Neoadjuvant immunotherapy with or without chemotherapy in locally advanced oral squamous cell carcinoma: Randomized, two-arm, phase 2 trial

Abstract

Patients with locally advanced oral squamous cell carcinoma (OSCC) have poor outcomes with standard care. Neoadjuvant therapy is shown to be effective for these patients. In the randomized, two-arm, phase 2, non-comparative trial, we investigate the efficacy and safety of the neoadjuvant programmed cell death 1 (PD-1) inhibitor camrelizumab with or without docetaxel-cisplatin-5-fluorouracil (TPF) chemotherapy in patients with resectable locally advanced OSCC. Patients with stage III-IVA OSCC receive neoadjuvant therapy with three cycles of camrelizumab (arm Cam) with or without two cycles of TPF chemotherapy (arm Cam+TPF), followed by surgery and adjuvant therapy. Major pathological response (MPR) is achieved in both arm Cam (5/34, 14.7%) and arm Cam+TPF (26/34, 76.4%). With a median follow-up of 32 months, the 2-year event-free survival (EFS) rate of arm Cam and Cam+TPF is 52.9% and 91.2%, respectively. This work demonstrates feasibility and safety for immunochemotherapy in the neoadjuvant setting for OSCC. This study was registered at ClinicalTrials.gov (NCT04649476).

Keywords: PD-1 inhibitor; chemotherapy; neoadjuvant therapy; oral squamous cell carcinoma; pathological response.

Graphical abstract

Figure 1

Trial flow diagram A total of 68 patients were enrolled in this trial, with 34 patients completing the full treatment protocol in arm Cam, and 32 patients completing the full treatment protocol in arm Cam+TPF.

Figure 2

Patients’ tumor response in the subgroups (A) Major pathological response in a 53-year-old man (T3N2bM0) who received neoadjuvant camrelizumab mono-immunotherapy. Images of the left tongue (left), radiological images (middle), and pathological images (right) before (top) and after (bottom) neoadjuvant therapy were shown; scale bar, 100 μm; red line indicates the longest diameter of the tumor. (B) Waterfall plot showing patients’ pathological response and radiological response in arm Cam. (C) Major pathological response in a 56-year-old man (T3N0M0) who received neoadjuvant camrelizumab immunotherapy plus TPF chemotherapy. Images of the right tongue (left), radiological images (middle), and pathological images (right) before (top) and after (bottom) neoadjuvant therapy were shown; scale bar, 100 μm; red line indicates the longest diameter of the tumor. (D) Waterfall plot showing patients’ pathological response and radiological response in arm Cam+TPF. (E) Percentage of radiological response rate in resected tumor specimens. Median percentage: arm Cam −11.5% (range −100% to 32%) and arm Cam+TPF 31% (range −41% to 55%). (F) Percentage of pathological response rate in resected tumor specimens. Median percentage: arm Cam 1.22% (range 0%–99%) and arm Cam+TPF 95% (range 0%–100%). (G) Proportion of patients with radiological response in the subgroups. (H) Proportion of patients with pathological response in the subgroups. CPS, combined positive score; PR, partial response; SD, stable disease; PD, progressive disease; NA, not applicable; pCR, pathological complete response; MPR, major pathological response; pPR, pathological partial response; pNR, pathological non-response.

Figure 3

Follow-up and survival outcomes of patients (A) Swimmer plot showing patients’ survival outcomes. (B) Kaplan-Meier curve of overall survival for the patients in arm Cam and arm Cam+TPF. (C) Kaplan-Meier curve of event-free survival for the patients in arm Cam and arm Cam+TPF. (D) Kaplan-Meier curve of event-free survival for the pathological responders (pPR+MPR+pCR) and non-responders (pNR). Log rank p value was used to estimate the significance among groups. (E) Kaplan-Meier curve of event-free survival for the pathological responders (pPR+MPR) and non-responders (pNR) in arm Cam. Log rank p value was used to estimate the significance among groups. (F) Kaplan-Meier curve of event-free survival for the pathological responders (pPR+MPR+pCR) and non-responders (pNR) in arm Cam+TPF. Log rank p value was used to estimate the significance among groups. (G) Proportion of patients who had progressive events in different subgroups. p value was determined by two-sided Fisher’s exact test. PR, partial response; SD, stable disease; PD, progressive disease; pCR, pathological complete response; MPR, major pathological response; pPR, pathological partial response; pNR, pathological non-response.

Figure 4

Worse prognosis of non-responders in arm Cam+TPF (A) Spatial heatmaps showing expression intensity of EGFR, including Cam+TPF-pNR group (n = 2) and Cam-pNR group (n = 2); scale bar, 5 mm. (B) Dot plot showing expression levels of genes related to EGFR signaling pathway, including Cam+TPF-pNR group (n = 2) and Cam-pNR group (n = 2). (C) Heatmap showing the quantitative immunohistochemical results of post-therapy EGFR and downstream proteins in non-responders in arm Cam+TPF compared with arm Cam, including Cam+TPF-pNR group (n = 2) and Cam-pNR group (n = 7). (D) Follow-up of the two non-responders in arm Cam+TPF. Red circle indicates primary tumor, red arrow indicates tumor metastasis. EGFR, epidermal growth factor receptor; CT, computed tomography; MRI, magnetic resonance imaging; PET-CT, positron emission tomography and computed tomography.