Conditional cancer-specific survival analysis by HPV status in oropharyngeal squamous cell carcinoma.

Abstract

e18128 Background: HPV-positive (HPV+) oropharyngeal squamous cell carcinoma (OPSCC) is known to be associated with increased survival compared to HPV-negative (HPV-) OPSCC. However, conditional survival trends between these two subtypes have not been evaluated to date. Conditional survival provides useful prognostic information for patients who have survived following initial treatment, and may inform differential surveillance guidelines for HPV+ and HPV- OPSCC. We investigated and compared conditional survival of HPV+ OPSCC with HPV- OPSCC and relevant risk factors using a large national cancer database. Methods: We identified patients diagnosed with a single primary HPV+ and HPV- OPSCC confirmed by p16 immunotesting from 2018 to 2022 using the Surveillance, Epidemiology, and End Results (SEER) database. Demographic characteristics (race, sex, age, marital status, urbanicity), treatment modality (radiation, chemotherapy, surgery), and disease staging were collected. 4-year conditional cancer-specific survival (CSS) was calculated using competing risks analyses stratified by HPV status and disease staging. CSS hazard ratios (HR) were estimated. Results: Of 14,614 patients, 11,852 had HPV+ OPSCC and 2,762 had HPV- OPSCC. Compared to HPV- OPSCC patients, HPV+ OPSCC patients were slightly younger (mean age 61.8 vs. 62.6 years), more commonly male (87.6% vs. 77.1%), white (91.0% vs. 82.3%), married (63.3% vs. 51.3%), and more likely to present with regional disease (83.7% vs. 70.2%) rather than localized or distant disease. HPV-negative status was associated with significantly worse CSS (HR = 3.28, 95% CI = (2.99, 3.61), p<0.001) for OPSCC. Despite lower baseline survival, patients with HPV- disease and those with distant disease demonstrated larger gains in conditional CSS. Conditional CSS for localized HPV+ disease remained consistently high from 0 to 3 years post-diagnosis (96.4% to 99.7%) while localized HPV- disease demonstrated marked improvement over the same period (81.9% to 98.6%). HPV- OPSCC with distant disease at diagnosis showed the largest gain in 4-year CSS from 37.5% at diagnosis to 93.5% at 3 years after diagnosis. By 3 years post-diagnosis, conditional CSS converges across HPV status and disease stage. Conclusions: Conditional survival analysis shows that while HPV+ OPSCC has better baseline prognosis, cancer-specific survival of HPV- OPSCC patients who survive the first 3 years after diagnosis improves significantly and approaches that of HPV+ disease. However, in the initial period post-diagnosis, patients with HPV- OPSCC including those with localized disease exhibit significantly lower cancer-specific survival, suggesting that more intensive early surveillance strategies may be considered for this population. Further work is required to evaluate longer-term conditional survival patterns and additional risk factors for mortality in both HPV+ and HPV- OPSCC.