Abstract
Background/Aim: The treatment of patients with recurrent or metastatic head and neck squamous cell carcinoma (R/M HNSCC) has remained challenging. The effect of salvage chemotherapy (SCT) after nivolumab has been identified recently in other cancer types. The aim of this study was to examine the efficacy of SCT after nivolumab treatment in patients with R/M HNSCC. Patients and Methods: A retrospective study was conducted at four institutions in Japan. Fifty-six patients were enrolled in the study. Results: The overall survival (OS) in SCT patients was significantly longer than that in best supportive care (BSC) patients. In the SCT patients, the median OS, median progression-free survival (PFS) and objective response rate (ORR) were 7.3 months, 2.3 months and 36%, respectively. Prognostic factor for OS and ORR was performance score (PS) and previous radiation, respectively. Conclusion: SCT after nivolumab is associated with better clinical outcomes in patients with R/M HNSCC compared to those receiving BSC.
- Salvage chemotherapy
- Immune checkpoint inhibitor
- nivolumab
- head and neck carcinoma
- recurrent or metastatic
- best supportive care
Head and neck cancer accounts for 8% of all cancer cases worldwide; approximately 1,450,000 new cases are diagnosed, along with 500,000 cancer-related deaths registered annually (1). The prognosis of patients with recurrent or metastatic head and neck squamous cell carcinoma (R/M HNSCC) is generally poor, with a median overall survival (mOS) of <1 year. Moreover, identifying the optimal treatment for R/M HNSCC has remained challenging. The EXTREME study in 2008 showed a mOS of 10.1 months, median progression-free survival (mPFS) of 5.6 months and objective response rate (ORR) of 36% in the first-line setting (2). This regimen was combined with cetuximab, 5-fluorouracil and cisplatin/carboplatin (3). Further, the CheckMate-141 study in 2016 with nivolumab indicated a mOS of 7.5 months, mPFS of 2.0 months, and ORR of 13.3% in the second-line setting (4). Additionally, nivolumab was associated with better clinical outcome in recurrent HNSCC patients in the real-world setting (5, 6). However, phase III trials for the third-line setting have not been described in patients with R/M HNSCC. In the recent KEYNOTE-048 study, pembrolizumab treatment with or without chemotherapy in patients with positive programmed death-ligand 1 (PD-L1) expression showed significantly improved overall survival (OS) compared to that observed in the EXTREME regimen (7). Hence, it is necessary to consider treatment measures after immune checkpoint inhibitors (ICIs).
Few recent studies have reported that salvage chemotherapy (SCT) after ICIs might be useful in patients with non-small-cell lung cancer (NSCLC). In the two CheckMate studies comparing nivolumab with docetaxel in the second-line setting, nivolumab was found to confer longer OS than docetaxel (8, 9). In the OAK study, patients treated with atezolizumab in the second- or third-line setting followed by chemotherapy showed significantly longer OS than those treated without chemotherapy (10). Additionally, the KEYNOTE-024 study showed that the PFS was significantly longer in NSCLC patients treated with ICIs followed by chemotherapy than in those treated with chemotherapy followed by ICIs (11). However, the effect of treatment with SCT compared to that with BSC remained to be studied in patients with HNSCC.
In this multicentre study, we have retrospectively examined the efficacy of SCT after treatment with nivolumab in R/M HNSCC patients. This is the first trial to compare the effect of SCT with BSC in patients with R/M HNSCC. We also identified the predictors of treatment outcomes in these patients.
Patients and Methods
Patients. This study was conducted at four facilities in Japan: International University of Health and Welfare Mita Hospital, Tokyo Medical University, Kitasato University School of Medicine, and Tokyo Medical University Hachioji Medical Center. We retrospectively reviewed the clinical records of patients with R/M HNSCC treated with nivolumab between May 01, 2017 and August 31, 2018. The inclusion criteria were: age ≥18 years; treatment with nivolumab; at least one measurable lesion evaluated by computed tomography (CT) or magnetic resonance imaging (MRI); Eastern Cooperative Oncology Group (ECOG) performance status (PS) of 2 or less, and adequate hematologic, renal, and hepatic functions; and ability to provide written consent. One hundred patients were treated with nivolumab during this period. Of these, 12 were histologically diagnosed with non-SCC, 10 died before disease assessment, and 22 continued nivolumab, and thus were excluded; therefore, 56 patients were enrolled in the study.
This study was approved by the ethics committee of each facility (No. 5-18-71, T2018-0021 and B18-256).
Study design. The primary endpoint was OS. Secondary endpoints were: ORR, complete response (CR) + partial response (PR); disease control rate (DCR: CR + PR + stable disease); PFS; and safety. OS was examined for all patients and SCT patients. PFS, ORR, and adverse events (AEs) were analysed for SCT patients. Additional endpoints were the prognostic factors affecting OS, PFS and ORR. According to the Response Evaluation Criteria in Solid Tumors, version 1.1., tumour response was assessed using CT or MRI before treatment and every 4 to 8 weeks until disease progression or treatment discontinuation. The evaluation of AEs was based on the Common Terminology Criteria for Adverse Events, version 4.0.
Statistical analysis. OS, PFS, and follow-up period were estimated using the Kaplan–Meier method. The relationship between age, sex, ECOG PS, R/M status, previous radiation therapy, response to chemotherapy before nivolumab, response to nivolumab, and the number of nivolumab courses were assessed using the univariate Cox regression model. p<0.05 was considered statistically significant. All statistical analyses were performed using the EZR software, version 1.35 (Saitama Medical Center, Jichi Medical University, Saitama, Japan).
Study flow chart of 100 patients receiving nivolumab. SCT: Salvage chemotherapy; BSC: best supportive care.
Results
Characteristics of all eligible patients. The flow chart of the study is provided in Figure 1. The characteristics of the 56 patients included in the study are summarized in Table I. The patients showed a median follow-up period of 9.3 months (95%CI=7.2-13.7) and a median age of 67 years. Forty-seven patients were men (84%). The hypopharynx was the most frequent primary tumor site, followed by oropharynx and oral cavity. Most of the patients had an ECOG PS of 0 or 1. Thirty-nine patients showed locoregional recurrence with or without distant metastases. All patients were segregated to compare the characteristics between the SCT (n=25) and BSC (n=35) groups. No statistically significant difference was found between the two groups.
Efficacy for all eligible patients. The median OS was 5.8 months (95%CI=3.6-7.7) for all the patients. SCT patients showed significantly longer OS (7.8 months) than BSC patients (3.5 months; p=0.0028) (Figure 2). The univariate analysis performed to identify prognostic factors indicated that patients receiving SCT after nivolumab have longer OS than those not receiving SCT (HR=0.3, 95%CI=0.1-0.7, p=0.004) (Table II).
Characteristics of SCT patients. The characteristics of 25 SCT patients are summarized in Table I. The median follow-up period was 10.0 months (95%CI=5.8-NA). The median duration between the last injection of nivolumab and start of SCT was 17 days (range=12-89 days). Cetuximab (Cmab) + paclitaxel (PTX) was the most frequently administered regimen in 16 patients, followed by EXTREME regimen in 4 patients, S-1 in 3 patients, and PTX in 2 patients.
Patient characteristics (N=56) N (%).
Treatment efficacy in SCT patients. The Kaplan–Meier plots for OS and PFS are shown in Figure 3A and B, respectively. The mOS was 7.3 months (95%CI=4.5-14.3), and mPFS was 2.3 months (95%CI=1.8-3.8). There were significant differences in the OS between patients with ECOG PS0 and PS1 (p=0.032) (Table III). The ORR was 36% (95%CI=18-57), while DCR was 56% (95%CI=35-75) (Table IV). Previous radiation therapy was found to decrease the ORR in patients receiving SCT (p=0.041) (Table V).
Adverse events in SCT patients. All grade and grade 3-4 AEs were reported in 24 (96%) and 15 (60%) patients, respectively (Table VI). The most common grade 3-4 AEs were anemia (36%), neutropenia (16%), hypomagnesemia (8%), hyponatremia (8%), and hypokalemia (8%). There were no grade 5 AEs and infusion reactions. There was no relapse of immune-related AEs during the treatment period.
Kaplan–Meier curve for the overall survival of SCT and BSC patient. Black line: BSC patients, dashed line: SCT patients. SCT: Salvage chemotherapy; BSC: best supportive care.
Discussion
The effect of chemotherapy after nivolumab treatment has been identified recently (8-11). Our study identified a significant difference in the OS between SCT and BSC patients after nivolumab treatment. This is the first report to suggest SCT after nivolumab treatment to improve the clinical outcomes in patients with R/M HNSCC. These findings were previously reported for NSCLC. Costantini et al. in their retrospective study with 303 patients showed that after treatment with ICI, mOS of SCT patients was statistically longer than that of BSC patients (12). Additionally, the efficacy of SCT after administration of atezolizumab was defined through the OAK phase-III trial. Patients receiving SCT showed significantly prolonged OS than those receiving BSC. In case of head and neck cancer, Saleh et al. reported that mOS was 7.8 months for R/M HNSCC patients treated with SCT after ICIs (13). Thus, the mOS of patients in our study was comparable to that reported by Saleh et al. Furthermore, our results showed that prolonged OS was achieved only in patients receiving SCT after nivolumab treatment.
In our SCT group, univariate analysis showed that better ECOG PS was associated with longer OS, which is consistent with Saleh et al.'s findings who reported that the ECOG PS and best response to ICIs were prognostic factors for enhanced OS in R/M HNSCC patients undergoing SCT after nivolumab treatment (13). Moreover, the EXTREME trials reported that better PS was associated with longer OS than poor PS. Furthermore, PS generally associates with OS in patients receiving any treatment.
Univariate analyses of prognostic factor for the OS.
Further, previous reports have described that ORR and mPFS were 10%-39% and 2.1-4.6 months, respectively, in the second-line setting (14-16). Thus, the results of our study are consistent with those of previous studies. We found that previous radiation therapy was a prognostic factor for decreased ORR in patients with SCT. Daniel et al. reported that radiation-induced anemia had a negative impact on treatment outcomes, especially with combined chemotherapy (17). Therefore, radiation-induced anemia may reduce the effect of chemotherapy.
Here, we recorded different extent of all grade and grade 3-4 AEs in patients receiving SCT. There are no previous reports about AEs in R/M HNSCC patients receiving SCT. The OAK and CheckMate studies for NSCLC patients demonstrated no increase in AEs in patients who received SCT after ICIs (8-10), whereas our study identified anemia as the most frequent grade 3-4 AEs in R/M HNSCC patients receiving SCT. The percentage of AEs was higher in our study than that in the EXTREME study and another report by Hitt et al. (2, 18). This may be due to the increased number of patients who previously received chemotherapy and were in poor health condition. However, all adverse events were tolerable and manageable.
Kaplan–Meier curves for (A) OS and (B) PFS of SCT patients. OS: Overall survival; PFS: progression-free survival.
Our study has several limitations. First, because of the retrospective nature, we could not compare the efficacy of SCT with that of other treatments. Moreover, there were no previous reports for treatment with SCT after nivolumab in R/M HNSCC in 2017; however, we intend to follow-up this retrospective analysis with a prospective study to validate the study findings. Second, due to the small sample size, we could not perform a multivariate Cox regression analysis. We conducted this study at 4 institutions; there were only 25 patients receiving SCT because of delayed treatments for the rare cancer. Thus, there is a need for further studies with large patient cohorts to validate of our findings.
Univariate analyses of prognostic factors for OS and PFS.
In conclusion, our findings suggest that SCT after nivolumab treatment is associated with better clinical outcomes in patients with R/M HNSCC. The results suggest that better PS is associated with prolonged OS in patients receiving SCT. Furthermore, previous radiation therapy was related to a decreased ORR. These findings should aid in deciding the treatment after administration of ICIs. Moreover, our results would help in selecting patients for suitable SCT treatment.
Efficacy of SCT.
Univariate analyses of prognostic factor for the ORR.
Summary of adverse events (N=25) N (%).
Acknowledgements
The Authors thank Editage (www.editage.jp) for English language editing.
Footnotes
Authors' Contributions
IO, TM, and CF designed the study. TM, YT, KM, and CF contributed to the collection and interpretation of the data. CF, TO, HS, KT, TK, TY, KH, GO, and HT contributed to data collection and patient management. CF was a major contributor in writing the manuscript. All Authors read and approved the final manuscript.
Conflicts of Interest
The Authors declare that they have no competing interests related to this study.
- Received July 11, 2020.
- Revision received July 20, 2020.
- Accepted July 21, 2020.
- Copyright© 2020, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved