Alternative therapeutic targets in patients treated with immune checkpoint inhibitors for advanced urothelial cancer: Real-world data of German comprehensive cancer centers (ATTICI biomarker study).

e16583 Background: Erdafitinib is the first targeted therapy for inoperable or metastatic urothelial carcinoma (mUC). Its use requires FGFR2/3 alterations in tumor tissue and failure of immune checkpoint inhibitor (ICI) therapy. Though, timely molecular testing in non-clinical trial settings can be challenging. This study explores the molecular background of mUC patients (pts) treated with ICI in a real-world setting, focusing on FGFR alterations and other targetable molecular changes. Methods: Following IRB approval, clinical data on mUC pts treated with ICI (01/13-09/24) and with available FFPE tumor samples from 9 German academic centers were documented. Data included demographics, clinical characteristics, treatment, and survival. FFPE samples prior to therapy underwent INVIEW Oncoprofiling NGS analysis, filtering for oncogenic variants. Interpretation used dbNSFP, OncoKB, ClinVar, and variant effect predictors. Further, IHC for potential therapeutic markers (Her2, EGFR, FGFR3, Nectin4, Trop2, MTAP, AR and ER status, FRalpha, Claudin18.2) was performed. Results: 182/251(72.5%) pts had clinical data, sufficient tissue samples for NGS, and consent available for analysis (median age 68 (IQR 60-77) years, 66.5% male. ECOG PS 0, 1, 2+ and missing in 30, 22, 13, and 35%). 82.4% of pts had UICC stage IV disease at start of ICI treatment (1st line 37%, 2 nd line 49%). mFU was 22.2 months. PFS and OS was 4.4 (95%CI 2.9-5.9) and 16.8 (95%CI 13.2-20.3) months, respectively. 22% responded (radiographic complete or partial response) to ICI monotherapy, while 45% progressed. On average, 9 oncogenic variants were found per patient. Potentially, targetable FGFR variants were present in 12/182 (6,6%), ERBB2 variants in 12/182(6.6%), TP53 mutations in 84/182 (46.2%), and SPEN mutations in 64/182 (35.2%). Other variants included NBN, APC, and KMT2C. By IHC, overexpression of Her2, EGFR, FGFR3 and Trop2 was noted in 53/139 (38.1%), 18/142 (12.4%), 67/140 (47.9%), and 80/140 (57.1%), respectively. CPS was ³10 in 10/142 (7.0%). Nectin4 was positive (H-Score > 100) in 31,7%, MTAP loss was noted in 88/142 (62.0%). AR and ER positivity and FRalpha were rare, and Claudin18.2 was not expressed. Comparing ICI-responders and non-responders, FGFR (15/45 33% vs. 47/78 60%) and HER2 (19/45 33% vs. 36/78 46% overexpression was more frequent in non-responders; other molecular parameters assessed did not differ between these groups. Conclusions: Frequency of FGFR alterations was low, though further targetable gene alterations as well as overexpression of alternative ADC target Her2, especially in ICI non-responders was detected. Frequent loss of MTAP expression points towards PRMT5 inhibitor therapy in some mUC pts. Our data favours early, comprehensive and consistent testing of pts with a mUC at the start of ICI therapy.