589 Background: Serum tumor markers have been the cornerstone in the detection and surveillance of testicular germ cell tumors (GCT). However, the sensitivity of specificity of STM for detecting GCT is poor. Circulating tumor DNA (ctDNA) has demonstrated ability to detect GCT in both primary and post-chemotherapy retroperitoneal lymph node dissection (RPLND). Herein, we sought to determine the ability of ctDNA to predict GCT on P-RPLND by comparing patients’ RPLND histology with their ctDNA status stratified by seminoma and non-seminomatous germ cell tumor (NSGCT). Methods: Patients undergoing primary (P-RPLND) had prospectively collected plasma ctDNA. Patients were stratified to seminoma or NSGCT based on orchiectomy and STM. The association between ctDNA positivity and P-RPLND histology was assessed. Sensitivity (SN), specificity (SP), and positive (PPV) and negative predictive values (NPV) were calculated for ctDNA to detect GCT on RPLND. Results: 64 patients undergoing P-RPLND had a pre-operative ctDNA with a median age of 37 (IQR:29-42). Plasma was collected on average, 8 days pre-operatively. 27 (42.2%) patients underwent P-RPLND for seminoma and 37 (57.8%) for NSGCT. In the seminoma group, 16 and 11 patients had CS 2A and 2B disease. 24 (89%) of these patients had a (+) ctDNA. All seminoma patients had GCT present on P-RPLND. Of NSGCT patients, 10 had CS 1, 21 patients had CS 2A and 6 had CS 2B disease. 23 (62%) patients had a (+) ctDNA. 2 NSGCT patients had (+) ctDNA and negative RPLND. One of these patients developed a lung recurrence and the second had a rising ctDNA with negative surveillance imaging at 9 months. The SN, SP, PPV, and NPV in the NSGCT cohort were 88%, 85%, 91%, and 79% and in the seminoma cohort were 89%, 0%, 100%, and 0%. The 0% SP and NPV in the seminoma cohort were due to the lack of patients with negative RPLND histology. Conclusions: This is the first study evaluating test statistics of ctDNA in relation to RPLND histology, when stratified by seminoma or NSGCT. These findings suggest that ctDNA may be a useful tool in predicting retroperitoneal histology in patients being considered for P-RPLND. Persistently positive ctDNA after RPLND may be indicative of disease outside of the retroperitoneum. Contingency tables for ctDNA ability to predict GCT in patients undergoing P-RPLND, stratified by seminoma and NSGCT. All P-RPLND (N=64) Disease positive (n=51) Disease negative (n=13) Test Positive (n=47) 45 2 PPV=96% Test Negative (n=17) 6 11 NPV=65% SN=88% SP=85% NSGCT only (N=37) Test Positive (n=23) 21 2 PPV=91% Test Negative (n=14) 3 11 NPV=79% SN=88% SP=85% Seminoma only (N=27) Test Positive (n=24) 24 0 PPV=100% Test Negative (n=3) 3 0 NPV=0% SN=89% SP=0% SN: Sensitivity; SP: Specificity; PPV: positive predictive value; NPV: negative predictive value.
Thakker et al. (Sun,) studied this question.