Genome-wide aneuploidy detected in circulating cell-free DNA of patients with metastatic urothelial cancer as a predictive biomarker of response to pembrolizumab.

4584 Background: Pembrolizumab is the standard of care second-line treatment of metastatic urothelial cancer patients. However, success rates are low and there is risk for significant toxicity. Therefore, reliable biomarkers to select patients who will or will not benefit from pembrolizumab are urgently needed. An earlier study showed that a high aneuploidy score (i.e. Z-score ≥ 5) at baseline in cell free DNA (cfDNA), based on the modified fast aneuploidy screening test-sequencing system (mFast-SeqS), associated with poor response to pembrolizumab. However, a low aneuploidy score (Z-score < 5) at baseline was not discriminative. The aim of this study was to evaluate early during treatment (after the first cycle pembrolizumab), whether the cfDNA based mFast-SeqS aneuploidy scores can discriminate patients with expected benefit from therapy from those who will not benefit. Methods: The mFast-SeqS to assess aneuploidy was performed on blood samples collected from 74 patients with metastatic urothelial cancer (mUC) at baseline and after cycle 1 of pembrolizumab. Responders were defined as patients with stable disease (SD) or partial/complete response (PR, CR) according to RECIST v1.1. and continued treatment beyond 6 months from initiation of pembrolizumab. Non-responders did not fulfil these criteria. Results: Paired samples were available in 60 out of 74 patients. At baseline 25% (n=15) of patients had a high aneuploidy score and 75% (n=45) had a low aneuploidy score. Two patients (13%) with a high baseline score were responders and 13 (87%) were non-responders. In all responders, there was a marked decrease in aneuploidy score after one cycle of pembrolizumab (p=0.005), irrespective of their baseline score. About half of patients (47%) with a low baseline score were non-responders and in those patients there was a significant increase of the aneuploidy score post cycle 1, when compared to the responders (p=0.039). Conclusions: The aneuploidy score based on mFast-SeqS may stratify mUC patients for expected lack of benefit early during treatment (after 1 cycle) with second line pembrolizumab, providing an opportunity to switch treatment and to spare patients from potential unnecessary toxicity. Clinical trial information: NCT03263039 .