Dynamic Changes in Circulating Tumor DNA During Immunotherapy for Head and Neck Cancer: SHIZUKU-HN Study

Abstract

Immune checkpoint inhibitors (ICIs) are effective in treating recurrent/metastatic head and neck squamous cell carcinoma (HNSCC), but only 20% of patients achieve durable responses. This study evaluated circulating tumor DNA (ctDNA) as a real-time biomarker for monitoring treatment response in HNSCC. The SHIZUKU-HN study prospectively collected and analyzed serial plasma samples (n = 27) from HNSCC patients undergoing ICIs, using Guardant360 to assess ctDNA variant allele frequency (VAF) and genetic mutations. Tumor volumes were quantified using 3D reconstruction of CT scans, and data from Japan's C-CAT database (n = 2255) provided insights into ctDNA testing in HNSCC. C-CAT data showed that ctDNA testing was underutilized, performed in only 7% of head and neck cancer cases. In SHIZUKU-HN, mean VAF significantly correlated with tumor volume (Spearman's ρ = 0.70, p = 0.001), often preceding radiographic progression. BRAF and APC mutations disappeared in partial responders, while GNAS mutations varied. EGFR and PIK3CA amplifications, detectable via ctDNA but missed in tissue biopsies, indicated emerging resistance mechanisms. The SHIZUKU-HN study demonstrates the potential of ctDNA as a dynamic biomarker in HNSCC, offering early insights into treatment efficacy and informing personalized ICI therapy.

Keywords: Center for Cancer Genomics and Advanced Therapeutics (C-CAT); circulating tumor DNA (ctDNA); dynamic monitoring; head and neck squamous cell carcinoma (HNSCC); immune checkpoint inhibitor (ICI); liquid biopsy; variant allele frequency (VAF).

Figure 1

(a) Distribution of panel tests based on tissue (blue) and blood (red) samples in the C-CAT database. (b) Representation of the head and neck squamous cell carcinoma (HNSCC) cohort within the C-CAT database. (c) Frequently detected gene variants in tissue and blood samples from HNSCC cases. CGP, comprehensive genomic profiling; C-CAT, Cancer Center for Advanced Cancer Genome Therapy; HNC, head and neck cancer; HNSCC, head and neck squamous cell carcinoma.

Figure 2

Correlation Between ctDNA and tumor volume, and comparison with tissue-based analysis. (a) Correlation between tumor volume (cm³), as determined by CT volumetric analysis, and mean VAF across 27 serial plasma samples from 10 patients. Pathogenic VAF (pVAF) represents VAF for pathogenic mutations, while all VAF (aVAF) includes all somatic mutations, encompassing variants of unknown significance (VUS) and synonymous mutations. (b) CtDNA dynamics for each gene variant, alongside the corresponding radiographic response, illustrated for a progressive disease case (HN05). Asterisks on CT images indicate target lesions. (c) Spider plot depicting mean VAF dynamics during ICI treatment, with color-coded radiographic responses: PD (red), SD (green), and PR (yellow). VAF, variant allele frequency; VUS, variants of unknown significance.

Figure 3

Study design and sequencing outcomes at our institution. (a) Detected plasma variants through next-generation sequencing (NGS), color-coded by ICI response, sample collection point, and type of genetic alteration. (b,c) Dynamic ctDNA patterns for each gene variant and the corresponding radiographic response, highlighting HN01 and HN12, which demonstrated partial response (PR). HN03 showed progressive disease (PD). Asterisks on CT images indicate target lesions. HN01 is shown as a PET/CT image to clarify tumor lesions. R/M HNSCC, recurrent/metastatic head and neck squamous cell carcinoma; NGS, next-generation sequencing.

Figure 4

Timing of plasma collection and body tumor volume estimation. (a) Patients monitored by ctDNA analysis at our institution; 12 patients with recurrent/metastatic head and neck squamous cell carcinoma (R/M HNSCC) who received immune checkpoint inhibitors (ICIs) were enrolled; 2 patients were excluded due to inadequate baseline samples. (b) Plasma ctDNA analysis using Guardant360. Plasma samples were obtained before ICI infusion (baseline), at 4 weeks, 6 months, or at progression of disease (PD). (c) Tumor volume measurement using computed tomography (CT) volume analysis on an image analysis workstation (syngo.via). Tumor contours were set in the axial and sagittal planes as shown. ctDNA, circulating tumor DNA; ICIs, immune checkpoint inhibitors; CT, computed tomography.