CLONEVO: Preoperative abemaciclib for cisplatin-ineligible muscle-invasive bladder cancer (MIBC) with molecular response assessment.

Abstract

4520 Background: Up to 40% of MIBC patients are ineligible to receive standard neoadjuvant cisplatin-based chemotherapy creating a significant unmet need. Based on our prior findings of frequent cell cycle alterations, we conducted the first window-of-opportunity, investigator-initiated trial of the CDK4/6 inhibitor abemaciclib (abema) followed by radical cystectomy (RC) in MIBC (NCT03837821). Methods: Eligibility was MIBC appropriate for RC and cisplatin-ineligibility or refusal. Planned treatment was abema (200mg BID PO) for 4-8 weeks prior to RC. We planned to enroll 20 patients (accounting for 20% attrition). 16 evaluable patients provided 80% power to detect 0.75 effect size (α = 0.05, r = 0.5 between pairs). Whole-exome (WES) and RNA sequencing of pre- and post-abema tissues and serial evaluation of ctDNA WES were performed on Caris Life Sciences' platform. Results: 20 patients received abema for a median of 36 days. Median age was 73, 16/20 were males, and 5/20 had cT4. 3 didn't undergo RC, and 1 withdrew consent. Abema resulted in pathologic complete response in 18.8% (3/16) and downstaging in 31.3% (5/16). No unexpected safety signals were detected. Grade 3 abema-related adverse events included anemia (4/20), abdominal pain (1/20) and diarrhea (1/20). Imaging Mass Cytometry of pre- and post-abema tissues showed a significant reduction in RB1 phosphorylation after abema confirming on-target activity. Variant allele frequency of somatic mutations significantly decreased after abema by 20.5% (p = 0.04), confirming its role in decreasing tumor burden. Serial ctDNA showed a significant reduction in tumor fraction (TF) following abema by 28.6%. Post-TURBT pre-abema TF increased but rapidly decreased within 2 weeks of abema (19.36%), confirming TF reduction was driven by abema not TURBT. Patients with CCND1 amplification had the most significant decrease in TF (63.8%) highlighting CCND1 as a potential response biomarker. Abema significantly downregulated MKI67, CCNA2, and PCNA proliferation markers with log-fold changes of -1.2, -0.7, and -0.6. Gene set enrichment analysis showed significant downregulation of E2F targets and G1/S transition pathways. Patients who achieved pathologic downstaging had significant decrease in E2F pathway activity (-1.6 vs. -0.4, p = 0.01) confirming that abema suppressed E2F-dependent cell proliferation. Interestingly, abema significantly inhibited homologous recombination repair of double-strand DNA break (DSBs) (FDR = 0.001), particularly TOPBP1 and RAD51. Conclusions: This first trial of short-term preoperative abema in MIBC demonstrated promising efficacy and tolerability while modulating cell cycle-dependent pathways. Our findings support future trials investigating sequential abema with antibody-drug conjugates such as enfortumab vedotin, where abema's effects on DSBs repair augment treatment response. Clinical trial information: NCT03837821 .