Sensitivity of In-Situ Hybridization (ISH) for High-Risk Human Papillomavirus (HR-HPV) in Head & Neck Squamous Cell Carcinoma (HNSCC)
TM Stevens, ST Dunn, S Caughron, Z Gatalica. Creighton University, Omaha, NE; The University of Oklahoma Health Sciences Center, Oklahoma City, OK; Yellowstone Pathology Institute, Billings, MT
Background: High risk HPV has been implicated in the etiology of a subset of HNSCC, particularly oropharyngeal squamous cell carcinomas. It has also been reported that HPV-related HNSCC carry a more favorable prognosis and are more radiosensitive than their HPV-negative HNSCC counterparts. HR-HPV has been detected in 20-90% of HNSCC, but is generally found at lower transcript levels than in cervical squamous cell carcinomas. Accordingly, a highly sensitive and specific assay for HPV detection is necessary. ISH for HR-HPV allows for direct visualization of HPV integration, thus circumventing the contamination problems inherent in other molecular assays. The sensitivity of ISH remains untested in HNSCC, however.
Design: In this study, 101 HNSCC were tested for high risk HPV by in-situ hybridization method (ISH iViewTM Blue Plus, Ventana, Tucson, AZ, performed at Creighton Medical Laboratories) and results compared to PCR (YPI) and Linear Array® (Roche Diagnostics, Branchburg, NJ, performed at University of Oklahoma).
Results: ISH for HR-HPV had sensitivities of 74 and 59%, when compared to PCR and Linear array, respectively (McNemar's test P>0.05 and <0.05 for PCR and Linear array, respectively). The negative predictive value of ISH for HR-HPV was 80% and 61%, as compared to PCR and Linear array methods, respectively. The specificity of ISH for detection of HR-HPV in HNSCC was 85% and 86%, respectively, when compared with PCR and Linear array. Tumor necrosis and the presence of bacterial contamination were identified as potential sources of false-positive ISHs. HR-HPV was detected in 53, 43, and 39% of HNSCC by Linear array, PCR, and ISH, respectively.
Conclusions: Compared to PCR and Linear array detection methods, ISH for HR-HPV in HNSCC shows good specificity. When compared against both PCR and Linear array as “gold” standards, ISH for HR-HPV in HNSCC has a relatively low NPV and low sensitivity, suggesting possibly that the more sensitive PCR and Linear array methods are more likely to detect clinically irrelevant “bystander” HPV DNA in HNSCC. In summary, ISH for HR-HPV is a highly specific test and allows for direct visualization of integrated HPV DNA thus allowing confident detection of biologically relevant HPV infection in HNSCC. In this series, the percentage of HNSCC harboring HR-HPV DNA is similar to previous reports.
Monday, March 22, 2010 1:00 PM
Poster Session II # 254, Monday Afternoon