Reoperative Retroperitoneal Lymph Node Dissection for Metastatic Germ Cell Tumors: Analysis of Local Recurrence and Predictors of Survival

Abstract

While reoperative retroperitoneal lymph node dissection results in durable long-term survival, outcomes are comparatively worse than in patients who undergo initial adequate resection. We identified predictors of cancer specific survival and correlated technical aspects of initial resection to local recurrence in patients treated with repeat retroperitoneal lymph node dissection. We reviewed subsequent data on 203 patients treated with reoperation for recurrent retroperitoneal germ cell tumor after initial retroperitoneal lymph node dissection with local relapse. We used multivariate Cox proportion hazard models for cancer specific survival and multivariate logistic regression for local recurrence. The only 2 factors associated with local recurrence at lymph node dissection were incomplete lumbar vessel division at initial resection (p<0.01) and teratoma histology in the reoperative pathology specimen (p=0.01). Median followup was 73 months. Initial survival analysis including preoperative variables indicated that active cancer at initial operation (p=0.04), increased serum tumor markers (p=0.02), M1b stage (p<0.01) and salvage chemotherapy (p=0.01) were independent predictors of worse cancer specific survival. After introducing the final pathological data from reoperation into the final multivariate model only active cancer at reoperation (p<0.01), M1b stage (p=0.01) and salvage chemotherapy before reoperation (p=0.02) retained the association with worse oncologic outcomes. Tumor biology and inadequate surgical technique (incomplete lumbar ligation) are associated with local recurrence after initial retroperitoneal lymph node dissection. Decreased cancer specific survival is expected in this population, mostly driven by active cancer in the final pathology specimen.