|Home | About | Journals | Submit | Contact Us | Français|
Several studies have found human papillomavirus virus (HPV) in tissue from head and neck squamous cell carcinomas (HNSCCs), although the number of positive cases varies greatly from study to study. The extent and molecular epidemiology of HPV in HNSCC were assessed within cases drawn from southeast Scotland by performing broad‐spectrum, real‐time HPV polymerase chain reaction (PCR) on DNA extracted from 100 cases of HNSCC in formalin‐fixed, paraffin wax‐embedded material. All HPV‐positive specimens were genotyped and sampled by laser capture microdissection. Pure samples of tumour, and, where possible, dysplastic and normal epithelium were then submitted for further HPV PCR and genotyping to investigate the sensitivity of the technique in small tissue samples. 10 of 100 cases tested positive for HPV, with 8 of these being derived from Waldeyer's ring. HPV DNA was found in adjacent epithelium in two of four cases where this was available. These findings confirm that HPV is likely to be involved in a subset of HNSCC in this population and that successful amplification of HPV nucleic acid is possible even using small amounts of paraffin wax‐embedded tissue.
Involvement of persistent human papilloma virus (HPV) infection in the development of squamous carcinoma of the cervix is now established. Accumulated data have also shown the presence of HPV DNA in head and neck squamous cell carcinomas (HNSCCs), suggesting a similar aetiological role for HPV in these tumours. Specifically, several studies have implicated the oropharynx and, in particular, Waldeyer's tonsillar ring as a predilection site for HPV‐associated carcinogenesis,1,2,3 and have suggested that these tumours represent a distinct clinical subgroup of squamous carcinomas with an improved prognosis.4 This evidence has led to suggestions that HPV DNA detection might be used in the clinical management of HNSCC5 provided that the detection technology was sufficiently robust for preserved pathological specimens.
In all, 101 archival cases of HNSCC were selected, composed of 25 women and 76 men aged 21–87 years (mean age 61.45 years). The cases included 64 laryngeal carcinomas (mean (SD) age 62.63 (10.73) years) and 36 carcinomas from around Waldeyer's ring (tonsil, base of tongue, posterior pharynx and uvula; mean (SD) age 61.06 (11.19) years). A total of 24 controls of benign tonsillar tissue were also identified, comprising 14 females and 10 males aged 14–87 (mean 35.1) years.
Thick sections were cut from each paraffin wax block and DNA was extracted using lysis buffer containing proteinase K at 50°C for 4 h. Real‐time PCR incorporating HPV‐specific broad‐spectrum primers was performed. Full details of the methods are as described by Cubie et al.6 A 150 nucleotide product was amplified using the GP5+6+ primers7 on a LightCycler. A positive HPV product was identified via Tm of generated amplicon visualised by incorporation of Sybr Green dye during amplification. Any positive sample identified by the real‐time protocol was subjected to the Roche line blot assay for genotyping, which involved amplification using the PGMY primer set8 designed to generate a 450 nucleotide amplicon. Control primers for the amplification of the β globin gene were also included in the PCR to assess the integrity of the cellular extract. Amplicons were then hybridised to a nylon strip containing 27 immobilised type‐specific HPV probes and 2 levels of β globin control probes according to the technique described by Gravitt et al.9
Each tumour that tested positive for HPV DNA was microdissected using laser capture microscopy. Sections of thickness, 7 µm were cut, dried and counterstained with Mayer's haematoxylin. Tumour epithelial cells were microdissected on to caps (CellPix II, Arcturus), DNA was extracted as above and PCR was repeated on these samples. For each case, two separate samples composed purely of tumour cells were taken, one containing approximately 100 cells and one with approximately 300 cells. Tissue was also taken from adjacent normal or dysplastic epithelium where applicable.
None of the 24 controls tested positive for HPV DNA by real‐time PCR whereas 10 of 101 (10%) cases did. Two cases from the laryngeal group yielded weak HPV DNA PCR‐positive results, which could not be repeated. These two cases did not test HPV positive after application of the PGMY primer set, and therefore could not be genotyped. Eight cases derived from Waldeyer's ring contained HPV DNA by real‐time PCR. All eight were strong, repeatable positives and genotyping was possible in every case.
The mean age in the 10 positive cases was 50.6 years (95% CI 44.11 to 57.09). The mean age in the 91 HPV‐negative cases (excluding controls) was 63.3 years (95% CI 61.17 to 65.5). This age difference is significant (unpaired Student's t test, t=−3.69, p<0.001).
HPV genotyping showed that 7 of 8 strong‐positive cases contained HPV type 16. The eighth strong‐positive case contained HPV type 11. No mixed infections were detected (all 8 positives tested positive for β globin; table 11).
HPV DNA was not detected in microdissected tissue from the two weakly positive carcinomas from the larynx. Of the eight cases from Waldeyer's ring, five tested positive for HPV DNA when microdissected (table 22).). HPV DNA was detected in one of three microdissected samples of adjacent dysplastic epithelium. The adjacent epithelium in case 7 had a non‐dysplastic, warty appearance and this epithelium and the background normal epithelium contained HPV DNA.
A previous study of HPV in head and neck carcinoma in a UK population concentrated on oral lesions, with HPV found in up to 46% of normal and tumour tissues.10 We have investigated the presence of HPV in tonsillar and laryngeal carcinomas, and our findings replicate the results of similar studies performed in other countries, confirming that any future HPV‐directed diagnostic and treatment strategies for HNSCC are relevant to this UK‐based population. We have confirmed that HPV seems to have a predilection for involvement in carcinomas arising in Waldeyer's ring, suggesting that detection of HPV in this subset of tumours may be of clinical relevance not least because HPV‐positive carcinomas seem to have a better prognosis, possibly because of a greater sensitivity to radiation treatment than conventional HNSCC.11,12 Paradoxically, a recent study found that HPV‐positive HNSCCs were associated with earlier nodal metastasis and suggested that screening high‐risk populations for HPV may be valuable in preventing premalignant lesions progressing to carcinoma.13
Microdissection confirmed that HPV DNA was present in carcinoma cells. The detection of HPV in some of the small microdissected samples has shown the sensitivity of these techniques in formalin‐fixed paraffin wax‐embedded material, which is of relevance, as biopsy specimens submitted from head and neck sites may be limited in size.
The presence of HPV DNA in abnormal epithelium adjacent to the invasive carcinoma in two cases and normal epithelium in one case suggests that the detection of HPV in the epithelium of Waldeyer's ring could indicate an increased risk of neoplastic transformation with possible future diagnostic utility, as is being developed for the cervix. Previous studies that investigated the presence of HPV DNA in normal head and neck epithelium found wide‐ranging results,14 and therefore further work in this area is necessary before HPV testing can be used clinically to identify patients who are potentially at risk of head and neck malignancy.
This work was supported by a grant from the Lothian University Hospitals NHS Trust Cancer Services Committee. We thank Dominic Rannie for his guidance in the use of the laser capture microscope.
HNSCC - head and neck squamous cell carcinoma
HPV - human papillomavirus virus
PCR - polymerase chain reaction
Competing interests: None.
Ethical approval: This study received ethical approval from the Lothian Research Ethics Committee, Edinburgh, UK.