Subsequent chemotherapy with paclitaxel plus cetuximab-based chemotherapy following immune checkpoint inhibitor in recurrent or metastatic squamous cell carcinoma of the head and neck

Background Immune checkpoint inhibitors (ICIs) are essential in treating recurrent/metastatic squamous cell carcinoma of the head and neck (R/M SCCHN). However, the overall response rate (ORR) is limited to 10-20%, and subsequent chemotherapy is critical to maximizing the subjects’ prognosis. Methods We retrospectively reviewed 59 patients with R/M SCCHN treated with paclitaxel+cetuximab (PE)-based chemotherapy (PCE, paclitaxel+carboplatin+cetuximab; or PTX+Cmab, paclitaxel+cetuximab) following disease progression after either pembrolizumab or nivolumab monotherapy. Results Of 59 patients, 15 were treated with pembrolizumab, with an ORR of 13.3%, and the remaining 44 with nivolumab, with an ORR of 11.4%. All patients in the pembrolizumab cohort had platinum-sensitive disease. Following ICI treatment, 19 patients were treated with PCE and the remaining 40 with PTX+Cmab. PE-based chemotherapy induced favorable and prompt tumor shrinkage even in cases where ICI was not effective, with a median change in the summed dimensions of target lesions of -43.4%, resulting in an ORR of 62.7%. Median time to response was 1.8 months. The patients in the pembrolizumab cohort appeared to have a numerically higher response rate than those receiving nivolumab (80.0% vs. 56.8%). For the 59 patients, progression-free survival and overall survival, calculated from the initiation of PE-based chemotherapy, were 4.6 months and 17.1 months, respectively. Grade ≥3 adverse events occurred in 40.7%, and no treatment-related death was observed. Conclusion PE-based chemotherapy following ICI is encouraging for its robust antitumor efficacy in R/M SCCHN.


Introduction
The treatment of recurrent/metastatic squamous cell carcinoma of the head and neck (R/M SCCHN) was significantly advanced with introduction of the anti-EGFR antibody cetuximab (Cmab) in 2005 (1), and further improved with the advent of immune checkpoint inhibitors (ICIs) (2,3).Among these, the anti-programmed cell death protein 1 (PD-1) antibodies pembrolizumab and nivolumab have become the standard systemic therapy for R/M SCCHN.Although these agents provide a durable response and favorable safety profile, overall response rates (ORR) to them as monotherapy are generally modest (around 10-20%) (2,3).Moreover, the KEYNOTE-048 study revealed a reverse relationship between relatively short progression-free survival (PFS) and longer PFS2 or overall survival (OS) with pembrolizumab monotherapy, indicating the importance of subsequent therapy following ICI (4).Several plausible explanations for the robust efficacy of subsequent therapy have been offered, including the additive or synergistic immunologic interplay between the two, based on the long half-life of ICI as prior therapy (5).To date, however, no standardized regimen following ICI has yet been established.
Against this background, taxane-containing regimens are now considered the most promisi ng therapeutic options.Conventionally, paclitaxel (PTX) plus Cmab-based (PE-based) chemotherapy, such as the combination of PTX, carboplatin (CBDCA), and Cmab (PCE), as well as the combination of PTX and Cmab (PTX+Cmab), have played a critical role in R/M SCCHN (6,7).Indeed, these combinations are reported to show robust efficacy (8-10)-primarily PTX+Cmab after nivolumab-probably due to the augmentation of anti-tumor immunity by both agents (11,12).Further, the recent subgroup analysis of the KEYNOTE-048 study revealed that, in a post-pembrolizumab monotherapy setting, non-taxane-containing 2nd line therapy might be associated with unsatisfactory PFS2 compared with taxane-containing therapy, again indicating a positive interaction between PE-based chemotherapy and ICI.However, no study has specifically focused on regimens following 1st line pembrolizumab monotherapy.
Here, we report the overall efficacy and safety of PE-based chemotherapy following ICI therapy.We then drill down into regimens by preceding ICI treatment line and discuss their expected roles as a component of sequential therapy in this patient population.

Patient
We retrospectively reviewed the patients who had received PEbased chemotherapy (PCE or PTX+Cmab) following ICI (pembrolizumab or nivolumab) monotherapy at National Cancer Center East between May 2017 and April 2022.Other inclusion criteria included (1) histologically proven SCC; (2) primary tumor location in the larynx, oral cavity, oropharynx, or hypopharynx; and (3) tumor recurrence or initial distant metastasis.To extract patients with these conditions, we used a computer-managed search system based on the prescribed regimens, and we then collected their clinical data from each medical record.Platinumsensitive was defined as disease progression or recurrence after 6 months from the end of platinum chemotherapy, and platinumrefractory was defined as occurring within 6 months.

Treatment
For pembrolizumab monotherapy, all patients were treated with a three-week cycle of pembrolizumab 200 mg/m2 on day 1.Nivolumab was administered to patients at 240 mg/m2 on day 1 in a two-week cycle.
For the PCE regimen, patients were treated with a three-week cycle of PTX with 100 mg/m2 and CBDCA with the area under the concentration-time curve (AUC) of 2.5 on days 1 and 8.For the PTX+Cmab regimen, patients were treated with a four-week cycle of PTX 80 mg/m2 on days 1, 8, and 15 of each cycle.Cmab was given in an initial dose of 400 mg/m2 followed by 250 mg/m2 weekly until disease progression or unacceptable toxicities in both regimens.In the PCE regimen, patients were treated with Cmab alone as maintenance therapy after six cycles of the PCE combination phase.Dose modification was allowed in accordance with the patient's systematic status and treatment-related adverse events.In the PTX+Cmab regimen, patients were treated with PTX until disease progression or unacceptable toxicities occurred.Computed tomography (CT) or magnetic resonance imaging (MRI) to evaluate treatment efficacy was carried out approximately every 1-3 months, or sooner if the physician in charge deemed it necessary.

Evaluation of efficacy and statistical analysis
Clinical outcomes of the PE-based chemotherapy are presented as overall response rate (ORR), disease control rate (DCR), time to response (TTP), progression-free survival (PFS), overall survival (OS), and adverse events (AEs).Clinical tumor response was evaluated according to the Response Evaluation Criteria in Solid Tumors (RECIST) ver.1.1 guidelines, including complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD).ORR and DCR were defined as a proportion of CR and PR, and of CR, PR, and SD, respectively.AEs were evaluated by CTCAE ver.5.0.Time to response was measured from the first day of PE-based chemotherapy administration until the first date of evaluation by imaging in the patients who appear PR or CR by PEbased chemotherapy.PFS was measured from the first day of PEbased chemotherapy administration until the date of tumor progression or death from any cause.OS was calculated from the first day of PE-based chemotherapy to the date of death from any cause.Further, PFS2 were measured from the first day of ICI monotherapy administration until the date of tumor progression by subsequent chemotherapy, herein PCE or PTX+Cmab, or death from any cause.Moreover, time from the first day of ICI monotherapy administration until date of death from any cause was defined as OSici.The Kaplan-Meier method was used to assess these prognostic data.Univariate analysis was performed using log-rank test.All analyses were performed using SPSS ver.28 (IBM Corp., Armonk, NY, USA) with a data cut-off of August 31, 2022.

Patient characteristics
Fifty-nine patients (pembrolizumab as prior ICI in fifteen patients and nivolumab as prior ICI in forty-four patients) treated with PE-based chemotherapy following ICI monotherapy were identified in this study (Supplementary Figure 1).Baseline characteristics of the entire population are presented in Table 1.Most patients were male (78.0%), and median age was 63 years.The most common primary tumor site was the oral cavity (37.3%), followed by the oropharynx (32.2%).The most common disease distribution at the initiation of PE-based chemotherapy was local/ regional metastasis only (42.4%), and platinum-refractory disease was observed in 54.2%.Nineteen patients (32.2%) were treated with PCE and the remaining forty (67.8%) with PTX+Cmab.Further, the different backgrounds of the pembrolizumab and nivolumab cohorts are shown in Supplementary Table 1.All patients in the pembrolizumab cohort harbored platinum-sensitive and PD-L1positive (≥1 in CPS) disease, versus only the minority of patients in the nivolumab cohort (100% vs. 27.3%,p<0.01).PCE tended to be selected for patients who were treated with pembrolizumab as prior therapy, compared with nivolumab (73.3% vs. 32.2%,p<0.01).

PE-based chemotherapy following ICI
Regarding the efficacy of PE-based chemotherapy in the entire population, CR, PR, SD, and PD were obtained in 3 (5.1%),33 (57.6%), 7 (11.9%),and 13 (22.0%)patients, respectively, with an ORR and DCR of 62.7% and 74.6% (Table 2).The median change in the summed dimensions of target lesions was -43.4% (range: 65.9% to -100%), and the ORRs of PCE and PTX+Cmab were very similar (63.2% in PCE and 62.5% in PE) (Figure 1A and Table 2).In addition, tumor response was rapid, evidenced by a median time to response of 1.8 months (range: 0.5-4.3)(Figure 1B).A large majority of responders to the paclitaxel plus cetuximab-based chemotherapy had no response to immune checkpoint inhibitor monotherapy: of the 38 patients classified with disease progression following immune checkpoint inhibitor, 25 responded to the PEbased chemotherapy (Supplementary Figure 2).Also, the duration of ICI treatment success did not appear to be associated with PEbased chemotherapy duration: even patients who experienced rapid tumor progression in the ICI phase achieved long disease control with PE-based chemotherapy following failure (Figure 2).In the PE-based chemotherapy phase, median length of hospitalization throughout the treatment period was 0 days (range: 0-33), and 44 patients (74.6%) received all treatment with the regimen on an outpatient basis.Fifty patients experienced disease progression on PE-based chemotherapy as of data-cutoff, and 41 (82.0%) switched to subsequent treatment (Supplementary Figure 3).With a median follow-up of 12.4 months from the initiation of PE-based chemotherapy, PFS and OS were 4.6 months (95%CI: 3.3-5.9)and 17.1 months (95%CI: 12.1-22.1),respectively (Figures 3A, B).
These results clearly indicate a significant advantage in PFS and OS (p < 0.01, both) when compared to the outcomes observed with other regimens besides PE-based chemotherapy (n = 55) for subsequent chemotherapy following ICI treatment during the same observation period (Supplementary Figure 4).In addition, from the initiation of ICI, PFS2 and OSici were 10.3 months (95% CI: 7.8-13.8)and 23.3 months (95%CI: 17.1-34.6),respectively (Figures 3C, D).
In Supplementary Figure 7 and Supplementary Table 2, We focused on the platinum sensitivity and compared the efficacy of PE-based chemotherapy between platinum sensitive and platinum refractory patients.All platinum refractory patients were treated with nivolumab as prior immune therapy and none were treated with PCE regimen as subsequent chemotherapy after ICI.There was a difference between platinum sensitive and refractory patients in the number of previous lines of systemic therapy before ICI for R/M SCCHN, however no difference was founf in PFS, OS, PFS2, and OSici.Acute toxicities experienced during the PE-based chemotherapy are listed in Table 4. Grade 3 or worse AEs were observed in 24 patients (40.7%), of which leukopenia (11.7%), neutropenia (6.9%), mucositis (6.8%), and pneumonia (6.8%) were most common.There was no treatment-related death throughout treatment.
In Table 5, the univariate analysis showed that number of lines of ICI were prognostic factors for PFS.OS did not differ significantly but tended to be worse later in the line of ICI administration.In addition, univariate analysis showed a significant difference between PS 0,1 and PS 2. On the other hand, there was no significant difference in PFS and the results were almost equal, suggesting that PE-based regimen can be expected to have a good therapeutic outcome even if PS is not favorable.
Figure 4 shows representative images from a patient with metastatic oral gingival squamous cell carcinoma who had a profound and durable tumor response by PCE.The patient experienced multiple subcutaneous and retroperitoneal metastases after initial surgery with marginal mandibulectomy and bilateral neck dissection (pT4aN2c, 16 months before recurrence) and additional neck dissection with adjuvant radiotherapy for late cervical lymph node recurrence (one month before recurrence).Pembrolizumab monotherapy was initiated as CPS in tissue newly obtained from a subcutaneous metastasis was around 50, and the patient had few tumor-related symptoms.However, the disease progressed within one week after the first administration, and subcutaneous metastasis on the hip ruptured and caused pain (Figure 4A).We immediately switched the patient to PCE and found that all recurrent lesions had macroscopically shrunk within a few days.Further, PR by RECIST criteria was identified by a CT scan taken on day 14 after PCE initiation (Figure 4B).Eventually, the patient obtained near-CR, and 25 months later is currently being treated with Cmab maintenance therapy.Further, pathological complete response of the buttock tumor with the presence of multinucleated giant cells was proven in surgery for wound closure (Figure 4C).

Discussion
This study focused on PE-based chemotherapy as subsequent treatment following ICI in R/M SCCHN.Results showed a rapid (median time to response: 1.8 months) and favorable tumor response (ORR: 62.7%) irrespective of the efficacy of ICI, with a manageable safety profile.This study is the first detailed report of clinical outcomes with the therapeutic sequence of pembrolizumab followed by PCE in these patients.
Beginning even before the emergence of immunotherapy in clinical practice, PE-based chemotherapy was recognized as a treatment option for R/M SCCHN with encouraging survival outcomes as well as favorable tumor shrinkage.As in platinum-refractory cases, PE regimens have been widely used, and ORRs of 34-54% have been reported in patients without prior ICI therapy (6,(13)(14)(15).In contrast, several retrospective studies reported the more robust antitumor efficacy of PE-based chemotherapy (ORR; 41-70%) following ICI treatment (8,10), probably due to favorable interaction between ICI and the regimen.In our study, the ORR with PE-based chemotherapy was also favorable, reaching 62.7%.In addition, two Phase II trials of ICI plus Cmab in R/M SCCHN have been reported, with response rates of 37% for pembrolizumab plus Cmab and 45% for pembrolizumab plus Cmab (16,17).Considering the high response rate in this study, the combination of paclitaxel can be expected to add approximately 20% to the response rate.Notably, the favorable ORR values on combination of taxane plus Cmab in these studies, including our present study (56.8%), might obviate the need for clinical trials to evaluate the significance of adding Cmab to taxane, particularly as taxane plus Cmab is a widely accepted therapeutic option in this setting (11,12).In this study, survival from the start of ICI was also very long (OSici: 23.3 months).This result is almost equivalent to the median OS of 21.9 months in the TPExtreme study, in which patients received cetuximab, docetaxel, and cisplatin followed by subsequent chemotherapy (18).On the other hand, the present results were longer than those of the median OS (14.9 months) in patients who received chemotherapy without ICI as a subsequent chemotherapy.These results suggest that immune checkpoint inhibitor therapy at any stage of the treatment sequence is important in the treatment of R/M SCCHN.
The PCE regimen has shown reliable efficacy primarily in platinum-sensitive settings.For example, the CSPOR HN02 study showed an ORR of 40.0% and median OS of 14.7 months in patients without prior immunotherapy.To date, however, the actual clinical efficacy of the PCE regimen following prior ICI therapy has not been reported in detail; accordingly, our present study is the first to provide reference data for clinically significant variables.Given that the KEYNOTE-048 study resulted in pembrolizumab monotherapy becoming a standard of care for patients with PD-L1-positive disease, and that subsequent taxanecontaining therapy might maximize prognosis, we believe that PCE is appropriate for subsequent chemotherapy following pembrolizumab.Numerically speaking, the high tumor response (ORR of 80.0%, and 86.6% achieved at least a 30% reduction in the summed dimensions of target lesions, as shown in Figure 1B) and rapid tumor shrinkage presents a reliable strategy for overcoming   was not possible because only patients with disease progression on ICI were included in the study.Second, in the pembrolizumab cohort, no conclusive finding was obtained on the validity of choosing the PEregimen over other Cmab-based chemotherapy (i.e., EXTREME regimen).Larger, background-matched comparisons or prospective trials may help answer this question.

Conclusion
PE-based chemotherapy provides favorable and rapid antitumor efficacy in patients with R/M SCCHN as subsequent chemotherapy following ICI.It is therefore an encouraging treatment option in patients with R/M SCCHN who fail or do not respond to ICI Moreover, the high tolerability and outpatient availability of this regimen should greatly benefit this patient population.
The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers.Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

TABLE 1
Patient characteristics.

TABLE 1 Continued
CharacteristicNumber of patients, (%) ECOG, Eastern Cooperative Oncology Group; CI, confidence interval; CPS, combined positive score; ORR, overall response rate; PFS, progression-free survival; ICI, immune check point inhibitor; R/M HNSCC, recurrent/metastatic squamous cell carcinoma of the head and neck; PTX+Cmab, paclitaxel and cetuximab; PCE, paclitaxel+carboplatin +cetuximab.Platinum-sensitive was defined as disease progression or recurrence after 6 months from the end of platinum chemotherapy, and platinum-refractory was defined as occurring within 6 months.;ORR, proportion of CR+PR.

TABLE 2
Clinical response by PE-based chemotherapy.

TABLE 3
Clinical response by PE-based chemotherapy according to prior ICI therapy.

TABLE 4
Selected toxicity during PE-based chemotherapy.The study has several limitations.First, although we found that subsequent PE-based chemotherapy proved to be effective even in cases where ICI was not effective, prognostic analysis related to ICI treatment