High-risk human papillomavirus in oral squamous cell carcinoma of young patients



The aim of this study was to investigate a possible relation between oral squamous cell carcinoma (SCC), the presence of high-risk human papillomavirus (HR-HPV) DNA and p16 expression in young patients. Paraffin-embedded tumor blocks from 47 oral SCC of young (≤40-year old) patients were evaluated. The presence of HPV DNA in tumor specimens was analyzed by polymerase chain reaction (PCR) using GP5+/GP6+ generic primers (L1 region) followed by dot blot hybridization for HPV typing. When necessary, the HPV16 positivity was confirmed by PCR HPV16 E7-specific primers. Cases involving young patients were compared with 67 oral SCC from patients ≥50-year old (controls). Demographic and clinical data were collected to analyze patient outcomes. p16ink4 expression was evaluated by immunostaining of tissue microarrays. HPV16 was detected in 22 (19.2%) cases; 15 (68.2%) young and 7 (31.8%) control patients, a statistically significant difference (p = 0.01). In 1 (1.7%) young group specimen, HPV DNA 16 and 18 was detected. p16 expression was observed in 11 (25.6%) cases from the young group and in 11 (19.6%) controls (p = 0.48). Association between HPV and p16 was verified, and it was statistically significant (p = 0.002). The higher prevalence of high-risk HPV types, especially HPV16, may be a contributing factor to oral carcinogenesis in younger individuals.

Oral squamous cell carcinoma (SCC) affects mainly men between the fifth and sixth decades of life1 and is rare in young patients (≤40-year old); however, incidence in this age group has increased in several countries over the last two decades.2, 3 In older patients, the main risk factors are tobacco and alcohol consumption.1 However, in younger patients, the role of these factors is uncertain due to the short time of exposure.4, 5 Some studies have reported that high-risk human papillomavirus (HR-HPV) infection might exert an important role in carcinogenesis in this group.6–11 HR-HPVs, especially HPV16 and HPV18 are considered a major cause of certain human cancers; they are responsible for all cervical cancers and about 50% of other anogenital cancers.12 In oral SCC, their prevalence and importance are controversial.6–11, 13–17 HPV infection occurs more frequently in young individuals than in older7, 18, 19 and it is correlated to sexual behavior.18, 19

The HPV oncoproteins E6 and E7 stimulate cell proliferation by activating cyclins E and A and inactivating p53 and retinoblastoma (Rb) cell cycle proteins. The absence of functional pRb can upregulate p16ink4 expression.12 p16ink4 expression has been used to determine the presence of biologically active HPV in head and neck (HN) SCC.20 Reports suggest that tumors with p16ink4 expression and HPV positivity show improved survival rates,18, 21, 22 although consensus has yet to be achieved.16

The aim of this study was to determine HR-HPV prevalence and investigate a possible relation between the presence of HR-HPV DNA, p16 expression and clinical outcomes in oral SCC of young patients (≤40-year old) as compared with tumors from a control group (≥50-year old).

Material and Methods

Patients, clinical background and treatment

A retrospective study was performed analyzing 114 patients with oral SCC treated from 1970 to 2006 at the Head and Neck Surgery and Otorhinolaryngology Department of the A.C. Camargo Hospital, São Paulo, Brazil. Demographic (age, sex and race), lifestyle (smoking habit and alcohol consumption), clinical (tumor site and clinical stage), treatment and pathological factors (histological grade) were analyzed. The clinical characteristics of these patients were obtained from their medical records and the tumors were restaged according to the 2002 version of the American Joint Committee on Cancer (TNM) classification23 and grouped as early (clinical Stages I–II) or advanced clinical stages (clinical Stages III–IV). In cases of locoregional disease recurrence, all were histopathologically confirmed. Histological grade was defined by the classification proposed by the World Health Organization1 as well differentiated (Grade I), moderately differentiated (Grade II) or poorly differentiated (Grade III). Patients with carcinoma in situ, verrucous carcinoma, SCC of lip, oropharynx and who received any prior treatment were excluded. This study was approved by the Human Research Ethics Committee of A.C. Camargo Hospital.

DNA extraction

Sections (5 μm) of paraffin-embedded tumors were collected in 1.5-ml microtubes. Tissues contained >80% tumor cells previously confirmed by hematoxylin–eosin (HE) stained sections. The microtome blade used was disposable and all tools and the surrounding area were cleaned with xylene and ethanol after processing each sample to avoid contamination among samples. DNA was extracted according to standard protocols.24 DNA quality was assessed by amplification of a fragment of the human β-globin gene using PCO3+/PCO4+ primers, resulting in a 110 base pairs DNA fragment.25

HPV detection

All samples positive for β-globin were checked for HPV DNA. Of the 114 good-quality DNA samples, 47 were from the young group (≤40-year old) and 67 from the control group (≥50-year old). The presence or absence of HPV DNA was established using generic primers GP5+/GP6+ specific for the L1 gene of several HPVs.26 Polymerase chain reaction (PCR) products were submitted to electrophoresis on 7% polyacrylamide gel, followed by silver staining.27 Specimens were considered positive for HPV DNA when they presented a band around 150 base pairs, and gave positive signals on genotyping by dot blot hybridization using radioactive probes.28 A total of 17 HPV types were analyzed: HPV6, HPV11, HPV16, HPV18, HPV31, HPV33, HPV35, HPV39, HPV42, HPV45, HPV51, HPV52, HPV53, HPV54, HPV55, HPV56 and HPV58. Samples with readings considered doubtful were submitted to PCR specific for HPV16.


Briefly, tissue sections were preheated at 56°C for 24 hr, deparaffinized and rehydrated in graded ethanol solutions. Antigen retrieval was conducted in a pressure cooker containing 10 mM citric acid solution (pH 6.0). Tissues sections were quenched of endogenous peroxidase with 3% hydrogen peroxidase (Merck, Brazil) for 20 min at room temperature, followed by a washing step with 10 mM phosphate-buffered saline (pH 7.4) for 5 min. Incubations with the primary antibodies against p16ink4 protein (Dako, E6H4 clone; Glostrup, Denmark) at 1:50 dilution were performed overnight at 4°C. Reactions were developed with Post Primary Block for 30 min at 37°C (NovoLink Max Polymer; Novocastra, New Castle, UK) and were visualized using the chromogen diaminobenzidine (DakoCytomation, Carpinteria, CA). Sections were counterstained with Carazzi's hematoxylin. Positive and negative controls were included in all reactions.

Statistical analysis

Associations between variables in contingency tables were analyzed by the Chi square test and a 5% significance level was considered for all statistical tests. The follow-up time was defined as the interval between the onset of treatment and the date of death or the last information for censored observations. The disease-free interval was measured from the date of treatment to the date when the first recurrence was diagnosed. Disease-free survival (DFS) and overall survival (OS) probabilities were estimated by the Kaplan–Meier method and the log-rank test was applied to assess significance of differences among actuarial survival curves. Statistical analyses were performed using the STATA program, version 10.0 (College Station, TX).


A total of 47 patients in the young group and 67 in the control group were studied (Table 1). The mean (range) age was 34.0 (20–40; standard deviation ±5.4) and 63.1 (50–82; SD ±8.9) years and the male:female ratio was 1:1.7 and 1:3.1 in the young and control groups, respectively. In both groups, the tongue (49.0% vs. 55.2%) was the most frequently involved site, followed by floor of the mouth (29.8% vs. 16.4%). Eight of the young patients (17.0%) had no tobacco exposure, whereas in the control group, 14 (20.9%) were nonsmokers (p = 0.58). Alcohol consumption was similar in both groups (p = 0.36).

Table 1. Demographic and clinical characteristics of 47 young and their control patients
inline image

Advanced clinical stage (CS III and IV) was observed in 38 (80.9%) and 50 (74.6%) young and control patients, respectively (p = 0.43). In the young group, surgery associated with postoperative radiotherapy was the treatment approach used on 26 patients (55.3%), surgery alone was performed on 14 (29.8%), a combination of surgery, radiotherapy and chemotherapy was used on 4 (8.5%) and radiotherapy alone was used on 3 (6.4%) patients. In the control group, surgery and radiotherapy was performed on 38 (56.7%) patients, surgery alone on 23 (34.3%), radiotherapy alone on 4 (6.0%) and 2 (3.0%) were treated by a combination of surgery, radiotherapy and chemotherapy (Table 1).

The presence of HPV DNA was investigated in all 114 cases of oral SCC. The overall prevalence of HPV was 19.2% (22/114 cancers). HPV16 was detected in 15 (68.2%) tumors from young patients and 7 (31.8%) from control tumors, a statistically significant difference (p = 0.01; Table 2). One case involving a young patient showed HPV16 and HPV18 DNA. Low-risk HPVs were not detected in any tumor.

Table 2. Prevalence of high-risk HPV in 114 OSCC by age group and correlation between HR-HPV DNA detection and p16 immunostaining
inline image

Koilocytosis characterized by perinuclear cytoplasmic haloes, nuclear dysplasia and binucleation were observed in 33 (28.9%) cases. In the young group, it was verified in 16 (48.5%) specimens; in one case, it was present only in the epithelium without dysplasia. In the control group, 17 (51.5%) tumors showed koilocytosis; thus, no statistically significant difference was observed between the groups (p = 0.31).

Expression of p16 was evaluated by immunohistochemistry in 99 cases, 43 from young patients (43.4%) and 56 controls (56.6%). Nuclear/cytoplasmic p16 expression was observed in 22 (22.2%) cases, 11 cases (25.6%) from the young group and in 11 (19.6%) controls (p = 0.48; Fig. 1). In one HPV positive case, p16 staining could not be performed. From 21 cases with HPV positive, 10 (47.6%) showed p16 immunoexpression and from 78 HPV negative, 66 (84.6%) cases had no p16 immunoreaction and this association was statistically significant (p = 0.002; Table 2). The same results were observed in the stratified analysis by group and in the young group, 50% of HPV positive were p16 negative.

Figure 1.

p16ink4 immunolabeling in oral SCC of young patient with HPV16 positive (Carazzi's hematoxylin counterstain, ×100).

The median follow-up periods (ranges) were 22.2 (3.5–330.4) months in the young group and 29.6 (1.0–180.3) months in the control group. No differences were observed in OS between the young and control groups (logrank, p = 0.32). A trend for better survival was observed among patients with HPV positive tumors independent of age. However, OS rates were not statistically different (Fig. 2). In addition, the young HR-HPV negative group had a tendency to show worse OS than control; in 10 years, it was 8.5% compared with 32.2% of the control (Table 3).

Figure 2.

Kaplan–Meier OS curve stratified by HR-HPV positive. (a) Young and control patients grouped (logrank, p = 0.12). (b) Young group (logrank, p = 0.10). (c) Control group (logrank, p = 0.15).

Table 3. Overall and disease free survival probability according to study group
inline image

The median time to relapse was 14.2 and 24.0 months in the young and control groups, respectively. Sixty-two recurrences were reported, 29 (46.8%) among young patients and 33 (53.2%) among the controls (p = 0.23). HPV positive did not improve DFS in the young group (p = 0.38) or among the controls (p = 0.25; Table 3).


The role of the presence of HR-HPV in the etiology of oral SCC is still a matter of debate especially among patients 40-year old or under. Kreimer et al.29 conducted a systematic review of HPV types detected by PCR technique in HNSCC worldwide. They identified 2,642 cases of oral SCC reported in 60 studies from 26 countries, of which ∼24% were HR-HPV, HPV16 was the most prevalent type (16%) and coinfection (3.6%) was uncommon. This is in agreement with the results presented in this work, where HPV16 was identified in 19.2% of the tumors and one case showed also HPV18 positivity. HPV prevalence is similar in Europe (16.0%) and North America (16.1%); however, it is higher in Asia (33.0%).29 A review of the literature uncovered no report specifically concerning South America. Paraffin-embedded specimens and fresh frozen tissues do not contribute to variability in HPV detection. Differences seem to be related to a small number of cases (<100) and poor quality biopsy specimens could lead to false negatives.29 Kantola et al.13 did not detect HPV in the biopsies of tongue analyzed, probably because of a small amount of tumor cells. The strength of this study is the size of the cohort and the homogeneity of tumor site: 114 SCC of the oral cavity from surgical tumor blocks with greater than 80% tumor cells were selected.

Several studies have suggested that HR-HPV could contribute to carcinogenesis in at least some cases of oral SCC.6–11 However, other authors have suggested that it is unlikely that HPV plays a significant role in the etiology, pathogenesis and clinical outcomes of oral SCC.13–17 In this study, HPV positivity was not correlated with sex, lesion site, clinical stage, histopathological grade, in agreement with other reports.6, 16 Many patients who were HR-HPV DNA positive reported tobacco and alcohol exposure as risk factors;16, 21, 30 however, Gillison et al.31 failed to verify any association between HPV16 and these factors in HNSCC. In the present work, 66.3% of patients, whose tumors had HPV16, were current smokers, suggesting that HPV positive alone is insufficient to result in malignant transformation,20, 21 although it could contribute to this process, especially in young smokers. HNSCC from traditional risk factors and HR-HPV could be molecularly homogeneous, as both result in the inactivation of p53 and pRb pathways.20, 31

HR-HPV DNA is more frequent in oral SCC in young patients than in those more than 60-year old,7, 18, 19 probably because of differences in sexual behavior.19, 32 Smith et al.33 identified the same type of HPV in genital and oral mucosa, although the transmission route remains unclear. Infection is most probably when the first intercourse occurs at an early age and with behavior involving multiple sex partners.32, 33 Open mouth kissing and oral sexual habits have been associated with oral HPV infection, especially in patients less than 30-year old.34, 35 Oral cancer was correlated with age at sexual debut in women.32 Previous studies focusing on young patients reported 0–12.5% of cases were HPV16 positive; however, these two reports had a very low number of cases, 15 and 8, respectively, and the latter considered young patients as individuals who were 45-year old or under.14, 17 In this study, HPV16 DNA was identified in 22 (19.2%) cases of oral SCC, 15 (68.2%) in tumors from young patients and 7 (31.8%) from control tumors, a statistically significant difference. In oropharyngeal SCC, HR-HPV could shorten the period for malignant clone formation, contributing to carcinogenesis in younger individuals.14

The presence of HPV is more frequent in oropharyngeal carcinomas than those affecting the oral cavity,14, 29, 32 probably because of the juxtaposition between SCC epithelium and lymphatic tissue similar to the female genital tract, invaginating crypts that provide a large epithelial surface and nonkeratinized mucosa that may facilitate viral access to basal cells.29, 32 In addition, unidentified cell membrane receptors with specific affinity for viral molecules may be present in the tonsils.14 However, an ultrastructural study described similarities between genital and buccal mucosa regarding histology, permeability and lipid compositions of the two epithelia.36

The immunohistochemical technique for p16ink4 can be used as a simple test to determine the presence of biologically active HPV in tumors,20 particularly in oropharyngeal cancers20, 37; however, in oral SCC, this correlation is controversial.16, 38 Association between HPV and p16 was verified in this study; however, in the young group, 50% of HPV positive were p16 negative. Weinberger et al.20 reported that tumors with this phenotype do not maintain transcriptionally active HPV; moreover, only the presence of E6 and E7 mRNA in a tumor is indicative of an active infection.39 RNA extraction, however, could not be performed in the specimens selected for this study.

It has been proposed that tumors with HPV positivity show improved OS rates18, 21, 22 although consensus regarding this matter has yet to be achieved.11 In this study, improved OS was observed in tumors with HPV positive in both groups; however, this difference was not statistically significant, possibly because of the low number of cases were HPV positive. Regarding DFS, no differences were verified between the groups, similar to that reported by Ritchie et al.18 The presence of HPV in HNSCC had been related to improved prognosis, especially when chemo or radiotherapy is used,22, 37, 40 probably because of the absence of field cancerization or enhanced chemoradiation sensitivity.40, 41

In summary, p16 immunoexpression was related to HPV DNA positive in oral SCC. HPV16 was significantly more prevalent in tumors of younger patients. Further studies to establish the risk of HPV presence in the development of oral SCC of young individuals are warranted.