Date Published: February 1, 2019
Publisher: Public Library of Science
Author(s): Ksenija Božinović, Ivan Sabol, Zoran Rakušić, Antonia Jakovčević, Mario Šekerija, Juraj Lukinović, Drago Prgomet, Magdalena Grce, Scott M. Langevin.
Head and neck squamous cell carcinoma (HNSCC) is the sixth most common malignancy worldwide. Main HNSCC risk factors are tobacco, alcohol, and high-risk human papillomavirus (HPV). HPV+ oropharyngeal squamous cell cancer (OPSCC) usually have different etiology, increasing incidence and often show an improved survival when compared to HPV-negative cases. The objective of the current study was to retrospectively examine the influence of HPV on the survival of OPSCC patients in a non-Western population setting.
We determined the presence of HPV DNA and/or RNA in 99 formalin-fixed paraffin embedded (FFPE) tissue samples of OPSCC patients treated between 2002 and 2015. Patients were compared based on laboratory, demographic, clinical, life style characteristics and survival.
HPV RNA was found in 29.3% cases. However, groups based on HPV data (either RNA, DNA or retrospectively collected p16 staining) did not show significant differences. Overall, 5-year survival was 30% with minimal influence of the HPV positivity.
The HPV influence on survival of OPSCC patients is not identical between populations probably due to other factors overshadowing the HPV effect. This should be taken into account when treatment or policy decisions are made in each particular setting.
Head-and-neck cancer (HNC) is a group of malignancies that most commonly arise from the upper aerodigestive tract mucosa or lining of the head-and-neck regions . It is the sixth most common malignancy worldwide. Most HNCs (95%) develop from squamous cell epithelia and are further characterized according to their primary site of origin. Most common sites are oral cavity, oropharynx, pharynx, larynx, and sinonasal tract . Globally, head-and-neck squamous cell carcinoma (HNSCC) accounts for approximately 550,000 cases annually  and in Croatia, there were estimated 750 cases in 2014 .
From a total of 99 OPSCC patients, 59 (59.6%) were HPV DNA negative and 40 (40.4%) HPV DNA positive, of which 11 (27.5%) were RNA negative, and 29 (72.5%) RNA positive. Statistical analysis was done on those 3 subgroups: HPV DNA/RNA negative (N = 59), DNA positive/RNA negative (N = 11), and DNA/RNA positive (N = 29). In addition, the tests were also performed on the whole HPV DNA positive subgroup, including both RNA negative and RNA positive samples (N = 40), but no additional statistical significance was found. Patient’s characteristics are presented in Table 1. Male patients were predominant in all groups, and account for 82.8% cases. Tonsil and base of tongue were the most common subsites, together accounting for more than 86% of cases. There were no statistically significant differences between the HPV groups in terms of age at the time of diagnosis (Fig 1) or gender. There were 14% never smokers and 13% never drinkers without significant differences between groups. There was no correlation between HPV and p16 results with Pearson’s correlation coefficient r = 0.14 for comparison with HPV RNA. The p16 positive staining was weakly statistically associated with patient’s older age (t-test, P = 0.042) and with cancer higher grade (Mann-Whitney, P = 0.017).
In this study, we performed HPV DNA and RNA analysis of oropharyngeal cancer patients treated within a 15-year period. The overall results indicate that 29.3% of cancer cases had active HPV E6 mRNA transcription and were thus likely HPV driven. The HPV DNA positivity, on the other hand, was shown in 40.4% patients, which is comparable with previous reports [13,16]. Previous studies [11,28–30] have already concluded that HPV DNA presence is not always sufficient to attribute cancer to HPV and that mRNA or combination of assays should be used. Unexpectedly, the correlation between p16 and HPV data was low with correlation coefficients r = -0.108 for comparison with HPV DNA and r = 0.146 for comparison with RNA suggesting that p16 is not a suitable replacement for HPV testing in the current setting. Some other studies have also shown that p16 testing might not be an accurate biomarker for oropharyngeal cancers, since the presence of p16 has been detected in HPV–HNSCC . However, literature also suggests that p16 positive, but HPV negative HNSCC, share some common characteristics, like favorable prognosis, with HPV+ HNSCC [5,30,31]. However, in our study there were also several HPV+ yet p16 negative tumors that negatively affected the correlation. Such samples were also seen in the study of Jordan et al. . However, recent Cancer Genome Atlas project analysis  indicated that CDKN2A gene (encoding p16 protein) is often deleted or mutated in smoking related HNSCC cancers. Such deletions would preclude p16 overexpression even if HPV is transcriptionally active.
In summary, this study provides the baseline relevant data for treatment of OPSCC patients in Croatia. Eventual policy and treatment decisions in similar regions should take into account the particularities of each population. Other factors like advanced stage, patient age or still highly prevalent smoking and drinking in Croatia might be overshadowing the positive effect of HPV seen in Western populations. Current data indicates that HPV, as a favorable prognostic marker, should not be considered to outweigh other relevant factors in a particular population until other socio-epidemiological changes evident in Western populations are also observed.