Introduction
Cutaneous squamous cell carcinoma (cSCC), the second most common type of non-melanoma skin cancer, accounts for 20% of all skin cancer deaths [
1]. Patients with distant metastatic cSCC have mortality rates of ≥ 70% [
1,
2]. Approximately 1.8 million cases of cSCC were diagnosed worldwide in 2017, with elderly patients at particularly high risk given that cumulative sun exposure is the primary risk factor for development [
3,
4]. The symptoms of cSCC and the high morbidity associated with standard treatment for the disease can result in pain, scarring, disfigurement and anxiety, any of which can affect patients’ overall health-related quality of life (HRQoL) [
5,
6]. Patients with cSCC may experience negative psychosocial and mental health effects as a result of treatments that alter their physical appearance or impede their ability to speak, swallow or see, particularly in those with advanced cSCC in the anatomic regions above the clavicle [
6‐
9].
There remains an unmet need for effective and tolerable therapies that also maintain physical functioning and HRQoL for patients whose tumours are unresectable due to anatomical location or recurrence within a previously irradiated field. Treatment guidelines for patients with recurrent or metastatic (R/M) cSCC recommend therapy with programmed death 1 (PD-1) immune checkpoint inhibitors (i.e. cemiplimab or pembrolizumab) if curative surgery or radiotherapy is not feasible, or platinum-containing chemotherapy or epidermal growth factor receptor inhibitors (e.g. cetuximab) if the patient is ineligible for immunotherapy [
10,
11]. The PD-1 inhibitors cemiplimab and pembrolizumab were included in the treatment paradigm based on trials showing effective anti-tumour activity and acceptable tolerability in locally advanced or metastatic cSCC [
12‐
14].
Pembrolizumab has demonstrated extended survival for several advanced malignancies, including R/M head and neck squamous cell carcinoma [
15‐
19] and Merkel cell carcinoma, which, like cSCC, is an ultraviolet-exposure–driven malignancy [
20]. The phase 2 KEYNOTE-629 study is a two-cohort trial designed to evaluate the efficacy and safety of pembrolizumab in patients with unresectable locally advanced or R/M cSCC [
14]. In an interim analysis of the R/M cSCC cohort of KEYNOTE-629 (
n = 105; data cutoff, 8 April 2019), pembrolizumab in heavily pre-treated patients demonstrated anti-tumour activity, durable and rapid responses, encouraging survival outcomes and manageable safety [
14]. The objective response rate was 34.3% [95% confidence interval (CI) 25.3–44.2], and the disease control rate was 52.4% (95% CI 42.4–62.2); median duration of response was not reached (range 2.7–13.1+ months). Median progression-free survival was 6.9 months (95% CI 3.1–8.5), and median overall survival was not reached (95% CI 10.7 months–not reached). Adverse events were consistent with the established safety profile of pembrolizumab. The patients with R/M cSCC in this study were primarily elderly [median age, 72 years (range 29.0–95.0)] and most were heavily pre-treated, indicating that pembrolizumab is efficacious and well tolerated in a relatively fragile population [
14]. The clinical benefit of pembrolizumab demonstrated in KEYNOTE-629 resulted in US Food and Drug Administration regulatory approval for the treatment of R/M cSCC not curable by surgery or radiation therapy [
21,
22].
We present results of the pre-specified exploratory HRQoL analyses for patients with R/M cSCC treated with pembrolizumab from the first interim analysis of KEYNOTE-629.
Methods
Study Design and Patients
KEYNOTE-629 (NCT03284424) is an open-label, single-arm, phase 2 study of pembrolizumab in patients with unresectable locally advanced or R/M cSCC. The study is ongoing at 39 centres in nine countries. Detailed methods and eligibility criteria for KEYNOTE-629 are published [
14]. Briefly, eligible patients were adults with histologically confirmed cSCC as the primary site of malignancy that was metastatic (i.e. disseminated disease) and/or unresectable (i.e. not curable by surgery or radiation therapy) and with an Eastern Cooperative Oncology Group performance status of 0–1. Pembrolizumab 200 mg was administered intravenously every 3 weeks until disease progression, unacceptable toxic effects or study withdrawal for up to 35 cycles (approximately 2 years).
The study protocol and amendments were approved by the appropriate ethics review committees [names listed in Electronic Supplementary Material (ESM) Table S1]. All patients gave written informed consent. The study was conducted in accordance with the protocol, its amendments and the ethics principles outlined in the Declaration of Helsinki.
HRQoL Assessments
HRQoL assessments involved the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30) [
23,
24] and European Quality of Life 5-Dimension 5-Level (EQ-5D-5L) [
25,
26]; questionnaires were administered electronically prior to other study procedures at baseline, at weeks 3 and 6, and then every 6 weeks through the first year, every 9 weeks until the end of treatment, and at the 30-day safety follow-up visit.
The EORTC QLQ-C30, the most commonly used cancer-specific instrument to assess HRQoL, is psychometrically and clinically validated [
23,
24]. It comprises five functional dimensions, nine items that measure symptoms or cancer-specific concerns, and a global health status/quality of life (GHS/QoL) scale [
24]. The EQ-5D-5L is a standard instrument used to measure health status consisting of two parts: a graded visual analogue scale (VAS) on which patients rate their perceived health status from 0 (worst health you can imagine) to 100 (best health you can imagine), and a descriptive system comprising five health state dimensions (i.e. mobility, self-care, usual activities, pain/discomfort and anxiety/depression) rated on a five-level descriptive scale (no problems, slight problems, moderate problems, severe problems, extreme problems/unable to) that is used to generate a health state profile for each patient [
25,
26]. For the descriptive system, each health state can be assigned an index or utility score based on societal preference weights.
Compliance with the HRQoL questionnaires was defined as the proportion of patients who completed the HRQoL assessment among those who were expected to complete the questionnaires at a given time point, excluding patients missing by design. Patients who were missing by design included those who died; who discontinued due to an adverse event, clinical progression, progressive disease, physician decision or withdrawal from study; who completed study treatment; for whom the clinic visit was not reached or scheduled; and for whom translation was not available. The completion rate for the HRQoL questionnaires was defined as the proportion of patients who completed the HRQoL assessment among all patients in the HRQoL analysis population.
End Points
The pre-specified exploratory HRQoL end points were change from baseline in the EORTC QLQ-C30 GHS/QoL, functioning, and symptom scores, and change from baseline in the EQ-5D-5L health status score. Mean changes from baseline in HRQoL scores were primarily evaluated at week 12 to ensure an adequate completion rate (i.e. ≥ 70% of patients completed the HRQoL assessment among all patients in the HRQoL analysis population). Analyses of mean change from baseline in QLQ-C30 GHS/QoL and physical functioning scores were also summarised for patients who were on study and able to complete questionnaires at each time point through week 48.
Responses for each of the GHS/QoL, physical functioning and symptom scales of the EORTC QLQ-C30 were calculated by averaging the items within the scales and transforming the scores linearly (scales range from 0 to 100) [
24]. Symptom scales for which higher scores represented higher symptom burden were reverse-scored to simplify presentation, such that a higher score for GHS/QoL, functioning and symptom scales denotes better HRQoL. Responses to the EQ-5D-5L VAS were scored from 0 (worst imaginable health) to 100 (best imaginable health) [
26]. EQ-5D-5L responses were weighted and aggregated into utility scores based on the US algorithm, with scores ranging from < 0 (where 0 is equivalent to death) to 1 (equivalent to full health) [
26,
27].
A ≥ 10-point change from baseline in the EORTC QLQ-C30 GHS/QoL, functioning and symptom scores was considered clinically meaningful [
28]. The overall proportions of patients with improved, stable and deteriorated GHS/QoL and physical functioning scores relative to baseline were summarised. Overall improvement was defined as a ≥ 10-point increase in score from baseline at any time during the trial, with confirmation at the next visit. For patients who did not achieve improved HRQoL scores, any of the following were defined as stable scores: improvement (≥ 10-point increase in score) confirmed by a < 10-point change at the next visit, < 10-point change in score confirmed by a < 10-point change at the next visit, or a < 10-point change in score and confirmed by an improvement (≥ 10-point increase in score) at the next visit. For patients without improved or stable scores, deterioration was defined as a ≥ 10-point decrease in score from baseline. In addition, an exploratory sub-analysis of the proportion of patients with improved, stable or deteriorated GHS/QoL and physical functioning scores was conducted in patients classified as responders based on objective response [patients with complete response (CR) or partial response (PR)] in comparison with non-responders [patients with stable disease (SD) or progressive disease (PD)], as assessed by blinded independent central review using Response Evaluation Criteria in Solid Tumours version 1.1 (RECIST v1.1). For the EQ-5D-5L, a ≥ 7-point change from baseline in VAS and a ≥ 0.06 change from baseline in EQ-5D-5L utility index was considered to be a minimally important difference [
29].
Statistical Analyses
The HRQoL analysis population included all patients who received ≥ 1 dose of pembrolizumab and had HRQoL assessments at both baseline and at least once after baseline. The exact binomial method was used to assess the overall proportion of patients with improved, stable or deteriorated GHS/QoL and physical functioning relative to baseline. No imputation for missing data was performed. HRQoL analyses were conducted using the interim analysis data cutoff of 8 April 2019.
Discussion
Efficacious therapies with manageable tolerability that do not negatively affect HRQoL in patients with R/M cSCC are needed. Pembrolizumab demonstrated durable anti-tumour activity and acceptable safety in patients with R/M cSCC in the first interim analysis of KEYNOTE-629 [
14]. However, because patients with R/M cSCC tend to be fragile due to advanced age, it is particularly important to assess the impact of pembrolizumab treatment on HRQoL in this population. In this pre-specified exploratory HRQoL analysis of KEYNOTE-629, patients treated with pembrolizumab who remained on study at week 12 demonstrated stable overall GHS/QoL, functioning and symptom scores; in addition, a correlation was observed between HRQoL benefit and response to pembrolizumab. Furthermore, the descriptive trend in stable GHS/QoL and physical functioning scores extended for as long as 48 weeks. These findings support the benefit of pembrolizumab monotherapy in this elderly patient population.
Growing evidence from immunotherapy trials has shown that monotherapy with PD-1 inhibitors provides efficacy and safety benefits while maintaining or improving HRQoL in several tumour types [
30]. The results of KEYNOTE-629 and a phase 2 trial of cemiplimab further support the benefits of PD-1 inhibitors in cSCC, with both studies reporting durable responses with stable or improved HRQoL in patients with advanced cSCC receiving treatment with a PD-1 inhibitor [
13,
14,
31]. In the exploratory analysis of HRQoL in the phase 2 trial of cemiplimab, patients with unresectable locally advanced or metastatic cSCC had a clinically meaningful reduction in pain at cycle 5 of cemiplimab (administered at 3 mg/kg every 2 weeks for 12 cycles or 350 mg every 3 weeks for 6 cycles), with most patients having stable or a trend toward improved HRQoL scores across the other EORTC QLQ-C30 domains [
31]. The generally stable or improved HRQoL with pembrolizumab and cemiplimab is particularly notable given the advanced and primarily elderly populations that were studied. The limitations of the current HRQoL analysis include the single-arm design of KEYNOTE-629. To our knowledge, HRQoL analyses have not been published for other treatments that are commonly used for R/M cSCC, such as epidermal growth factor receptor inhibitors, and therefore, the effect of these agents on HRQoL is unclear.
Acknowledgements
The authors thank the patients and their families and all investigators and site personnel. In addition, the authors thank Zhi Jin Xu for statistical analysis.
Author Contributions
Conceptualisation: Olga Vornicova, Diana Chirovsky, Ramona F. Swaby; Methodology: Brett G. M. Hughes, Olga Vornicova, Abhishek Joshi, Josep M. Piulats, Diana Chirovsky, Pingye Zhang, Burak Gumuscu, Ramona F. Swaby, Jean-Jacques Grob; Formal analysis and investigation: Brett G. M. Hughes, Rene Gonzalez Mendoza, Nicole Basset-Seguin, Olga Vornicova, Abhishek Joshi, Nicolas Meyer, Florent Grange, Josep M. Piulats, Jessica R. Bauman, Diana Chirovsky, Pingye Zhang, Burak Gumuscu, Ramona F. Swaby, Jean-Jacques Grob; Writing—original draft preparation: Brett G. M. Hughes, Abhishek Joshi, Nicolas Meyer, Josep M. Piulats, Jessica R. Bauman, Diana Chirovsky, Pingye Zhang, Burak Gumuscu, Ramona F. Swaby, Jean-Jacques Grob; Writing—review and editing: Brett G. M. Hughes, Rene Gonzalez Mendoza, Nicole Basset-Seguin, Olga Vornicova, Jacob Schachter, Abhishek Joshi, Nicolas Meyer, Florent Grange, Josep M. Piulats, Jessica R. Bauman, Diana Chirovsky, Pingye Zhang, Burak Gumuscu, Ramona F. Swaby, Jean-Jacques Grob; Funding acquisition: None. Resources: Brett G. M. Hughes, Olga Vornicova, Jean-Jacques Grob; Supervision: Brett G. M. Hughes, Olga Vornicova, Josep M. Piulats, Jessica R. Bauman, Diana Chirovsky, Burak Gumuscu, Ramona F. Swaby, Jean-Jacques Grob.
Disclosures
Brett G. M. Hughes reports serving as an advisor for Bristol Myers Squibb, Roche, AstraZeneca, Pfizer, Eisai, Takeda, and Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Inc., Kenilworth, NJ, USA (MSD), and receiving an institutional research grant from Amgen. Rene Gonzalez Mendoza has nothing to disclose. Nicole Basset-Seguin and Olga Vornicova report being investigators for MSD. Jacob Schachter reports receiving personal fees from Bristol Myers Squibb. Abhishek Joshi reports receiving grants from Novartis, Roche, Ipsen, AstraZeneca, and Bristol Myers Squibb. Nicolas Meyer reports being an investigator and/or consultant and/or speaker and/or received research grants from Bristol Myers Squibb, MSD, Roche, Novartis, Pierre Fabre, Merck GmbH, AbbVie, and Sun Pharma. Florent Grange has nothing to disclose. Josep M. Piulats reports receiving grants from MSD, Bristol Myers Squibb, BeiGene, and Janssen, and personal fees from MSD, Bristol Myers Squibb, BeiGene, Roche-Genentech, Janssen, Astellas, and Sanofi-Genzyme. Jessica R. Bauman reports serving as a consultant/advisor for Pfizer, AstraZeneca, Kura, and Bayer. Diana Chirovsky reports employment at MSD and is a shareholder in Merck & Co., Inc., Kenilworth, NJ, USA. Pingye Zhang and Burak Gumuscu report employment at MSD. Ramona F. Swaby was an employee of MSD at the time of the conduct of the study; she reports current employment at Prelude Therapeutics, 200 Powder Mill Road, Wilmington, DE 19803, USA. Jean-Jacques Grob reports serving as an advisor for MSD, Bristol Myers Squibb, Novartis, Roche, Sanofi, Pierre Fabre, Merck, Pfizer, and Sun Pharma and reports receiving travel funding or speaker fees from MSD, Bristol Myers Squibb, Novartis, and Pierre Fabre. Ramona F. Swaby was an employee of Merck & Co., Inc., Kenilworth, NJ, USA at the time the study was conducted.