Advancing Education, Research, and Quality of Care for the Head and Neck oncology patient.
Background: HPV-negative (HPV-) cancers account for approximately 75-80% of all head and neck cancers (HNC) in the USA. HPV- HNC are associated with poor outcomes and high rate of recurrence. The detection of molecular residual disease (MRD) may help personalize treatments, detect relapse earlier, determine response to treatment, all of which can lead to improved clinical outcomes. Recently, we published a clinical validation of a tissue-agnostic, genome-wide methylome enrichment MRD assay for HNC (Liu, Annals of Oncology, 2024). In this analysis, we report the performance of the assay in the participants with HPV-HNC.
Methods: We evaluated the MRD detection test performance in patients with HPV- HNC (n=73) including HPV- oropharynx cancer (OPC) (n=10), HPV status unavailable but morphology representing HPV-OPC (n=1), and other HNC sites with known low HPV prevalence hence untested for HPV status (n=62). Blood collection time points include at diagnosis, and approximately 3, 12, and 24 months after curative intent treatment. All samples were analyzed with a bisulfite-free, non-degradative genome-wide DNA methylation enrichment platform using 5-10 ng of cfDNA isolated from plasma. The sensitivity and specificity of the MRD detection test to detect recurrence in all HPV- patients and subsets including early, late-stage, locoregional and distant recurrences were analyzed. The Kaplan-Meier method was applied to compare recurrence-free survival (RFS) between the MRD-detected and MRD-not-detected groups, and statistical significance was evaluated using a two-sided log-rank test. A P-value of less than 0.05 was considered statistically significant. The hazard ratio was calculated using the Cox proportional hazards model. Fisher’s exact test was performed to compare test performance across subgroup strata.
Results: With a median follow up of 5.1 years, the MRD detection test had 92% sensitivity and 91% specificity for detection of recurrence in surveillance monitoring. Test performance did not vary across subgroups. Sensitivity was not statistically significantly different between early- and late-stage patients (P>0.99) or between patients with locoregional and distant recurrence (P=0.49). Specificity remained the same for locoregional and distant recurrence patients and did not differ significantly across early- and late-stage patients (P = 0.62). Patients with MRD detection showed significantly worse 5-year RFS (15% vs. 96%, hazard ratio 42.3 [9.8, 182.3], P < 0.0001). The significantly poor RFS persisted across all subgroups with MRD detection. The lead time between MRD positivity and clinical recurrence was up to 14.9 months, with a mean lead time of 4.2 months.
Conclusions: This analysis demonstrates the robustness of using a blood-based, tissue-agnostic genome-wide methylome enrichment platform for ctDNA quantification and prognostic prediction in patients with HPV- HNC. The MRD detection test showed high sensitivity for identifying recurrence across early and late-stage patients, as well as locoregional and distant recurrence locations, highlighting the broad applicability. Given the poor outcomes typically seen in HPV-HNC patients, the ability to detect molecular residual disease is critical for informing treatment decisions and improving clinical outcomes, as early identification of MRD may enable timely therapeutic intervention to prevent recurrence.