Homologous recombination deficiency (HRD) of high grade serous ovarian tumors from the NOVA Phase III clinical study Keith Wilcoxen,1 Christopher Neff,2 Victor Abkevich,2 Joshua Timothy Jones,2 Kathryn Kolquist,2 Michael Mirza,2 Jerry Lanchbury,2 Keith Mikule,1 Shefali Agarwal,1 Anne-Renee Hartman,2 Alexander Gutin,2 and Kirsten Timms.2 1TESARO, Inc., Waltham, MA, USA 2Myriad Genetics, Inc., Salt Lake City, UT, USA Introduction Evaluation of Phase 3 (NOVA) patient tumors non-gBRCAmut (n=211 analyzed to date) DDR Ovarian cancer is characterized by a high degree of genomic instability caused by deficiencies in DNA repair. Cells can develop these homologous recombination (HR) deficiencies through the loss of function or inactivation of genes involved in DNA repair, such as BRCA1 and BRCA2. Ovarian cancers known to have these HR deficiencies have been shown to benefit from therapy with DNA-damaging agents, such as platinum and PARP-inhibitors. Homologous recombination deficiency (HRD = defined as the sum of LOH, LST and TAI genomic defects), assessed across >500 ovarian tumors exhibits a bio-modal distribution allowing a clear differentiation of HR proficient and HR deficient tumors. Genome wide analysis was conducted on tumors obtained from patients enrolled in the NOVA study, a phase 3 clinical trial evaluating the PARP inhibitor niraparib as a maintenance treatment in patients with platinum sensitive ovarian cancer. Homologous recombination deficiency (HRD), sequence analysis of 43 genes involved in DNA damage response and other measures of genomic instability were evaluated. LOH score only Niraparib in 2nd Line (Recurrent) Ovarian Cancer Maintenance (NOVA) BRCA mutant BRCA wt LOH only cutoff of ≥7 High Grade Serous Ovarian Cancer, Platinum Sensitive, Relapsed Number of tumors Response to Platinum Treatment n=490 Non-gBRCAmut gBRCAmut 2:1 Randomization 2:1 Randomization Niraparib 300 mg QD Placebo Niraparib 300 mg QD Placebo 0-1 2-3 4-5 6-7 8-9 10-11 12-13 14-15 16-17 18-19 20-21 22-23 24-25 26-27 28-29 30-31 32-33 34-35 36-37 38-39 40-41 42-43 44-45 46-47 48-49 50-51 52-53 54-55 56-57 58-89 60-61 62-63 64-65 66-67 68-69 70-71 72-73 74-75 76-77 78-79 80-81 82-83 84-85 86-87 88-89 90-91 92-93 94-95 96-97 98-99 100 LOH Score n=207 n=103 n=120 n=60 BRCA deficient BRCA intact Endpoint Assessment Endpoint Assessment HRD score = LOH+LST+TAI HRD cutoff of ≥ 42 Primary Endpoint: PFS; >90% power to detect 4.5 month improvement (HR 0.50 ) Non-gBRCAmut cohort endpoint assessed hierarchically to control type 1 error: HRD + population first, followed by entire population #Number of tumors BRCA mutant BRCA wt Number of tumors gBRCAmut (n=136 analyzed to date) HRD score Homologous recombination proficient (HRD-) Homologous recombination deficient (HRD+) LOH score only HRD Score ≥7 = cutoff 95% of BRCAmut HRD distribution in NOVA tumors (n=347 analyzed) Number of tumors HRD score = LOH+LST+TAI gBRCAmut tBRCAmut HRD+ BRCAwt HRD- Conclusions Methods An assay has been developed that provides a quantitative continuous measure of genomic scarring and BRCA1/2 sequencing from tumor tissue in one test. Genomic analysis of 347/490 patient tumor samples in the NOVA study has been conducted, indicating 100% of the gBRCAmut cohort and 55% of the non-gBRCAmut cohort are HRD positive as defined by the Myriad HRD test. The use of three algorithms of DNA damage allow a clearer differentiation of the HRD population in ovarian cancer than the use of the single LOH algorithm. Deleterious or suspected deleterious mutations, with confirmed loss of both tumor alleles, were detected in 12 additional DNA damage response genes in 35 tumors. HRD scores in tumors with DDR gene mutations showed a similar range of HRD scores to that observed in BRCA1/2 mutants (25-82 compared to 25-88). Patients: The NOVA study is a Phase 3, multicenter, randomized, double-blind, placebo-controlled study of niraparib as maintenance in platinum sensitive ovarian cancer patients who have either gBRCAmut or a tumor with high-grade serous histology and who have responded to their most recent chemotherapy containing a platinum agent. Sample Analysis: DNA was extracted from formalin fixed paraffin-embedded (FFPE) tumor tissue and used to create libraries that were hybridized to a custom Agilent SureSelect capture array carrying probes for 54,091 single nucleotide polymorphism sites distributed across the human genome, as well as probes targeting 43 genes involved in DNA repair, including BRCA1 and BRCA2. The captured and enriched DNA was sequenced on an Illumina HiSeq 2500 sequencer. Sequences covering SNP positions were used to generate allelic imbalance profiles. Measures of genomic instability, including determination of a HRD score (integer value of 0-100), were calculated using allelic imbalance profiles and determination of loss of heterozygosity by ASCN. HRD Score: The HRD score is the unweighted sum of LOH (number of subchromosomal LOH regions longer than 15 Mb), TAI (number of regions with allelic imbalance that extend to one of the subtelomeres but do not cross the centromere), and LST (the number of break points between regions longer than 10 Mb after filtering out regions shorter than 3 Mb). Tumor samples were defined as HR deficient if they produced a score ≥42 and/or BRCA1/2 mutation. LOH Score # of tumors HRD score = LOH+LST+TAI ≥ 42 = cutoff 95% of BRCAmut HRD Score Number of tumors Shift to higher scores non-gBRCAmut cohort gBRCAmut cohort Total 211 136 tumor BRCA1/2mut 18% (39/211) tumor BRCA1mut 71% (97/136) HRD+ (BRCAwt) 37% (77/211) tumor BRCA2mut 29% (39/136) HRD- 45% (95/211) NA HRD+ 55% (116/211) 100% (136/136) HRD Score (LOH+TAI=LST) Analysis of LOH score only from tumors obtained from gBRCAmut cohort patients indicates a broad score distribution. Utilizing the sum of three algorithms to score HRD (LOH, LST, TAI) results in a shift to higher HRD scores for BRCAmut tumors in NOVA. A cutoff of 7 or greater for LOH score alone captures 95% of gBRCAmut tumors in NOVA. A cutoff of 42 or greater for HRD scores captures 95% of gBRCAmut tumors in NOVA. HRD+ tumors are defined as those with an HRD score of ≥ 42 or a deleterious or suspected deleterious mutation in BRCA1/2 as detected from tumor tissue Patient eligibility criteria for the NOVA study are consistent with enrichment for patients with tumors having high HRD scores due to platinum sensitivity Patients with HRD positive tumors consist of 100% of the patients in the gBRCAmut cohort and 55% of the patients in the non-gBRCAmut cohort based on this preliminary analysis of 347 pts. References