A Pathology-Based Mathematical Model Predicts Nodal Metastasis in Squamous Cell Carcinoma of the Penis.
Alcides Chaux, Gustavo Guimaraes Cardozo, Fernando Soares, Isabella da Cunha, George J Netto, Antonio L Cubilla. Johns Hopkins University, Baltimore, MD; Hospital do Cancer A. C. Camargo, São Paulo, Brazil; Instituto de Patología e Investigación, Asunción, Paraguay
Background: Inguinal nodal metastasis is the strongest predictive factor for survival in patients with penile squamous cell carcinomas (SCC). Accurate tools are needed for defining which patients will benefit with a groin dissection.
Design: Clinical and pathological data from 192 patients treated for penile SCC were evaluated. Among multiple factors, using binary logistic regression, it was found that clinical status of lymph nodes (CN), tumor histological subtype (HS), histological grade (HG), and perineural invasion (PNI) were significant predictors of nodal involvement. The following logistic regression model (LRE) was constructed using these covariates,
LRE = -4.542 + 1.829*CN + 1.777*HS + 2.536*HG + 1.389*PNI
Where each covariate assume a value of 1 if present and of 0 if absent. The probability P(M) of nodal metastasis in any given patient was calculated using the following equation
P(M) = 1/(1 + e-LRE)
Where e is the Euler's number (e ≈ 2.71828) and LRE is the value obtained using the first equation.
Results: Using a cut-off point of 40% for P(M) 78.9% of patients with negative and 79.7% with positive inguinal nodes were correctly classified. Overall accuracy was 79.2% (95% confidence interval 72.7–84.7%). Most false positive patients presented enlarged inguinal nodes while in most of false negative cases inguinal nodes were clinically impalpable.
Conclusions: This equation estimates the probability of inguinal nodal metastases in patients with penile SCC and it is based on the significance of pathological prognostic factors in resected specimens. The method accuracy is about 80% for all cases. The calculations, which can also be made using a handheld calculator, take only a few seconds in a standard personal computer. No specialized or paid software is needed.
Category: Genitourinary (including renal tumors)
Tuesday, March 1, 2011 1:00 PM
Poster Session IV # 88, Tuesday Afternoon