Understanding the impact of high-risk human papillomavirus on oropharyngeal squamous cell carcinomas in Taiwan: A retrospective cohort study

2020 
Background In North America and Western Europe, human papillomavirus (HPV)-driven oropharyngeal squamous cell carcinoma (OPSCC) has increased dramatically over the past 40 years, whereas HPV-negative OPSCC, typically associated with alcohol and tobacco as etiological factors, has declined. In Taiwan, the OPSCC rate is increasing; however, there is limited understanding of the role of HPV, as tobacco, alcohol, and betel quid use are still very prominent. Here we investigated the involvement of HPV and its prognostic implications for OPSCC in Taiwan. Methods and findings We studied a retrospective cohort of 541 OPSCCs undergoing care between 1998 and 2016 at the Chang Gung Memorial Hospital-Linkou, Taiwan. Clinical and risk exposure data were retrieved from hospital charts. Most cases were males (94%) and had concomitant (87%) exposure to alcohol (51%), tobacco (83%), and betel quid (65%). Formalin-fixed, paraffin-embedded tissue sections were immunostained for p16, a surrogate for active HPV, and DNA was tested for HPV detection and genotyping by Multiplex PCR-MassArray. Tumors with p16 and/or HPV DNA positivity were identified as HPV-positive. HPV status was assigned to 528 tumors. The prevalence of HPV-positive OPSCC was 28.4% (150/528), with a strong correlation between p16 and HPV DNA results (F < 0.0001). We observed an incremental trend for HPV-positive OPSCC over 18 years. HPV16 alone was found in 76.9%, and HPV58 in 7.5% tumors. Among males more tumors were HPV-negative than positive, but among the 34 females more tumors were HPV-positive (62% vs. 38%, p < 0.0001), more commonly from the tonsils (p < 0.01), and less prone to use tobacco, alcohol, and betel quid (p < 0.0001). For all patients, HPV-positive OPSCC was associated with higher age at diagnosis (55.5 vs. 52.7 years, p = 0.004), and lower T-stage (p = 0.02). HPV-positivity in OPSCC tumors was an independent predictor of better overall survival (OS) (hazard ratio [HR] 0.57 [95% CI 0.41-0.80], p = 0.0009), and disease-free survival (DFS) (HR 0.53 [95% CI 0.40-0.72], p < 0.0001) in multivariable estimations for up to 5-year post-diagnosis. Alcohol consumption was strongly associated with higher risk of recurrence and death (OS: HR 2.02 [95% CI 1.52-2.68], p < 0.0001; DFS: HR 1.70 [95% CI 1.31-2.21], p < 0.0001). No predictive associations were found for smoke or betel quid. Kaplan-Meier's OS and DFS were longer in HPV-positive cases (p < 0.0001), but this advantage decreased for alcohol users (p < 0.0001). Lastly, HPV-positive OPSCC had the best outcomes in non-drinkers, non-smokers, and non-betel quid chewers, but the prognostic benefit of HPV persisted in the presence of these risk factors. Conclusions Consistent with the increasing role of HPV on OPSCC globally, HPV is an important etiological factor in more than one-fourth of OPSCC cases from Chang Gung Hospital. Like in Western countries, HPV provides considerable independent survival benefits to OPSCC, but the added prognostic value is reduced by exposure to the risk factors, alcohol, smoking, and betel quid. There is a need to consider HPV as an etiologic factor in treatment and in cancer-prevention policies in Taiwan.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    93
    References
    0
    Citations
    NaN
    KQI
    []