Delayed Response After Confirmed Progression (DR) and Other Unique Immunotherapy-Related Treatment Concepts in Cutaneous Squamous Cell Carcinoma

Non-melanoma skin cancers are one of the most common cancers diagnosed worldwide, with the highest incidence in Australia and New Zealand. Systemic treatment of locally advanced and metastatic cutaneous squamous cell carcinomas has been revolutionized by immune checkpoint inhibition with PD-1 blockade. We highlight treatment issues distinct to the management of the disease including expansion of the traditional concept of pseudoprogression and describe delayed responses after immune-specific response criteria confirmed progressive disease with and without clinical deterioration. We term this phenomenon “delayed response after confirmed progression (DR)”. We also discuss the common development of second primary tumors, heterogeneous disease responses, and expanding clinical boundaries for immunotherapy use.


INTRODUCTION
Non-melanoma skin cancers (NMSC) predominantly basal cell carcinomas (BCC) and cutaneous squamous cell carcinomas (CSCC) are the most frequently diagnosed cancer in North America and Australia/New Zealand (1). Although most are resectable, the morbidity related to disease is significant and accounts for the most common cancer-related cause for hospitalization in Australia exacerbated by the multiplicity of skin cancer excisions (2)(3)(4). Approximately 5% of CSCC recur or metastasize leading to death or management associated with significant morbidity due to disease occurrence on sun-exposed areas such as the face, head and neck (5,6).
We report our immunotherapy management experiences unique to CSCC that challenge and expand current clinical concepts in practice.

EXPANDING ON THE CONCEPT OF PSEUDOPROGRESSION -"DELAYED RESPONSES AFTER CONFIRMED PROGRESSION (DR)" AND RESPONSE AFTER CLINICAL DETERIORATION
The Response Evaluation Criteria in Solid Tumors (RECIST) criteria is a validated measure for the standardized evaluation of cancer therapies, determined by the assessment of the change of tumor burden with treatment (12). There are notable limitations of the RECIST guidelines in patients treated with immunotherapy, given that "pseudoprogression" can occur with an increase in tumor size due to inflammatory cell infiltrates followed by tumor reduction (13), and improved OS can occur without RECIST defined reduction in tumor measurements (14)(15)(16). Thus, RECIST criteria have been modified to include immune-related response criteria (irRC), immune-related response evaluation criteria in solid tumors (irRECIST), modified RECIST1.1 for immunotherapy (iRECIST), and immune-modified RECIST (imRECIST) (17)(18)(19)(20).
Pseudoprogression occurs in under 10% of all cancers treated with immunotherapy, with an incidence in head and neck cancer of approximately 1% (21)(22)(23)(24). To address pseudoprogression, the irRC, irRECIST and iRECIST require the use of confirmatory imaging at least 4 weeks after initial progression is documented (we will refer to this collectively as iCPD, immune confirmed disease progression as per iRECIST), a minimum size increase of >5x5mm 2 or 10mm, an increase in the sum of tumor measurements from the nadir of 20-25%, and incorporation of new lesions in the sum of tumor dimension measurements before confirmation of progressive disease (PD) (17)(18)(19). The imRECIST criteria is similar, but permits the best response assessment to occur after observation of progressive disease, which avoids underestimation of survival rates (20). All criteria recommend that treatment beyond progression should only occur if a patient's performance status and disease-related symptomatology are stable.
We describe two cases to redefine our understanding of pseudoprogression, with delayed disease response observed after iCPD and after observation of clinical deterioration. We introduce new terminology to capture the phenomenon as "delayed response after confirmed progression (DR)". We also discuss a case of tumor response observed in a patient with clinical deterioration, with treatment beyond progression.

Case One
This case illustrates the observation of DR. Patient One was diagnosed with a T2N0M0 occipital scalp/vertex CSCC excised with clear margins in September 2017. In July 2018, a 35mm intransit recurrence was resected with an involved margin. Reexcision and ipsilateral neck dissection demonstrated a 2.5mm residual CSCC 1.2mm from the deep margin, and 1/41 nodes involved with extranodal extension to the surgical margin. Adjuvant radiotherapy of 66 Gy/33# was completed in October 2018. Within three weeks, biopsy of multiple new in-field intransit CSCC and an out-of-field intramuscular lesion confirmed recurrent disease. Following multidisciplinary meeting (MDM) discussion, he was referred for consideration of immunotherapy for his recurrent CSCC.
The patient was enrolled on the NCT02760498 to receive cemiplimab 400mg Q4W, with RECIST 1.1 measurements determined by radiological assessments as per protocol. The progression of lesions documented by photography, and radiological assessments are summarized in Figures 1A, B. Four weeks after initial dose, Patient One reported worsening disease-related pain. The largest CSCC lesion had increased by 10mm with new ulceration (black arrow), other baseline lesions had also increased in size with ulceration, while at least three new lesions greater than 10mm had developed with multiple other smaller lesions visible. This corresponded to unconfirmed progressive disease (iUPD). Four weeks later prior to his third dose, ongoing progression was noted with further ulceration and coalescence of lesions (white checked arrow), and increasing size of new nodules by more than 10mm (black spotted arrow) confirming iCPD. At that time point, RECIST 1.1 assessment by imaging confirming PD ( Figure 1B). Approval for treatment beyond progression was granted. By the next visit, all lesions had improved clinically with a reduction in size and improvement in pain. The patient completed 12 months of therapy with a complete clinical response. To date, the patient remains in clinical and radiological remission.

Case Two
This case illustrates DR. Patient Two had a long history of multiple NMSC lesions being excised from the head, neck, and chest, including a Merkel cell carcinoma (MCC). Patient Two had a CSCC lesion from the left clavicular area that required multiple re-excisions over a six-month period before clear margins were achieved in January 2018. In May 2018, a left infraclavicular chest recurrent CSCC was excised that involved   Figure 1D) demonstrated complete resolution of the pons lesion, stable orbital lesion and "not measurable" ulcerated left scalp lesion.
Two of the patient cases presented introduce a new phenomenon we have defined as a "delayed response after confirmed progression (DR)". That is, the observation of a clinical and radiological response after iCPD. In both cases, clinical deterioration was observed early in treatment but after treatment beyond progression, durable clinical and radiological improvement was obtained. In one case, change of tumor evolution with the cessation of new lesions developing may have been the only indication to herald DR. Review of our institutional experience of patients treated with cemiplimab on trial (up to 15/SEP/2020) in the advanced setting who have received more than one dose, demonstrates an estimated incidence of 2/39 (5%) of DR without any cases of "traditional pseudoprogression". We have observed DR occurring late in treatment courses and "traditional pseudoprogression" for nontrial patients. We also raise the concept that clinical deterioration may not militate against tumor response. In distinction to the three cases discussed where all patients experienced diseaserelated deterioration, the KEYNOTE-629 study which used RECIST 1.1 criteria for response evaluation reported that 29 clinically stable patients received treatment beyond progression (10). Of these, 12 patients continue on therapy and eight patients have developed responses (1 complete response, 7 partial responses according to irRECIST ; FIG S4 (10)). Therefore, our observation of DR and scenarios discussed further below, caution against a nihilistic approach to CSCC patient management when using immunotherapy. These cases highlight the important consideration of the timing and method of disease response assessments and the limitations of most criteria to capture pseudoprogression/DR, response rates and best overall response (BOR) assessments as part of clinical trial reporting. Use of the RECIST 1.1 framework will have not accurately reflected BOR for these patients, with the known limitation of most frameworks to act as a surrogate for progression-free and overall survival (25). Although the timing of imaging assessments are necessarily at a time point to permit sufficient receipt of therapy to assess response, these scans may not contemporaneously reflect disease evolution and response. Further, data from trials that utilize immunotherapy in the neoadjuvant setting prior to definitive surgery suggests that complete pathological responses may be seen only as partial response on imaging (26)(27)(28)(29). This highlights the clinical need for research into the predictive and prognostic role of imaging techniques in patients treated with immunotherapy for CSCC and the need to identify relevant molecular liquid biopsy biomarkers for disease surveillance (30)(31)(32).

SECOND PRIMARY TUMORS ON IMMUNOTHERAPY
Multiplicity is common for patients with NMSC, with~74% of all skin cancers being excised from patients with multiple NMSC lesions particularly involving the head and neck region (2, 3). We have observed that patients on immunotherapy can develop both CSCC and BCC as second primary tumors (SPT) despite responsive disease elsewhere, likely due to field cancerization (33). The mechanism by which SPT escape treatment control have not yet been elucidated and are likely to illuminate molecular mechanisms behind immune escape.
From a clinical management perspective, it is critical that SPT development is not mistaken as treatment failure given that these lesions can resolve with ongoing therapy or can be managed with local therapy. As a general principle, the SPT should be observed for a period whilst continuing immunotherapy and if regression or stability does not occur then local therapy can be pursued with a view of continuing systemic therapy for control of the other immunotherapy-responsive disease. On retrospective review of our patients treated with cemiplimab on trial, of eight patients who developed biopsy proven SPT resistant to immunotherapy, seven patients with CSCC and six with BCC required formal excision and/or radiotherapy for local management with all but one patient having more than one lesion. We have not identified any situation for which the new SPT has resulted in metastatic disease or has heralded the development of disease progression in existing lesions. Molecular profiling of these SPTs that develop on immunotherapy is important to define disease heterogeneity and to identify the likely numerous mechanisms of immune escape.

HETEROGENEOUS RESPONSES ON IMMUNOTHERAPY
Discordant immunotherapy responses can be observed between existing lesions, where concurrent local therapy for immunotherapy-resistant lesions may be warranted to secure control. Given the common multiplicity of NMSCs occurring in the same patient (2), awareness of response heterogeneity is key to avoid inappropriate early cessation of immunotherapy.

Case Four
This case demonstrates heterogeneous responses of two baseline lesions to immunotherapy. Patient Four was a 70 year old man with a multiply recurrent CSCC of his left forearm which had required six re-excisions. The largest resected recurrence was a 70x50mm spindled/poorly differentiated CSCC, up to 8mm in depth, with lymphovascular and perineural invasion. Adjuvant 56Gy/28# of radiotherapy was completed in September 2018, with a truncated course due to toxicity. In May 2019, Patient Four developed a painful locally recurrent CSCC of his left elbow with bone on view that measured 41x32mm on MRI scan, and a small right cheek lesion suspected to be a separate primary CSCC. Following MDM discussion, surgery for the cheek lesion was planned prior to immunotherapy for the elbow lesion. Wide local excision of the cheek lesion on the 26/JUN/ 2019 revealed three areas containing moderately differentiated CSCC spanning over 26mm, 32mm and 40mm with a transected deep margin, multiple CSCC tissue deposits of 2mm to 8mm, with perineural invasion abutting the margin. Re-excision was performed on 03/JUL/2019 revealing multiple foci of moderately differentiated CSCC deposits with perineural invasion, vascular tumor emboli, and tumor 0.2mm from a margin. Patient Four consented to participation the NCT02760498. At baseline, the exophytic left elbow lesion clinically measured 70x35mm. Prior to the third dose (Day 57), a near complete clinical response of the left elbow CSCC was observed with a residual superficial ulcer measuring 10mm accompanied by an improvement in pain. MRI scan confirmed the lesion had reduced to 32x13mm. A month later (Day 87) a right cheek nodule recurred and given its persistence, resection was performed on Day 106 demonstrating an 11mm CSCC with ≤1mm margins. Ongoing immunotherapy secured a complete clinical response for the left elbow lesion by Day 141 with imaging showing no identifiable tumor (nominal RECIST measurement of 5mm). However, three further excisions (Day 147, Day 188, Day 218) were required to manage the cheek lesion which recurred twice during adjuvant radiotherapy planning. Palliative radiotherapy of 36Gy in 6# to the cheek lesion was completed on Day 283. Most recent imaging demonstrated a complete metabolic response (Day 335) and ongoing radiological PR (nominal 5mm, Day 447) of the left elbow disease. However, since Day 394 clinical recurrence of the right cheek nodule has been progressive. Due to the absence of local therapies available to effectively treat the recurrent right cheek CSCC, Patient Four decided to cease further immunotherapy to pursue best supportive care.
Tumoral heterogeneity leading to discordant treatment response is a known therapeutic hurdle that can contribute to disease progression with the development of clonal resistance (34)(35)(36)(37)(38). In the era of immunotherapy which can secure durable disease control, understanding the contribution of local therapy towards overall disease control and survival is crucial but poorly understood. It has been observed in patients with melanoma treated with immunotherapy who received local therapy to progressive lesions in order to achieve no evidence of disease, that those who had local therapy to new lesions had poorer survival compared to those who had local therapy to progressive pre-existing metastases (PFS 6% vs 70%, p=0.001) (39). Consideration of the anatomical site of oligoclonal resistant disease will have clinical (e.g. brain versus lung) and therapeutic implications (e.g. surgery versus radiotherapy). Patient Four's case demonstrates that discordant immunotherapy responses can be observed between lesions, where concurrent local therapy may be warranted in an attempt to secure control of immunotherapyresistant lesions.

EXPANDING CLINICAL BOUNDARIES -ACTIVITY IN ADVANCED, FUNGATING, DISFIGURING DISEASE AND GOOD TOLERANCE DESPITE MULTIPLE COMORBIDITIES
It is essential to consider patient factors, treatment morbidity and goals of management in oncological care. Comorbidities and ECOG performance status guide management to optimize patient outcomes that may focus on quality of life or may be driven to obtain survival benefits or both. It is well recognized that a large disparity exists between trial patients versus "real-world" patients, who often are older, of minority groups, and with comorbidities that may interfere with assessment of therapeutic efficacy or toxicity (40) (41,42). In the context of the functionally and cosmetically sensitive anatomical region of the head and neck, it is crucial to define that immunotherapy is generally tolerable and associated with improved quality of life (22,43,44). In CSCC, Maubec et al. have reported improved health-related quality of life for patients with immunotherapy-responsive disease (11), and use of PD-1 blockade has been demonstrated to be tolerable with side effects reported similar to other checkpoint inhibitors and with the ability to secure durable disease control (7,9,45). Noteworthy in the trial reports, is the median age of patients being 71-80 years with the oldest patients being 99 years old (7,10,11).
Our institutional experience in the trial and "real-world" setting is that checkpoint inhibitor therapy is exceptionally well tolerated by patients with CSCC. Generally, few treatment contraindications exist and few comorbidities raise concern including extreme age, dialysis, other synchronous malignancies requiring treatment, and poor ECOG performance status. This is a paradigm change in our approach to patients with CSCC and is paramount given the dramatic response rates achieved by therapy, providing symptomatic and durable control. This is in stark contradistinction to our approach with mucosal head and neck cancer patients. As illustrated in Figure 2, there are not many clinical situations in which disease is considered "too advanced" for immunotherapy. That is, the stage of disease, location of disease, and extent of disease does not militate against response to immunotherapy. Anecdotally, the extent of re-epithelialization following response can be impressive creating complexity around the timing of reconstructive surgery if pursued.

CONCLUSION AND FUTURE DIRECTIONS
Use of immunotherapy has revolutionized the care of patients with advanced CSCC, leading to a paradigm shift in the selection of patients for treatment with the expectation of response and durable control even in the advanced recurrent or metastatic setting. We have focused on describing unique clinical concepts related to the treatment of CSCC with immunotherapy including the phenomenon of delayed response after confirmed progression (DR), observation of tumor responses despite clinical deterioration in iCPD, and the need to consider flexible treatment approaches in patients with multiple NMSC.
An improved understanding of CSCC will undoubtedly enhance patient selection for therapy as ongoing clinical research efforts investigate the role of immunotherapy in the adjuvant and neoadjuvant settings. Few cancer registries collect data on CSCC or advanced CSCC, limiting our understanding of the spectrum of disease, burden of need, morbidity and costs related to treatment. The therapeutic advances necessitate rapid development of real-time methods to assess tumor response (e.g. liquid biopsy, imaging or combination approaches) that are more informative than current imaging modalities and response criteria. Translational research will be crucial to molecularly define the clinical spectrum of CSCC (46,47), and identify reliable predictive and prognostic markers to therapy, including mechanisms of immune evasion. Specifically, comprehensive profiling of immunotherapy exposed tumors, including single cell sequencing approaches, will be important to further clarify inter-tumoral, intra-tumoral and tumor microenvironment molecular processes that underpin the described clinical concepts (48,49).

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article. Further inquiries can be directed to the corresponding author.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Peter MacCallum Cancer Centre Ethics Committee. Written consent has been provided by the patient with reidentifiable images.

AUTHOR CONTRIBUTIONS
Concept and design -AL, DR. All authors contributed to the article and approved the submitted version.

ACKNOWLEDGMENTS
Regeneron and Sanofi have reviewed the manuscript prior to publication but have not contributed or modified the manuscript in any substantial way. We thank all the patients and families, who have received treatment at our center. We thank our clinical trial team members who have contributed significantly to patient care.