Quantitative circulating tumor DNA (ctDNA) assessment in patients (pts) with advanced urothelial carcinoma (UC) treated with pembrolizumab (pembro) or platinum-based chemotherapy (chemo) from the phase 3 KEYNOTE-361 trial.

person Thomas Powles Barts Cancer Institute, Queen Mary University of London, London, United Kingdom info_outline Thomas Powles, Wayne Yen-Hwa Chang, Yoshiaki Yamamoto, José Muñoz-Langa, Felipe Reyes, Avivit Peer, Evan Y. Yu, Graham Cohen, Anja Lorch, Abhishek Bavle, Blanca Homet Moreno, Julia Markensohn, Mackenzie Edmondson, Cai Chen, Razvan Cristescu, Carol Elaine Pena, Jared Lunceford, Seyda Gunduz

Authors person Thomas Powles Barts Cancer Institute, Queen Mary University of London, London, United Kingdom info_outline Thomas Powles, Wayne Yen-Hwa Chang, Yoshiaki Yamamoto, José Muñoz-Langa, Felipe Reyes, Avivit Peer, Evan Y. Yu, Graham Cohen, Anja Lorch, Abhishek Bavle, Blanca Homet Moreno, Julia Markensohn, Mackenzie Edmondson, Cai Chen, Razvan Cristescu, Carol Elaine Pena, Jared Lunceford, Seyda Gunduz Organizations Barts Cancer Institute, Queen Mary University of London, London, United Kingdom, Taipei Veterans General Hospital, Taipei, Taiwan, Yamaguchi University Hospital, Yamaguchi, Japan, Hospital Universitario Y Politécnico La Fe, Valencia, Spain, Fundación Arturo López Pérez, Santiago, Chile, Rambam Health Care Campus, Haifa, Israel, Fred Hutchinson Cancer Center and University of Washington, Seattle, WA, Life Groenkloof Hospital, Mary Potter Oncology Centre, Groenkloof, Pretoria, Gauteng, South Africa, Universitätsspital Zürich, Zürich, Switzerland, Merck & Co., Inc., Rahway, NJ, Istinye University Liv Hospital, Istanbul, Turkey Abstract Disclosures Research Funding Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA Background: ctDNA is emerging as a potential biomarker of disease in early-stage bladder cancer but is less understood in advanced UC. We present a retrospective analysis of pre-treatment and on-treatment ctDNA data by clinical outcomes with pembro monotherapy versus chemo in pts with advanced UC from the phase 3 KEYNOTE-361 trial (NCT02853305). Methods: Pts with previously untreated advanced UC were randomly assigned 1:1:1 to pembro + chemo, pembro alone, or chemo alone. Tumor tissue mutations were evaluated with whole exome sequencing of tumor and matched normal DNA. Plasma ctDNA was evaluated using the Guardant Health (GH) Guardant 360 LDT assay. ctDNA changes from pre-treatment cycle 1(C1) to on-treatment cycle 2 (C2) were quantified by maximum variant allele frequency (maxVAF) of tumor tissue–specific mutations (tumor-informed [TI] approach) or GH molecular response (MR) score. Association between C2/C1 ratios or baseline TI maxVAF and clinical outcomes (ORR, PFS, and OS) were evaluated. Nominal statistical significance of maxVAF was prespecified at 0.05 (hypothesis, negative). Results: ctDNA samples from 263 pts (n = 131, chemo; n = 132, pembro) were analyzed. Clinical characteristics and baseline ctDNA levels within arms were similar. Lower C1 maxVAF was associated with improved ORR, PFS, and OS in the pembro arm ( P < 0.01) and was robust to adjustment for TMB and PD-L1, but not in the chemo arm ( P > 0.05). Larger C2 reductions in ctDNA levels were observed in the chemo vs pembro arm (median ratio of C2/C1 TI maxVAF, 0.03 vs 0.71, respectively); similar observations were made for MR score. ctDNA reductions were associated with improved ORR, PFS ( P < 0.001), and OS ( P < 0.01) for TI maxVAF in the chemo arm; MR score was significantly associated with ORR and PFS ( P < 0.01). Associations were stronger in the pembro arm ( P < 0.001) and were robust to adjustment for TMB and PD-L1. ctDNA changes from C1 to C2 did not show independent explanatory value for OS when RECIST v1.1 response status was added as a variable. Conclusions: ctDNA levels at baseline appear prognostic for pembro. Reductions in ctDNA during the first treatment cycle were associated with outcome. Distinct patterns of ctDNA response were observed with the different treatments, and stronger associations with long-term clinical outcome were observed with pembro. Association of OS for both pembro and chemo was not retained when adjusting for tumor response. Early ctDNA dynamics did not offer additional benefit in predicting outcome beyond radiographic assessment of tumor size change. Clinical trial information: NCT02853305.

Clinical