Compared to husbands of women without a diagnosis of anogenital (HPV-associated) cancer, the relative risk estimates for base of tongue and tonsillar cancers were elevated in the husbands of women diagnosed with invasive anogenital cancer [5]. However, the RR estimates increased from 1969–2001 to 2002–2015, and peaked at 9.44 in the youngest couples (< 50 years of age) diagnosed with TSCC/BOTSCC (husbands) and anogenital cancer (wives) in 2002–2015. These results are in line with our a priori hypothesis that increased exposure to hrHPV infections following the sexual revolution, results in an increased risk of familial OPC some decades later (2002–2015).
The identified increase in the familial RR estimates over calendar time fit the continuous increase in the incidence and prevalence of HPV in TSCC and BOTSCC in Sweden [10]. It is noteworthy that in the Nordic countries, the hrHPV prevalence in the fertile-aged population increased somewhat beyond the onset of the HPV vaccination era [11, 12]. Thus, contemporary fertile-aged individuals have a highly increased risk of familial HPV-associated OPC [4].
The mostly HPV-associated OPCs, i.e., tonsillar and base of tongue cancers are anatomically differentiated from non-HPV-associated OPCs [13]. Furthermore the clinical features of the HPV-associated OPCs also differ from non-HPV-associated OPCs [14]. These facts further emphasize the importance of studying these two HPV-associated OPCs (TSCC/BOTSCC) separately in the familial setting. The Swedish high quality registry data (sizeable cohort with ample power for over time analysis) provided a unique opportunity for this. As our study is anonymous and registry-based it has its limitations, most notably the HPV infection status of the cases remains unknown. However, 70% of the cases identified during the last 10 years were HPV positive [15].
It has been established that the prevalence of oral HPV infections is higher in women with cervical HPV infection [16]. The higher the prediagnostic viral load, the higher the risk of both cervical carcinoma in situ and invasive cervical cancer [17, 18]. Thus, the transmissible virus possibly was acquired from women later diagnosed with in situ cervical carcinoma, but due to the long incubation time of occult TSCC/BOTSCC increased risk in the husbands of these women was not detectable yet. On the other hand, approximately half of the carcinoma in situ cases do not develop into invasive cervical cancer and the possibility of subtle intertypic differences between the oncogenicity of HPV 16 types transmitted by women with cervical carcinoma in situ as compared to women with invasive anogenital cancer is worth noting. It might partially explain the apparent difference in the familial risk estimates between the two types of couples: husbands with OPC and wives with cervical carcinoma in situ or women with invasive anogenital cancer.
As a limitation of the study we note the small case numbers especially in the invasive cancer group and age-strata. Nonetheless, the test over the strata-specific relative risks showed a significant result. The population-based Swedish family data linked to cancer diagnoses does not have data on risk-taking sexual behaviour which could not be adjusted for. Finally, the role of chance cannot be totally ruled out.
HPV vaccination will eventually alter the epidemiological situation. However, because of the long incubation period from chronic HPV infection to OPC, and the changes in sexual practices, especially the increase in the practice of oral sex, the RR estimates for male familial OPCs may still be on the rise. How to tackle the significant time-dependent increase in the familial risk of tonsillar and base of tongue cancers in males warrants further studies.