Methods: HPV DNA was identified using a multiplex tandem HPV E6 polymerase chain reaction (PCR) assay and p16 expression by semiquantitative immunohistochemistry.
Results: p16 distribution was essentially bimodal (42 % of cancers had ≥70 % positive staining, 52 % <5 % positive, 6 % between 5 and 70 %). Cancers with 5 to <50 % staining had similar characteristics to the p16 negative group, and cancers with 50 to <70 % staining were consistent with the ≥70 % group. Using a p16 cut-point of 50 %, there were 25 % HPV DNA positive/p16 negative cancers and 1 % HPV DNA negative/p16 positive cancers. HPV DNA positive/p16 negative cancers had outcomes similar to HPV DNA negative/p16 negative cancers.
Conclusions: 50 % is a reasonable cut-point for p16; HPV DNA positive/p16 negative OSCCs may be treated as HPV negative for clinical purposes; HPV DNA/p16 testing may add no prognostic information over p16 alone.