Comparative analysis of human papilloma virus detection methods and prognostic significance of transcriptionally active viral load in oropharyngeal squamous cell carcinoma

Background

Human papilloma virus (HPV)-related head and neck squamous cell carcinoma (HNSCC) incidence is rising. Despite the prognostic and therapeutic implication of HPV status, standardised detection methods are lacking. This study compared multiple HPV detection methods and evaluated the prognostic value of viral transcriptional activity in archival HNSCC tissues.

Methods

HPV-DNA status and subtyping was determined by GP5+/GP6 + PCR. Multiplex immunohistochemistry (p16INK4) and chromogenic in-situ hybridization (CISH for E6/E7 mRNA) was performed on tissue microarrays (TMAs) from a retrospective cohort of HNSCC (n = 207). Viral load was quantified using QuPath and an independent deep-learning HPV prediction algorithm was validated on H&E-stained tissue cores.

Results

The multiplex approach accurately detected active HPV in HNSCC primary tumours and metastases, including archival tissue > 20 years old. HPV-DNA was detected exclusively in oropharyngeal SCC and was absent in all other head and neck subsites. Receiver operating curve analysis confirmed all three tested methods match the diagnostic performance of the gold standard HPV-DNA PCR. Furthermore, viral load was identified as an independent predictor of overall survival, maintaining prognostic significance even after adjusting for age, stage and lifestyle factors.

Conclusions

This study underscores the importance of detecting transcriptionally active HPV to better clinical decision-making of HNSCC patients and supports broader adoption of CISH alongside emerging AI-based analysis in clinical practice.