2835 - Genomic analysis of urothelial cancer and associations with treatment choice and outcome

Background

The use of next generation sequencing (NGS) is in a constant rise, without clear evidence to its utility in guiding treatment choices in urologic malignancies. Here our aim was to retrospectively assess the utility of NGS in selecting advanced treatment lines and define the response rates (RR) to chemo and IO in a real-life cohort.

Methods

mUC patients receiving first-line chemo were included. Paraffin-embedded pathological samples were subjected to FoundationOne genomic analysis (Roche) and to PD-L1 IHC staining using the FDA-approved Ventana’s companion diagnostic test. Reports were anonymized and analysis was performed using Excel (Microsoft Office). Response was defined as either a clinically or radiologically-significant improvement.

Results

60 pts were included in this analysis with a median age of 69, of which 87.7% were men and 70% were current or past-smokers. RR to first line chemo or IO was 66% (95% CI 50-82) and 33% (95% CI 12-64), respectively. RR to second-line IO was 14% ((95% CI 4-40). Median TMB was 10 Mut/Mb (range 1-62) with no difference between smokers and non-smokers. 31% pts had tumor PD-L1 staining of > 5% and 3.7% were MSI-high. 107 genes were altered across all pts. The median number of genomic alterations (mutations (muts), amplifications (amps), deletions (dels) and re-arrangements) was 7 (range 2-17). The most frequent were TERT promotor and TP53 mutations, occurring in in 39 (69.6%) and 30 (53.6%) pts, respectively. FGFR3 alterations were found in 9 (15%) pts (7 mut and 2 fusions), 12 pts (20%) had alterations in ERBB2/Her2 (6 amps, 3 muts, 3 equivocal amps), BRCA1 mutations were found in 6 (10%) pts, and TSC1 alterations were found in 6 (10%) pts (4 muts and 2 dels). Altogether, about 40% of patients had potentially actionable alterations; of these, to date, the treatment of 8 patients (13%) was consequently altered. The RR to IO was non-significantly higher in pts with TMB>10 (29% VS 20%, p = 0.1) across all pts in this cohort.

Conclusions

RR to first-line chemo is higher than to first-line IO, and the association between TMB and response to IO is still debatable. NGS can yield potentially ’actionable’ alterations in about 40% of UC pts, and can change the treatment paradigm in the third of them.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

On behalf of the Israeli GU oncology group.

Funding

Has not received any funding.

Disclosure

D.L. Sarid: Honoraria (institution), Advisory / Consultancy, Travel / Accommodation / Expenses: Roche. All other authors have declared no conflicts of interest.