Secondary retroperitoneal lymph node dissection in the management of recurrent germ cell testicular tumors

All patients undergoing redo postchemotherapy (PC) RPLND between 1975 and 1993 were retrospectively reviewed. Statistical analysis was done comparing outcomes of this "redo" group to "primary" PC RPLND. Also, patients operated initially at Indiana University vs. elsewhere were compared. Intraoperative technical considerations in the redo group were noted. Survival at two and five years was significantly less in redo PC RPLND than in primary PC RPLND (p=0.00001). In redo cases, viable cancer was diagnosed in 37% of resected specimens, teratoma in 40%, and necrosis in 23%. Redo PC RPLND was extensive and entailed resection of adjacent organs in 34% of cases. In-field relapse was significantly higher in the subset of patients operated initially outside IUMC (p=0.000001). Failure to completely resect tumor at the initial PC RPLND was associated with increased morbidity owing to the need for redo RPLND with its attendant demands for wide tumor clearance. Redo RPLND is a technically extensive operation and one third of patients required adjacent organ resection. Patients with cancer at redo PC RPLND have an compromised prognosis (34-46%) compared to those with cancer found at primary PC RPLND (70%). (p=0.00001). Patients referred for PC RPLND after primary chemotherapy (n-404) without other associated risk factors represent only 50% of our cases ("clean" subgroup). These have a much better prognosis than our redo group (n-212, p=0.0001). Only 5% of our primary PC RPLND cases (20/404) in this "clean" subgroup have died of cancer. Therefore, the best chance for surgical cure is the first chance. If the PC RPLND surgery is recognized as incomplete (n=92), or if there is subsequent local relapse requiring redo (n=212), the outcomes are significantly impacted in terms of relapse and survival (p=0.0001).