Neoadjuvant with low-dose radiotherapy, tislelizumab, albumin-bound paclitaxel, and cisplatin for resectable locally advanced head and neck squamous cell carcinoma: phase II single-arm trial

Abstract

Although pathological complete response (pCR) and major pathological response (MPR) rates of neoadjuvant immunotherapy combined with chemotherapy in head and neck squamous cell carcinoma (HNSCC) trials remain suboptimal, emerging evidence highlights the synergistic potential of combining low-dose radiotherapy with immunotherapy to promote the efficacy of immunotherapy. This phase II, open-label, single-arm, multicenter trial (NCT05343325) enrolled 28 patients with untreated stage III-IVB HNSCC (NeoRTPC02). Patients received neoadjuvant low-dose radiotherapy, the programmed death-1 (PD-1) inhibitor tislelizumab, albumin-bound paclitaxel, and cisplatin for two cycles, followed by radical resection ~4 weeks after treatment completion. The primary endpoint, pCR rate, was achieved in 14 of 23 patients (60.9%; 23/28, 82.1% of the total cohort underwent surgery). Secondary endpoints included MPR rate (21.7%, 5/23), R0 resection rate (100%), and objective response rate (64.3%; 18/28). Treatment-related adverse events were manageable, with grade 3 or 4 treatment-related adverse events occurring in 10 (35.7%) patients. No surgical delays were observed. Single-cell RNA sequencing revealed remodeling of the HNSCC tumor microenvironment, which may correlate with improved clinical outcomes. This trial met the pre-specified primary endpoint, demonstrating a high pCR rate with promising efficacy and manageable toxicity in locally advanced HNSCC.

Fig. 1. Clinical trial flow chart.

In the study, 37 patients were assessed for eligibility, 28 patients were enrolled and included, and 23 patients underwent radical surgical resection.

Fig. 2. Waterfall plot of characteristic of radiographic response (n = 28) and pathological response (n = 23) profiles.

Each column indicates one patient, ranking from highest to lowest rate of radiographic response. Corresponding sequence numbers of patients are labeled below. The dashed horizontal line denotes 30% radiographic response, indicating the cutoff points for PR and SD (Left). Each column indicates one patient. Two dashed horizontal lines denote 50% and 90% pathological response, indicating the cutoff points for NPR and PPR, PPR and MPR. Gray vertical axes represent patients who did not undergo surgery (Right). Source data are provided as a Source Data file.

Fig. 3. Swimming plot of overall survival for individual patients (n = 28).

Each bar indicates one patient. Source data are provided as a Source Data file. Source data are provided as a Source Data file.

Fig. 4. Magnetic resonance images (T1 enhanced sequence images) and H&E staining images (20×) of primary site tumor before and after two courses of neoadjuvant therapy.

a The number 10 patient with a pCR of pathological response and a CR of radiographic response. b The number 06 patient with a pCR of pathological response and a PR of radiographic response. c The number 02 patient with a pCR of pathological response and a SD of radiographic response. Scale bar = 100 μm. Source data are provided as a Source Data file.

Fig. 5. Biomarker analyzes of treatment and pathological.

a–c Bar plots to show the expression of CD4, CD8 and CD20 in tissues before and after neoadjuvant treatment (n = 46). Statistical significance was determined using two-sided paired t-tests. d Bar plots showing the pretreatment levels of CD20 in tumor tissues (n = 23). Data are presented as mean values ± standard error of the mean (SEM). Error bars represent the SEM. Statistical significance was determined using a two-sided Wilcoxon rank-sum test. e, f The CD20 expression in tissues before and after treatment of patients with pCR and PPR. Scale bar, 100 μm and 50 μm. g This representative figure shows multiplex immunofluorescence (CD20 in red, DAPI in blue) demonstrating the spatial distribution across pathological response groups post-treatment: pCR (left), MPR (middle), and PPR (right) (Scale bar=50 μm). Top: Pre-treatment. Bottom: Post-treatment. Source data are provided as a Source Data file.

Fig. 6. The immune cell landscape in HNSCC patients receiving neoadjuvant treatment with LDR combined with αPD-1 and chemotherapy.

a UMAP plot depicting all cells from HNSCC patients, with each cell type indicated by a different color. b UMAP plot showing the subpopulations of T lymphoid cells and NK cells from HNSCC patients, also color-coded by cell type. c Cell Type Infiltration Percentage Stacked Bar Chart shows the percentage distribution of CD8⁺ T cell subpopulations. d Violin plot that shows treatment group differences in MS4A1 expression within B cells. Violin plots depict the distribution of expression levels, with the top and bottom edges indicating the maximum and minimum values, respectively, white dots representing median values, and black bars showing the interquartile range (IQR, 25th to 75th percentiles). Statistical significance was determined using a two-sided Kruskal-Wallis test followed by Dunn’s post-hoc test (****p < 0.0001). e UMAP plot of myeloid cell subpopulations from HNSCC patients, color-coded by cell type. f Violin plot illustrating the functional characteristic scores across monocyte-macrophage subpopulations. Violin plots depict the distribution of expression levels, with top and bottom edges indicating the maximum (100th percentile) and minimum (0th percentile) values, respectively, white dots representing median values (50th percentile), and black bars showing the interquartile range (IQR, 25th to 75th percentiles). g Cell Type Infiltration Percentage Stacked Bar Chart shows the percentage distribution of myeloid cell subpopulations. h Kaplan-Meier curve that demonstrates the correlation between the infiltration of Macro_APOE and the overall survival (OS) of TCGA-HNSC patients (n = 500). Statistical significance was determined using a two-sided log-rank test, with Bonferroni correction (α = 0.05/5 = 0.01) for multiple comparisons.