p16 expression as a prognostic and predictive marker in high-grade localized osteosarcoma of the extremities: an analysis of 357 cases

The potential prognostic and predictive value of p16 in high-grade localized osteosarcoma of the extremities has been recently investigated in small series of cases, and the results from different studies were somewhat controversial. A retrospective immunohistochemical analysis of p16 expression was performed in a series of 357 patients, included in different neoadjuvant chemotherapy protocols from 1986 to 2010, to explore its potential prognostic and predictive value. Immunohistochemistry was performed with a commercially available p16 monoclonal mouse antibody. Follow-up data were available in all cases with a median of 120 months. Positivity for p16 was detected in 70.6% (252/357) of cases. The p16 expression did not differ by age, sex, tumor site, histologic subtype, tumor volume, surgical margin, serum alkaline phosphatase levels, and lactate dehydrogenase levels. In the different chemotherapy protocols included, the incidence of p16 expression was similar. The absence of p16 expression was significantly associated with an adverse disease-free survival (P = .04) and overall survival (P = .05) when compared with the presence of p16 expression. At the multivariate Cox regression analysis, p16 expression lost its prognostic significance. Multivariate logistic regression analysis showed that as p16 expression was the only statistically significant parameter to predict the pathological response to neoadjuvant chemotherapy treatment with an odds ratio of 3.025 (P < .001) for “good” chemotherapy response. Our data indicate that the negative expression of p16 is associated with a reduced rate of good response to primary chemotherapy and to a worse probability of survival, although it was not confirmed as an independent prognostic biomarker after multivariate analysis.