1. IntroductionA major etiologic pathway of penile squamous carcinoma (PSCC) involves high-risk human papillomavirus (HPV). HPV infection occurs in about 30–50% of penile cancers but varies significantly via histologic subtype [1,2,3]. There are >150 various HPV genotypes, however only a few (HPV 16, 18, 31, 33, 45, 52) have been shown to be associated with PSCC with HPV 16 being the most prevalent genotype [1,2,3]. These high risk HPV genotypes are thought to be involved in the carcinogenesis of penile cancer through the activity of viral E6 and E7 oncoproteins that are capable of binding to and inactivating the p53 and the retinoblastoma-1 tumor suppressor proteins (Rb), respectively [4]. E7’s inhibition of the Rb pathway leads to increased expression of the p16INK4a (CDKN2A) protein (also known as p16). Overexpression of p16 can be detected by immunohistochemical (IHC) staining and has been found to be a reliable marker for high-risk HPV infection in oropharyngeal SCC and to a lesser extent in PSCC [5,6,7,8,9]. This study aims to provide additional evidence regarding the prognostic implications of assessing HPV and p16 status in a cohort of men with PSCC and 5-year follow-up. 4. DiscussionPSCC is a rare but lethal cancer whose biology is driven, at least in part, by differences in high risk HPV status [16,17,18]. Subsequent to high risk HPV infection, the p16INK4a protein (p16) is upregulated and has been detected in high risk HPV related malignancies utilizing IHC [2,5,6,8,9,17]. The p16 protein is not only a surrogate for HPV expression but has become a reproducible, reliable, and cost-efficient predictor of prognosis following a cancer diagnosis in oropharyngeal (OPC) SCC [8,19,20]. The correlation between p16 and HPV has been studied in a more limited fashion in PSCC and a similar strong correlation to that of OPSCC was observed (88%) [6,9]. To validate this correlation in PSCC, we investigated the correlation between p16 and HPV and the prognostic significance of p16 in a larger cohort of patients with PSCC and median 5.8-year (95% Cl; 4.7, 7.5) follow-up.The expression of p16 was found to be superior to all other clinical, HPV genotyping, or IHC parameters for predicting prognosis for patients with OPSCC [21,22]. We found a similar prognostic effect with p16+ PSCC in our cohort. Patients with p16+ tumors exhibited an improvement in 3-year and 5-year CSS, respectively (92% vs. 65% and 88 vs. 58%; p = 0.004). These findings were clinically relevant and feasible to document utilizing IHC and a binary cut off value (i.e., >75%) for p16 status. Of note, p16 positivity may exert its prognostic significance among patients with more advanced PSCC (i.e., pT ≥ 3, pN ≥ 2) as the correlation with better survival outcomes was significant among those with advanced versus lower stage disease (pT ≤ 2, pN ≤ 1). Among advanced disease PSCC patients the use of multimodality therapy is certainly more prevalent. It is plausible but not proven that p16 as a potential marker of HPV driven malignancy may correlate with response to multimodal therapies such as chemoradiation or chemotherapy. In one recent study patients with advanced PSCC that were HPV+ exhibited significantly greater overall survival when treated with radiotherapy when compared with the HPV− cohort [23].Our study highlights the importance of p16 staining in PSCC and how this relatively simple methodology using IHC compares to the more costly HPV testing technologies. To our surprise HPV status was not shown to have the same prognostic ability as p16 staining in the overall cohort but was significantly associated with an overall survival benefit among patients with advanced primary tumor stage. A recent meta-analysis by Sand and colleagues of PSCC studies [24] that assessed HPV and p16 expression in a cohort of 649 men that included 20 studies aligns with our findings. They found that p16 was associated with positive prognostic value for CSS with a hazard ratio of 0.45 (95% CI; 0.30 to 0.69) as was high risk HPV expression but with a smaller protective effect (hazard ratio = 0.61, 95% CI; 0.38 to 0.98). The observation that high risk HPV expression did not exhibit the same predictive effect as p16 in our cohort may be explained by p16′s downstream role in cell cycle control, and that p16 may serve as a biomarker for the degree that the tumor is HPV driven in PSCC. However, caution is advised as some OPSCC studies have found high p16 expression in a small subset of tumors where HPV-DNA was not detected [25,26]. Interestingly, these patients with high p16 expression and undetected HPV-DNA had poorer survival curves similar to alcohol and tobacco related OPSCC and not HPV-driven OPSCC. Whether the p16+, HPV negative phenotype has a better or poorer prognosis than the concordant phenotype clearly deserves further study in PSCC. In our 143 patient samples, HPV and p16 status were discordant in 20 samples. All HPV testing was initially run using the Cobas system and a subset that gave invalid results [27] were analyzed using the alternative RNAscope analysis. These two assays are significantly different. Cobas is an automated system that extracts DNA from FFPE tissue slices and uses real-time PCR amplification of the L1 gene from 14 high-risk HPV genotypes [28], while RNAscope is an in situ hybridization technique that stains E6 and E7 mRNA of high-risk HPV genotypes directly within FFPE tissue sections [29]. However, in our small sample set we observed no significant difference in discordance between HPV and p16 status utilizing either Cobas or RNAscope assays. Thus, it is unlikely that the assays used resulted in the discordance noted. A novel finding of our study is that the use of a binary cut off of 75% staining was a simple method to assess p16 positive or negative status. A stronger overall agreement between HPV and p16 status was found using our HS method versus the staining pattern method as measured by the kappa coeffecient. Considering its prognostic value we noted a stronger correlation with CSS (HR = 0.30; 95% CI; 0.13, 0.72; p-value 0.007) utilizing our HS method over a previously published staining pattern method (HR = 0.32; 95% CI; 0.11, 0.90; p-value 0.03) [13,14]. However, we acknowledge the lack of independent validation of our p16 HS method and evaluation of p16 staining by a single genitourinary pathologist. While we showed that p16 IHC with a simple 75% positive cut off to categorize cases at our center was a strong prognostic factor for survival we invite further validation from other experienced centers treating penile cancer patients. In this manner p16 staining determined by multiple pathologists to determine inter-observer concordance and prognostic capability would be optimal. Of note, such a validation study going forward will likely be feasible as the World Health Organization recently recommended that p16 status be reported as a “requirement” for PSCC pathology reporting [30].

The primary limitations of this study are its retrospective nature in the setting of a tertiary referral center and heterogeneous treatment regimens. To minimize these effects, we analyzed the importance of a variety of known prognostic covariates on survival outcome. Clinical and pathologic correlates of advanced disease were significant predictors of both OS and CSS. However, after adjusting for all variables, p16 remained an important independent predictor of both OS and CSS. Thus, the present study provides further evidence that p16 IHC is a rapid, easy to perform test that has a high sensitivity and high negative predictive value for assessing HPV status in PSCC. As such p16 status serves as an important prognostic marker and excellent first line test to potentially describe the biology driving a given patient’s PSCC.