Article
Method-related bias in prevalence of oncogenic human papillomavirus (HPV) subtypes in head and neck squamous cell carcinoma
Search Medline for
Authors
Published: | April 4, 2012 |
---|
Outline
Text
Head and neck squamous cell carcinoma (HNSCC) of the oropharynx (OSCC) is known to be at least partially triggered by oncogenic subtypes of papillomaviruses (HPV). In OSCC, research mostly focused on HPV16 but essential variations in prevalence are reported. It is not yet completely clear, if these differences have to be linked to varying methods used.
Before starting HPV detection in >300 HNSCC, 100 OSCC samples of our cohort underwent analyses in five test methods based on PCR technique for validation purposes. All DNA samples were obtained from biopsies stored in TRIzol at -80°C until preparation. 100 ng each of aliquoted DNA sample underwent analyses using the following kits: BMT HPV 9G™ (BMT; Biometrix Techn. Inc.), digene HPV Genotyping RH (Qiagen), Inno-LiPA HPV (Innogenetics Inc.), Cobas™ and Linear Array™ (LA; Roche).
Different HPV detection results were obtained; e.g. for LA and BMT (two commercially available tests certified for cervical cancer) we obtained discordant results in 31/100 OSCC. Using LA, HPV was identified in 34 OSCC samples (34%); HPV16 was predominant (31 OSCC, 31%). In addition three further genotypes (HPV18, 31 and 33) were found. Using BMT, HPV-DNA was detected in 44 OSCC (44 %). However, HPV16 was just found in 22. Furthermore, one additional genotype (HPV35) was found and HPV31 was not detected. Of 100 OSCC samples, 69 showed agreement of both assays. In 45 cases both testing systems showed negative and in 24 cases concordant genotype-specific results (21 samples for HPV16).
The discordance highlights the need for standardization of HPV-DNA detection. Heterogeneity of the testing systems may be a reason for observed deviations which will be outlined in detail.
Supported by: This work was supported by LIFE-007 D9 and LIFE-006 B7.