PARPi post PARPi: Clarity from the Clearity real world experience (164)

Abstract

<b>Objectives:</b> To describe clinical characteristics and oncologic outcomes in patients who received PARPi after prior PARPi failures. <b>Methods:</b> Patients with ovarian cancer who received PARPi after progression on prior PARPi therapy were identified in the Clearity Foundation Data Repository. Contingency tables of PARPi received, therapy line, and treatment response after first and second PARPi exposure, germline, and somatic testing were analyzed and compared using the Chi-square test. Time-to-event curves of time-to-next-therapy (TTNT) were generated and compared using the log-rank test. Findings with p<0.05 were considered significant. <b>Results:</b> Twenty unique patients comprised the cohort. Thirteen patients (65%) had <i>BRCA1</i> or <i>BRCA2</i> mutations diagnosed by germline or tumor testing. Four additional patients had genomic alterations in <i>HRR</i> genes. PARPi received included olaparib, rucaparib, veliparib, niraparib, and talazoparib, with olaparib most frequently received. Thirteen patients (65%) had a primary platinum-sensitive recurrence, and seven patients (35%) had platinum-resistant disease with a primary course of therapy. Patients received a median number of two therapy lines prior to 1^st PARPi exposure. Patients received a median number of five therapy lines prior to 2^nd PARPi exposure. Patients without a <i>BRCA</i> mutation were significantly more likely to be treated with a PARPi in combination with another drug (i.e., chemotherapy, anti-VEGF agent, immunotherapy, or small molecule TKI) than patients with a <i>BRCA</i> mutation at 1^st PARPi exposure (71% vs 23%, p=0.04) and 2^nd PARPi exposure (100% vs 46.2%, p=0.02). There was a significantly higher clinical benefit rate (CBR) to 1^st PARPi compared to 2^nd in the entire cohort (70% vs 25%, p<0.01). Those with <i>BRCA</i> mutations had a significantly higher CBR with 1^st PARPi exposure compared to those without BRCA mutations (85 vs 43%, p=0.04). However, CBR with 2^nd PARPi exposure was not significantly different between those with <i>BRCA</i> mutations and those without <i>BRCA</i> mutations (15% vs 43%, p=0.2). Median TTNT after 2^nd PARPi exposure did not differ between those with <i>BRCA</i> mutations and those without (HR: 4.8 [95% CI: 2.4-7.1] vs 3.4 [95% CI: 2.8-4.0] months) (Figure 1). Within the <i>BRCA</i> mutated cohort, median TTNT with 2^nd PARPi exposure was significantly shorter than TTNT with 1^st PARPi exposure (4.8 [95% CI: 2.4-7.1] versus 8.9 [95% CI: 2.9-14.9] months; HR: 4.6 [95% CI: 1.6-13.1], log-rank p=0.004). There was no significant difference in median TTNT for patients without <i>BRCA</i> mutations with 2^nd PARPi exposure compared to 1^st PARPi exposure (3.4 [95% CI: 2.8-4.0] vs 4.7 [95%CI 0.6-8.7] months). <b>Conclusions:</b> We did not find a clinical benefit to second PARPi exposure after prior PARPi failure, regardless of <i>BRCA</i> mutation status. Patients without a <i>BRCA</i> mutation were much more likely to receive PARPi in combination with another agent in real-world use, despite established clinical benefits.