Authors: Laura Marandino, Russell Madison, Daniele Raggi, Filippo Pederzoli, Marco Bandini, Patrizia Giannatempo, Alberto Briganti, Francesco Montorsi, Jeffrey S. Ross, Andrea Necchi
Published: 2021-03-02
DOI: 10.1200/jco.2021.39.6_suppl.479
Source: Full article
479 Background: MIBC is an aggressive disease and a limited number of patients (pts) will benefit from neoadjuvant chemotherapy (NAC) or immunotherapy. CGP may help identify prognostic factors, predictive biomarkers and targets for therapy. Methods: Hybrid-capture based CGP of treatment-naive tissue samples (n = 205) was performed on 3 cohorts of pts with MIBC: i) PURE-01 cohort (n = 144), treated with neoadjuvant pembrolizumab ii) NAC cohort (n = 31) iii) therapy-naïve radical cystectomy (RC) cohort (n = 30). Targetable gene alterations (GAs) were assessed according to the European Society for Medical Oncology (ESMO) Scale for Clinical Actionability of molecular Targets (ESCAT). A modified version including tumor mutational burden (TMB) (cut off 10 mut/Mb) was used. Logistic models were used to analyze associations with pathologic complete response (pCR, pT0N0) or downstaging (pT≤1). Kaplan-Meier method was used to estimate event free survival (EFS) and distant relapse free survival (DRFS). Univariable Cox regression analyses were run. Results: ESCAT tier ≥3A alterations were found in 164 (80%) pts in total and in 134 (65%) when TMB was excluded. 15% of pts had tier 1B GAs ( FGFR3 mutations [mut] or fusion), 54% tier 1C (TMB≥10 mut/Mb), and 16% had tier 2B GAs. Tier 3A GAs were found in 52% pts. ERBB2 amplification (8%) and ERBB3 mut (7%) were the most frequent tier 2B GAs. PIK3CA (24%) and ERBB2 mut (14%), followed by ATM (7%), were the most represented tier 3A GAs. In the PURE-01 cohort, TMB≥10 mut/Mb was significantly associated with pCR and downstaging at univariable and multivariable analyses. Univariable Cox regression analyses did not show association between TMB and EFS or DRFS. ERBB2/ ERBB3 GAs were associated with TMB≥10 (p = 0.003 in entire population, p = 0.048 in PURE-01 cohort). However, in the PURE-01 cohort there was no association between ERBB2/ ERBB3 GAs and response. In the NAC cohort, pCR rate was 27% in pts with TMB≥10 while none with TMB < 10 had pCR (p = 0.6). Downstaging was 55% in TMB≥10 vs 40% in TMB < 10 (p = 0.7). DRFS did not show significant difference according to TMB (p = 0.6). In the PURE-01 cohort, pCR rate and downstaging were numerically higher in pts with homologous recombination repair alterations (HRR+), but there was no significant association (p = 0.3). No association was found between HRR status and EFS or DRFS. In the NAC cohort pCR was 33% in HRR+ vs 19% in HRR- (p = 0.6). There was no difference in downstaging. At median follow-up of 19.2 months, mDRFS was longer in HRR+ vs HRR- (mDRFS NE vs 27.9 months p = 0.039). Conclusions: CGP identified GAs which potentially predict benefit from approved or investigational treatments or provide rational for clinical trial evaluation in a high proportion of pts. TMB≥10 was associated with response to pembrolizumab, while HRR GAs with DRFS after NAC. Cases with ERBB2/ 3 GAs featured higher TMBs but were not associated with response to immunotherapy.