This is the first report utilizing ACS NSQIP to review surgical approaches as well as the impact of trainee involvement on clinical outcomes. The increased complication rates and cost of healthcare might be mitigated by awareness, investment in surgical simulation laboratories, and competency assessment.
PURPOSE According to National Comprehensive Cancer Network guidelines, adjuvant chemotherapy (AC) has been advocated after primary retroperitoneal lymph node dissection (RPLND) to reduce the risk of relapse in pathologic nodal (pN) stage pN2 or pN3, whereas surveillance is preferred for pN1. We sought to explore the oncologic efficacy of primary RPLND alone for pathologic stage II in nonseminomatous germ cell tumors (NSGCTs) to reduce overtreatment with chemotherapy. METHODS Patients with pathologic stage II NSGCT after primary RPLND between 2007 and 2017 were identified. Patients were excluded for elevated preoperative serum tumor markers, receipt of AC, or if pure teratoma or primitive neuroectodermal tumor elements were found in the retroperitoneal pathology. RESULTS We identified 117 patients with active NSGCT in the retroperitoneum after primary RPLND. We excluded seven patients who lacked meaningful follow-up and 13 patients who received AC. There were 97 patients treated with RPLND alone: 41 pN1, 46 pN2, and 10 pN3. In total, 77 of 97 patients had not recurred after a median follow-up time of 52 months. The 2-year recurrence-free survival (RFS) was 80.3%, and the 5-year RFS was 79%. No differences in RFS were noted among nodal stage—pN1, pN2, and pN3—on Kaplan-Meier analysis. Lymphovascular invasion in the orchiectomy specimen, a high-risk pathologic feature, was also predictive of recurrence after primary RPLND. All 20 patients who recurred were treated with first-line chemotherapy and remained continuously disease free. CONCLUSION Most men with pathologic stage II disease treated with surgery alone in our series never experienced a recurrence. We did not observe a difference in recurrences between patients with pN1 and pN2. The recommendation for AC for pN2 disease may be overtreatment in most patients.
The prediction model was validated internally using the original dataset. The bootstrap split sample validation procedure was utilized for prediction model validation. A bootstrap sample consisted of a training set and a validation set with a 60:40 ratio in size of the original study sample, respectively. Patients were randomly selected into the training or the validation set with replacement. For each bootstrap sample, the selected prediction model was fit based on the training sample. The predicted probability of the outcome for each patient was computed for the validation sample by applying the model parameter estimates obtained from training sample. Area under ROC Curve (AUC) for the validation sample were then calculated based on the predicted probability and the actual event proportion. AUC for training set was also calculated to compare with AUC from validation set. A total of 200 bootstrap samples were created, and hence 200 repetitions of validation process were conducted. The mean AUCs of all bootstrap AUCs were calculated for training sets and validation sets. The bootstrap confidence bands of AUCs were created by selecting 2.5 th and 97.5 th percentiles of the 200 bootstrap AUCs for the training sets and the validation sets. The internal bootstrap validation procedure was performed using SAS 9.4 (Cary, NC).
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.