Neoadjuvant presurgical PD-1 inhibition in oral cavity squamous cell carcinoma

Abstract

Oral cavity squamous cell carcinoma (OCSCC) is a prevalent surgically treated subset of head and neck cancer with frequent recurrence and poor survival. Immunotherapy has demonstrated efficacy in recurrent/metastatic head and neck cancer. However, whether antitumor responses could be fostered by neoadjuvant presurgical immunotherapy remains unclear. Using a Simon's two-stage design, we present results of a single-arm phase-II trial where 12 patients with stage II-IVA OCSCC received 3 to 4 biweekly doses of 3 mg/kg nivolumab followed by definitive surgical resection with curative intent. Presurgical nivolumab therapy in this cohort shows an overall response rate of 33% (n = 4 patients; 95% CI: 12%-53%). With a median follow up of 2.23 years, 10 out of 12 treated patients remain alive. Neoadjuvant nivolumab is safe, well-tolerated, and is not associated with delays in definitive surgical treatment in this study. This work demonstrates feasibility and safety for incorporation of nivolumab in the neoadjuvant setting for OCSCC (ClinicalTrials.gov: NCT03021993).

Keywords: PD-1 therapy; head and neck cancer; neoadjuvant; neoadjuvant immunotherapy; nivolumab; oral cavity cancer.

Conflict of interest statement

C.M.P. is the co-founder of Ares Immunotherapy. R.S.L. received research or clinical trial support from Merck, Pfizer, BMS, and OncoSec. D.M.N. received research or clinical trial support from BMS. M.R.I.Y. received research or clinical trial support Merck and BMS.

© 2021 The Author(s).

Figures

Graphical abstract

Figure 1

Trial schema Patients with surgically resectable, locoregionally advanced oral cavity squamous cell carcinoma (OCSCC) stage II-IVA underwent baseline clinical and radiographic imaging and then received 3 mg/kg nivolumab every 2 weeks for 3 doses (green). Repeat post-nivolumab clinical and radiographic assessment between days 30 and 35 was then performed. If there was disease progression (red) by RECIST 1.1 criteria, patients proceeded directly to surgery between days 36 and 42. Conversely, if there was a response or stable disease (blue), patients received a single 4th dose of nivolumab 3 mg/kg on day 43 before proceeding to surgery on day 50 to 56. Pathologic response was determined by comparison of tumor size on final pathologic evaluation compared to tumor size on baseline radiographic imaging.

Figure 2

OCSCC response to neoadjuvant presurgical nivolumab Waterfall plot of pathologic response (percentage change from baseline imaging to pathologic evaluation) to neoadjuvant nivolumab (33% response rate). Dashed lines indicate RECIST 1.1 cutoffs for progression (>20% change, red bars), stable disease (+20% to −30% change, gray bars), and response (>30% reduction, blue bars). There were no significant differences in baseline characteristics between responders and non-responders (see also Table 1). ECOG, Eastern Cooperative Oncology Group; AJCC, American Joint Committee on Cancer Clinical Stage; LRR, locoregional recurrence; DM, distant metastasis.

Figure 3

Overall survival and disease-free survival (DFS) in OCSCC patients treated with neoadjuvant nivolumab (A and B) Overall survival for (A) all patients and (B) by response category displayed. Median overall survival for all patients, responders, or patients with stable disease was not reached, while median OS for progressors was 1.17 years. (C and D) DFS for (C) all patients or (D) by response category. Median overall DFS was 3.26 years, while DFS for responders was 1.9 years, for progressors was 2.0 years, and was unreached for patients with stable disease. n = 12 patients; n = 4 with partial response, 4 with stable disease, and 4 with progressive disease. Tick marks indicate censored data.

Figure 4

Clinical, radiographic, and pathologic features of response and progression after neoadjuvant nivolumab in OCSCC (A) Patient 6 had a right gingivobuccal sulcus lesion, which measured 4.3 cm on initial imaging (left), This tumor decreased in size visually (upper right) and radiographically (to 3.6 cm; lower right) on interval evaluation following nivolumab. The lesion had decreased to 3.0 cm at surgical resection, consistent with partial response. Hematoxylin and eosin (H&E) stain on surgical pathology indicating invasive squamous carcinoma with marked acute and eosinophilic inflammatory response. 20× magnification for the top image, scale bar, 50 μM; 40× magnification for the bottom image, scale bar, 20 μM. (B) Patient 3 had a right floor of mouth/alveolar ridge lesion measuring 3.2 cm on initial imaging (left) which increased to 3.7 cm on interval evaluation following nivolumab (right) and was 4.2 cm at surgical resection, consistent with progression. H&E stain on surgical pathology demonstrates tentacular stands of the squamous cell carcinoma with deficient inflammation. 20× magnification for the top image, scale bar, 50 μM; 40× magnification for the bottom image, scale bar, 20 μM. UL, upper lip; LL, lower lip; OT, oral tongue; TD, tongue depressor; AR, alveolar ridge. Arrow indicates tumor.

Figure 5

Immunological features of responsive and progressive tumors after neoadjuvant nivolumab (A and B) Surgical specimen from patient 9, who experienced the greatest reduction in tumor size post nivolumab, was evaluated to characterize immune infiltrate by Vectra Polaris. Focus on cytokeratin regions revealed limited PD-L1 expression, with high levels of CD8+ and CD4+ but low FoxP3+ infiltrates. (A) 10×, ruler, 100 μM and (B) 20×, ruler, 50 μM. (C and D) Surgical specimen from patient 4, who represents the most rapid progressor, was analyzed for immune infiltration. Multiplexed imaging indicated robust PD-L1 staining across 50% of the pathologic section corresponding with cytokeratin+ cells, with higher FoxP3+ infiltrates and evident immune exclusion from tumor areas relative to non-tumor regions. (C) 4×, ruler, 200 μM and (D) 20×, ruler, 50 μM. Magnification on Phenochart analysis software.