Non-small cell lung cancer (NSCLC) is the most common type of lung cancer, accounting for nearly 80% of all cases, and often presents in the locally advanced or metastatic settings (1). Currently, a combination of chemotherapy and/or immunotherapy (IO) is considered standard first-line treatment for metastatic NSCLC (mNSCLC) without driver mutations, often tailored based on a patient’s programmed death-ligand 1 (PD-L1) status (2). Before the advent of IO as first-line therapy, REVEL, a randomized, multicenter, phase III clinical trial, demonstrated improved progression-free survival (PFS), overall survival (OS), and quality of life (QoL) with ramucirumab and docetaxel (ram+doc) chemotherapy and antiangiogenic combination in patients whose disease progressed after platinum-based doublet first-line chemotherapy compared with docetaxel alone (3, 4). This combination was proposed utilizing ramucirumab as a complete human monoclonal IgG1 antibody with direct vascular endothelial growth factor receptor 2 (VEGFR2) antagonism given already known improved outcomes with docetaxel in platinum-resistant disease (5). Indeed, this biological rationale of overcoming the demonstrated immunosuppressive effect of VEGF has been proven in other lines of therapy and in combination with immunotherapy, such as the IMPOWER150 study which demonstrated improved outcomes for nonsquamous mNSCLC when combining atezolizumab, bevacizumab, and platinum-doublet chemotherapy in the first-line setting (6). There have been studies that have shown promising results in other forms of platinum-resistant tumor histologies, namely, urothelial and gastric cancers (7, 8). There exist data mostly limited to retrospective cohort analyses in East Asia and Europe discussing responses to ram+doc treatment in patients pretreated with IO-based therapy; however, any synergistic benefit has not been proven for patients with mNSCLC (9). Therefore, our study aims to clarify the efficacy and outcomes of this combined therapeutic approach in patients with paclitaxel-resistant mNSCLC.

We performed a retrospective cohort study among all patients with mNSCLC treated at the University of Miami Sylvester Comprehensive Cancer Center. We retrospectively identified all patients with mNSCLC whose disease demonstrated progression after IO-based therapy and then received ram+doc as a subsequent line of therapy between January 1^st, 2010 and March 1^st, 2020. A total of thirty-nine patients were identified whom met these inclusion criteria. We subsequently excluded 4 patients with EGFR or ALK driver mutations from our analysis. As such, thirty-five (n = 35) patients were included in our final analysis. We assessed patients’ PFS and OS after ram+doc treatment utilizing the Kaplan-Meier method as primary outcomes. We compared our center’s retrospective data with those from REVEL data utilizing a simulation study via Wilcoxon test. Since REVEL data was not available for reproduction, we used median PFS and median OS and corresponding 95% CIs to estimate the distribution of median survival time for our dataset compared to that of REVEL using an approximate Bayesian computation (ABC) approach. We also collected information on adverse events (AEs) during ram+doc treatment as a secondary outcome. This study was approved by the University of Miami Institutional Review Board eProst #20170427.

In our retrospective analysis, we found a statistically longer mPFS and mOS for patients with mNSCLC treated with ram+doc after progression on IO at our center compared to the original cohort reported in the REVEL study, which first analyzed this combination chemotherapy and antiangiogenic-therapy regimen in 2014 ( Figures 1A, B ). Our results did not show a difference in these primary outcomes based on tumor histology.

These findings are striking in that they demonstrate a possible synergy between IO pre-treatment and exposure to second-line (or beyond) antiangiogenic-therapy, namely, ramucirumab, in combination with docetaxel. These findings have been echoed in other retrospective cohorts in East Asia. For example, a retrospective review of 99 patients in multiple Japanese centers found a statistically significant mPFS response of 5.9 months in those pre-treated with IO compared to those who did not have IO exposure (2.6 months) (11). These findings were further bolstered in a post-hoc sub-group analysis of the original REVEL study in which East Asian patients demonstrated a mPFS of 4.88 months and mOS of 15.4 months (12).

Furthermore, our cohort did not demonstrate unexpected and/or significant adverse events, thereby supporting the safety of ram+doc as combination therapy. This observation also mirrors outcomes in other retrospective cohorts, such as a review of 77 patients in Germany that did not demonstrate unexpected toxicities, i.e., no more than 9.09% febrile neutropenia, fatigue, mucositis, stomatitis, or ileus (13). Similarly, another retrospective analysis in Japan estimated up to 16.7% total (and 11.1% Grade 3 or more) pneumonitis, which is more frequent than the 2.9% of pneumonitis as well as COPD exacerbation observed in our cohort (14).

Our findings also echo activity among a similar combination of docetaxel with nintedanib, a tyrosine kinase inhibitor that has activity against multiple kinases including VEGF. This combination was originally approved based on the LUME-Lung1 study which demonstrated improved PFS and OS compared to docetaxel alone, particularly in adenocarcinoma histology, compared to docetaxel alone in the second-line setting (15). Real-world outcomes mirror those reported at our institution after treatment with chemo- and immunotherapy. Specifically, the VARGADO cohort demonstrated a mPFS of 6.4 months, with a 1-year OS rate of 52% in the third-line setting (16). Furthermore, another German cohort reported a mOS of 8.4 months in adenocarcinoma histology specifically (17).

Two strengths of our study are its location and demographics. Specifically, we report herein the first such retrospective analysis consisting of North American patients, with a majority of patients whom identified ethnically as Hispanic/LatinX. Additionally, our cohort’s outcomes rank among the longest PFS and OS benefits recorded with post-IO ram+doc exposure to date. This result will require additional study with similar ethnic and geographic cohorts.

Our analysis has several limitations. First, this is a single-center, retrospective analysis, and as such these observations should be confirmed in a prospective fashion. The Phase II Lung-Map S1800A study evaluated ramucirumab with pembrolizumab combination therapy compared to standard of care chemotherapy, of which two-thirds of the control arm received ram+doc, and was found to demonstrate an OS benefit (17). Post-hoc analyses will be required to understand the true PFS and OS estimates seen in this sub-group, however. Additionally, the TREAT-LUNG observational study reported preliminary data for second- and third-line docetaxel vs. ram+doc in patients previously treated with both platinum-based chemotherapy and IO with a subset of patients demonstrating long-term responses (i.e., plateaus in Kaplan-Meier plots) (18). These findings merit closer attention once presented formally in the literature. The greatest limitation of our study is its small size. For example, a larger Japanese cohort of 1,439 patients utilized a propensity score analysis and did not find a PFS or OS advantage with this treatment strategy (19).

Overall, our institution’s experience with this combination chemo- and antiangiogenic-therapy strategy adds to the data related to ram+doc after IO exposure in mNSCLC. Interpretation should be limited given its retrospective timeframe and single-center patient population.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by University of Miami IRB Eprost 20170427. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

First authorship: SK. Equal contribution: KG, KK, DK, GL, ER. Senior/last authorship: RD. All authors contributed to the article and approved the submitted version.