Validation of a prediction model for avoiding post-chemotherapy retroperitoneal lymphadenectomy in patients with metastatic nonseminomatous germ cell cancer.

Abstract

Post-chemotherapy residual masses (PCRMs) may contain persistent cancer or teratoma in more than 50% of patients with metastatic non-seminomatous germ cell tumours (mNSGCTs). Retroperitoneal lymph node dissection (RPLND) is curative, but controversy exists about selection criteria for surgery. A validated prediction model by Vergouwe et al (2007) based on over 1000 patients was evaluated at our centre. mNSGCT patients treated with RPLND for PCRMs were identified from an electronic database. Typographical errors in the model were identified and corrected using their 2003 publication, but retaining the 2007 coefficients. Six clinical variables were included in the model and the calculated probability of benign tissue was compared with pathology. "Benign tissue only" was considered a positive test outcome in patients with a predicted probability of "benign tissue only" greater than 70%. Fifty-two (52) mNSGCT patients between 1980 and 2014 were evaluable. Median age was 32 years (range 17-52) and International Germ Cell Consensus Classification (IGCCC) prognostic stages were: good 46.2%, intermediate 32.7%, and poor 21.2%. Most patients received bleomycin/etoposide/cisplatin (BEP) chemotherapy and full bilateral RPLND. Pathology showed residual cancer or teratoma in 31 patients (59.6%) and benign findings in 21 patients (40.6%). Positive and negative predictive values and accuracy were 100%, 69%, and 73%, respectively. "Benign tissue only" was found in 100% of patients in whom this was predicted using our pre-determined criteria. This study involved a limited number of patients, but confirms the potential value of the Vergouwe et al model. Routine use of this prediction model in clinical practice should be tested for mNSGCT patients with PCRMs.