Single-center prospective study on the efficacy of nivolumab against platinum-sensitive recurrent or metastatic head and neck squamous cell carcinoma

Nivolumab, an immune checkpoint inhibitor, is beneficial to patients with platinum-refractory recurrent or metastatic head and neck squamous cell carcinoma (R/M-HNSCC). However, platinum-sensitive R/M-HNSCC has not yet been studied. Hence, in this prospective study, we evaluated the efficacy and safety of nivolumab in patients with platinum-sensitive R/M-HNSCC. This prospective single-arm study was conducted in a single institution in Japan. Patients with platinum-sensitive R/M-HNSCC (defined as head and neck cancer that recurred or metastasized at least 6 months after platinum-based chemotherapy or chemoradiotherapy) were enrolled. The primary endpoint was overall survival (OS). The secondary endpoints were progression-free survival (PFS), overall response rate (ORR), immune-related adverse events (irAEs), and quality of life (QOL). This study was registered at the University Hospital Medical Information Network Clinical Trials Registry (UMIN000031324). Twenty-two patients with platinum-sensitive R/M-HNSCC were enrolled. The median OS was 17.4 months, and the 1-year OS rate was 73%. The median PFS was 9.6 months, 1-year PFS rate was 48%, and ORR was 36%. Sixteen irAEs were recorded in 12 patients; however, no grade 4 or 5 irAEs were observed. The QOL assessments revealed that nivolumab did not decrease the QOL of patients. Nivolumab is effective against platinum-sensitive R/M-HNSCC with acceptable safety.


Results
Patient characteristics. From February 16, 2018, to December 31, 2019, 61 patients with recurrent or metastatic head and neck cancer received at least one cycle of nivolumab. Among these, 39 patients were excluded owing to platinum resistance (N = 35), non-squamous cell carcinoma (N = 3), and age > 75 years (N = 1). Twentytwo patients who met the inclusion criteria were enrolled in this study. The data cutoff date was December 31, 2020. Detailed data of all patients are provided in the Supplementary Information. Table 1 lists the age, sex, smoking history, drinking history, Eastern Cooperative Oncology Group performance status (ECOG PS), tumor-node-metastasis (TNM) classification and staging at recurrence or metastasis, and programmed cell death-ligand 1 (PD-L1) expression. The median age of patients was 63 (range, 42-74) years. The primary tumor sites included the nasopharynx (N = 6) and maxillary sinus (N = 4). PD-L1 expression was < 1% in 5 patients, ≥ 1% in 16 patients, ≥ 20% in 12 patients, ≥ 40% in 8 patients, and unmeasured in 1 patient. Treatment modalities prior to nivolumab administration are presented in Table 2. Previous exposure to platinum included induction chemotherapy (docetaxel/cisplatin/5-fluorouracil, N = 3), definitive concurrent chemoradiotherapy (N = 13), postoperative chemoradiotherapy (N = 6), and systemic therapy for recurrence or metastasis (cetuximab/cisplatin/5-fluorouracil, N = 1). In addition, one patient underwent a duplicate platinumcontaining regimen. Cetuximab was used prior to nivolumab treatment as a definitive concurrent radiotherapy (N = 1) and as a systemic therapy for countering the recurrence or metastasis of disease (cetuximab/paclitaxel, N = 3; cetuximab/cisplatin/5-fluorouracil, N = 1). The median period between the date of the last administration of platinum and the date of disease progression was 351 (range, 201-1,081) days, whereas the median period between the date of the last administration of platinum and the date of the first administration of nivolumab was 397 (range, 211-1,110) days. The median number of cycles of nivolumab administration was 9 (range, 1-57).
Efficacy. After data cutoff, the median follow-up period for all patients was 14.9 (range, 3.1-33.7) months.

Safety.
Quality of life assessments. Functional scales (physical, role, emotional, cognitive, and social functioning) and global health status were assessed using the European Organization for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire (QLQ) Core 30 Module (QLQ-C30). Domain scales (pain, swallowing, sense problems, speech problems, trouble with social eating, trouble with social contact, and reduced sexuality) were assessed using the EORTC QLQ Head and Neck Cancer Module (QLQ-H&N35). Detailed data regarding the quality of life (QOL) scores are provided in the Supplementary Information. The data of 19 patients, excluding the 3 patients who died by week 8, were analyzed. Figure 3 shows the QOL scores relative to the baseline value at 8 and 16 weeks after the initiation of nivolumab. There was no significant change in the QOL of patients treated with nivolumab.

Discussion
The aim of this study was to determine the efficacy and safety of nivolumab in patients with platinum-sensitive R/M-HNSCC through a prospective study. The main outcomes of this study were the median OS of 17.4 months, median PFS of 9.6 months, ORR of 36%, and DCR of 68%. Although the number of patients in the study by primary site was too small to ensure the certainty of the results, there was no clear effect of the primary site on efficacy.
Considering that the prognosis of patients with R/M-HNSCC is poor at less than 1 year 1-3 , it is evident that treatment with nivolumab improved the OS. To the best of our knowledge, this is the first prospective study to demonstrate the efficacy of nivolumab in patients with platinum-sensitive R/M-HNSCC only.
No prospective study has examined the effect of nivolumab in patients with platinum-sensitive head and neck cancer or with other carcinomas. The concept of platinum-resistant and platinum-sensitive cancer was originally used in the treatment of recurrent ovarian cancer and has recently been applied to R/M-HNSCC 11,12 . In recurrent ovarian cancer, patients with platinum-sensitive cancer are more likely to respond to subsequent chemotherapy and have a better prognosis than those with platinum-resistant cancer 11,12 . For R/M-HNSCC, the effects of the EXTREME regimen on the OS of two groups of patients with R/M-HNSCC, namely patients with platinum-resistant and platinum-sensitive R/M-HNSCC, have been compared 6 19 . In our institution, we followed up with the patients every month after completing the local treatment for head and neck cancer and performed imaging studies every 3-4 months. In patients with suspected recurrence, imaging was performed at relatively short intervals. In this study, tumor burden and growth rate were not measured. However, R/M-HNSCC was likely diagnosed at a relatively early stage, resulting in the immediate administration of nivolumab. We believe this is one of the reasons for the favorable OS and PFS observed in this study. Unlike multicenter studies, our study did not require a specific time for the enrollment of patients, which enabled the rapid initiation of treatment, an advantage of single-arm studies conducted in a single institution.
In a sub-analysis of data of CheckMate-141, Ferris et al. compared the OS of patients with and without prior exposure to cetuximab 20 . They reported that patients without prior exposure were at a lower risk of death from nivolumab use. Cetuximab binds to the epidermal growth factor receptor on tumor cells but promotes the growth of immunosuppressive regulatory T cells in the tumor microenvironment 21 . Patients whose lesions worsen after cetuximab treatment are less likely to respond to immunotherapy owing to the proliferation of immunosuppressive regulatory T cells, such as regulatory T cells and myeloid-derived suppressor cells 21,22 . In our study, five patients (22.7%) with a history of cetuximab treatment before nivolumab treatment could not attain an endpoint for previous cetuximab use. Therefore, it is not known whether previous cetuximab use had any effect on patient response. The CheckMate-141 QOL assessments showed that nivolumab does not reduce the QOL (compared with other chemotherapy or molecularly targeted drug treatments) 4,23 . This observation is not limited to R/M-HNSCC; it has also been reported for melanoma, non-small cell lung cancer, and renal cell carcinoma [24][25][26] . Our results also showed no reduction in the QOL of patients after treatment with nivolumab.
Among the 22 patients, 16 irAEs occurred in 12 patients. Grade 3 or higher irAEs were liver dysfunction (N = 3) and hypothyroidism (N = 1). Nivolumab was discontinued due to irAEs in three patients with grade 3 liver dysfunction. After treatment with steroids, all patients showed an improvement in liver dysfunction. There were three other patients who discontinued nivolumab due to the deterioration of their general condition. Safety was not a significant problem.
There were some limitations to the study. This was a single-arm study, and although the enrollment number was set at 50 patients, it was not reached. Sixty-one patients with R/M-HNSCC received nivolumab within the target period, but 35 were excluded due to platinum resistance. The effects of nivolumab in patients with   Table 4. Immune-related adverse events. www.nature.com/scientificreports/ platinum-resistant and platinum-sensitive R/M-HNSCC should be further examined in a multicenter prospective study with a higher number of patients. Therefore, as the next step of the study, we are investigating the results of treatment in platinum-sensitive and platinum-resistant groups in a multicentered study. The results of this study will be published separately. In addition, the outcomes of nivolumab and pembrolizumab as first-line treatments for patients with platinum-sensitive R/M-HNSCC with a history of platinum use should be compared in future studies. These studies are necessary to establish more effective treatment strategies for patients with R/M-HNSCC.
In conclusion, our results suggest that nivolumab is highly effective in treating platinum-sensitive R/M-HNSCC, with no reduction in the QOL following treatment. These results may have implications for the current treatment regimen for R/M-HNSCC.

Patients and methods
Study design. This was a single-institution, open-label, single-arm, phase II interventional study conducted in Japan to evaluate the efficacy of nivolumab in patients with R/M-HNSCC. The study was approved by the ethics committee of Tokyo Medical University (SH3946) and registered at the University Hospital Medical Information Network Clinical Trials Registry prior to the recruitment of patients (UMIN000031324; date of first registration;15/02/2018). The study adhered to the tenets of the Declaration of Helsinki, and informed consent was obtained from all the patients.

Patients.
Patients with R/M-HNSCC were screened for eligibility for this study. The number of patients expected to be enrolled was 50, which was calculated based on the number of patients with R/M-HNSCC treated in our institution (approximately 60 patients per year), the expected rate of enrollment (40%), and the study period to enroll patients (2 years). The main exclusion criteria were prior exposure to nivolumab; platinum refractoriness, defined as disease progression within 6 months after the last administration of platinum; nonsquamous cell carcinoma; and age over 75 years. The complete inclusion and exclusion criteria are provided in the Supplementary Protocol.
Outcomes and assessments. The primary endpoint was OS. The period of OS was defined as the duration between the date of nivolumab initiation and the date of the last follow-up or the patient's death, whichever occurred first. The secondary endpoints were PFS, ORR, and QOL. The period of PFS was defined as the duration between the date of nivolumab initiation and the date of objective disease progression or patient's death from any cause, whichever occurred first. Tumor response was assessed by three radiologists in our institution according to the Response Evaluation Criteria in Solid Tumors Guideline (version 1.1) 27 . The QOL was assessed using the EORTC QLQ-C30, a basic quality of life questionnaire used for patients with malignancies, and the EORTC QLQ-H&N35, a disease-specific questionnaire 28,29 . The EORTC QLQ-C30 and QLQ-H&N35 were translated into Japanese and answered by the patients prior to treatment initiation and at 8 and 16 weeks after treatment initiation. These scores ranged from 0 to 100, with higher scores indicating higher functioning and symptom burden, except for the global health status items. Higher scores indicate lower functioning and symptom burden. TNM was classified using the Union for International Cancer Control stage, 7 th edition 30 . The discontinuation criteria for nivolumab were specified in the protocol based on the grade of irAEs, but no weight loss criterion was set 4 . The irAEs were assessed using the Common Terminology Criteria for Adverse Events version 4.0 31 .
Administration of nivolumab and PD-L1 measurement. Nivolumab was administered as a single intravenous dose (3 mg/kg body weight) every 2 weeks. A single treatment cycle lasted 2 weeks, and imaging assessment was carried out every 4-8 cycles. Treatment was continued until the confirmation of objective disease progression, the occurrence of unacceptable toxicity, or other reasons assessed by the attending physician. However, even after confirming progression based on clinical or imaging findings, treatment was continued when the attending physician thought that clinical benefit was likely 32 .
PD-L1 was measured in CheckMate-141, and Dako 28-8 antibody (Dako, Carpinteria, CA) was used for immunohistochemical assay. The scoring systems include the Tumor Proportion Score (TPS), which measures PD-L1 on tumor cells, and the Combined Positive Score (CPS), which measures PD-L1 expressed on not only tumor cells but also macrophages and lymphocytes. In this study, the measurement was performed using the TPS, and not the CPS 32 .
Statistical analysis. OS and PFS were estimated using the Kaplan-Meier method. QOL assessments were conducted using a repeated measures linear mixed model, with repeated measures covariance structure as a composite symmetry, each QOL score as the dependent variable, time of measurement as a fixed factor, and subjects as a variable. In this case, the least squares mean and the 95% CI at each measurement point were calculated. The estimated mean and 95% CI were also calculated for changes based on the pre-nivolumab scores (baseline), and the significant change from the pre-nivolumab scores was tested. No correction for multiplicity was used. Statistical analysis for OS and PFS was performed using EZR, a statistical software that extends the functions of R and R commander; it is available on the Jichi Medical University Saitama Medical Center website 33 . SPSS statistics version 22.0 (IBM Japan, Ltd., Tokyo, Japan) was used for the statistical analysis of QOL. Results with P < 0.05 were considered statistically significant.

Data availability
The datasets generated in the current study are available from the corresponding author on request.