective response (34). A phase II study prospectively evaluated EGFR mutant NSCLC patients treated with erlotinib or gefitinib and noted that 83% achieved a PR and 11% SD without a statistically significant difference in progression-free survival (PFS) (6.6 months) or overall survival (OS) (15.9) between the two TKIs (35).

Dose escalation has also been examined as a potential strategy to increase CNS permeability and overcome resistance. In a small but compelling retrospective case series of nine EGFR mutant lung cancer patients with brain or leptomeningeal metastases that occurred despite conventional dosing of an EGFR inhibitor, patients were treated with high dose "pulsatile" erlotinib (1,500 mg weekly) and a CNS partial response rate of 67% (6 of 9 patients) was noted; however, median time to CNS progression was only 2.7 months (36). Following this, a phase 1 study in 34 patients with EGFR mutant lung cancer treated with escalating pulse doses of erlotinib found the maximum tolerated dose to be 1,200 mg given on days 1 and 2, with 50 mg given on days 3–7 weekly and it should be noted that 32% of patients had BM at study entry and none of these patients had progression of an untreated CNS metastasis or new CNS lesions while on study (37).

The role of erlotinib in leptomeningeal disease has also been examined. A retrospective review of 25 NSCLC (9 with exon 21 EGFR mutation and 8 with exon 19 deletion) patients with leptomeningeal (LM) carcinomatosis treated with either erlotinib or gefitinib demonstrated that those treated with erlotinib had a cytologic conversion rate of 64.3% compared to 9.1% with gefitinib (38), suggesting greater activity of erlotinib over gefitnib in the setting of LM disease. In another series of NSCLC patients with leptomeningeal metastasis who had failed gefitinib treatment, all 6 patients with an EGFR mutation-derived clinical benefit with 3 PRs and 3 with SD (39). 1 patient whose tumor did not harbor an EGFR mutation developed progressive disease as the best response.

## Gefitinib

Gefitinib, another first-generation EGFR TKI, is a substrate for the *P*-glycoprotein efflux pumps and the drug has a brain penetration rate of only 1% (40, 41). There have been many retrospective reviews of NSCLC patients with BM treated with gefitinib. An old retrospective study of 14 NSCLC patients with BM observed 1 CR (6%) and 5 PRs (33%); this was done prior to the understanding of the role of EGFR mutation status on response to targeted therapies (42). Another report on 15 patients found an ORR of 60% (43). In 2009, a retrospective study of 23 Korean never-smoker patients with lung adenocarcinoma and brain metastasis without prior whole brain radiotherapy (WBRT) found that gefitinib or erlotinib without WBRT resulted in an intracranial response rate of 73.9%, noting that the prevalence of EGFR mutations in Korean non-smoker NSCLC population is high (44). Following this, Zhang et al. retrospectively reviewed 43 Chinese EGFR mutant NSCLC patients with BM treated with gefitinib or erlotinib until extracranial lesion progression; an intracranial lesion ORR of 57% and PFS of 9.3 months was observed, with no statistically significant difference in OS between gefitinib versus erlotinib (45).

Multiple prospective studies have also shown efficacy of gefitinib in NSCLC patients with BM. In 2004, Ceresoli reported on 41 NSCLC patients with BM treated with gefitinib including 18 patients with prior WBRT and observed a partial response rate of only 10% (46). Chiu et al. conducted a prospective study in 57 unselected NSCLC with BM patients observing an ORR of 33% and PFS of 5 months (47). Similarly, a 2007 study by Wu et al. examined 40 unselected NSCLC with BM patients (23 with prior WBRT) and found a 32% ORR and PFS of 9 months (48). However, a phase II study in 21 Chinese NSCLC patients with BM treated with prior WBRT reported a much higher 81% ORR and a PFS of 10 months (49).

Subsequent studies focused on gefitinib's efficacy in NSCLC with BM patients who harbored EGFR-activating mutations. Iuchi reported in 2013 on a phase II trial of 41 Japanese lung adenocarcinoma patients with BM showing a brain metastasis ORR of 87.8% with 13 CRs (50). Stereotactic radiation and WBRT were required in 20 patients (50). Patients with exon 19 deletions had a statistically significant PFS and OS advantage compared to L858R mutations (50). While the results of many of these studies appeared to be promising, the results and thus, the potential efficacy of first-generation EGFR TKIs in patient with BM, need to be interpreted with caution given the small numbers of patients in these predominantly retrospective reports.

#### Icotinib

Icotinib, a first-generation TKI approved in China, has a median CSF penetration rate of 6.1% (51). The BRAIN study was a multicenter, open-label, parallel randomized controlled trial of 176 Asian EGFR mutant NSCLC patients with at least three brain lesions; patients treated with icotinib had a median intracranial PFS of 10.0 months compared to 4.8 months in those treated with whole brain irradiation plus concurrent or sequential chemotherapy, translating to a 44% risk reduction from intracranial progression or death, and making this a potentially promising option (52).

## SECOND-GENERATION TKIs

## Afatinib

Afatinib is an oral second-generation TKI which selectively and irreversibly blocks EGFR, HER2, and HER4 kinase activity (53–55). The LUX-Lung 3 was a phase III trial of front-line afatinib in EGFR mutant advanced or metastatic NSCLC (56). Subgroup analysis of 35 patients with asymptomatic BM showed a PFS of 11.1 months versus 5.4 months with cisplatin and pemetrexed (57). The LUX-Lung 6 study was an open label randomized, multicenter phase III trial of Asian patients with EGFR mutant advanced or metastatic lung cancer (58). Prespecified subgroup analysis of 46 asymptomatic BM patients revealed that PFS was improved from 8.2 to 4.7 months in those treated with gemcitabine and cisplatin (58). Those who received whole brain radiation therapy appeared to have better PFS benefit than those who did not receive radiation; however, among BM patients, rates of CNS progression were similar to chemotherapy in both the LUX-Lung 3 (45 vs 33% chemotherapy) and LUX-Lung 6 (21.4 vs 27.8%) (58).

Another study through the afatinib compassionate use program examined 100 NSCLC patients with BM and/or leptomeningeal disease who had progressed on platinum chemotherapy and a first-generation EGFR TKI, 74% of whom had a documented EGFR mutation (59). Median time to treatment failure was 3.6 months and was similar to a matched group of 100 patients without CNS metastasis; 35% had cerebral response and one heavily pretreated patient with impressive leptomeningeal response and neurological recovery had a CSF concentration of nearly 1 nM (59).

The LUX-Lung 7 study, which was an international, open label, randomized phase II trial comparing afatinib to gefitinib in EGFR mutant advanced or metastatic NSCLC patients, noted that PFS was longer with afatinib (11 months) than gefitinib (10.9 months) but not statistically significant among the subgroup of 26 patients (16%) with asymptomatic brain metastasis (60). Thus, despite its promise as a second-generation irreversible EGFR targeted agent, afatinib did not pan out to be significantly superior to the first-generation agents in systemic disease (or CNS disease) and its use is limited by its greater toxicity profile, except in some of the less common EGFR mutations where data suggest better efficacy.

## Dacomitinib

Dacomitinib, another second-generation TKI with activity against EFGR, HER2, and HER4 was studied in two double blind, multicenter, randomized phase III trials: BR.26 and ARCHER 1009 (61, 62). In BR.26, dacomitinib did not improve OS in patients who had previously received gefitinib or erlotinib; routine brain imaging was not done in this study (61). Similarly, in Archer 1009, dacomitinib was not superior to erlotinib in advanced or metastatic NSCLC and only 2% of patients in this trial had brain metastasis at baseline (62).

## THIRD-GENERATION TKIs

#### Osimertinib

Osimertinib is an oral, irreversible EGFR TKI that targets the classical activating mutations as well as the gatekeeper resistance mutation, i.e., T790M (63). Preclinical models showed that osimertinib had greater penetration of the murine blood–brain barrier (BBB) than gefitinib, rociletinib, or afatinib and increased exposure by labeled radiography in cynomolgus monkey brains (64).

AURA 3 was a randomized, international, open-label, phase II trial of T790M positive advanced NSCLC patients who had progressed on front-line EGFR TKI therapy (63). 144 patients had CNS metastases and those who received osimertinib had a longer median PFS (8.5 months) than platinum-pemetrexed chemotherapy (4.2 months) with a hazard ratio of 0.32 (63). A pooled analysis of AURA extension and AURA2 trials in 50 patients with asymptomatic BM found a CNS ORR of 54% to osimertinib treatment with 12% having complete CNS response, with benefit also noted in patients who had not received prior radiotherapy to the brain (65).

The BLOOM study was a phase I trial of patients with CSF cytology confirmed leptomeningeal disease (66). Preliminary results of 32 treated patients (23 evaluable) found 10 had radiographic improvement and 13 with SD; additionally 7 of 8 symptomatic patients improved and, the geometric mean decrease in CSF EGFRm DNA copy was 57% (66).

The FLAURA study is a phase III study in 556 EGFR mutant (exon 19 del or L858R) advanced NSCLC patients which randomized patients 1:1 to a standard of care EGFR TKI (erlotinib or gefitinib) or osimertinib (67). Patients with neurologically stable CNS metastases were allowed on this study, accounting for 21% of patients on this study (67). Front-line treatment with osimertinib resulted in improved median PFS (18.9 months) compared to standard EGFR TKI therapy with erlotinib or gefitinib (10.2 months); median OS data were not mature at the time of the PFS analysis (67). It should be noted that ORR of known/treated CNS metastasis at trial entry was 77% in the osimertinib-treated patients compared to 63% in the standard EGFR TKI patients (68). Response duration lasting >or = 6 months was noted in 88% of patients on the osimertinib arm, with a CR rate of 18%, while no CRs were observed in the arm with standard EGFR TKI. This has led to the FDA approval of osimertinib in the United States as an option for the front-line treatment of EGFR-mutated NSCLC harboring exon 19 deletions or exon 21 L858R mutations.

#### Rociletinib

Rociletinib (CO1686) was a unique, oral, irreversible TKI designed for NSCLC patients with activity against activating EGFR mutations (L858R and Del 19) and the gatekeeper resistance mutation T790M. The CNS activity of rociletinib was poor compared to systemic disease (69, 70). Camidge et al. reported that 22 of 42 patients continued rociletinib for an average of 120 days after CNS disease progression, treated with brain radiation (70). The development of this drug has since been halted given the high risk-to-benefit ratio related to the hyperglycemia resulting from blockade of the insulin growth factor receptor. Importantly, rociletinib appears to have poor brain penetration with most patients on rociletinib coming off the drug for CNS progression. In a study of the clinical activity of osimertinib in 45 EGFR-mutant NSCLC patients previously treated with rociletinib, subsequent treatment with osimertinib still achieved a brain disease control rate (response + SD) of 88% (71).

## MISCELLANEOUS TKIs

#### AZD3759

AZD3759 in an oral EGFR TKI designed for CNS penetration with a ratio of unbound brain to unbound plasma concentration of 0.65 (72, 73). AZD3759 caused tumor regression in leptomeningeal and brain metastasis mouse models (73). Preliminary results of the phase I BLOOM study of 38 EGFR-mutant NSCLC with BM or leptomeningeal metastasis (LM) treated with AZD3759 showed an intracranial ORR of 63% and extracranial ORR of 50% (74). Trough CSF concentrations were above the IC90 for pEGFR (74). Further development of this drug is on hold given the highly promising results with osimertinib in the FLAURA trial.

## Tesevatinib

Tesevatinib (KD019) is a novel, oral, reversible TKI, which inhibits EGFR, HER2/neu, and Src family nonreceptor tyrosine kinases. Preclinical studies demonstrated good blood–brain penetration of tesevatinib with brain/blood radioactivity of 1 at 6–24 h and brain/plasma ratio of 2.3–4.4 from 1 to 24 h after a dose of tesevatinib (75). The studies also reported good antitumor activity with extended median survival time by 20% in preclinical mice models. Considering the preclinical results, Berz D et al. enrolled NSCLC pts with EGFR activating mutation and BM (*n* = 4) or leptomeningeal metastases (LM) (*n* = 3) which had progressed after prior EGFR TKI therapy (76). The authors used RECIST 1.1 for BM measurement and response evaluation. Symptomatic LM disease was diagnosed with CSF cytology or MRI finding and response was measured by improvement in symptoms, CSF cytology, and/or MRI findings. One patient with BM had 19% reduction of target lesion on day 23, and another patient had 57% reduction of target BM with resolution of LM symptoms and MRI findings on day 41. Grade ≥ 3 AEs were QTc prolongation, hypokalemia, dehydration, UTI, and ALT elevation.

## COMBINATORIAL/ALTERNATIVE TREATMENT APPROACHES

## Combination EGFR TKI and Radiotherapy

Brain metastasis are resistant to systemic chemotherapy due to the BBB which restricts passage of small, non-polar molecules, or those with receptor-mediated transport (77). Thus research has investigated whether radiation can enhance TKI efficacy. In preclinical models, EGFR TKIs have been shown to increase radiation responses by promoting radiation-induced apoptosis as well as inhibiting cellular cycling, DNA damage repair, accelerated repopulation, and angiogenesis (78–80).

Concurrent use of EGFR TKIs during radiotherapy remains in question. A retrospective study of 44 EGFR-mutant NSCLC treated with concurrent radiotherapy evaluated adverse events (AEs) (81). The most common AEs were rash (50%), anorexia (18%), and diarrhea (15%) with two patients having grade ≥ 3 rash (81). Radiation-related AEs included hydrocephalus (2 patients), pneumonitis (3 patients, one grade ≥ 3), myocarditis (1 patient), radiodermatitis (3 patients), laryngopharyngitis (2 patients), esophagitis (2 patients), and enteritis (1 patient) (81). A metaanalysis of 9 retrospective studies and 1 randomized controlled trial examining WBRT with EGFR TKI versus WBRT alone or EGFR TKI therapy alone included 1,041 unselected NSCLC patients with BM (82). In comparing combination therapy versus EGFR TKI alone the hazard ratios showed improved intracranial PFS with EGFR TKI alone (82). In comparing combination therapy versus WBRT alone the combination therapy had significantly improved OS (HR 0.52), intracranial PFS (HR 0.36), and extracranial PFS (HR0.52) (82). In addition, another metaanalysis of 15 studies including 3 phase II and 1 phase II trials in 1,552 unselected NSCLC patients with BM found that combination radiotherapy and EGFR TKI had improved response rate and disease control rate than radiotherapy alone or chemotherapy (83). Combination therapy significantly prolonged time to CNS progression (HR 0.56) and median OS (HR 0.58) but increased AE including rash (83).

A 2015 meta-analysis examined 12 observational studies that exclusively included EGFR-mutant NSCLC patients with brain metastasis (84). The analysis found that upfront cranial radiation improved intracranial PFS and 2-year OS but more neurological AEs were noted (84). One retrospective review by Gerber et al. examined 222 EGFR-mutant NSCLC BM patients treated with erlotinib, or WBRT or stereotactic radiation (SRS) (85). Patients treated with SRS had an OS of 64 months which was statistically significantly longer than the erlotinib group with median OS of 26 months (85). The results are likely biased due to the selection of patients with lower intracranial disease burden for the SRS approach. The median time to intracranial progression was understandably longer in the WBRT arm than the upfront erlotinib arm (24 vs 16 months; *p* = 0.04) (85). Another multiinstitutional analysis of 351 EGFR-mutant NSCLC patients with BM compared treatment with SRS followed by EGFR TKI, WBRT followed by EGFR TKI, or EGFR TKI followed by radiotherapy (SRS or WBRT) at intracranial progression (86). Those receiving SRS upfront had improved OS (46 months) compared to those receiving upfront WBRT followed by TKI (30 months), or upfront EGFR TKI (25 months) (86).

## Immunotherapy

PD-1 blockade has revolutionized the treatment of lung cancer and has been shown to have intracranial responses. However, many of the landmark immunotherapy studies have excluded EGFR-mutant NSCLC patients or patients with BM. Early analysis from a non-randomized, open-label, phase II trial showed 33% brain metastasis response rate among 18 NSCLC with BM (87). However, only one patient in this study had EGFR mutation (87). *In vitro* studies have shown that PD-L1 protein expression is higher in EGFR-mutant NSCLC cell lines than in EGFR wt and expression of mutated EGFR can induce PD-L1 expression (88, 89). In NSCLC, estimates of brain metastasis PDL1 positivity (PDL1 tumor cell expression exceeding 5%) have ranged from 12 to 52% (90–92) but this has not been well characterized in the EGFR-mutant population.

Given the potential for intracranial activity the question may arise if checkpoint inhibition has a role in EGFR-mutant NSCLC with BM. While there is a paucity of data for checkpoint inhibitors in this population, some extrapolation from EGFRmutant NSCLC is possible. A meta-analysis of Checkmate 057 (nivolumab), Keynote 010 (pembrolizumab), and POPLAR (atezolizumab) showed that immune checkpoint inhibition prolonged OS over docetaxel in EGFR wt but not EGFR-mutant NSCLC (93). As checkpoint inhibition does not appear superior to chemotherapy EGFR-mutant NSCLC, immunotherapies use in the EGFR-mutant NSCLC with BM population is likely equally reserved. Nevertheless, one should consider immunotherapy in later lines of therapy.

## Combinational EGFR TKI and Anti-Angiogenic Therapy

Several studies have looked at combining EGFR TKI with vascular endothelial growth factor directed monoclonal antibody therapy (94). The BELIEF trial was an international, multicenter, single-arm phase II trial of 109 treatment-naïve, advanced or metastatic, EGFR-mutant, lung adenocarcinoma patients treated with the combination erlotinib and bevacizumab (95). 37 patients (33%) harbored T790M mutations and 21 (19%) had brain metastasis; the median PFS was 13.2 months overall and 8.8 months for patients with brain metastasis (95). One of the greatest concerns with bevacizumab use among brain metastatic patients has been CNS hemorrhage. While CNS hemorrhage carries high morbidity and mortality, the incidence of CNS hemorrhage among bevacizumab-treated patients is less than 0.2% (96). Ongoing studies are investigating the combination of osimertinib and bevacizumab in EGFR-mutant NSCLC with BM (NCT02971501).

## CONCLUSION

In conclusion, patients with EGFR-mutant NSCLC are continuing to live longer with median overall survival of 30.9 months and nearly 15% of patients are alive at 5 years (97). As patients live longer, most of these patients are likely to develop BM and we will need optimal therapies with low toxicity to manage the BM. Based on the summary of literature to date (**Table 1**), it is the expert opinion of the authors that a CNS-active TKI such as osimertinib is the EGFR TKI of choice in newly diagnosed

## REFERENCES


advanced NSLC harboring exon 19 deletions or exon 21 L858R mutations, given not only its CNS response rate but the durability of the CNS control, in addition to compelling data with tripling of the median PFS. It would be hard to argue against our opinion that osimertinib is the drug of choice in patients with and without BM. That said, the data regarding the use of upfront SRS followed by EGFR TKI needs to be taken into account in personalizing treatment options for patients with EGFR-mutant NSCLC and BM. The EGFR-mutant NSCLC patient who presents with a solitary brain metastasis should still be considered for surgical resection followed by CNS-active EGFR TKI therapy such as osimertinib. Selected patients with CNS oligometastatic disease with large volume BM that are symptomatic may benefit from the Magnuson approach of using upfront SRS while those with military or multiple, small, and especially asymptomatic BM may be able to delay the need for radiation with the use of upfront EGFR TKIs such as osimertinib. Whole brain radiation should be an option that is reserved for refractory BM that have progressed beyond SRS and systemic therapies, thus delaying the onset of neurocognitive decline that almost inevitably follows such an approach. Novel WBRT techniques such as hippocampal sparing (RTOG 0933) or use of drugs such as memantine (RTOG 0614) may further help to reduce the long-term neurotoxicity of WBRT in these patients (98).

## AUTHOR CONTRIBUTIONS

WK drafted epidemiology, prevalence, TKI, radiotherapy, immunotherapy, and anti-angiogenic therapy sections. NS provided literature review and drafted TKI sections. DS provided concept, design for paper, scientific review, and interpretation of data. All authors contributed to manuscript revision, read, and approved the submitted version.


cancer: visualization by positron emission tomography in a patient harboring a mutation in the epidermal growth factor receptor. *J Thorac Oncol* (2011) 6(7):1287–9. doi:10.1097/JTO.0b013e318219ab87


non-small cell lung cancer patients with brain metastases: results of phase I dose-escalation study. *Lung Cancer* (2016) 96:93–100. doi:10.1016/j.lungcan. 2016.04.003


**Conflict of Interest Statement:** WK and NS do not have personal, professional, or financial relationships that constitute conflict of interest. DS has received honoraria for her role as advisory board member or speaker for Bristol Myers-Squibb, Takeda Oncology, and Astra Zeneca.

*Copyright © 2018 Kelly, Shah and Subramaniam. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Brain Metastases in Oncogene-Addicted Non-Small Cell Lung Cancer Patients: incidence and Treatment

#### *J. Remon1 \* and Benjamin Besse2,3\**

*1Medical Oncology Department, Hospital Universitari Vall d'Hebron, Barcelona, Spain, 2Cancer Medicine Department, Institut Gustave Roussy, Villejuif, France, 3University Paris-Sud, Orsay, France*

Brain metastases (BM) are common in non-small cell lung cancer patients including in molecularly selected populations, such as *EGFR*-mutant and *ALK*-rearranged tumors. They are associated with a reduced quality of life, and are commonly the first site of progression for patients receiving tyrosine kinase inhibitors (TKIs). In this review, we summarize incidence of BM and intracranial efficacy with TKI agents according to oncogene driver mutations, focusing on important clinical issues, notably optimal first-line treatment in oncogene-addicted lung tumors with upfront BM (local therapies followed by TKI vs. TKI monotherapy). We also discuss the potential role of newly emerging late-generation TKIs as new standard treatment in oncogene-addicted lung cancer tumors compared with sequential strategies.

#### *Edited by:*

*Marco Lucchi, Università degli Studi di Pisa, Italy*

#### *Reviewed by:*

*Minesh P. Mehta, Baptist Health South Florida, United States Emilio Bria, University of Verona, Italy*

*\*Correspondence:*

*J. Remon jremon@vhio.net; Benjamin Besse benjamin.besse@gustaveroussy.fr*

#### *Specialty section:*

*This article was submitted to Thoracic Oncology, a section of the journal Frontiers in Oncology*

*Received: 29 December 2017 Accepted: 14 March 2018 Published: 11 April 2018*

#### *Citation:*

*Remon J and Besse B (2018) Brain Metastases in Oncogene-Addicted Non-Small Cell Lung Cancer Patients: Incidence and Treatment. Front. Oncol. 8:88. doi: 10.3389/fonc.2018.00088*

Keywords: brain, metastases, non-small cell lung cancer, EGFR, ALK

## INTRODUCTION

Lung cancer is the leading cause of cancer-related deaths worldwide (1). Because of the lack of screening programs in most countries, more than half of non-small cell lung cancer (NSCLC) patients are diagnosed at an advanced stage1 . The brain is a common metastatic site in this population, with 30% of patients developing brain metastases (BM) during the course of their disease, with the brain being the only site of metastatic disease in 51% of these cases. Median delay between diagnosis of the primary tumor and development of BM is 11 months. Up to half of cases, patients present with synchronous diagnosis of BM at the time of diagnosis of the primary lung tumor (2). Ironically, the lifetime incidence of BM is increasing due to prolonged survival seen in NSCLC patients thanks to new systemic therapies and improved neuro-imaging techniques (3). Unfortunately, prognosis associated with BM remains poor with reports of median overall survival (OS) between 3 and 14.8 months (4), and compared to other metastatic sites, BM are responsible for a major decrease in quality of life (5).

The discovery of targetable genomic alterations in approximately 30% of advanced NSCLC tumors, mainly adenocarcinomas, has altered the therapeutic landscape and outcome of many of these subgroups of NSCLC patients (6, 7). In the recent era of personalized treatment targeting these alterations, prognosis of NSCLC patients with BM has improved significantly achieving a median OS of nearly 4 years (8). The question whether BM harbor distinct genetic alterations beyond those observed in primary tumors has not been definitively addressed. Recent data with

<sup>1</sup>Available from: www.seer.cancer.gov (Accessed: March 23, 2018).

whole-exome sequencing in 86 patient-matched BM (including 38 NSCLC patients) reported 53% of cases with potentially clinically informative alterations in BM that were not detected in the matched primary-tumor sample (9). However, these findings have a number of technical limitations and are yet to be supported by clinical evidence. On the other hand, response rates (RR) to targeted therapies in molecularly defined NSCLC patients are typically similar in central nervous system (CNS) and extra-CNS disease, arguing for fewer molecular discordances between the primary tumor and CNS metastases, at least for actionable mutations. This is an important issue to resolve for determining the best treatment strategies for managing BM.

One important consideration, when interpreting CNS efficacy with tyrosine kinase inhibitors (TKIs) in molecularly selected NSCLC patients, is the inherent limitation of the standard RECIST criteria for the measurement of baseline CNS disease and response (10). This assessment does not account for potential pseudo-progression correlating with radionecrosis and non-viable tumors in patients who have received brain radiotherapy (11). New imaging tests might offer better characterization of CNS progression vs. pseudo-progression (12). While the systemic efficacy of TKI in oncogene-addicted NSCLC has been well established, their intracranial efficacy is today less well validated for a number of reasons. Brain imaging during followup is often optional in clinical trials, MRI is not commonly used compared to the less sensitive CT scan, and patients with BM are often excluded from NSCLC trials, and when they are accepted, BM is not a stratification criteria. The CNS is shielded by the blood–brain barrier (BBB) and is considered a "pharmacological sanctuary." The key molecular properties that influence the BBB are the P-glycoprotein (P-gp) or breast cancer resistance protein substrate nature of the TKIs, their molecular weight, polar surface area and lipophilicity index (LogP) (13). These factors may explain why only 2% of small-molecule drugs are able to effectively cross the BBB (14), likely explaining why the CNS is a frequent site of failure after clinical benefit with some TKIs.

In this review, we summarize the incidence of BM in oncogene-addicted NSCLC patients and CNS efficacy for personalized treatment in these different sub-populations. We also evaluate new challenges such as the value of upfront personalized treatment vs. radiotherapy in oncogene-addicted NSCLC patients with BM at baseline, and administration of more potent drugs upfront vs. sequential treatment.

#### EGFR-MUTANT NSCLC PATIENTS

Within the lung cancer population, activated epidermal growth factor receptor (*EGFR)* mutations occur in 10% of Caucasians and 50% of Asians (15). There are several classes of activating somatic *EGFR* mutations, with in-frame deletions in exon 19 (ELREA, *Del1*9) and single-point mutations in exon 21 (*L858R*) being the most common. These mutations predict sensitivity to first- and second-generation EGFR TKIs, such as erlotinib, gefitinib, or afatinib. RRs and progression-free survival (PFS) with EGFR TKIs have proven superior to standard first-line platinum doublet chemotherapy, making them the current upfront standard of care (16). Recently, osimertinib a third-generation EGFR TKI, showed a significant improvement in PFS compared with standard of care (erlotinib or gefitinib) as first-line treatment, making it a new treatment option in the first-line setting (17).

#### Incidence of BM in EGFR-Mutant NSCLC

The baseline incidence of BM in *EGFR*-mutant NSCLC is similar to that of other oncogenic driver mutations, ranging from 23 to 32% (18–20). The cumulative incidence increases over time (19, 21), with a 2-year actuarial risk of CNS progression of approximately 15–20% when patients received standard of care EGFR TKIs (21, 22). BM development on EGFR TKI treatment is significantly more common among patients with baseline BM (2-year cumulative incidence of 47% among patients with pre-existing BM compared to 11% among those with no prior BM; *p* = 0.003) and correlates with a worse outcome (21, 23, 24). Literature reporting the risk of cumulative incidence of brain progression according to *EGFR* mutation subtype is contradictory, some studies reporting higher cumulative risk among *Del19-*mutant tumors (21), and others among *L858R*-mutant tumors (22, 24).

Although, it has been suggested that *EGFR* mutations appear early during multistep carcinogenesis and may even be associated with an increased propensity for metastatic cell to spread into the brain (25), the lifetime risk is confounded by this molecular subgroup's longer survival. However, some reports suggest that the incidence of BM is higher in *EGFR*-mutant patients compared to *EGFR*-wild type (31.4 vs. 19.7%, odds ratio 1.86, 95% CI: 1.39–2.49; *p* < 0.001) (18), but it could be explained by inability of first-generation EGFR TKIs to cross BBB, reported in up to 60% of patients (26, 27). The high incidence and significant rate of CNS failure highlights the need for additional strategies to prevent CNS progression.

#### Treatment With EGFR TKIs

First- and second-generation EGFR TKI brain penetration potential, measured by the unbound brain-to-plasma ratio, termed K*p*,uu, is very low (28), indicating that penetration into the brain is diffusion-limited or low passive BBB permeability (13). However, the importance of the BBB for intracranial tumors is debated. Retrospective observational and phase II studies have reported activity with erlotinib and gefitinib in *EGFR*-mutant patients with BM (29–34). Two studies with erlotinib showed intracranial RRs of 58 and 82% and intracranial PFS of 10.1 and 11.7 months (29, 30). Gefitinib achieved an intracranial RR of 88% and intracranial PFS of 14.5 months, with a time to salvage brain radiation from diagnosis of 17.9 months (32) (**Table 1**).

*In vivo* studies in NSCLC mice showed that afatinib penetrated the BBB and cerebrospinal fluid (CSF) levels correlated with plasma levels (35). In a compassionate-use program including 31 patients, afatinib demonstrated a 35% CNS response in molecularly non-selected patients who had previously failed TKI therapy, with a median time to treatment failure of 3.6 months (36). In a combined dataset *post hoc* analysis in 81 *EGFR*-mutant NSCLC patients with BM (30% had received brain radiotherapy) in the first-line LUX-Lung 3 and LUX-Lung 6 phase III clinical trials, afatinib significantly improved PFS (8.2 vs. 5.4 months, hazard ratio (HR) 0.50, *p* = 0.0297) and RR (21 vs. 5%, *p* = 0.0027), although without a significant difference in OS (22.4 vs*.*



*icDOR, intracranial duration of response; icRR, intracranial response rate; icPFS, intracranial progression-free survival; NA, not available; NR, not reached.*

*a Systemic RR and progression-free survival (PFS).*

*bPatients should have at least 3 metastatic brain lesions.*

*c In 50 evaluable patients.*

*dIn 30 evaluable patients with osimertinib.*

25.0 months; HR 1.14, *p* = 0.64) compared with platinum-based chemotherapy (37). The magnitude of the PFS benefit was suggested being increased, for patients who had received prior whole brabin radiotherapy (WBRT, *n*= 24; 13.8 vs*.* 4.7 months; HR 0.37, *p* = 0.07). Evaluation of intracranial response was not assessed as a separate endpoint in these trials (37), however, these results suggest that asymptomatic BM are not a limitation for upfront treatment with an EGFR TKI (**Table 1**).

Icotinib, another EGFR TKI only available in Asia, gave an intracranial RR of 65% and median PFS of 10 months in treatmentnaïve *EGFR*-mutant patients with at least three BM (38). AZD3759 is a novel reversible EGFR TKI, only active against sensitizing *EGFR* mutations, which was designed to effectively cross the BBB and achieves high drug-free exposure in the brain. In a phase I trial, it achieved an intracranial RR of 83% among 18 EGFR TKI treatment-naïve patients with evaluable BM (28) (**Table 1**).

The substitution of threonine to methionine at amino acid position 790 (*T790M*) in exon 20 of the *EGFR* gene reduces firstgeneration EGFR TKI binding by enhancing the ATP binding affinity of the kinase domain of the *EGFR*-mutant receptor (39). It accounts for acquired resistance in approximately 50–60% of patients (40, 41). Knowledge of acquired resistance mechanisms to EGFR TKIs was one of the triggers behind the development of third-generation EGFR-TKIs, such as osimertinib, which are active against *exon19* and *21* mutations as well as the *T790M* mutations. Osimertinib was the first such agent to receive FDA and EMA approval (in November 2015 and February 2016, respectively) for metastatic *EGFR*-mutant and acquired *EGFR T790M*-mutant NSCLC patients progressing on or after EGFR TKI therapy2,3.

The rate of acquired *T790M* mutations is discordant between intracranial and extracranial metastases. In a study of 78 *EGFR*mutant patients who had undergone re-biopsy after TKI failure, only 17% of CNS lesions were *T790M* mutated compared to 41% of systemic lesions (42), suggesting that the selection pressure is lower intracranially owing to the lower EGFR TKI concentrations in CSF compared to serum concentrations (42, 43). Preclinical data demonstrated greater penetration and brain exposure with osimertinib than with gefitinib, rociletinib, or afatinib (44).

Central nervous system activity of osimertinib was reported in pretreated *T790M*-positive NSCLC patients in the AURA study phase II extension component (45), the phase II AURA2 trial (46), and was recently confirmed in the phase III AURA3 trial (47) and the first-line FLAURA trial (17). In the pooled analysis of the two phase II trials (*N* = 411), osimertinib demonstrated an overall RR of 66% and median PFS of 11 months (48). In the pre-specified subgroup analysis of CNS response in this pooled analysis among 128 patients with CNS metastases at baseline, 50 were evaluable for CNS response. CNS response and DCR were 54 and 92%, respectively, and CNS response was observed regardless of prior radiotherapy. Median CNS duration of response (DOR) was not reached and at 9 months 75% of patients were estimated to remain in response. Median CNS PFS was not reached (49), with 1-year PFS of 56% (50). In the AURA3 trial, osimertinib demonstrated significantly greater efficacy in RR (71 vs. 31%) and PFS (10.1 vs. 4.4, HR 0.30, 95% CI: 0.23–0.41, *p* < 0.001) than platinum-pemetrexed chemotherapy, in 419 *T790M*-positive NSCLC patients who had progressed on firstline EGFR TKIs (47). Among 116 patients from the AURA3 trial with BM (measurable or not), PFS was longer with osimertinib compared to chemotherapy (11.7 vs. 5.6 months, HR 0.32; 95% CI: 0.15–0.69) and cumulative incidence of CNS progression at 6 months was lower with osimertinib compared to chemotherapy (11.5 vs. 28.2%) (51). Among 46 patients with evaluable BM, the intracranial RR was 70% with osimertinib compared with 31% with chemotherapy, with a median DOR of 8.9 vs*.* 5.7 months, respectively (51) (**Table 1**). It has been proposed that BM may not develop secondary resistance mutations to EGFR TKIs that develop during extracranial progression, due to reduced drug penetration of the BBB (52). However, CNS efficacy with osimertinib reported in AURA3 trial, appears to contradict this theory. In the CNS full analysis set (*N* = 128) from FLAURA trial, osimertinib reported improved CNS RR (66 vs. 43%) and longer CNS PFS (NR vs. 13.9 months, HR 0.48, 95% CI: 0.26–0.86, *p* = 0.04) and reduced the risk of CNS progression compared with the standard of care. Among evaluable CNS evaluable patients (*N* = 41), osimertinib improved the CNS RR (91 vs. 68%) with similar DOR compared with the standard of care (15.4 vs. 18.7 months) (53).

In light of the reduced CSF concentrations of EGFR TKIs, various studies have examined administration of high doses in an attempt to achieve therapeutic levels (54–57). "Pulsatile" erlotinib at 1,500 mg given weekly resulted in an intracranial RR of 67% with a median time to CNS progression of 2.7 months in nine patients (55). In a phase I trial, pulse and daily low-dose erlotinib prevented progression of untreated or new CNS metastases, without improving extracranial outcome compared with

<sup>2</sup>Available from: http://www.fda.gov/MedicalDevices/ProductsandMedical-Procedures/InVitroDiagnostics/ucm301431.htm (Accessed: March 23, 2018). 3Available from: http://www.ema.europa.eu/docs/en\_GB/document\_library/ EPAR\_-\_Product\_Information/human/004124/WC500202022.pdf (Accessed: March 23, 2018).

standard-dose (58). However, the limited number of patients, the short follow-up period and the fact that half of the patients with baseline BM had already been treated are confounding factors that prevent any conclusions being reached regarding efficacy in the CNS with this strategy.

#### Combined EGFR TKIs and Antiangiogenics

Activity and an acceptable safety profile of bevacizumab, an anti-VEGF monoclonal antibody, have been reported in NSCLC patients with asymptomatic and untreated BM (59). Moreover, in a preclinical model of lung adenocarcinoma, bevacizumab prevented BM formation (60). In a phase II trial in 154 Asian *EGFR*mutant NSCLC patients, the addition of bevacizumab to erlotinib as first-line treatment significantly improved PFS compared to erlotinib alone (16.0 vs. 9.7 months, HR 0.54; 95% CI: 0.36–0.79, *p* = 0.0015) (61), leading to EMA approval of the combination in this population in April 2016. Two ongoing phase III trials evaluating erlotinib combined with ramucirumab (NCT02411448) or bevacizumab (BEVERLY study, NCT02633189) compared with erlotinib, will hopefully further validate this strategy.

In the single arm phase II BELIEF trial in 109 Caucasian *EGFR*mutant NSCLC patients, combined erlotinib and bevacizumab gave median PFS and OS of 13.2 and 28.2 months, respectively. However, the primary endpoint was only met in baseline *T790M*positive tumors with a median PFS of 16 months, whereas in *T790M*-negative tumors, median PFS was 10.5 months (62). In the subgroup of patients with pretreated BM (*N* = 21), median PFS was 8.8 months. The efficacy of this combination in the BM population does not appear to be superior to standard EGFR TKI therapy (37), however only 21 patients with BM were included in the BELIEF trial. Results from the ongoing randomized phase II BRILLANT trial (NCT0265536), testing bevacizumab plus erlotinib vs. erlotinib in BM *EGFR*-mutant patients, should reveal the efficacy of this combination in this population. Also the combination of osimertinib and bevacizumab in *EGFR*-mutant NSCLC patients and BM is currently assessed in a phase II trial (NCT02971501).

## ALK-REARRANGED NSCLC PATIENTS

Anaplastic lymphoma kinase (*ALK)* rearrangements result from inversions or translocations on chromosome 2 and are present in ~5% of NSCLC tumors, with no apparent differences in incidence according to race. Crizotinib was the first treatment to be approved in this population achieving a median PFS of 10.9 vs*.* 7.0 months with platinum-pemetrexed chemotherapy in the front-line setting in the phase III PROFILE 1014 study (63). In the subsequent phase III ASCEND-4 trial in *ALK*-positive (by central immunohistochemistry) NSCLC patients, upfront ceritinib, a second-generation ALK TKI, gave a median PFS of 16.6 vs*.* 8.8 months with platinum-pemetrexed chemotherapy (64). Based on these results, the FDA approved ceritinib as first-line treatment in *ALK*-positive NSCLC patients in May 2017. More recently, the phase III ALEX trial demonstrated a significant improvement in PFS with alectinib (a second-generation ALK TKI) compared with crizotinib (25.7 vs. 10.4 months, HR: 0.50, 95%CI: 0.36–0.70, *p* < 0.001) by independent review, and with a better toxicity profile, as first-line treatment in *ALK*-positive NSCLC patients (65). The EMA and FDA approved alectinib as first-line treatment in 12 October 2017 and in 6 November 2017, respectively. Treatment strategies in this population are provided below and in **Figure 1**.

## Incidence of BM in ALK-Positive NSCLC

In *ALK*-positive NSCLC patients, CNS metastases affect from 24 to 42% of patients (19, 65–68) with risk increasing over time, reaching 58% at 3 years (19). In this population, median OS after development of BM was 49.5 months, with no survival differences detected according the number of BM (single vs. more than one BM) (69), confirming the prolonged survival of *ALK*-positive NSCLC patients with BM. However, the CNS is a common site of progression with crizotinib; in patients with known BM (treated or untreated), the CNS was a site of new lesions or non-target progression in 70% of cases of progression during crizotinib treatment. In patients without BM at the time of crizotinib initiation, 20% subsequently experienced CNS progression (68, 70). It remains controversial whether this increased risk was an expression of the natural *ALK*rearranged disease course independent of the therapy received, or if, as in *EGFR*-mutant NSCLC patients, it is related to low CSF penetrance of ALK TKIs. Crizotinib is a substrate for the ATP-binding cassette (ABC) drug efflux transporters, P-gp and ABC subfamily G member 2, and has been associated with poor accumulation of the drug in the brain, a CSF-to-plasma ratio of 0.0026 reported in a case study (71). In support of this, ABCB1−/− and ABCG2−/− mice had a 25- to 70-fold higher brain concentration following oral administration of crizotinib compared to wild type (71).

Nonetheless clinical evidence for crizotinib CNS efficacy has been reported. A pooled retrospective analysis of crizotinib efficacy in *ALK*-positive NSCLC patients with BM from the PROFILE 1005 phase II and PROFILE 1007 phase III trials has been reported (68). At baseline, 31% of patients (275 of 888) had asymptomatic BM. Analytic subgroups were stratified according to prior brain radiotherapy (60%) or not. The intracranial disease control rate (DCR) at 12 weeks was similar in these two groups at 62 and 56%, respectively. Of note, previously treated patients demonstrated higher CNS objective RR with crizotinib (33 vs*.* 18%, respectively), as well as prolongation of the median time to intracranial progression (13.2 vs*.* 7.0 months, respectively; **Table 2**) (68). Intracranial efficacy of crizotinib in treatmentnaïve *ALK*-positive patients was studied in the PROFILE 1014 trial (70). Of 343 patients, 79 (23%) had treated BM at baseline. Compared to chemotherapy, crizotinib demonstrated longer PFS (9.0 vs. 4.0 months; HR 0.40, 95% CI: 0.23–0.69, *p* < 0.001) and a better RR (77% vs. 28%, *p* < 0.01). Crizotinib achieved a 12-week intracranial DCR of 85% and median intracranial time to progression of 15.7 months in patients with treated BM (**Table 2**). CNS progression as the only site of progression on crizotinib was reported in 38% of patients with treated BM at baseline and 19% without (70). In the randomized phase III ALEX trial, crizotinib was used as the standard of care in the control arm and brain MRIs were mandatory at baseline and during follow-up (65). Among the 22 patients with measurable BM

at baseline, crizotinib achieved an intracranial RR of 50% and a median duration of intracranial response of 5.5 months. For the seven patients previously treated with brain radiotherapy, crizotinib gave an intracranial RR of 71% and median DOR of 17.3 months (72). Despite these data suggesting intracranial efficacy with crizotinib, especially among previously-treated BM patients, recent data showed that the cumulative incidence rate of CNS progression at 12 months was consistently higher with crizotinib compared with alectinib (32 vs*.* 4.6% respectively in patients without BM at baseline; HR 0.14, 95% CI: 0.06–0.33, *p* < 0.0001) (65, 72), suggesting that the risk of BM progression may correlate more closely with ALK TKI subtype and not the natural *ALK*-rearranged disease course.

#### Treatment With Novel ALK TKIs Ceritinib

Ceritinib is a second-generation ALK inhibitor that is 20 times as potent as crizotinib. It is effective in *ALK*-positive patients upfront and patients who progress while on crizotinib, including patients with BM (73). In the phase III ASCEND-4 trial, among 121 *ALK*-positive TKI-naïve NSCLC patients with BM, first-line treatment with ceritinib improved PFS (10.7 vs. 6.7 months, HR 0.70) compared to chemotherapy. Intracranial RR in 22 patients with measurable BM at baseline was 73%, with a median duration of intracranial response of 16.6 months (64). The CNS efficacy of ceritinib in crizotinib-pretreated and ALK-naïve patients was tested in the phase I ASCEND-1 trial (73), as well as the phase II ASCEND-2 (74), and ASCEND-3 (75) trials (**Table 2**). In the phase III, ASCEND-5 trial in previously treated (chemotherapy and crizotinib) *ALK*-positive NSCLC patients, ceritinib compared with chemotherapy significantly improved PFS across all patient subgroups, including in 133 patients with BM at baseline (56% previously treated with brain radiotherapy), from 1.5 months to 4.4 months (HR 0.54, 95% CI: 0.36–0.80). Among the 17 patients with measurable BM, ceritinib gave a 35% intracranial RR and median duration of intracranial response of 6.9 months (76). Nonetheless, despite these second-generation more potent ALK TKIs, BM remained the main site of progression among patients with BM at baseline (76). Based on ceritinib efficacy, an international prospective phase II open-label study is ongoing (ASCEND-7, NCT02336451) specifically evaluating the anti-tumor activity of ceritinib in *ALK*-positive NSCLC patients with BM or leptomeningeal disease (previously treated with radiotherapy or not).

Gastrointestinal toxicity by ceritinib may reduce treatment compliance. The ASCEND-8 (NCT02299505) aimed to evaluate whether administering ceritinib, 450 or 600 mg, with a low-fat meal may enhance gastrointestinal tolerability vs. 750 mg fasted while maintaining similar exposure in 267 treatment-naive ALKpositive NSCLC (neurologically stable BM were stable). The study demonstrated similar efficacy in terms of ORR and DCR with less frequent dose reductions/interruptions and higher relative dose



*icRR, intracranial response rate; icTTP, intracranial time to progression; s/icPFS, systemic/intracranial progression-free survival (PFS); icDOR, intracranial duration of response; CT, chemotherapy; NA, not available; NR, not reached.*

*a ALK TKI naïve and chemotherapy-naïve or up to three lines of chemotherapy with progression during or after the last chemotherapy regimen.*

*bData reported for previously untreated BM/previously treated BM.*

*c 12-week intracranial disease control rate.*

*dResults expressed as ALK inhibitor-naïve patients, ALK inhibitor-pretreated patients.*

*e Patients receiving 180 mg/day.*

intensity. Ceritinib administered at 450 mg fed dose demonstrated an ORR of 78% and a median PFS of 17.6 months, suggesting this dose as a potential new treatment regimen. However, fed doses of ceritinib in patients with BM were not reported to provide a clear recommendation in this subset (77, 78).

#### Alectinib

Alectinib is a potent ALK TKI, active against several *ALK* mutations that confer resistance to crizotinib (79). It is able to penetrate the CNS and activity is expected based on animal models showing high brain-to-plasma ratios (0.63–0.94) and activity in intracranial tumor implantation models. Unlike crizotinib and ceritinib, preclinical studies suggest that alectinib is not a substrate of P-gp, a key drug efflux pump typically expressed in the BBB, and that it has greater CNS activity than other ALK TKIs (80). In the clinic, alectinib gave an intracranial RR of 52% in 21 crizotinib-resistant patients with baseline BM treated in a phase I trial (81). Alectinib was approved by the FDA in 2015 for *ALK*-positive crizotinib-resistant NSCLC patients based on two phase II clinical trials demonstrating a systemic objective RR of 50–52% (82, 83). A pooled analysis evaluating systemic efficacy of alectinib in both phase II trials enrolling 225 *ALK*-positive crizotinib-resistant NSCLC patients has been performed. Alectinib gave a systemic RR of 51%, and median PFS and OS of 8.3 and 26 months, respectively (84), with 11% of patients having CNS as the only site of progression (85). Intracranial efficacy of alectinib in this population was assessed in 136 crizotinib-resistant patients with BM (37% with measurable disease and 70% previously treated). Intracranial RR in the whole population was 43% (36% in previously irradiated vs. 59% in patients without prior radiation) with a median DOR of 11.1 months. For patients with measurable disease (*N* = 50), the intracranial RR was 64%, with complete response in 22%, and median intracranial DOR was 10.8 months (85). The phase III ALUR trial comparing alectinib with chemotherapy in 107 previously treated (chemotherapy and crizotinib) *ALK*-positive NSCLC patients, reported improved outcome with alectinib (PFS 9.6 vs. 1.4 months, HR 0.15; 95% CI: 0.08–0.29; *p* < 0.001) (86). Among 76 patients with baseline BM, alectinib achieved an intracranial RR of 36% reaching to 54% among the 40 patients with measurable BM (87). These results endorse preclinical data showing promising CNS efficacy profile with alectinib.

In the phase III ALEX trial, 303 previously untreated *ALK*positive (by immunohistochemistry) NSCLC patients were randomized to receive either alectinib (600 mg twice daily) or crizotinib (250 mg twice daily). Crossover was not allowed. As mentioned previously, PFS was significantly longer with alectinib than with crizotinib and delayed the onset of BM (65). In the ALEX trial, 122 out of 303 (40%) patients had asymptomatic BM at baseline. Alectinib achieved an intracranial RR of 59% with a systemic PFS similar to that reported in the whole population (HR 0.40, 95% CI: 0.25–0.64, *p* < 0.0001) (65), with a median PFS of 14 months among patients with BM at baseline who had not received previous radiotherapy (72). Among the 21 patients with measurable BM, alectinib gave an intracranial RR of 81% and median DOR of 17.3 months (65) (**Table 2**). Patients with previously-irradiated BM measurable lesions had higher intracranial RR (86 vs. 79%) and intracranial DOR (not reached vs. 17.3 months), compared with patients without prior radiotherapy (72). Similarly, the phase III J-ALEX trial in 207 Japanese *ALK*-positive NSCLC patients demonstrated the superiority of alectinib in terms of PFS over crizotinib (HR 0.34, 95% CI: 0.17–0.71, *p* < 0.0001), and delayed risk of CNS progression in patients with BM at baseline (HR 0.16, 95% CI: 0.02–1.28) and those without (HR 0.41, 95% CI: 0.17–1.01). Among 43 patients with BM at baseline, alectinib significantly improved systemic PFS over crizotinib as first-line treatment (HR 0.08; 95% CI: 0.01–0.61) (88).

#### Lorlatinib

Lorlatinib (PF06463922) is a selective, potent, brain-penetrant next-generation ALK and ROS1 TKI, active against most known resistance mutations (79, 89). Lorlatinib was tested in a phase I trial in 54 pretreated or treatment-naïve *ALK*- (*N* = 41) or *ROS-1* (*N* = 12) positive NSCLC patients (11% treatment-naïve, 52% two or more previous TKIs, and 72% with BM). Patients reached a RR of 46% in the *ALK*-positive population irrespective of the number of prior ALK TKI therapies, and median PFS and DOR of 9.6 and 12.4 months, respectively. Lorlatinib was highly active in the CNS, including intracranial responses in 8 of 19 (42%) *ALK*positive patients with baseline measurable BM, in over a half of whom two or more previous ALK TKIs had failed (90) (**Table 2**). The recommended dose for the phase II trial was 100 mg/day.

In the phase II trial (91), lorlatinib conferred a clinically meaningful benefit, including substantial intracranial efficacy ranging from 42 to 75% in patients with advanced *ALK*-positive disease who were treatment-naïve or who had received a range of prior ALK inhibitors and/or chemotherapies (**Table 2**). In the treatment-naïve cohort (*N* = 30), lorlatinib achieved an RR of 90%, neither PFS nor DOR were reached, while the intracranial RR was 75% among eight patients with BM at baseline. Among the 111 heavily pretreated (two or three previous TKI with or without chemotherapy) patients, lorlatinib reached an overall RR of 39% with median PFS of 6.9 months, and 48% intracranial response among 48 patients with BM at baseline (91). Lorlatinib received breakthrough therapy designation in April 2017 for *ALK*-positive patients previously treated with at least one ALK TKI. Based on these results, the ongoing randomized phase III CROWN trial (NCT03052608) is assessing the efficacy of lorlatinib compared to crizotinib as first-line treatment in *ALK*-positive NSCLC patients. Asymptomatic and pretreated BM are not exclusion criteria.

#### Brigatinib

Brigatinib is another new ALK TKI (also active against *ROS1*, *EGFR-T790M*, *IGFR*, and *FLT3* mutations) with a broader spectrum of preclinical activity than ceritinib and alectinib against known crizotinib-resistant *ALK*-mutants (79, 92). Brigatinib was granted break-through therapy designation by the FDA in October 2014 on the basis of its early phase I/II trial data (93). In the phase I trial, among 71 crizotinib-resistant *ALK*-positive NSCLC patients treated with brigatinib the confirmed RR was 62% with a median PFS of 13.2 months. Among 46 patients with BM at baseline, the RR was 53% and 35% for those with measurable (*n* = 15) and non-measurable (*n* = 31) intracranial metastases, respectively. The median intracranial PFS and DOR in this population was 15.6 and 18.9 months, respectively. The recommended dose for the phase II study was determined to be 180 mg/day with a 7-day lead-in at 90 mg to reduce the risk of pulmonary toxicity (93).

In the phase II trial, 222 crizotinib-refractory *ALK*-positive NSCLC patients were randomized to brigatinib 90 mg/day (arm A) or 180 mg/day with a 7-day lead-in at 90 mg (arm B) (94), and updated results were recently presented (95). By independent review, the RR was 51 and 55%, in arms A and B, respectively, and PFS was 9.2 and 16.7 months, respectively, while OS was not reached in arm A and was 27.6 months in arm B. This is the longest PFS in crizotinib-resistant tumors reported with new ALK TKIs. Based on these results, the FDA approved brigatinib in crizotinib-pretreated patients in 28 April 2017. Among the 154 patients with BM at baseline (69%), intracranial RR (by independent-review) in patients with measurable disease (*N* = 44) was 50 and 67% in arm A and B, respectively. For patients with active BM (*N* = 34) RRs were similar to those with baseline BM, 47 and 73% in arms A and B, respectively. Median intracranial PFS was 12.8 and 18.4 months in arms A and B, respectively (95). The intracranial efficacy of brigatinib compares favorably with other second-generation ALK TKIs (74, 85) (**Table 2**). Brigatinib is currently being investigated in a randomized phase III ALTA-1L (NCT02737501) trial comparing brigatinib vs. crizotinib in *ALK*-positive TKI-naïve patients. Asymptomatic and pretreated BM are not exclusion criteria. This trial allows crossover from crizotinib to brigatinib and may help to elucidate whether a sequential strategy is better than upfront brigatinib.

#### Ensartinib

Efficacy of ensartinib (X-396) 225 mg/day in an expansion study has been reported. Forty of the 80 enrolled patients were evaluable for response, achieving 58% partial responses (88% in eight crizotinib-naïve patients, and 64% in 22 crizotinib-resistant) (96). Updated results among 15 TKI-naïve patients showed an 80% RR and median PFS of 23.8 months (97). CNS responses [(60% partial responses) were observed in both crizotinib-naïve and crizotinib-resistant populations, with a median DOR of 5.8 months (98)]. The ongoing phase III XALT3 (NCT02767804) will compare ensartinib with crizotinib as first-line treatment (previous chemotherapy allowed).

#### OTHER MOLECULAR ALTERATIONS: ROS1, RET, BRAF, AND NTRK

#### ROS1 Rearrangements

*ROS1* rearrangement occurs in approximately 1 to 2% of NSCLC patients. Compared with ALK rearrangements, ROS1 rearrangements are associated with lower rates of extrathoracic metastases, including fewer BM at initial metastatic diagnosis (19 vs. 39%, *p* = 0.033) (99), however *ROS1* does increase the likelihood of BM (100).

In 50 *ROS1-*positive NSCLC patients, crizotinib achieved an RR of 72% and median PFS of 19.2 months (101). Based on these results, the FDA and EMA approved crizotinib for treatment of *ROS1*-positive NSCLC patients in March and August 2016, respectively. Recently, a phase II trial in 32 Asian *ROS1*-positive NSCLC patients, ceritinib gave an RR of 62%, median PFS of 9.3 months (19.3 months among 30 crizotinib-naïve patients), and median OS of 24 months. Among eight patients with BM, intracranial RR with ceritinib was 63% (102). In a phase I trial with lorlatinib, 12 *ROS1*-positive NSCLC patients achieved an intracranial RR of 50% (80% among five patients with target lesions) and median systemic PFS and DOR of 7 and 12 months, respectively (90) (**Table 2**).

In a phase II study in 47 *ROS*-positive NSCLC patients (28% TKI-naïve, 64% one previous TKI and 8% two or more previous TKIs) treated with lorlatinib, the RR was 36%, with a 45% DCR at 24 weeks, and median PFS and DOR of 9.6 and 13.8 months, respectively (91). Among the 25 patients with BM at baseline, intracranial RR was 56%.

Entrectinib is another ROS1 TKI (also active against *ALK* and *NTRK*) specifically designed to cross the BBB. In a phase I/II trial, entrectinib (600 mg QD) achieved a RR of 78% and median PFS of 29.6 months among 32 treatment-naïve *ROS1*-positive NSCLC patients. The intracranial RR was 83% among 11 patients with BM at baseline (103, 104). Pending questions are the best treatment sequential strategy and whether *ROS1*-positive NSCLC patients with BM should be treated upfront with entrectinib. Given the low *ROS*-1 incidence, it is difficult to perform a randomized trial comparing different treatment strategies.

#### RET Rearrangements

In NSCLC, *RET* rearrangements occur in 1 to 2% of unselected cases and 16% of NSCLC tumors that lack other oncogenic drivers. They are more common in adenocarcinomas and in never or lighter-smokers (105, 106). *RET*-rearranged NSCLC patients benefit from pemetrexed-based chemotherapy to a comparable extent as *ALK*- and *ROS1*-rearranged patients (107). Multikinase inhibitors, such as cabozantinib (108) and vandetanib (109, 110) in phase II or retrospective studies (105), have limited efficacy, with RR between 18 and 53%, median PFS between 2.3 and 4.5 months (105, 108–110), and median OS of 6.8 months (105). It has been speculated that the type of fusion partner may play a role in determining treatment response (109); however, this was not validated in the retrospective study (105).

Baseline BM incidence in *RET*-rearranged NSCLC is 27%, without differences in age, smoking status or fusion-partner type. Lifetime incidence of BM in *RET*-rearranged NSCLC patients is 49%. In 37 patients treated with multikinase inhibitors with activity against RET, there were no significant differences in median PFS (2.1 vs. 2.1, *p* = 041) or median OS (3.9 vs. 7.0 months, *p* = 0.10) in patients with BM (*N* = 10) vs*.* without (*N* = 27) (111). In the phase II trial with cabozantinib, baseline untreated BM were present in five patients. Cabozantinib achieved intracranial disease control in two patients with measurable disease (−34 and −1%). Brain progression during TKI treatment may be less common than in other oncogenic alterations. Of 22 patients who discontinued cabozantinib, BM was the cause in only 10% of cases (111). Similarly, intracranial responses have been reported with alectinib, one patient responding after escalating alectinib to 900 mg twice daily (112). The efficacy of alectinib (900–1200 mg/day) as first-line treatment in *RET*-positive NSCLC patients will be assessed in a multi-cohort phase II/III B-FAST trial (NCT03178552). Treated and asymptomatic BM will be allowed. LOXO-292, another RET TKI has reported tolerability and efficacy in *RET*-dependent cancers even in progressive BM after alectinib (113).

In a phase I trial, vandetanib and everolimus showed antitumor activity in *RET*-positive NSCLC patients with BM (114, 115). The short-term outcomes with multikinase inhibitors with activity against RET compared to EGFR/ALK TKIs in *EGFR*mutant/*ALK*-rearranged NSCLC, strongly suggest that there is a need for more selective and potent RET targeted agents as monotherapy or in combination in order to enhance activity (116).

#### BRAF-Mutants

The combination of the BRAF inhibitor dabrafenib with the MEK inhibitor trametinib was approved by the FDA and EMA based on clinical activity in 57 pretreated *BRAF-V600E*-mutated NSCLC patients (1–2% of lung adenocarcinoma patients) following a phase II trial giving an RR of 67% and median PFS and OS of 8.6 and 18.2 months, respectively, however no data regarding CNS efficacy are available (117). Similar outcomes were recently replicated among 36 TKI-naïve BRAF (*V600E*)-mutant NSCLC patients (118). In melanoma BM patients, this combination has reported intracranial responses (119), making it highly probable that activity will be observed in NSCLC patients, although this needs to be validated.

#### NTRK Rearrangements

Fusions involving the genes *NTRK1, NTRK2,* and *NTRK3* are oncogenic drivers. They encode the proteins TRKA, TRKB, and TRKC, respectively, and play roles in neuronal development, cell survival, and cellular proliferation (104). These fusion genes have been detected in a variety of tumors including lung in up to 3% of cases, using different assay (NGS or FISH-based) (120). Entrectinib has reported efficacy in *NTRK*-positive tumors, including NSCLC patients, with a median PFS of 15.6 months (104) and also intracranial activity (104, 120), confirming that entrectinib crosses the BBB. Larotrectinib (LOXO 101) is a pan-TRK TKI. In a phase I clinical trial with 55 *NTRK*-positive solid tumors (five NSCLC patients), larotrectinib achieved an RR of 78% across a wide range of ages and tumor types (121). CNS efficacy of this agent remains unknown.

#### UPFRONT TKIs vs. UPFRONT RADIOTHERAPY IN ONCOGENE-ADDICTED NSCLC

In oncogene-addicted NSCLC, TKIs have clearly demonstrated increased CNS efficacy, including with next-generation TKIs, which are more potent than first-generation TKIs. Most data have been generated in *EGFR*- or *ALK*-positive patients, although similar outcomes are expected with other druggable alterations. Nonetheless, alternative treatment options exist in this group such as surgery, WBRT or stereotactic radiosurgery (SRS) (122), and the optimal treatment combination or sequence remains unclear.

#### Sequential Strategies

A systematic review and meta-analysis of 12 non-comparative studies in 363 *EGFR*-mutant NSCLC patients with BM, showed evidence that upfront radiotherapy (SRS or WBRT) improved survival outcomes (123). However, this study is based on published data and not on individual patient data limiting its validation. This study also reported that radiotherapy caused more neurological adverse events relative to EGFR TKIs alone. In a retrospective multi-institutional analysis in 351 *EGFR*-mutant TKI-naïve NSCLC patients with BM, median OS for three alternative strategies, SRS followed by an EGFR TKI (*n* = 100), WBRT followed by an EGFR TKI (*n* = 120), or an upfront EGFR TKI (*n* = 131), was 46, 30, and 25 months, respectively (*p*< 0.001) (124). In a multivariate analysis, SRS and WBRT vs. EGFR TKI were associated with improved OS, but not with median time to intracranial progression, suggesting that an upfront EGFR TKI and deferred radiotherapy is associated with inferior OS. SRS followed by EGFR-TKI resulted in the longest OS and allowed patients to avoid the potential neurocognitive sequelae of WBRT. However, the retrospective setting meant that data for quality of life and chronic neurocognitive assessments, extracranial disease burden were unavailable, and randomized study design was not used, all of which can be considered as limitations of this analysis. In addition, it is likely that there were a higher number of oligo-metastatic patients in the SRS arm, in whom there is not an urgent need for a TKI to control the extracranial disease, which would generate a major bias.

In a retrospective study (*n* = 97), intracranial PFS was improved in patients who received upfront radiotherapy followed by icotinib compared to those receiving icotinib alone, although without OS improvement (125). However, the absence of randomization makes it difficult to draw a conclusion. On the other hand, in a phase III trial, upfront icotinib (*N* = 85) compared with WBRT (30 Gy) plus chemotherapy (*N* = 91) in *EGFR*-mutant patients with at least three BM significantly improved intracranial PFS (10.0 vs*.* 4.8 months; HR 0.56, 95% CI: 0.36–0.90; *p* = 0.014), intracranial RR (67.1 vs. 40.9%, *p* < 0.001), and systemic RR (55.0 vs*.* 11.1%, *p*< 0.001), with a better toxicity profile. Median OS had no significant difference between the arms (18.0 vs. 20.5 months; HR 0.93, 95% CI: 0.60–1.44, *p* = 0.734) (38).

Any of these studies evaluated radiotherapy strategies compared to third-generation EGFR TKIs, so, a prospective randomized trial evaluating intracranial progression after SRS (to avoid the potential neurocognitive sequelae of WBRT) followed by third-generation EGFR TKI vs. third-generation EGFR TKI followed by SRS is needed. The clinical question has also been raised as to whether SRS as consolidative treatment in brain residual disease after EGFR TKI response could improve intracranial PFS in this population or whether this radiotherapy should be only administered in cases of progression on an EGFR TKI.

#### Concomitant Strategies

In a recent meta-analysis, radiotherapy plus EGFR TKIs resulted in a superior RR and DCR, and markedly prolonged the CNStime to progression and OS of NSCLC patients with BM (126), although patients were not selected according to EGFR status. The role of combining an EGFR TKI with WBRT was investigated in a single arm phase II trial of 40 patients (17 *EGFR*-mutant) (127). Patients received erlotinib 150 mg/day for 1 week, followed by erlotinib with concurrent WBRT (2.5 Gy/day, 5 days per week, to 35 Gy) and underwent formal cognitive testing before enrollment and at each follow-up visit. In the *EGFR*-mutant subset, patients had longer OS compared to wild-type *EGFR* (19.1 vs. 9.3 months, respectively). Erlotinib was well tolerated in combination with WBRT with no unexpected cases of neurotoxicity.

In a retrospective study in 133 *EGFR*-mutant patients with BM, radiotherapy (WBRT, SRS) and EGFR TKIs (erlotinib, gefitinib) improved median cranial PFS (16.0 vs. 11.5 moths, *p* = 0.017) and median OS (22 vs. 15 months, *p* = 0.015) compared with EGFR TKIs alone (128). On the contrary, in another retrospective cohort of 230 *EGFR*-mutant BM NSCLC patients, the addition of WBRT to EGFR TKIs compared to EGFR TKIs alone did not result in significant differences in intracranial PFS (7.4 vs. 6.9, *p* = 0.23) or systemic PFS (7.9 vs. 7.5, *p* = 0.55), and combined treatment was associated with worse survival (26.4 vs. 21.6 months, *p* = 0.049) (129). These results should be interpreted with caution given the sample sizes, absence of evaluation of side effects and non-randomized study design. While it can be argued that EGFR TKIs can be safely administered with concurrent WBRT (although for *ALK*-positive patients no data are available), high level evidence to support this is lacking, and concomitant strategies are not overtly recommended in either clinical guidelines (122) or in a recent systematic review (130). In cases of asymptomatic BM patients, given the unclear potential synergistic cognitive toxicities caused by combined therapies, WBRT or SRS should be delayed when other effective systemic therapies are available. A recent systematic literature review about results of combined irradiation and targeted therapies has been also recently published (131).

#### Continuing TKIs With and Without Local Therapy

Many strategies to treat CNS disease in *ALK*-positive NSCLC patients have been reported as case reports, such as high-dose crizotinib with a limited intracranial PFS of 1 month (132) and high-dose pemetrexed in combination with high-dose crizotinib with overall stable cerebral disease for 7 months. However, it remains unknown whether the response is attributable to one or both drugs given at high dose (133). In preclinical models, for enhancing CNS drug penetration, P-gp inhibitors such as elacridar, increased the intracranial concentration of crizotinib ~70-fold (134).

In 120 *ALK*-positive NSCLC patients continuing crizotinib beyond initial progression (51% with brain progression), longer median survival was reported compared with patients who received other chemotherapy (16.4 vs*.* 5.4 months) (135), although this benefit could also be related to local therapies and more indolent disease in the crizotinib arm. Treating isolated CNS progression with local therapies (surgery and/or radiotherapy) while continuing crizotinib could be viewed as an acceptable option (136). In the PROFILE 1014 study, among 25 patients with intracranial progression on crizotinib, 19 received radiotherapy, while continued crizotinib achieved a median treatment time beyond progression of 5.1 months, which was longer than the 2.9 months achieved with crizotinib beyond progression among patients with extracranial progression (70). In a retrospective single-institution study, local therapy (either surgery or radiotherapy) for BM in *EGFR*-mutant (17 treated with erlotinib) or *ALK*-rearranged (38 treated with crizotinib) NSCLC patients and CNS progression allowed continuation of therapy for an additional 7.1 months (137).

Recent studies have reported that *ALK*-positive NSCLC patients with BM treated with SRS and/or WBRT and TKIs have prolonged survival (68, 69, 138). Given the extended OS for *ALK*-positive patients and frequent need for repeated courses of CNS radiotherapy, SRS is the preferred strategy for minimizing cerebral toxicity. Synergistic efficacy of crizotinib and radiotherapy could be explained by increased BBB permeability and decreased P-gp expression following irradiation (139). These results suggest intracranial interventions and TKIs beyond progression are of value in patients with asymptomatic and limited CNS progression on a TKI. This SRS strategy is being validated in an ongoing phase II clinical trial (NCT02314364) among oncogene-addicted *(EGFR-, ALK-, ROS1-*positive*)* NSCLC patients with up to four BM.

The promising CNS activity of the next-generation TKIs suggests that switching targeted agents may be a reasonable alternative to local therapies. However, prospective data are needed to determine which strategy offers the best OS, intracranial control rate, quality of life and therapeutic ratio, taking into account the number of BM and whether patients are symptomatic at the time of progression.

#### Second- or Third-Generation TKIs Upfront or Sequentially

In *EGFR*-mutant NSCLC patients, osimertinib has reported higher intracranial activity compared with chemotherapy (51) and first-generation EGFR TKIs (17), and longer delay of onset of BM (51). However, lack of stratification according the presence of BM, no reported survival benefit with osimertinib and no prospective validation of this efficacy, limit interpretation. Nonetheless, preclinical data strongly support the increased intracranial efficacy of osimertinib compared with other EGFR TKIs (44). The ongoing phase II APPLE trial (NCT02856893) assesses the optimal strategy for delivering osimertinib in *EGFR*mutant NSCLC patients and will prospectively validate the efficacy of osimertinib among those patients with BM at baseline (stratification criteria and brain MRI will be performed at baseline) and also the time to radiological brain progression respect to with first-generation EGFR TKI (gefitinib) (140).

In *ALK*-positive NSCLC patients, based on this significant PFS improvement with alectinib and the delay of CNS progression compared with the current standard first-line crizotinib in *ALK*-positive NSCLC patients, alectinib has became a new standard treatment, and is approved by the EMA and FDA. However, it has not yet been demonstrated whether upfront treatment with second-generation ALK TKIs impact OS compared with sequential treatment strategies (**Figure 1**). In *ALK*-positive NSCLC patients, 4-year OS was 57% with upfront crizotinib in the randomized phase III PROFILE 1014 trial (*N* = 172), and was 70% with alectinib (300 mg twice daily) among 43 Japanese patients included in a phase II trial (141). Although data are immature, no survival benefit has been reported with upfront alectinib compared with crizotinib in the ALEX trial (65). Also, in PROFILE 1014, patients who received crizotinib followed by another ALK-TKI had longer OS compared with those randomized to chemotherapy followed by no ALK-TKI or other treatment (who had the poorest OS), suggesting a potential benefit of sequential strategies (142).

In a multicentre retrospective study, OS in patients treated with crizotinib followed by alectinib tended to be longer than in patients treated with alectinib alone (143) and median OS up to 50 months has been reported in patients who receive sequential strategies with upfront crizotinib (144, 145). A French nationwide retrospective cohort (CLINALK study) with 318 *ALK*positive NSCLC patients reported that patients who received next-generation ALK TKIs after crizotinib progression (ceritinib, alectinib, lorlatinib; *N* = 84) had improved OS, reaching a median of 89.6 months (146). Large-scale prospective studies are needed to confirm these preliminary observations.

Finally, each ALK TKI is associated with a distinct spectrum of *ALK*-resistant mutations, and the frequency of mutations increases significantly after treatment with second-generation ALK TKIs (20% with crizotinib vs. 53% with alectinib) (79). It is important to note that there are few new ALK TKIs that may overcome alectinib resistance, and efficacy is dependent on the acquired *ALK* mutation subtype upon progression on alectinib (79). Lack of a tissue biopsy for molecular profiling at progression and limited access to new ALK TKIs worldwide might limit access to subsequent therapies in alectinib-resistant diseases. Validation of liquid biopsies for dynamic markers of TKI efficacy (147) as well as predictive markers for personalized treatment at progression on ALK TKIs is also a challenge. On the other hand, the high CNS response and the delay in the onset of BM with alectinib, which could have a positive impact on patients' quality of life, might justify first-line treatment with alectinib in this population.

## CONCLUSION

Brain metastases are common in NSCLC including in molecularly selected populations, and are associated with a reduced quality of life. A multidisciplinary approach is the optimal strategy in oncogene-addicted NSCLC patients with BM. Based on the available clinical data and long OS in patients with asymptomatic synchronous BM at diagnosis, upfront treatment with TKIs alone should be considered with close CNS surveillance for early intervention in patients with an inadequate CNS response. This strategy may defer CNS radiotherapy and avoid long-term neurologic sequelae associated with local therapies. For patients with symptomatic BM, initial TKI therapy is an option, especially in *EGFR*-mutant and *ALK*-positive NSCLC patients treated with new *EGFR* and ALK TKIs based on their higher CNS efficacy. In other cases, sequential treatment initiated with local therapy followed by a TKI is appropriate. For patients who experience CNS progression with controlled extracranial disease while on TKI treatment, local therapy (preferably SRS) followed by the same TKI is an option in patients with a limited number of lesions or who are asymptomatic. In cases of multiple CNS progression, a switch to another TKI with higher CNS-penetration activity with or without WBRT is appropriate.

## REFERENCES


## AUTHOR CONTRIBUTIONS

Review concept, design, acquisition, analysis, and interpretation of data performed equally by BB and JR.


and asymptomatic brain metastases: a phase II study (CTONG-0803). *Ann Oncol* (2013) 24:993–9. doi:10.1093/annonc/mds529


non-small-cell lung cancer (AURA2): a multicentre, open-label, single-arm, phase 2 study. *Lancet Oncol* (2016) 17:1643–52. doi:10.1016/ S1470-2045(16)30508-3


lung cancer and activating EGFR mutations (BELIEF): an international, multicentre, single-arm, phase 2 trial. *Lancet Respir Med* (2017) 5:435–44. doi:10.1016/S2213-2600(17)30129-7


plasma T790M in EGFR-mutant NSCLC patients. EORTC 1613. *Clin Lung Cancer* (2017) 18:583–8. doi:10.1016/j.cllc.2017.02.005


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Copyright © 2018 Remon and Besse. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Response to HeR2 Inhibition in a patient With Brain Metastasis With eGFR tKI acquired Resistance and an HeR2 amplification

*Arenda D. Meedendorp, Arja ter Elst, Nils A. 't Hart, Harry J. M. Groen, Ed Schuuring and Anthonie J. van der Wekken\**

*University of Groningen and University Medical Center Groningen, Groningen, Netherlands*

#### *Edited by:*

*Lizza E. L. Hendriks, Maastricht University Medical Center, Netherlands*

#### *Reviewed by:*

*Timothy F. Burns, University of Pittsburgh Cancer Institute, United States Luigi Formisano, Vanderbilt University Medical Center, United States*

> *\*Correspondence: Anthonie J. van der Wekken a.j.van.der.wekken@umcg.nl*

#### *Specialty section:*

*This article was submitted to Thoracic Oncology, a section of the journal Frontiers in Oncology*

*Received: 10 January 2018 Accepted: 04 May 2018 Published: 22 May 2018*

#### *Citation:*

*Meedendorp AD, ter Elst A, 't Hart NA, Groen HJM, Schuuring E and van der Wekken AJ (2018) Response to HER2 Inhibition in a Patient With Brain Metastasis With EGFR TKI Acquired Resistance and an HER2 Amplification. Front. Oncol. 8:176. doi: 10.3389/fonc.2018.00176*

A 62-year-old man was referred to our university hospital for treatment of advanced adenocarcinoma of the lung after disease progression on two lines of EGFR TKI and one line of chemotherapy. Fluorescent *in situ* hybridization analysis upon progression showed an *HER2* amplification. At our weekly Molecular Tumor Board (MTB), a decision was made to treat this patient with afatinib, which resulted in a partial response. However, progression was observed with a facial nerve paresis due to a metastasis in the skull. A biopsy of a location in the thorax revealed the presence of an EGFR-T790M mutation associated with acquired resistance, after which treatment with osimertinib was started. After 6 months, disease progression was observed, and a new biopsy was taken from the pelvic bone, which revealed the original amplification of *HER2* together with the EGFR-L858R mutation, the EGFR-T790M mutation was not detected. The MTB decided to treat the patient with trastuzumab/paclitaxel. A partial response was observed in different bone lesions, while the skull metastasis with ingrowth in the brain remained stable for 6 months. Because of progression of the bone metastases after 6 months, a biopsy of a lesion in the thorax wall was taken. In this lesion, the EGFR-T790M mutation could be detected again. The MTB advised to start treatment with a combination of osimertinib and afatinib. This resulted in an impressive clinical improvement and a partial response of the bone metastases on the most recent 18-fluorodeoxyglucose positron emission tomography and computer tomography-scan. In conclusion, adjusting treatment to the mutational make-up of the tumor is a great challenge. For optimal treatment response multiple biopsies and re-biopsy upon progression are imperative. As more genes are investigated, treatment decision becomes increasingly difficult, therefore, expert opinions from an MTB is essential.

Keywords: EGFR, HER2, non-small-cell lung carcinoma, TKI, brain metastasis, Molecular Tumor Board

## INTRODUCTION

Treatment of driver mutations cannot be based on large clinical trials or high levels of evidence at all times. However, a Molecular Tumor Board (MTB) can help in making treatment decisions based on databases, case reports, xenograft models, and cell lines. Here, we present such a case.

A 62-year-old man was referred to our university hospital for treatment of advanced adenocarcinoma of the lung after disease progression on two lines of EGFR TKI and one line of chemotherapy in September 2015 (**Figure 1**).

Four months prior, in April 2015, he was diagnosed with an adenocarcinoma of the left lung with multiple bone metastases in sternum, ribs, and vertebrae. A biopsy from a metastasis in the left femur showed a mutation in the *EGFR* gene: c.2573T>G; p.(L858R). He was initially treated with gefitinib. After 2 months, the patient showed progression of bone metastases; the same EGFR mutation was found in a biopsy of a rib metastasis, without additional mutations in other mutational hotspots (e.g., *BRAF*, *KRAS*, *HER2*, *KIT*, *ALK*, *NRAS*, *PDGFRA*, *PIK3CA*, and *MET*). Therefore, at that time, carboplatin and pemetrexed were provided and because of pain, local irradiation of a sternal metastasis was applied. After two cycles of chemotherapy, the patient showed disease progression and was referred to our hospital. Because of lack of new treatment options, this patient was discussed in the Groningen MTB consisting of pulmonary oncologists, pathologists, clinical molecular biologists in pathology, general oncologists, and a structural biologist (www.moloncopath.nl). The MTB advised to determine the HER2-copy number status as a possible resistant mechanism for EGFR TKI. Fluorescent *in situ* hybridization (FISH) on a biopsy of a subcutaneous thoracic metastasis revealed *HER2* amplification and treatment with afatinib (dual EGFR and HER2 inhibitor) 30 mg QD was started in October 2015. Evaluation by 18-fluorodeoxyglucose positron emission tomography and computer tomography (18-FDG-PET-CT) showed after 6 weeks a significant partial response with disappearance of the FDG activity of the bone metastases and after 4 months in the left upper lobe a single FDG-positive lesion was left. This lesion was irradiated by means of stereotactic ablative radiotherapy (1× 20 Gy), and afatinib was continued. Treatment with afatinib was well tolerated with minor skin rash; patient showed clinical improvement: he had less pain and more energy. Nine months after start of afatinib, progressive disease was again noticed. Growth of the primary tumor in the left upper lobe, a new ipsilateral pulmonary lesion and multiple new bone metastases including the skull, with ingrowth into the brain, causing paralysis of the right facial nerve (**Figure 1**). Sequence analysis of a new right-sided rib lesion showed the known L858R *EGFR* mutation and an additional T790M mutation.

Because of the novel T790M, afatinib was discontinued and replaced by osimertinib 80 mg QD (1). Eight weeks after start of osimertinib a PET-CT showed a response of most lesions except for a growing lesion in the pelvic region and the skull with ingrowth in the brain. Clinically there was, however, temporary improvement of the patient's ability to move his right eyelid and right corner of the mouth, which had been paralyzed due to ingrowth of a skull metastasis into the brain and right facial nerve. A biopsy was performed of a growing FDG-positive lesion in the left pelvic bone that showed adenocarcinoma with the known L858R *EGFR* mutation, but the previously found T790M mutation was not present in this location (no biopsy of the skull metastasis available). The MTB advised to perform immunohistochemistry on Her2Neu (positive in agreement with HER2 amplification) and to determine *MET* amplification (negative by FISH). Based on these findings, it was decided to discontinue osimertinib because of the loss of the T790M mutation and to start a combination of paclitaxel 90 mg/m2 on days 1, 8, and 15, and trastuzumab 4 mg/kg on days 1 and 15, in cycles of 4 weeks, because trastuzumab is an HER2 antibody. Radiotherapy 1× 8 Gy was given on the pelvic lesion because of localized pain. 18-FDG-PET-scan after four cycles, paclitaxel and trastuzumab showed again a partial tumor response. No major side effects were observed although symptoms of the

Figure 1 | Computed tomography images showing the thoracic wall lesions of the adenocarcinoma of the lung and MRI images showing the metastasis in the skull with ingrowth in the brain at different time points in combination with treatment started at those time points. Abbreviation: PD, progressive disease. White circles indicate lesion, which showed progression.

paralysis of the right facial nerve did not improve further; it remained stable during the course of therapy. The patient underwent plastic surgery on his right eyelid, which improved the closure of his right eye.

Two months after the fourth and last cycle, the patient presented with a subcutaneous metastasis on his forehead. Afatinib 30 mg QD was started, because this treatment worked before, pending results of a new biopsy. The biopsy, however, yielded insufficient material for mutation analysis, and re-biopsy was scheduled. In the meantime, 18-FDG-PET-scan showed multiple FDG-positive bone lesions (partly new lesions), some close to the myelum, and the patient was admitted to the hospital for radiotherapy on cervical and thoracic vertebrae. Afatinib was discontinued. Biopsy of a new lesion in the thoracic wall showed an *EGFR*-L858R, T790M mutation, and *HER2* amplification. There were no other hotspot mutations in *EGFR*, *BRAF*, *KRAS*, *ERBB2 (HER2)*, *ALK*, *PIK3CA*, or *MET* detected. The case was again reviewed by MTB. It was decided to treat the patient with afatinib 30 mg QD as well as osimertinib 80 mg QD at alternating days, to address the T790M mutation as well as the *HER2* amplification resistance mechanism. Since the start of this latest treatment regimen, the subcutaneous skull metastasis disappeared, and the patient experienced less pain, regained his energy, and was able to walk outdoors again. The most recent 18-FDG-PET-CT-scan, 4 months after the start of this latest treatment regimen, showed again a partial response of the bone metastases again. Two months later (December 2017) progression of disease was observed, and the performance status deteriorated. Patient insisted to take a new biopsy from a new thoracic wall metastasis. Mutations analysis showed the known *EGFR* L858R and T790M mutations together with a new mutation in *HER2*: L755S. However, his condition got worse in short time, and he died in January 2018. An overview of the clinical findings, the mutational status at different time points and the given treatment regimens is provided in **Table 1**.

## BACKGROUND

## EGFR

The incidence of *EGFR* mutations in advanced stage adenocarcinoma of the lung in Caucasian patients is 10–15 and 40–60% in Asian patients (2). In the north of the Netherlands, the incidence is 9% (3). L858R mutation in exon 21 of the EGFR kinase domain is the main hotspot mutation in the *EGFR* gene and accounts for 35–45% of *EGFR* mutations (4, 5). L858R mutation increases the kinase activity of EGFR, leading to hyperactivation of downstream signaling pathways improving cell survival and proliferation (6). An EGFR TKI is the preferred first-line treatment in patients with activating *EGFR* mutation in non-small-cell lung carcinoma (NSCLC) (4). Gefitinib and erlotinib are firstgeneration TKI and registered as first-line treatment in patients with metastatic NSCLC with a tumor harboring an activating *EGFR* mutation within the European Union and according to the Dutch guideline for treatment of NSCLC (7). These small molecules bind competitively and reversibly to the adenosine triphosphate (ATP) binding site of the tyrosine kinase domain of


EGFR. This prevents the autophosphorylation of the TK, blocks the activation of the EGFR signal transduction, inhibits tumor cell proliferation, and induces cell cycle arrest and apoptosis (8). The majority of patients will progress after 9–12 months of treatment due to various mechanisms of intrinsic or acquired resistance to first-generation EGFR TKIs (9).

#### EGFR T790M

The most common mechanism of acquired TKI resistance is the acquisition of a single recurrent missense mutation within exon 20, the T790M mutation (10). This mutation leads to the substitution of threonine by methionine at position 790, which encodes part of the kinase domain of the receptor and results in increased affinity for ATP (11). The T790M mutation can be detected in about 60% of tissue biopsy samples taken after acquired resistance (12, 13). As residue threonine at position 790 (T790) is located at the entrance in the back of the ATP binding cleft, substitution of residue threonine at position 790 with a bulky methionine (resulting in T790M) may cause steric interference with binding of TKIs (14). Irreversible inhibitors overcome this resistance simply through covalent binding (15). Osimertinib is registered for the treatment of NSCLC with an *EGFR* T790M mutation. It is a selective third-generation TKI which targets the ATP binding site of EGFR *via* irreversible covalent bond formation. In contrast to many other TKI, osimertinib penetrates the blood–brain barrier (16, 17). Osimertinib improves overall survival and progression-free survival in T790M-positive NSCLC patients with and without brain metastases (18, 19). Acquired resistance to osimertinib may be caused by primary coexistence of tumor cell populations with and without T790M mutation due to EGFR C797S mutation. Tumor progression can be explained by growth of the T790M negative population, while the tumor cells expressing T790M mutation are effectively suppressed by osimertinib (20).

#### HER2

Overexpression of *HER2* induces cell transformation and tumorigenic growth and is clinically associated with resistance to erlotinib (21). *HER2* amplification is detected in a subset of EGFR TKI resistant lung tumors. *HER2* amplification and T790M mutation are thought to be mutually exclusive (22). However, in our patient *HER*2 amplification as well as T790M mutation appeared in the same biopsy of a new lesion in the thorax wall. Afatinib is an ATP-competitive aniline-quinazoline derivate which covalently binds to EGFR, HER2, and HER4 and irreversibly inhibits HER-family phosphorylation and signal transduction (23). As second generation TKI it is highly potent, irreversible dual EGFR/HER2 tyrosine kinase inhibitor, including the oncogenic *EGFR*-L858R mutation (23, 24). Afatinib is registered for advanced NSCLC with *EGFR* mutations. Clinical benefit of afatinib seems less in patients with *de novo* T790M mutations (25). Although afatinib is equally potent against wild-type EGFR and EGFR harboring the T790M mutation, in patients the dose is lower due to toxicity constraints (26).

Trastuzumab, a humanized monoclonal antibody against HER2, has been reported to be effective in *HER2*-positive NSCLC *in vitro* and in case reports (4, 27, 28).

## DISCUSSION

Here, we describe a patient with EGFR mutant advanced NSCLC with recurrent episodes of disease progression due to subsequent mutant clones. Yu et al. selected 155 patients with lung adenocarcinomas and acquired resistance to erlotinib or gefitinib. These patients underwent a re-biopsy. The most common finding was a T790M mutation. They also found transformation to small cell lung carcinoma, *MET* amplification and *HER2* amplification (10). Sequist et al. described a wide variety of gained and lost *EGFR* mutations in a patient population with acquired drug resistance. They recommended reassessing cancers by taking new biopsies of growing lesions in patients with progressive disease after an initial response to TKI treatment (29). Following this strategy, we observed HER2 amplification and T790M mutation at different time points under the selective pressure of different EGFR TKI treatment. Of note, the occurrence of both aberrations at the same time has not been described earlier.

After discussion in the MTB about the most suitable therapy, as mentioned in the background, treatment was adjusted accordingly. Case evaluation by a multidisciplinary MTB is important to benefit from individualized genetic data and maximize clinical impact (30–32). MTB interprets results of routine molecular NGS-testing with those of other techniques, for example, immunohistochemistry, FISH, DNA methylation testing, and multiplex ligation-dependent probe amplification. NGS testing is not only performed on biopsies but currently also from tumor DNA in peripheral blood. The spectrum of molecular markers is constantly growing. Patients who progress after an EGFR TKI should undergo a re-biopsy to perform molecular analysis specifically looking for acquired mechanisms of resistance, such as *EGFR* T790M mutation. This approach can influence the next therapeutic step or reveal alternative EGFR TKI resistance mechanisms such as transformation to small cell lung cancer or bypass tracks that could potentially be addressed in clinical trials (11). Our patient responded well to subsequent treatment based on aberrations found in NGS, IHC, and FISH after discussion in the MTB. We observed that a change in treatment gave a short-lasting clinical improvement of several months and tumor response of fast growing new metastases. Osimertinib alternating with afatinib for T790M in EGFR and HER2 amplification was very effective in decreasing tumor sites. However, we expected that our patient would have immense toxicity of skin rash and diarrhea, but toxicity was not more than CTC grade 1. Longlasting tiredness grade 1 was the most prominent side effect.

## CONCLUDING REMARKS

This case report shows the importance of re-biopsy of growing lesions in lung cancer patients with metastatic progressive disease under targeted therapies. Mutation status can vary under selection pressure of these drugs, and knowledge of these changes makes it possible to adapt treatments. This patient also exemplifies the importance of having a multidisciplinary expert team (MTB) to give rational treatment advice in cancer patients with uncommon mutations or combinations of mutations causing complex resistance mechanisms.

## ETHICS STATEMENT

This case report was written and offered for publication with written informed consent from the patient. The patient gave written informed consent in accordance with the Declaration of Helsinki.

## AUTHOR CONTRIBUTIONS

AM: organized the relevant information about the patient; wrote all paragraphs of the article, did the literature research, and

## REFERENCES


applied changes brought in by the other authors. AE, HG, and ES: supervised and corrected the manuscript. NH: supervised and gave advice. AW: main supervisor; supervised and corrected the manuscript.

## ACKNOWLEDGMENTS

Written informed consent to write and publish this case report was obtained from the patient. We thank the patient for his permission to publish this case report.


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The handling Editor declared a past co-authorship with one of the authors AW.

*Copyright © 2018 Meedendorp, ter Elst, 't Hart, Groen, Schuuring and van der Wekken. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Management of Leptomeningeal Metastases in Non-oncogene Addicted Non-small Cell Lung Cancer

Ana Turkaj <sup>1</sup> , Anna M. Morelli <sup>1</sup> \*, Tiziana Vavalà<sup>2</sup> and Silvia Novello<sup>1</sup>

<sup>1</sup> Department of Oncology, University of Torino, Ospedale San Luigi Gonzaga, Orbassano, Italy, <sup>2</sup> SC of Oncology, ASL CN1, Ospedale Civile di Saluzzo, Saluzzo, Italy

Brain metastases in non-small cell lung cancer (NSCLC) patients are more often detected due to imaging modalities improvements but also emerge because of improved treatments of the primary tumor which lead to a longer survival. In this context, development of leptomeningeal metastases (LM) is a devastating complication and its prognosis remains poor despite advances in systemic and local approaches. Histology characterization of NSCLC and molecular expression influence LM management. For those with "oncogene addiction," new generation epidermal growth factor receptor (EGFR) and anaplastic lymphoma kinase (ALK) tyrosine kinase inhibitors (TKIs) were developed to strongly penetrate the blood-brain barrier (BBB) with the aim to prevent central nervous system cancer dissemination, eventually impacting on LM appearance and its subsequent management. Systemic chemotherapy, often combined with intrathecal chemotherapy (when possible), was one of common indications for lung cancer patients affected by LM, without driver mutations and a good performance status but currently, with the advent of innovative systemic approaches treatment solutions in this subgroup of patients are rapidly evolving. Whole brain radiation therapy (WBRT) is the conventional treatment for patients with brain metastases. Furthermore, modern radiation techniques, as stereotactic radiotherapy (SRT), improve outcomes in those cases with a limited number of lesions. However, LM represent a minority of CNS metastases and few literature data are available to drive the radiotherapy approach. Considering all relevant progress made in this setting, after a literature review, the aim of this paper is to discuss about recent developments and therapeutic options in LM management of non-oncogene addicted NSCLC.

Keywords: non-small cell lung cancer (NSCLC), brain metastases, leptomeningeal metastases, chemotherapy, intra-thecal chemotherapy, immunotherapy, radiotherapy

## INTRODUCTION

Non-small cell lung cancer (NSCLC) is characterized by a high incidence of central nervous system (CNS) dissemination, with approximately 40% of patients developing brain metastases (BM) in the course of their disease (1, 2) and leptomeningeal metastases (LM) appearance in a smaller percentage (3–5%) (3). Particularly, LM incidence among NSCLC patients is 3.8% more frequent in adenocarcinoma subtype, with about a third of those patients having concomitant BM (4). LM usually manifest as a late complication, which have been reported to emerge as late as up to 112

#### Edited by:

Deepa S. Subramaniam, MedStar Georgetown University Hospital, United States

#### Reviewed by:

K. Shilo, The Ohio State University, United States Timothy F. Burns, University of Pittsburgh Cancer Institute, United States

> \*Correspondence: Anna M. Morelli a.mariamorelli@gmail.com

#### Specialty section:

This article was submitted to Thoracic Oncology, a section of the journal Frontiers in Oncology

Received: 29 March 2018 Accepted: 04 July 2018 Published: 27 July 2018

#### Citation:

Turkaj A, Morelli AM, Vavalà T and Novello S (2018) Management of Leptomeningeal Metastases in Non-oncogene Addicted Non-small Cell Lung Cancer. Front. Oncol. 8:278. doi: 10.3389/fonc.2018.00278

**84**

months after initial diagnosis (5). LM cases are increasing in incidence as a result of improved survival in subgroups of patients with targetable mutations treated with molecular therapies, but also because of modern neuro-imaging tools, able to clearly identify even small foci of meningeal dissemination (6–8). Median survival of NSCLC patients affected by LM is particularly poor even with signs of improvement, from a historical median survival of 1–3 to 3–11 months with novel therapies and integration of local and systemic treatments (9). Specific treatments of LM depends on histology characterization of NSCLC, molecular expression, time of appearance and patient's performance status. Molecularly targeted therapies and immunotherapy showed antitumor activity for brain metastases although effectiveness in cytologically confirmed and symptomatic LM is still limited or unknown (3). Systemic and intrathecal (IT) chemotherapy with site-specific radiotherapy are usually applied, particularly in non-oncogene addicted NSCLC, while up to one-third of the patients are treated with best supportive care alone. Despite the lack of the standard treatments, active treatments have been associated independently with longer overall survival (OS) (10). Recent advances in the understanding of LM biology in NSCLC patients along with the development of highly active targeted drugs for tumors with specific genetic alterations, helped to redefine the prognosis in this subgroup of patients and the same evolution is largely awaited in those NSCLC patients without oncogene addiction (5, 11–14). Based on literature review, this paper aims to discuss about recent developments and therapeutic options of LM in NSCLC patients without driver mutations.

## BIOLOGY OF BLOOD BRAIN BARRIER AND DRUG DELIVERY

The blood–brain barrier (BBB) is constituted by a continuous stratum of endothelial cells connected by tight junctions surrounded by pericytes and perivascular end feet of astrocytes, thus being a highly selective barrier which separates systemic circulation from cerebrospinal fluid (CSF) (15).

BBB maintains CNS homeostasis by enabling the transport of selected substances only, through a combined result of influx and efflux mechanisms. Therapeutic efficacy in this area is determined by whether drug concentrations can be achieved in the CSF and these differ as a result of multiple conditions. The ability of a drug to cross the BBB is substantially improved by particular physic-chemical properties, including low potential for active efflux, few rotatable bonds, small polar surface area, few hydrogen bond donors (9). For instance, anticancer therapeutics (tyrosine kinase inhibitors or chemotherapeutic agents), are substrates of efflux transport proteins, such as Pglycoprotein, which is responsible for the transport of most drugs outside the intracranial region. Most chemotherapy agents have low CSF concentrations, with relevant liquor permeability reported only for temozolomide, methotrexate and topotecan (16–19), however predicted CNS penetration does not necessary correlate with known response rates to chemotherapeutic agents. Moreover, free diffusion of molecules across the BBB requires both lipophilicity and low molecular weight (less than 0.5 kDa): chemotherapy drugs are usually larger than 150 kDa, hydrophilic and frequently protein-bound molecules, therefore unable to penetrate an intact BBB (20– 23). In this context, growing scientific evidence highlights that CNS metastases cause BBB interruption; this process, probably due to tumor neo-angiogenesis, lead to generate new vessels lacking of structural and physiological features of normal BBB, thus favoring the passage of drugs into the brain (24, 25). The same hypothesis emerges after whole brain radiation (WBRT) approach, thus providing a biologic rationale for using concomitant or sequential systemic and local treatments in these cases (26, 27).

## DIAGNOSIS OF LEPTOMENINGEAL METASTASES

LM involve penetration of inner layers of meninges and subarachnoid space in which CSF circulates. Its diagnosis is specifically based on three different assessments: clinical signs and symptoms, CSF cytological examinations and neuroradiological imaging.

Early clinical presentation can be subtle and may include headaches and back pain, cranial nerve deficits, cauda equine symptoms or signs, visual disturbances, diplopia, hearing loss and neurocognitive syndromes. In later stages, symptoms related to elevated intracranial pressure could occur (28, 29). Cytological identification of malignant cells in CSF is the gold standard for diagnosis of LM. The sensitivity of the initial lumbar puncture was reported to be as low as 50%, with a potential increase to 75–85% with a second CSF analysis (30). A meningeal biopsy is rarely needed to confirm a clinical suspect. A recent study performed by Jiang et al. demonstrated that the use of nextgeneration sequencing (NGS) performed on cerebrospinal fluid circulating tumor cells (CSF-CTC) may be a more sensitive and an effective way to diagnose LM, serving also as a liquid biopsy for gene profiles in NSCLC patients with LM (31). Besides clinical and cytological diagnosis, brain and spine imaging are able to identify involved sites, even in cytology-negative cases (32).

Magnetic resonance imaging (MRI), ideally with a 3 Tesla scanner is the most useful imaging modality for the detection of LM. Both T1 with and without contrast enhancement and high resolution T2 fluid attenuated inversion recovery sequences postcontrast are important in establishing a radiological diagnosis of LM.

Particularly the disruption of BBB in presence of CNS metastases is often evidenced by peritumoral edema and accumulation of contrast during the imaging scans and, as more recently observed, penetration of CNS metastases is identified by nuclear medicine tracers, such as 18-Sodium Fluoride (33).

The EANO-ESMO clinical practice guidelines propose to classify LM by using two major criteria, being "type I" those LM when the diagnosis has been verified citologically or histologically and "type II" in the absence of verification. While on the basis of the neuroimaging findings: linear leptomeningeal disease (type A), nodular leptomeningeal disease (type B), both (type C) or neither nor, e.g., no neuroimaging evidence of LM except possibly hydrocephalus (type D) (34).

## TREATMENT OPTIONS OF LEPTOMENINGEAL METASTASES

Treatment of LM is preferentially multidisciplinary and mostly indicated when diagnosis is unequivocal or symptoms are strongly suggestive, in case of negative cytology.

The goals of treatment in patients with LM are to improve or stabilize the neurologic status of the patient, maintain or regain quality of life and optimally to prolong survival together with marginal toxicity. Limited data are available to establish treatment recommendations in the management of LM: no randomized trials proved a survival benefit of a specific treatment modality and, accordingly, the optimal strategy is still poorly defined, particularly in non-oncogene addicted NSCLC. In this last setting of patients, palliative radiotherapy to symptomatic sites of disease, cytotoxic chemotherapies, intrathecal therapy (or a combination of these modalities) are traditionally considered, with the new innovative immunotherapy chance (35–39).

#### Radiation Therapy

In the era of improved radiation modalities, local treatment of BM is rapidly improving: a growing amount of data support an integrate use of WBRT or stereotactic radiosurgery (SRS) with systemic treatment in a variety of clinical scenarios, together with alternative radiation approaches, such as intensity modulated radiation therapy (IMRT) or even proton beam therapy, when applicable (40–42).

LM represent a minority of CNS metastases (11–20%) (43, 44) and less data are available to inform decisions about therapy. RT is mainly administered for symptoms alleviations, CSF flow correction or for debulking to facilitate chemotherapy.

WBRT is typically used in cases of concurrent brain metastases or major meningeal cerebral involvement (45). For BM dose and fractionation scheme is at the discretion of the treating radiation oncologist, though most commonly used dose and fractionation schemes are 20 Gy in 5 fractions of 4 Gy (standard schedule in Europe) and 30 Gy in 10 fractions of 3 Gy (46). For LM focal radiation therapy is recommended on symptomatic, bulky or obstructive sites and the dose depends on performance status (20–40 Gy in 5– 20 fractions), volume to treat and available techniques (47). Different studies reported a survival difference in favor of patients with better performance status after various treatments, including WBRT (48, 49). Gani et al. evidenced that WBRT alone in patients with LM from breast and lung cancer is an effective palliative treatment for patients unfit for chemotherapy and poor performance status (49). As tumor dissemination affects the whole CSF compartment, according to some studies, the complete craniospinal axis should be encountered as target volume. Favorable results have been reported in small series. To determine the effects of craniospinal irradiation (CSI) a retrospective study of 16 patients with LM (mostly from breast and lung cancer) was conducted by Hermann et al. In this study, ten patients were additionally treated with intrathecal methotrexate. The authors conclude that craniospinal radiotherapy is feasible and effective for palliative treatment of LM (36).

However, craniospinal irradiation (CSI) is generally not recommended, as it is assumed to cause substantial myelotoxicity. In fact, in his review, Chamberlain stated that "whole neuroaxis radiation of the craniospinal axis is rarely indicated in the treatment of LM in solid tumors" (50).

Conformal radiotherapy may help to limit bone marrow and neurotoxicity making focal radiotherapy better tolerated. Proton therapy is only available in a few centers, but this approach promises further reduction of toxicity and effectiveness from CSI (51, 52).

Focal RT administration in fractionated regimens, such as involved-field or stereotactic RT or administered in single fractions (radiosurgery), can be used to treat nodular disease and symptomatic cerebral or spinal sites (34). By contrast, stereotactic radiosurgery (SRS), which is a radiation therapy technique in which multiple focused radiation beams intersect over a target, results in delivery of a highly conformal, high-dose of radiation to the target and minimal radiation to the surrounding normal tissues. In patients with BM a Gamma Knife Radiosurgery (GKRS) is typically used, depending upon the volume and location prescription doses typically range from 15 to 24 Gy for single fraction session (53). GKRS allow to achieve high rates of local control, and is able to delay the need for WBRT thus avoid potential neurocognitive toxicities, although a phase 2 RTOG study suggested that concomitant administration of memantine together with WBRT may reduce and delay subsequent cognitive consequences (54). Few studies have reported on the role of SRS in the setting of LM (55, 56). In the small and heterogeneous study by Wolf et al. the prescription tumor margin dose was a median of 16 Gy (11–20 Gy) to the 50–80% isodose volumes. The authors suggested SRS capable to provide high rates of local control for restricted LM with a median survival of 10 months and with 60% of the population alive at 6 months. SRS for focal LM is preferable in those patients who are eligible for systemic therapy, including immuno-therapies and targeted therapies, which can potentially further prolong overall survival.

Involved field radiotherapy (IFRT) is considered to be the standard of care for palliative treatment of LM. Relevant to this, the US National Comprehensive Cancer Network (NCCN) 2017 guidelines for management of LM recommend Intrathecal Chemotherapy (IT) in combination with IFRT in patients with good prognosis disease (as defined by high performance status, non-fixed neurologic deficits, minimal systemic disease, and reasonable options for systemic disease treatment). Patients not meeting criteria for good prognosis are recommended to undergo IFRT to symptomatic sites or best supportive care (45). Up to now, no randomized clinical trial to assess the efficacy and tolerance of RT in LM have been conducted, however use of radiation therapy in NSCLC patients with LM, particularly in those not presenting driver mutations, needs to be better defined in clinical trials. Concomitant strategies with ITC are currently not considered as standard care due to the toxicity profile. Phase II clinical trial of combination therapy with involved field RT combined with concurrent intrathecal-MTX or intrathecal-ARA-C is currently underway (57).

**Table 1** summarize the published studies from 2000: different types of radiation modality on different histologies in patients with LM are included (36, 49, 56, 58–61).

In the last few years, understanding of immune system's role in the response to ionizing radiation is progressively raising, novel opportunities to study how to combine immunotherapy with radiation-induced cell killing are revolutionizing cancer treatment. Accumulating preclinical and clinical data showed that combination of radiation techniques with immunotherapy stimulates immune response, improves locoregional and distant control finally resulting in better OS. Radiation appears to stimulate the immune system through multiple mechanisms, including the increase of the tumor-associated antigens (TAAs) availability, improving antigen presentation and subsequent stimulation of effector T cells, and enhancing infiltration of dentritic cells and T cells into the tumor microenvironment. The limited evidence for immunotherapy to date in the treatment of BM and LM stems from the deliberate exclusion of patients with active brain metastases from many large randomized trial assessing drug efficacy (61–63). However, comparable efficacy of immunotherapy agents in the brain and at extracerebral sites with radiation therapy has been recently reported by an ongoing study from Goldberg et al. In this initial analyses 36 cases were considered, 18 with melanoma and 18 with NSCLC. Patients were treated with pembrolizumab 10 mg/kg every 2 weeks until progression, no target lesions were previously resected, 20 patients received some form of local CNS therapy prior to enrollment (9 and 7 lesions had been treated respectively with WBRT and SRS, 75 lesions were untreated). The primary endpoint was BM response rate and the initial results presented demonstrated a systemic benefit from immunotherapy in patients with metastatic melanoma and NSCLC (61). Combining immunotherapeutic agents with stereotactic radiosurgery appears to enhance both local and distant control, and result in better survival (42).

Finally, other types of RT in patients without obstruction to CSF flow, as radioimmunotherapy (RIT) consisting of intra CSF administration of radioisotopes like iodine-131 (131I) and yttrium-90 (90Y) with radiolabeled antibodies as HMFG1, 3F8 and 8H9 have been utilized, but further improvement in the pharmacokinetic modeling of CNS RIT modality should refine this emerging therapy to fit the clinical context (63, 64).

#### Systemic Therapy

In the last decade, the advent of EGFR- tyrosine kinase inhibitors (TKIs) has improved the prognosis of NSCLC patients harboring EGFR mutations on both CNS metastases and extracranial disease, by contrast non-oncogene addicted NSCLC prognosis remains extremely poor.

The role of chemotherapy for patients with CNS metastases from NSCLC has been neglected for years, because of prevailing belief that chemotherapeutic drugs cannot cross at all the BBB. A Platinum based-combination (preceded or not by a local radiation treatment) is the mainstay of treatment in

TABLE 1 | Selected trials published of radiation modality treatment with or without chemotherapy/immunotherapy association in patients with LM/BM [BM, brain metastasis; LM, leptomeningeal metastasis, Involved-field radiotherapy (IF-RT), Intrathecal chemotherapy (IC), Craniospinal Irradiation (CSI) Methotrexate (MTX), Non available (NA), Not reached (NR), Patients (pt)].



TABLE 2 | Platinum-based chemotherapy trials for CNS from NSCLC (BM, brain metastasis; LM, leptomeningeal metastasis).

advanced non-small cell lung cancer (aNSCLC) with BM and or LM NSCLC at diagnosis without oncogenic driver mutations or programmed death-ligand 1 (PD-L1) tumor proportion score (TPS) values ≥ 50% (65–73). **Table 2** summarize platinum-based chemotherapy for NSCLC patients with CNS (65–73).

Pemetrexed is a compound currently approved both in combination with platinum in first-line setting and as a single agent in maintenance or second line setting for the treatment of non-squamous cell carcinoma (74–77). Although a penetration of CNS of less 5%, pemetrexed demonstrated a consistent activity against BM from NSCLC with an intracranial RR of 30.8−41.9% and an overall clinical benefit of 63% without specific data for LM (72, 73, 78, 79) Bevacizumab is an anti-VEGF monoclonal antibody approved in patients affected by locally advanced or metastatic non-squamous cell carcinoma (80). Recent reports showed that bevacizumab improve intra-tumor penetration of other chemotherapeutic agents, such as carboplatin or paclitaxel by normalizing angiogenesis at the tumor site with a low incidence of CNS hemorrhage (81, 82). Additionally, bevacizumab appeared to alter neuroimaging characteristics of LM, confounded diagnosis and possibly also influenced the pattern of tumor spread of LM (83). Bevacizumab is also known as a "steroid-sparing" drug, that allows reduction in steroid dosage and achieves reductions in mass effect and peritumoral edema getting better control of neurological symptoms (84). In those patients with squamous histology and low PD-L1 expression, gemcitabine and taxanes are largely prescribed, also in patient with BM, but no specific data about LM are available (70, 85, 86).

For both squamous and non-squamous cell carcinoma without oncogene-addiction and TPS ≥ 50%, pembrolizumab represents the standard of care in the first-line setting. Data on the efficacy of immunotherapy for BM or LM are currently limited, because of the exclusion of these patients from clinical trials (38, 87–89). In the trial by Goldberg et al., already mentioned above, 36 patients with BM were treated with pembrolizumab. Among patients with NSCLC, BM response rate was 33% and treatment-related serious adverse events were rare. Several aspects of study population need to be considered looking at the results: patients eligible for this trial were those with BM < 20 mm, asymptomatic and not requiring corticosteroids to control neurologic symptoms, without autoimmune disease and with no prior treatment with agents targeting PD-1/PD-L1 (61). Only preliminary data are available on efficacy of anti-programmed cell death 1 (PD-1) agents (nivolumab, pembrolizumab) or anti-PD ligand 1 (atezolizumab) in NSCLC patients with BM. A review on efficacy and safety of nivolumab conducted by Dudnik in five patients with aNSCLC with new/progressing asymptomatic CNS metastasis (including patient with LM) suggested that immune-check inhibitors might have an intracranial activity. Two intracranial responses were observed, while stabilization of LM was achieved in 10 weeks (90). In the Italian Nivolumab Squamous NSCLC Expanded Access Program, 38 patients with treated and asymptomatic BM were included, none of them with LM. In this subgroup of patients immunotherapic agent obtained a disease control rate equal to 47.3% (91). On the other hand, Otsubo et al., reported a case report of the development of LM during a pronounced response of the primary tumor to pembrolizumab therapy in a NSCLC patient (92). Relevant to this, in the era of molecular oncology it will be important to consider genomic differences in systemic malignancies that can implicate a distinct immune response.

Looking at other type of compounds, a phase II trial with abemaciclib (orally bioavailable inhibitor of cyclin-dependent kinases 4 and 6) (93) is ongoing in patients with LM from NSCLC and solid tumors (94).

#### Intra-Thecal Chemotherapy

Intrathecal administration is the most common method to deliver chemotherapeutic agents in non-nodular and nonbulky LM in solid tumors, although efficacy compared to systemic administration and choice of regimen are poorly understood due to limited randomized controlled trials (95). Systemic chemotherapy, which may be combined with intrathecal chemotherapy, remains standard treatment for lung cancer patients with LM and a good-risk profile (45).

Several retrospective studies demonstrated a survival benefit from IT-therapy. Because of the paucity of available patients, LM studies often accrue multiple primary histologies. Most reports of intrathecal LM treatments include patients who simultaneously receive systemic agents (29, 95–98). Although the compounds routinely used for intra-CSF treatment do not have a key role as single agents for systemic treatment of most common cancers causing LM, three agents are commonly prescribed for the intrathecal treatment of LM: methotrexate (98), cytarabine (including liposomal cytarabine) (99, 100) or thioTEPA (8). Several schedules have been proposed, without agreement on optimal dose, frequency of administration or optimal duration of treatment. No intra-CSF agent has shown a significant survival advantage over another (8, 99, 100). Up to now most of the patients are treated until progression or for 1 year, if tolerated. In the absence of evidence from appropriate clinical trials, clinical symptoms, MRI and CSF findings, as well as tolerance of treatment, guide individual decisions on the duration of treatment (34). Pemetrexed is a newer generation multitargeted anti-folate agent and, compared with methotrexate, has better tolerability, exhibits a more favorable side effect profile, and possesses fewer-drug interactions (78). A phase I trial (NCT03101579) is ongoing to define safety profile and clinical response rate associated with this specific intrathecal therapy (101).

#### REFERENCES


## CONCLUSION

LM is undoubtfully a serious complication in NSCLC patients. Prognosis remains poor, even with the use of personalized treatments, principally due to low penetration into the CSF of currently used agents. To our knowledge no randomized trial has demonstrated a clear survival benefit of any single modality in the treatment of LM. The optimal treatment strategy involves a multi- disciplinary approach. The increasing prevalence of LM warrants further investigation into therapies and prognostic variables to serve as a guide for treatment recommendations and patients counseling. An improved understanding of the biologic mechanisms underlying tumor CNS metastases and novel available diagnostic tools will allow for patient-tailored treatment strategies. Future welldesigned randomized controlled studies are needed to evaluate the efficacy of chemotherapeutics, immunotherapies and radiation treatment in this specific subgroup of patients and, more specifically, in those without an "oncogene addiction."

## AUTHOR CONTRIBUTIONS

AT and AM wrote the manuscript with support from TV. SN supervised the manuscript.


multicentre randomised controlled trial. Lancet (2017) 389:255–65. doi: 10.1016/S0140-6736(16)32517-X


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Turkaj, Morelli, Vavalà and Novello. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.