Introduction

About 600,000 new head and neck cancers are diagnosed annually worldwide [1]. Beside tobacco smoking and alcohol consumption, human papillomavirus infection (HPV) was recognized as a leading etiologic factor—especially in oropharyngeal cancers [2, 3]. Currently, HPV-associated oropharyngeal cancers show a rapidly increasing incidence [4]. More than 150 HPV types have been identified from which type 16 is responsible for about 80–90% of all HPV-harbouring head and neck cancer cases [5].

For detecting HPV in head and neck cancers, numerous methods are available. HPV mRNA in situ hybridization (ISH) can be regarded as the gold standard as it is sensitive to transcriptionally active viruses and specific to the given type of HPV. The main disadvantage of this method is its high cost and limited efficiency in formalin-fixed, paraffin-embedded samples. HPV DNA polymerase chain reaction is considered to be sensitive, however, it gives a high rate of false positive cases as it shows positive reaction in HPV-harbouring but not HPV-driven oropharyngeal cancers. Immunohistochemical (IHC) detection of the tumor suppressor p16INK4 is regarded as a surrogate marker, given the sensitivity of 100% and specificity around 70–80%.

There are recommendations to use p16INK4 IHC for primary screening, and to use HPV DNA PCR for the p16INK4-immunolabelled cases. According to Sache et al., using this method 97% of sensitivity and 94% of specificity can be achieved [6].

In this study, we aimed to assess the HPV induced fraction of Hungarian head and neck cancer patients, and to compare p16INK4 IHC alone with the combined p16INK4IHC/HPV DNA PCR technique regarding prognosis, and prediction of therapeutic response. In this work, the term HPV-induced refers only to tumors showing both p16INK4-expression and HPV DNA presence.

Patients and methods

Patients

Archived, formalin-fixed, paraffin-embedded (FFPE) samples of 124 patients with head and neck squamous cell carcinoma (characteristic) were analysed in our study. Each patient underwent treatment between 2012 and 2014 at the Department of Oto-Rhino-Laryngology, Head and Neck Surgery, Semmelweis University (Table 1).

Table 1 Descriptive statistics of pateints involved in our study
Full size table

Tissue microarray (TMA)

TMA blocks containing 2 mm diameter cores were created using the TMA Master (3DHISTECH Kft, Budapest, Hungary). Tissue sections (4 μm) were cut on adhesion slides and used for immunohistochemical analysis.

P16ink4immunohistochemistry (ihc)

Immunohistochemical analysis was performed in the II. Department of Pathology, Semmelweis University.

BenchMark XT IHC/ISH (Roche, Germany) semi-automated device was used for immunohistochemical staining with the application of XT UltraView DAB v3 kit. The protocol of staining method was carried out as described previously [7].

Briefly, sections were incubated at 72 °C for 4 min. We used EZ Prep Solution (Ventana Medical Systems, Tucson, AZ) three times to remove paraffin. Cell conditioning solution pH 8 (Ventana) was used for heat induced epitope retrieval at 95 °C for 30 min followed by a heating at 100 °C for 4 min. Endogenous peroxidase activity was inhibited with one drop UV INHIBITOR (Ventana), which was applied at 37 °C for 6 min. Primary monoclonal antibody against p16INK4 (Clone CINtec E6H4, Ventana) was appliedat 37 °C for 32 min in a dilution of 1:100. After incubation with UV HRP UNIV MULT secondary antibody solution (Ventana) at 37 °C for 8 min, peroxidase activity was visualized with diaminobenzidine (DAB) chromogen (Ventana). Nuclear counterstaining was done with hematoxylin II (Ventana). All washing steps were performed with diluted Reaction Buffer Concentrate (Ventana).

Immunostained TMA slides were digitized using Panoramic Scan device (3DHISTECH, Hungary) and analyzed with Panoramic Viewer (3DHISTECH, Hungary) software. In each core, nests of at least 200 tumor cells were analyzed by two independent assessors. Cut-off for p16INK4-immunolabelling was set at 75% of cytoplasmic or nuclear staining[8]. Intensity of staining played no role in the evaluation. However, almost all p16INK4-immunolabelled samples showed a remarkably strong intensity.

HPV genotyping

Human DNA was extracted from FFPE tissue sections in cases showing p16-immunolabelling by QIAmp DNA FFPE Tissue Kit (Qiagen, Hilden, Germany) in line with the manufacturer’s instructions. High-risk HPV DNA detection was performed using CONFIDENCE™ HPV test (NEUMANN Diagnostics, Hungary) combined with genotyping for HPV 16, 18, 31, 33, 45, 52, 58. The sufficient amount of input DNA was controlled by fluorometric quantitation by Qubit™ Fluorometer (Invitrogen, Carlsbad, CA, USA) using Qubit™ dsDNA HS Assay Kit (Invitrogen) according to the manufacturer’s protocol.

CONFIDENCE™ HPV is a TaqMan®-based L1 region specific multiplex real-time PCR assay for viral DNA detection. The test detects HPV 16 and 18 separately and other high-risk types (HPV 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, 68) in a pooled manner. In case of high-risk HPV presence, genotyping was performed using type specific primers in separated reactions. The quantitative real-time PCR was carried out on QuantStudio™ 6 Flex platform in 384-well plate format.

Statistical analysis

Statistical analysis was performed using IBM SPSS Statistics for Mac v.22.0.0 (SPSS Inc., Chicago, IL, USA). Disease-specific survival (DSS) time was calculated from the time of diagnosis to the time of the last consultation. The event was defined to occur in cases of disease-specific death and if the patient was not tumor-free in the time of the last follow-up examination. For survival analysis, Kaplan–Meier estimations with log-rank tests, and Cox-regression analyses were used. Correlations were tested using Chi-square test/Fisher’s exact test with Monte Carlo simulation. 95% confidence interval was applied for all tests, and statistical significance was declared if the p value was under 0.05.

Results

Expression of p16INK4

Figure 1a, b show representative images of a p16INK4-immunolabelled and a non-p16INK4-immunolabelled case. The result of p16INK4 immunohistochemistry was available in 110 patients. Out of the 110 tumor samples, 19 cases (17.3%) proved to be p16INK4-immunolabelled.

Fig. 1
figure 1

Capture of representative p16-immunolabelled (a) and non p16-immunolabelled (b) tissue samples (Magnification: 40X.)

Full size image

The highest proportion of p16INK4-immunolabelled cases was observed in oropharyngeal tumors (38.1%), whereas other locations showed much lower (larynx: 4.8%, hypopharynx: 4.2%) or no (oral cavity: 0%) p16INK4-immunolabelling rate (Chi-square test: p < 0.001) (Table 2).

Table 2 Proportion of tumors according to anatomical localisation (HPV human papillomavirus)
Full size table

Involving all tumor locations, p16INK4 status was not significantly associated with disease-specific survival (DSS) (median survival: 17.5 vs. 30.3 months, p = 0.107 with the log-rank test; Cox-regression: HR: 0.5 (95% CI 0.212–1.181), p = 0.114) (Fig. 2a).

Fig. 2
figure 2

a Kaplan–Meier survival curve of all tumor locations included, based on p16 status. b Kaplan–Meier survival curve of oropharyngeal tumors according to p16 status

Full size image

In contrast, p16INK4-immunolabelled oropharyngeal tumors showed an improved DSS compared to the non-p16-immunolabelled oropharyngeal cancers (median survival: 30.3 vs. 8.8 months, p < 0.001with the log-rank test; Cox-regression: HR: 0.175 (95% CI 0.059–0.520), p = 0.002) (Fig. 2b).

We found no correlation between p16INK4 expression and T, N or M status, stage, tumor grade or patients’ age. This latter observation was true comparing both tumors of all regions and tumors of the oropharynx.

High-risk HPV DNA detection by PCR

The p16INK4-immunolabelled cases were tested for 7 high-risk HPV subtypes (HPV 16, 18, 31, 33, 45, 52 and 58) using DNA PCR method. Out of 19 cases, 9 tumors harboured HPV DNA (HPV-induced cases). HPV 16 was present in 8 cases, HPV 33 in one single case. All the HPV-containing samples originated from the oropharynx. This means that the HPV-induced rate of oropharyngeal cancers was 21.4% (9/42 patients). Thus, the specificity of p16 to detect oropharyngeal HPV presence was 56.3% (out of the 16 p16INK4-immunolabelled oropharyngeal tumors, 9 cases tested positive for HPV as well). DSS of patients with p16INK4-immunolabelled/HPV DNA-containing tumors (HPV-induced cancers) was significantly better when compared with that of patients having non-HPV-induced cancers (lack of p16INK4-immunolabelling and/or not HPV DNA harbouring) (median survival: 25.9 vs. 9.5 months, p = 0.024 with the log-rank test; Cox-regression: HR: 0.217, 95% CI 0.051–0.932, p = 0.04) (Fig. 3).

Fig. 3
figure 3

Kaplan–Meier survival curve of oropharyngeal cancers based on HPV status. HPV-induced p16-immunolabelled and HPV DNA-containing, non-HPV-induced non-HPV DNA-containing and p16 immunolabelled or not (HPV human papillomavirus)

Full size image

P16INK4-immunolabelling/HPV DNA presence and response to neoadjuvant chemotherapy

Of the 110 patients with available immunohistochemical staining, 32 patients received neoadjuvant chemotherapy (docetaxel-cisplatin-5-fluorouracil in 30 cases and cisplatin-5-fluorouracil in 2 cases). P16INK4-immunolabelled individuals showed a better response compared with the non-p16INK4-immunolabelled group (Fisher’s exact test: p = 0.025).

There was a difference between groups based on HPV status as well (Fisher’s exact test: p = 0.009) (Table 3).

Table 3 PD progressive disease, SD stable disease, PR partial remission, CR complete remission, HPV human papillomavirus
Full size table

P16INK4 -immunolabelling/HPV DNA Presence and Smoking

We investigated smoking habits among p16INK4-immunolabelled cases dividing them into HPV DNA-containing and non-HPV DNA-containing subgroups. Interestingly, none of the patients with HPV DNA-containing tumors were current smokers, and 55.4% of them had no history of smoking. On the other hand, 67.7% of the patients with tumors lacking HPV DNA belonged to current smokers and only 7% of them had no history of tobacco use.

There was a significant difference between the groups (Fisher’s exact test: p = 0.002) (Table 4). We have not found any difference in alcohol consumption between groups.

Table 4 Distribution of patients according to HPV status and smoking habits (HPV human papillomavirus)
Full size table

Discussion

In this study, we assessed the rate of HPV-induced and p16INK4-expressing tumors in Hungarian head and neck cancer patients with both p16INK4 immunohistochemistry and HPV DNA PCR method. We detected significant association between p16INK4-immunolabelling, response to chemotherapy and disease-specific survival. Besides, correlation between HPV DNA-containing tumors and smoking habits was revealed.

Numerous studies reported the growing prevalence of human papillomavirus in head and neck squamous cell cancers. Chaturvedi et al. [9] observed increasing incidence of oropharyngeal, oral cavity and lung cancer among women in contrast to young men, where oropharyngeal cancer is the only one of these three tumor types that shows an increasing incidence. This phenomenon underlines the potential effect of HPV infection on the increasing incidence of oropharyngeal cancers. According to the data published by Castellsagué et al., HPV-induced rate of oropharyngeal cancers was 24.3% assessed by HPV DNA and p16INK4 dual testing [10].

In our study, we found HPV-induced rate to be 21.4% in the oropharynx using p16INK4/HPV DNA PCR co-testing. Regarding other localizations, no p16INK4-immunolabelled/HPV DNA-containing tumor was detected.

However, there is still a lack of consensus in diagnostic methods to detect HPV[11, 12], Sache et al [6] found that the combination of p16INK4 IHC/DNA PCR showed the best specificity and sensitivity when analyzed eight possible assay/assay combinations. In our study, we used both of the recommended methods, and found a remarkably high proportion of p16INK4-immunolabelled but not HPV DNA-containing tumors. This can be explained either by somatic mutations of the retinoblastoma (pRb) signaling pathway associated with tobacco smoking or by carcinogenesis driven by non-analyzed HPV subtypes.

In both p16INK4-immunolabelled and HPV-induced cases, we found significantly better prognosis compared with non-p16INK4-immunolabelled and non-HPV-induced groups.

Improved outcomes were found among patients with HPV-associated tumors after induction chemotherapy or chemoradiation [13]. Considering the adverse effects of curative surgeries or chemoradiotherapy and the younger age of the patients with HPV associated tumors, there are several clinical trials searching for de-intensified therapeutic approaches with less toxicity for these patients. Ang et al. [14] used recursive-partitioning analysis, and classified oropharyngeal tumors into low/intermediate/high risk categories considering HPV status, pack-years of tobacco smoking, tumor stage, nodal stage. O’Sullivan et al. [15] recommended a risk classification system according to HPV status [15]. Both of these new classifications could help to decide whether treatment could be de-intensified without jeopardizing oncologic outcome. According to our results, p16INK4-immunolabelling divided prognostic groups more precisely than p16INK4/HPV DNA PCR double testing, implying that p16INK4-immunolabelling is of prognostic value in oropharyngeal cancers regardless of human papillomavirus status. Both p16INK4-immunolabelled and p16INK4-immunolabelled/HPV DNA-containing tumors had a better response to neoadjuvant chemotherapy.

Smoking and alcohol consumption are well known risk factors in head and neck cancers. Ng et al. [16] reported a large reduction in smoking habits, however, because of the growth of population, the absolute number of smokers increased significantly. This observation confirms that tobacco remained a considerable carcinogenic agent. Since head and neck cancers are etiologically heterogeneous, we have to consider the co-existence and pathogenic effect of HPV infection, tobacco use and also the effect of alcohol consumption. In our research, tobacco use was more frequent in non HPV-containing cases, but 44.4% of the patients with HPV-related tumors had a history of smoking. In these cases we have to count on both noxas. Interestingly, 93% of patients with p16INK4-immunolabelled but non-HPV DNA-containing tumors had a positive smoking history which underlines the possible mutagenic effect of tobacco on pRb signaling pathway.

Recent data of The Cancer Genome Atlas suggests [17] that there are significant differences regarding molecular alterations between HPV-induced and non-HPV-induced tumors. These differences in structural alterations or mutation signatures could distinguish subgroups in head and neck carcinomas. These results can help us designing more accurate clinical trials and help clinicians to select the best treatment option.

Further studies in a multi-institutional setting using a sufficiently large sample size are required to validate the independent prognostic value of p16INK4 protein level for introducing this test into the daily clinical practice.

Conclusion

P16INK4 immunohistochemistry is a possible, precise and widely affordable tool in prognostic and predictive characterization of head and neck squamous cell cancers. In comparison with p16INK4/HPV DNA PCR double testing, p16INK4 status alone proved to be a slightly more precise indicator of clinical outcome.