Prognostic risk factors in low stage nonseminumatous testicular germ cell tumors (NTGCT)

Objective: Percentage of embryonal carcinoma (%EC) and vascular invasion are known risk factors to predict pathological stage II disease in clinical stage I NSGCT. Proliferation markers, cell cycle regulators, proteases and cell to cell adhesion molecules have gained increased interest in some human malignancies. In NSGCT only few studies have been performed with regard to the clinical utility of pathohistological and immunohistochemical prognosticators for lymph node involvement Aim of our study was to determine if primary tumor tissue expression of Ki-67, PCNA, p53, bcl-2, cathepsin D and e-cadherin ,%EC and VI would distinguish pathological stage I from pathological stage II disease. Patients and Methods: All 149 patients underwent staging retroperitoneal lymphadenectomy (RPLND). Primary tumors were analyzed for percentage of tumor composed of EC, Yolk sac, seminoma, teratoma as well as presence of VI. Utilizing monoclonal antibodies nuclear PCNA, Ki-67, p53. bcl-2 and cytoplasmic cathepsin D and e-cadherin expression was measured immunohistochemically . Statistical analysis was performed by uni- (URA) and multivariate logitic regression analysis (MRA). Results:URA identified %EC (p<0.001) and presence of VI (p<0.0001), PCNA-EC (p<0.00l), p53-EC (p<0.03) and e-cadherin - EC (p<0.04) to be significant risk factors associated with pathological stage II; Ki-67 and bcl-2 did not correlate with pathologicalstage. MRA revealed only %EC (p<0.0001) and presence of VI (p<0.001) as significant independent prognosticators for lymph node metastases. Applying cut-off values of 45%EC and absence of VI in clinical stage I NSGCT, 91% of pathological stage I patients could be predicted correctly; applying cut-off values of >80%EC and presence of VI. 88% of pathological stage II were correctly predicted. Alltogether, 89% of all 149 clinical stage I NSGCT were correctly predicted using pathohistological risk factors. Discussion: Based on our results determination of %EC and presence/absence of VI correctly predicts not only pathological stage H patients but also pathological stage I patients. We suggest that all clinical stage I NSGCT should have their primary tumors analyzed for quantitative pathology in order to decide for RPLND or surveillance.

Duke Scholars

Author Judd Wendell Moul Urology