Human Papilloma Virus (HPV) is a well-studied risk-factor for the development of oropharyngeal squamous cell carcinoma (OPSCC). Although the HPV(+) OPSCC is known to have better prognosis than HPV(-) OPSCC, the impact on individual subsites remains undefined. This study aimed to characterize OPSCC survival stratified by HPV and oropharyngeal subsite.
Methods: National Cancer Data Base (NCDB) data was reviewed from 2004 to 2017. Subsites included base of tongue (BoT), tonsillar complex (TC), lateral and posterior oropharyngeal walls (LPOW), soft palate (SP), valleculae and anterior surface of the epiglottis (VE), and unspecified or overlapping (Other). Multivariate Cox Regressions were conducted with the following covariates: age (cutoff median of 60 years), sex, race, tumor grade, extracapsular extension, lymphovascular invasion, and AJCC pathologic TNM staging. Multivariate analyses were presented as hazard ratio [95% confidence interval]. Log Rank (Mantel-Cox) was used to compare overall survival at 2 years (2yOS) and 5 years (5yOS) between HPV status and oropharyngeal subsites.
Results: A total N of 58,712 patients had OPSCC with 39,624 HPV(+) patients and 19,088 HPV(-). When analyzed by subsite N (HPV(-):HPV(+)), there were BoT 22,520 (7304:15216), TC 28,822 (7526:21285), LPOW 872 (583:289), SP 1,630 (1286:344), VE 695 (412:283), and Other 4,184 (1977:2207). Multivariate analysis showed worse survival for older age (1.53[1.36,1.72]), lymphovascular invasion (1.53[1.35,1.73]), extracapsular extension (1.77[1.54,2.02]), higher pT staging, (T2 1.53[1.32,1.77], T3 3.01[2.53,3.60], T4 3.95[3.26,4.79]), and presence of metastases staging (3.72[2.13,6.50]). Overall survival for HPV(-) OPSCC was worse for BoT when compared to TC (p<0.001), although weaker differences in 2yOS and 5yOS was seen for HPV(+). HPV(+) OPSCC had better survival than HPV(-) OPSCC for all subsites, including BoT, TC, LPOW, SP, VE, and Other (p<0.001).
Conclusions: Survival for OPSCC differs not only by HPV status, but also by oropharyngeal subsites. Survival was worse for BoT when compared to tonsil for HPV(-) OPSCC, but survival only differed mildly between these subsites for HPV(+) OPSCC. SP and LPOW showed the worst 5yOS with a significant difference in HPV(+) tumors. In contrast, a larger difference in 2yOS and 5yOS was noted between these subsites for HPV(-) OPSCC. Overall survival between BoT and VE are similar for HPV(-), but not HPV(+). The findings show that OPSCC of each subsite is characterized by different prognoses, and HPV affects survival differently between subsites of the oropharynx. Future directions include examining survival differences for each subsite stratified by patient characteristics (e.g., tobacco exposure) and impact of treatment modalities.