A different entity: a population based study of characteristics and recurrence patterns in oropharyngeal squamous cell carcinomas

Scott Murray, Michael N. Ha, Kara Thompson, Robert D. Hart, Murali Rajaraman, Stephanie L. Snow

Cases of squamous cell carcinoma (SCC) of the oropharynx were compared with other head and neck cancer (HNC) anatomic subsites in patients treated at the provincial referral centre for HNC, the Nova Scotia Cancer Centre (NSCC). A retrospective chart review was performed on HNC patients assessed at the NSCC between 2010 and 2011. Patient demographics, disease characteristics, treatment details and outcomes, including recurrence rates and survival were collected. Data was collected on new and recurrent cases of HNC. This data was compared between the two types of HNC using chi-square tests for dichotomous categorical variables or Fishers exact test where appropriate. Wald test was used to compare categorical variables with 3 categories. Continuous variables were compared using the non-parametric Wilcoxon test. 318 charts were included in the analysis. 122 (38 %) were oropharyngeal squamous cell carcinomas (OPSCCs). In terms of disease characteristics, OPSCCs were more likely to be poorly differentiated/undifferentiated (n = 267, 49(40 %) vs 42(21 %), p < 0.001), non-keratinizing (n = 169, 25(20 %) vs 17(9 %), p < 0.001), greater than 2 cm (n = 253, 72(59 %) vs 78(40 %), p = 0.0061), stage 4 (n = 313, 55(45 %) vs 64(33 %), p = 0.0315) and have had locoregional nodal spread (n = 315, 103(84 %) vs 55(28 %), p < 0.001). In the subset of 57 patients that had p16 testing, OPSCCs were more likely to be p16(+) (37(30 %) vs 1(1 %), p < .001). There were no significant differences in terms of Charlson probability of 10 year survival, smoking or alcohol consumption although OPSCC patients were significantly less likely to have COPD as a co-morbidity (n = 318, 19(16 %) vs 53(27 %), p = 0.0175). Finally, OPSCCs had less chance for relapse than non-OPSCCs in both univariate (2.119 times less, p=0.0034) and multivariate (1.899 times less, p=0.0505) analyses along with a 1.822 times less overall mortality in a multivariae analysis (p=0.0408).  CONCLUSIONS: This analysis suggests that Nova Scotian OPSCCs should be considered distinct from other HNC lesions, most notably in terms of disease characteristics and prognosis. Specifically, despite a higher association with disease factors traditionally considered to be linked to poor prognosis, outcomes were actually superior in terms of relapse and overall mortality.