Purpose: The EORTC-90111–24111 phase II window study evaluated afatinib versus no preoperative treatment in patients with primary squamous cell carcinoma of the head and neck (HNSCC). We investigated afatinib-induced tumor and microenvironment modifications by comparing pre- and posttreatment tumor biopsies. Patients and Methods: Thirty treatment-naïve patients with primary HNSCC were randomized. Twenty-five patients received afatinib for 14 days before surgery (40 mg 1×/day) and 5 patients were attributed to the control arm. Biopsies were taken at work-up and during surgery. Good quality RNA samples were used for omics analyses. The control arm was enlarged by samples coming from our previous similar window study. Results: IHC analyses of afatinib-treated tumor biopsies showed a decrease in pEGFR (P ≤ 0.05) and pERK (P ≤ 0.05); and an increase in CD3+ (P ≤ 0.01) and CD8+ (P ≤ 0.01) T-cell infiltration, and in CD3+ (P ≤ 0.05) T-cell density. RNA sequencing analyses of afatinib-treated tumor samples showed upregulation of inflammatory genes and increased expression scores of signatures predictive of response to programmed cell death protein 1 blockade (P ≤ 0.05). In posttreatment biopsies of afatinib-treated patients, two clusters were observed. Cluster 1 showed a higher expression of markers and gene sets implicated in epithelial-to-mesenchymal transition (EMT) and activation of cancer-associated fibroblasts (CAF) compared with cluster 2 and controls. Conclusions: Short-term treatment with afatinib in primary HNSCC induces CD3+ and CD8+ tumor infiltration and, in some patients, EMT and CAF activation. These results open perspectives to overcome resistance mechanisms to anti-HER therapy and to potentiate the activity of immune checkpoint inhibitors.

Afatinib-induced tumor and microenvironment modifications in head and neck squamous cell carcinoma were evaluated by spatial transcriptomics in surgical specimens and RNA-sequencing in tumor biopsies of patients included in the EORTC-90111-24111 window-of-opportunity study. The aim was to explore tumor evolution and composition under anti-HER therapy. Based on our previous investigations by RNA-seq on tumor biopsies, surgical slides of ID08 and ID15 from the epithelial-to-mesenchymal (EMT) cluster and ID30 from the non-EMT cluster were investigated with spatial transcriptomics. Dimension reduction in ID30 revealed 14 clusters, with clusters overlapping three tumor nodules and the stroma. Differential expression analysis between tumor nodules showed enrichment of the hallmark EMT genelist, with 123 genes in common between the analyses. These genes were involved in PDGF and MET signaling pathways. By comparing gene expression in paired tumor biopsies and the 123 genes from differential analyses obtained in ID30, a list of 13 genes involved in cancer pathways and EMT emerged, which were also highly expressed in ID08 and ID15. These results show a progressive apparition of genes implicated in EMT, MET, and PDGF pathways in tumors after afatinib. Notably, a list of 13 genes emerged which may contain targets to prevent tumor evolution after anti-HER therapy.

Lists of expressed hallmark genesets in GSEA