Management of relapsed stage I nonseminomatous germ-cell tumor with retroperitoneal only relapse.

5035 Background: For patients (pts) with stage 1 NSGCT on active surveillance (AS), the most common site of relapse is the retroperitoneum (RP). For those with RP only relapse and normal tumor markers, treatment (tx) options include chemotherapy (chemo) or retroperitoneal lymph node dissection (RPLND). We describe the characteristics of pts on AS with stage I NSGCT who relapsed in RP only and were treated w/ either chemo or RPLND. Methods: The prospectively maintained Indiana University testicular cancer database was queried for pts with stage 1 NSGCT on AS w/ RP relapse only between 1990-2023. Pts were categorized into a chemotherapy or surgery group based on tx at relapse. Comparisons between groups were done using Chi-square tests for categorical variables or Wilcoxon test for continuous variables. Kaplan-Meier method was used to analyze progression-free survival (PFS) and overall survival (OS) using the log rank test to compare groups. Results: We identified 143 pts with stage 1 NSGCT on AS w/ disease relapse in the RP only. The median age at diagnosis was 28.8 yrs (range, 15.6-61.5). Predominant histology was embryonal in 49.0%, mixed in 16.1%, seminoma in 15.4%, teratoma in 10.5%, yolk sac tumor in 7.7%, and choriocarcinoma in 0.7%. IGCCCG risk was good in 140 pts (98.0%) and intermediate in 3 (2.1%). At time of RP relapse, 116 pts (81.1%) were treated with chemo and 27 (18.9%) w/ RPLND. For those treated w/ chemo, the median time to relapse on stage I AS was 5.2 months (0.5-259.2) vs. 9.0 months (1.68-75.24) for those treated w/ RPLND (p=0.01). Chemo regimens used were BEPx3 (82.8%), EPX4 (9.5%), BEPX3 +EPX1 (2.6%), and other (5.2%). 55 of the 116 pts (47.4%) treated w/ chemo required a post-chemo RPLND (pcRPLND). In those treated w/ surgery, RPLND path was pure GCT in 16 pts (59.0%), pure teratoma in 4 (14.8%), malignant transformation (MT) of teratoma in 3 (11.1%), both GCT and teratoma in 2 (7.4%), both teratoma and MT in 1 (3.7%), and normal in 1 (3.7%). Tumor markers were normal for all 27 pts treated w/ surgery. All but two pts treated w/ surgery had RP lymph nodes <3cm in size. With a median follow-up of 34.4 mos, 20 pts relapsed after initial tx: 17 (14.7%) in the chemo group and 3 (11.1%) in the surgery group (p=0.001). The 3 in the surgery group were treated w/ chemo at time of relapse; 2 of those pts remain NED. The other pt remains AWD w/ MT of teratoma. The 5-yr OS for pts treated w/ chemo was 93% vs 100% for surgery, and 5-yr PFS was 80% for chemo vs 74% for surgery. Conclusions: In pts on AS with stage I NSGCT w/ RP only relapse, there was no difference in 5-yr PFS or OS for those treated w/ RPLND vs. chemotherapy. Almost half of the pts treated w/ chemo needed a pcRPLND. When pts were carefully selected for RPLND at time of relapse, most were cured with single-modality therapy. In those treated w/ surgery, most had RP lymph nodes <3cm, all had normal tumor markers, and the median time to relapse while on AS for stage I disease was longer.