Patterns of genomic instability in > 2000 patients with ovarian cancer across six clinical trials evaluating olaparib Authors: Alan Barnicle, Isabelle Ray-Coquard, Etienne Rouleau, Karen Cadoo, Fiona Simpkins, Carol Aghajanian, Alexandra Leary, Andrés Poveda, Stephanie Lheureux, Eric Pujade-Lauraine, Benoit You, Jonathan Ledermann, Ursula Matulonis, Charlie Gourley, Kirsten M. Timms, Zhongwu Lai, Darren R. Hodgson, Cathy E. Elks, Simon Dearden, Coumaran Egile, Pierre Lao-Sirieix, Elizabeth A. Harrington, Jessica S. Brown Published: 2024 ABSTRACT BackgroundThe introduction of poly(ADP-ribose) polymerase (PARP) inhibitors represented a paradigm shift in the treatment of ovarian cancer. Genomic data from patients with high-grade ovarian cancer in six phase II/III trials involving the PARP inhibitor olaparib were analyzed to better understand patterns and potential causes of genomic instability.Patients and methodsHomologous recombination deficiency (HRD) was assessed in 2147 tumor samples from SOLO1, PAOLA-1, Study 19, SOLO2, OPINION, and LIGHT using next-generation sequencing technology. Genomic instability scores (GIS) were assessed in BRCA1 and/or BRCA2 (BRCA)-mutated (BRCAm), non-BRCA homologous recombination repair-mutated (non-BRCA HRRm), and non-HRRm tumors.ResultsBRCAm was identified in 1021/2147 (47.6%) tumors. BRCAm tumors had significantly higher GIS than non-BRCAm tumors (P < 0.001) and high biallelic loss (815/838; 97.3%) regardless of germline (658/672; 97.9%) or somatic (101/108; 93.5%) BRCAm status. In non-BRCA HRRm tumors (n = 121) a similar proportion were HRD-positive (GIS ≥ 42: 55/121; 45.5%) relative to HRD-negative (GIS < 42: 52/121; 43.0%). GIS was highly variable in non-BRCA HRRm (median 42 [interquartile range (IQR) 29–58]) and non-HRRm (n = 1005; median 32 [IQR 20–55]) tumors. Gene mutations with high GIS included HRR genes BRIP1 (median 46 [IQR 41–58]), RAD51C (median 58 [IQR 48–66]), RAD51D (median 62 [IQR 54–69]), and PALB2 (median 64 [IQR 58–74]), and non-HRR genes NF1 (median 49 [IQR 25–60]) and RB1 (median 55 [IQR 30–71]). CCNE1-amplified and PIK3CA-mutated tumors had low GIS (CCNE1-amplified: median 24 [IQR 18–29]; PIK3CA-mutated: median 32 [IQR 14–52]) and were predominantly non-BRCAm.ConclusionsThese analyses provide valuable insight into patterns of genomic instability and potential drivers of HRD, besides BRCAm, in ovarian cancer and will help guide future research into the potential clinical effectiveness of anti-cancer treatments in ovarian cancer, including PARP inhibitors as well as other precision oncology agents.Trial registrationThe SOLO1 trial was registered at ClinicalTrials.gov (NCT01844986) on April 30, 2013; the PAOLA-1 trial was registered at ClinicalTrials.gov (NCT02477644) on June 18, 2015 (retrospectively registered); Study 19 was registered at ClinicalTrials.gov (NCT00753545) on September 12, 2008 (retrospectively registered); the SOLO2 trial was registered at ClinicalTrials.gov (NCT01874353) on June 7, 2013; the OPINION trial was registered at ClinicalTrials.gov (NCT03402841) on January 3, 2018; the LIGHT trial was registered at ClinicalTrials.gov (NCT02983799) on November 4, 2016.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13073-024-01413-5. Background Homologous recombination repair (HRR) is a DNA repair pathway that acts on double-strand breaks and interstrand cross-links [1]. A deficiency in the HRR pathway—known as homologous recombination deficiency (HRD)—is associated with various tumor types, most prominently ovarian, breast, prostate, and pancreatic cancers [1]. The HRD phenotype is associated with various signatures of genomic instability [1], including loss of heterozygosity (LOH) [2], telomeric allelic imbalances (TAIs) [3], and large-scale state transitions (LSTs) [4]. The identification of homologous recombination deficient tumors can be achieved either by measuring these signatures of genomic instability (HRD testing) or by detecting mutations in the breast cancer susceptibility genes BRCA1 and/or BRCA2 (BRCAm) as well as by detecting other HRR pathway defects [1]. Poly(ADP-ribose) polymerase (PARP) inhibitors are a class of anti-cancer drugs that target tumors with HRD. The mechanism of action involves trapping of PARP at sites of single-stranded DNA breaks, resulting in replication-dependent DNA double-strand breaks that cannot be repaired accurately in HRD-positive tumors [5, 6]. Three different PARP inhibitors have received global regulatory approval as first-, second-, or later-line therapy in ovarian cancer, and some of these are also approved therapies in breast, prostate, and pancreatic cancers (see Additional file 1 for further details). High-grade serous ovarian cancer was considered a promising tumor type for PARP inhibitor therapy as it is characterized by a high frequency of genetic alterations in BRCA1 and/or BRCA2 and genomic instability [7]. In newly diagnosed patients with ovarian cancer, the benefit from maintenance PARP inhibitors has been greatest in patients with a BRCAm or HRD-positive tumors [8–11]. However, the drivers of HRD in ovarian cancer, other than mutations in BRCA1 and BRCA2, are not well understood. We analyzed genomic data from six olaparib global phase II/III studies (Additional file 1: Table S1) [8, 9, 12–22]. These included over 2000 patients with ovarian cancer in whom molecular profiling including genomic instability was assessed. Our objective was to describe patterns of genomic instability in BRCAm and non-BRCAm tumors and better understand drivers of HRD in the newly diagnosed and platinum-sensitive relapsed ovarian cancer (PSROC) populations. Methods Study design, participants, and tumor samples This was a pooled analysis of data from the SOLO1 [8, 13], PAOLA-1 [9, 22], Study 19 [14–16], SOLO2 [18, 21], OPINION [19, 20], and LIGHT [12, 17] clinical trials of olaparib in high-grade serous or high-grade endometrioid ovarian cancer. Details of these studies have been reported previously and are summarized in Additional file 1: Table S1. All trials and analyses were performed in accordance with the principles of the Declaration of Helsinki and Good Clinical Practice guidelines and were approved by the appropriate Institutional Review Boards. All patients provided written informed consent. Archival tumor samples, predominantly obtained at diagnosis, from each study were used, and genomic DNA derived from formalin-fixed paraffin-embedded tumor tissue or whole blood samples was analyzed (Additional file 1: Table S2). Definitions and assays Genomic DNA was used to assess tumor BRCAm (tBRCAm) status. Somatic BRCAm was defined as a positive tBRCAm status and negative germline BRCAm status; samples without a germline BRCAm status were classified as somatic/germline BRCAm status not determined. Genomic instability and BRCA1/BRCA2 mutation status were measured using a version of the Myriad MyChoice® CDx next-generation sequencing-based tumor test developed by Myriad Genetics (Myriad Genetic Laboratories, Inc., Salt Lake City, UT, USA) (see Additional file 1: Table S2 for further details of the assays used). Genomic instability was determined by measuring LOH (number of LOH regions longer than 15 Mb but shorter than the whole chromosome), TAI (number of regions with allelic imbalance that extend to one of the subtelomeres but do not cross the centromere and are longer than 11 Mb), and LSTs (number of break points between regions longer than 10 Mb after filtering out regions shorter than 3 Mb). LOH, TAI, and LST scores were combined to provide a total genomic instability score (GIS) of 0–100, with higher scores indicating greater genomic instability. A tumor was determined to be HRD-positive if it had a tBRCAm or met the previously established GIS cutoff of ≥ 42 [23]; 42 was used as the GIS cutoff for the MyChoice® HRD Plus assay (now known as the MyChoice® CDx assay) in the PAOLA-1 [9], OPINION [19], and LIGHT [12] studies. A tumor was determined to be HRD-negative if the GIS was < 42 with non-tBRCAm status. HRD status was unknown for cases where the GIS could not be determined due to assay failure and the tumor was non-tBRCAm status. For example, samples that were low tumor purity were failed to ensure that false negative results were not reported. Genomic instability was assessed in samples with BRCA1m, BRCA2m, and non-BRCA HRR mutations (non-BRCA HRRm), defined as loss-of-function mutations in 13 other genes involved in HRR: ATM, BARD1, BRIP1, CDK12, CHEK1, CHEK2, FANCL, PALB2, PPP2R2A, RAD51B, RAD51C, RAD51D, and RAD54L; PPP2R2A was included in the non-BRCA HRRm analysis because it was considered a non-BRCA HRRm gene in prespecified analyses of the included studies. Genomic instability was also assessed in samples with deleterious or suspected deleterious mutations other than BRCAm or non-BRCAm HRRm, including MAP3K1, MYC, RB1, KMT2D, CSMD3, SOX2, NF1, MAP2K4, NTHL1, PTEN, STAG2, PTPRD, TP53, FANCM, ERBB2, TSC1, MCL1, BLM, PIK3CA, MYH, NBN, CCNE1, ARID1A, NRAS, and KRAS, as determined by the Myriad tumor tissue gene panel (108 in total; full gene list is provided in Additional file 1: Table S3) and a Foundation Medicine gene panel (for Study 19 only). Gene-specific LOH was defined by Myriad Genetics as LOH localized to a given locus and included biallelic inactivation (locus contains at least one homozygous mutation or two or more deleterious mutations in a single gene; no wild-type allele detected), heterozygous (locus contains a single deleterious mutation or alteration in one allele; one functional copy remains), and unknown (locus could not be determined as biallelic or heterozygous). Biallelic loss was initially determined by the LOH status of the chromosomal region encompassing a gene. The LOH status was unknown if a boundary between two regions, one with LOH (classified as “biallelic loss”) and one without LOH (classified as “heterozygous”), was located inside the gene. In addition, the initial LOH status was unknown if the GIS analysis failed. Initial LOH calls were manually reviewed and corrected (e.g., a patient carrying a deleterious or suspected deleterious BRCAm with LOH in BRCA1/BRCA2 was considered to have a biallelic loss of BRCA1/BRCA2). Tumors with co-occurring BRCA1 and BRCA2 alterations or with co-occurring individual non-BRCA HRR alterations were excluded from individual gene analyses as gene-specific HRD and zygosity on the patient level could not be assessed. Tumors with a non-HRR alteration were reported as having that gene alteration even if there was a co-occurring non-HRR alteration; however, tumors with a non-HRR alteration and a co-occurring BRCA1 and/or BRCA2 alteration or non-BRCA HRR alteration were considered BRCAm and non-BRCA HRRm, respectively. Statistical analyses Statistical variance between groups was determined by applying an independent Student’s t-test, with a threshold of P < 0.05 regarded as statistically significant. Results Patterns of genomic instability in the analysis cohort Across the six studies, sequencing results were available for tumors from 2147 patients, with a valid GIS available for 1838 tumors. Where GIS was evaluable, the median GIS observed across studies was 54.0 (interquartile range [IQR] 31–66). Patterns of GIS were assessed for different baseline characteristics of interest. The observed range of GIS did not differ considerably by patient race (White, Black or African American, Asian, other/unknown), across tumor histologies (serous, endometrioid, other/unknown), or by primary tumor location (ovary, fallopian tube, peritoneal, other/unknown) (Additional file 1: Fig S1) where evaluable. However, in this analysis cohort, the sample size was limited beyond patients with White race or tumors with serous histology. GIS in BRCAm tumors Distinct patterns of GIS were observed by mutation status. Of the 2147 tumors with available sequencing results, 1021 (47.6%) had tBRCAm, including 692 (67.8%) with tBRCA1m, 323 (31.6%) with tBRCA2m, and 6 (0.6%) with both tBRCA1m and tBRCA2m; 1126 (52.4%) were non-tBRCAm (Fig. 1). The majority of tumors with tBRCAm had a valid GIS (n = 863 [84.5%]; n = 583 [84.2%] with tBRCA1m alone and n = 274 [84.8%] with tBRCA2m alone), and most had a GIS ≥ 42 (n = 810 [93.9%]; n = 559 [95.9%] with tBRCA1m alone and n = 245 [89.4%] with tBRCA2m alone) (Fig. 1). Where GIS was measurable in the overall study population (n = 1838), the distribution of GIS was bimodal (Fig. 2A). Those with tBRCAm predominantly had high GIS (median 62 [IQR 54–70]) and tBRCAm tumors had a significantly higher median GIS than non-tBRCAm tumors (median 34 [IQR 21–56]; P < 0.001) (Fig. 2B). This pattern was observed for both BRCA1m and BRCA2m patients; however, a significantly higher GIS was seen in BRCA1m (median GIS 64 [IQR 55–71]) than BRCA2m (median GIS 59 [IQR 50–67]) tumors (P < 0.001; Fig. 2C).Fig. 1GIS status in A tBRCA1m and tBRCA2m and non-tBRCAm and B non-BRCA HRRm and non-HRRm. BRCA1m, BRCA1 mutation; BRCA2m, BRCA2 mutation; GIS, genomic instability score; HRRm, homologous recombination repair mutation; t, tumorFig. 2GIS distribution A overall, B in tumors with and without tBRCAm, C in tumors with BRCA1m and BRCA2m and without tBRCAm, and D in BRCA1m and BRCA2m mutation subtypes. In panel A, distributions were based on 1838 tumors with measurable GIS (623 samples from PAOLA-1, 241 from OPINON, 344 from LIGHT, 226 from SOLO1, 210 from SOLO2, and 194 from Study 19). The dashed vertical line denotes the GIS cutoff of 42. In panels B and C, the box plot shows median (IQR) and whiskers indicate 1.5 times the IQR above Q3 and below Q1. The dashed horizontal line denotes the GIS cutoff of 42. Panel C excludes tumors with co-occurring BRCA1 and BRCA2 mutations. In panel D, 582 BRCA1m tumors and 273 BRCA2m tumors were evaluable. Excludes patients with BRCA1 or BRCA2 double-hit mutations (n = 2) and co-occurring HRRm and non-BRCAm genes (n = 4). The dashed vertical lines denote the GIS cutoff of 42. BRCA1m, BRCA1 mutation; BRCA2m, BRCA2 mutation; BRCAm, BRCA1 and/or BRCA2 mutation; GIS, genomic instability score; HRRm, homologous recombination repair mutation; IQR, interquartile range; Q, quartile; tBRCAm, tumor BRCAm GIS status in A tBRCA1m and tBRCA2m and non-tBRCAm and B non-BRCA HRRm and non-HRRm. BRCA1m, BRCA1 mutation; BRCA2m, BRCA2 mutation; GIS, genomic instability score; HRRm, homologous recombination repair mutation; t, tumor GIS distribution A overall, B in tumors with and without tBRCAm, C in tumors with BRCA1m and BRCA2m and without tBRCAm, and D in BRCA1m and BRCA2m mutation subtypes. In panel A, distributions were based on 1838 tumors with measurable GIS (623 samples from PAOLA-1, 241 from OPINON, 344 from LIGHT, 226 from SOLO1, 210 from SOLO2, and 194 from Study 19). The dashed vertical line denotes the GIS cutoff of 42. In panels B and C, the box plot shows median (IQR) and whiskers indicate 1.5 times the IQR above Q3 and below Q1. The dashed horizontal line denotes the GIS cutoff of 42. Panel C excludes tumors with co-occurring BRCA1 and BRCA2 mutations. In panel D, 582 BRCA1m tumors and 273 BRCA2m tumors were evaluable. Excludes patients with BRCA1 or BRCA2 double-hit mutations (n = 2) and co-occurring HRRm and non-BRCAm genes (n = 4). The dashed vertical lines denote the GIS cutoff of 42. BRCA1m, BRCA1 mutation; BRCA2m, BRCA2 mutation; BRCAm, BRCA1 and/or BRCA2 mutation; GIS, genomic instability score; HRRm, homologous recombination repair mutation; IQR, interquartile range; Q, quartile; tBRCAm, tumor BRCAm The higher GIS for both BRCA1m and BRCA2m (relative to non-tBRCAm) was consistent across the individual studies (Additional file 1: Fig. S2) and in patients with newly diagnosed ovarian cancer (Additional file 1: Fig. S3A) or PSROC (Additional file 1: Fig. S3B). GIS patterns were similar regardless of mutation subtype, including missense, splice, loss/rearrangement, nonsense, and frameshift/indel alterations (Fig. 2D). As detailed in the Methods section, the origin of BRCAm (germline or somatic) was assessed by germline testing of blood or by computational assessment of germline/somatic status of tBRCAm in cases where germline testing was not conducted. Across all studies, > 95% of germline BRCAm reported by blood testing were also reported based on tumor tissue testing (Additional file 1: Table S2). Of tBRCA1m or tBRCA2m samples with a valid GIS (n = 857), 693 (80.9%) were germline BRCAm, 113 (13.2%) were somatic BRCAm, and 51 (6.0%) could not be classified as germline or somatic and their origin were considered as undetermined. Median GIS was similar regardless of whether BRCA1m and BRCA2m were of germline origin (germline BRCA1m: median 64 [IQR: 56–71]; germline BRCA2m: median 59 [IQR 50–68]), somatic origin (somatic BRCA1m: median 62 [IQR: 56–71]; somatic BRCA2m: median 57 [IQR 44–65]), or undetermined origin (undetermined BRCA1m: median 63 [IQR 54–70]; undetermined BRCA2m: median 60 [IQR 50–67]) (Additional file 1: Fig. S3C). Gene-specific LOH (biallelic inactivation) was observed in almost all evaluable BRCAm tumors (815/838; 97.3%): 572/574 (99.7%) of those with tBRCA1m and 243/264 (92.0%) with tBRCA2m; monoallelic loss was seen in the remaining evaluable tumors (Additional file 1: Fig. S4A). The higher GIS seen in BRCA1 tumors may be partially explained by the slightly lower rate of monoallelic loss observed in BRCA1 than in BRCA2. When assessing gene-specific LOH and GIS patterns together in BRCAm tumors (n = 825), the GIS was high in tumors with biallelic inactivation (median GIS 62 [IQR 55–70]), which accounted for the majority of BRCAm (802/825, 97.2%) (Additional file 1: Fig. S4B). GIS was relatively lower in heterozygous BRCAm tumors (median GIS 39; IQR 33–53), which accounted for the minority of BRCAm (23/825, 2.8%) (Additional file 1: Fig. S4B). Evaluation of gene-specific LOH revealed that the rate of biallelic inactivation was high for tumors irrespective of the origin, with a germline (658/672; 97.9%) or somatic (101/108; 93.5%) BRCAm (Additional file 1: Fig. S4C). Gene-specific zygosity in patients with BRCA1m or BRCA2m by individual study is shown in Additional file 1: Fig. S5. GIS in non-BRCA HRRm tumors When assessing the distribution of GIS across datasets, three distinct clusters were identified based on BRCA, non-BRCA HRR, and non-HRR mutation status. As detailed above, GIS for tBRCAm tumors was predominantly high, while those patients with non-BRCA HRRm (n = 107; median 42 [IQR 29–58]) and those with non-HRRm (n = 868; median 32 [IQR 20–55]) had GIS patterns that were variable and predominantly low, respectively (Fig. 3A). Classifying evaluable patients from PAOLA-1, OPINION, LIGHT, and Study 19 (n = 1788) according to their HRD and HRR biomarker status demonstrated that HRD and HRRm assays are not interchangeable (Fig. 3B). It should be noted that the SOLO1 and SOLO2 trials were not included in this analysis as these studies only enrolled patients with BRCAm. Therefore, non-BRCA HRRm and non-HRRm status were not evaluable [8, 21]. Of the 1788 tumor samples from PAOLA-1, OPINION, LIGHT, and Study 19 with genomic data, 121 (6.8%) were identified as non-BRCA HRRm, which was smaller than the proportion identified as HRD-positive excluding BRCAm (319; 17.8%) (Fig. 3B). Overall, patients with a non-BRCA HRRm were not enriched in HRD-positive tumors and a similar proportion of non-BRCA HRRm tumors were HRD-positive (GIS ≥ 42 in 55/121; 45.5%) compared with HRD-negative (GIS < 42 in 52/121; 43.0%) or GIS unknown (14/121; 11.6%) states (Fig. 3B).Fig. 3GIS distribution A in tumors with tBRCAm, non-BRCA HRRm, and non-HRRm, B relative proportion of tBRCAm, non-BRCA HRRm, and non-HRRm status in the context of HRD-positive, HRD-negative, and HRD-unknown biomarker status in PAOLA-1, OPINION, LIGHT, and Study 19, and C GIS distribution in tumors with BRCA1m, BRCA2m, non-BRCA HRRm, and non-HRRm biomarker status. In panel A, distributions were relative to the number of tumors within each subgroup; 863 tumors were tBRCAm, 107 were non-BRCA HRRm, and 868 were non-HRRm. The dashed vertical line denotes the GIS cutoff of 42. In panel B, tumor HRD status was based on 1788 tumors. HRD-positive was defined as the presence of a tBRCAm and/or a GIS of ≥ 42, HRD-negative as a GIS of < 42 and absence of a tBRCAm, and HRD-unknown were cases where a GIS could not be determined. In panel C, the box plot shows median (IQR) and whiskers indicate 1.5 times the IQR above Q3 and below Q1. The dashed horizontal line denotes the GIS cutoff of 42. Tumors with co-occurring BRCA1 and BRCA2 mutations were excluded. BRCA1m, BRCA1 mutation; BRCA2m, BRCA2 mutation; BRCAm, BRCA1 and/or BRCA2 mutation; GIS, genomic instability score; HRRm, homologous recombination repair mutation; IQR, interquartile range; Q, quartile; tBRCAm, tumor BRCAm GIS distribution A in tumors with tBRCAm, non-BRCA HRRm, and non-HRRm, B relative proportion of tBRCAm, non-BRCA HRRm, and non-HRRm status in the context of HRD-positive, HRD-negative, and HRD-unknown biomarker status in PAOLA-1, OPINION, LIGHT, and Study 19, and C GIS distribution in tumors with BRCA1m, BRCA2m, non-BRCA HRRm, and non-HRRm biomarker status. In panel A, distributions were relative to the number of tumors within each subgroup; 863 tumors were tBRCAm, 107 were non-BRCA HRRm, and 868 were non-HRRm. The dashed vertical line denotes the GIS cutoff of 42. In panel B, tumor HRD status was based on 1788 tumors. HRD-positive was defined as the presence of a tBRCAm and/or a GIS of ≥ 42, HRD-negative as a GIS of < 42 and absence of a tBRCAm, and HRD-unknown were cases where a GIS could not be determined. In panel C, the box plot shows median (IQR) and whiskers indicate 1.5 times the IQR above Q3 and below Q1. The dashed horizontal line denotes the GIS cutoff of 42. Tumors with co-occurring BRCA1 and BRCA2 mutations were excluded. BRCA1m, BRCA1 mutation; BRCA2m, BRCA2 mutation; BRCAm, BRCA1 and/or BRCA2 mutation; GIS, genomic instability score; HRRm, homologous recombination repair mutation; IQR, interquartile range; Q, quartile; tBRCAm, tumor BRCAm Median GIS for non-BRCA HRRm (median GIS 42 [IQR 29–58]) was significantly lower than that for BRCA1m or BRCA2m (P < 0.001 in both instances) and significantly higher than that for non-HRRm (P ~ 0.003) (Fig. 3C). The difference in median GIS for non-BRCA HRRm between tumors from patients in response after first-line platinum-based chemotherapy and those with platinum-sensitive relapsed disease was not statistically significant (P = 0.12) (Additional file 1: Fig. S6A). Genomic instability for tumors with a non-BRCA HRRm was also assessed by individual study (Additional file 1: Fig. S6B). The median GIS for non-BRCA HRRm patients exceeded the cutoff for HRD-positivity only in OPINION (median GIS 54; IQR 38–68; n = 31) and was lower in PAOLA-1 (median GIS 38; IQR 24–54; n = 43), LIGHT (median GIS 38; IQR 23–53; n = 22), and Study 19 (median GIS 33; IQR 30–54; n = 11). To assess this finding further, we performed gene-by-gene–level analysis within the non-BRCA HRRm group. In relation to gene-specific LOH, CDK12 (20/20; 100.0%), BRIP1 (14/16; 87.5%), RAD51C (14/14; 100.0%), and RAD51D (12/12; 100.0%) genes had high levels of biallelic inactivation; the number of evaluable tumors with a non-BRCA HRRm other than in these genes was too low to identify any trends in gene-specific LOH (Fig. 4A). GIS was also assessed for individual non-BRCA HRR genes. Tumors with mutations in five non-BRCA HRR genes (BRIP1, RAD51C, RAD51D, PALB2, and FANCL) had a median GIS ≥ 42 (Fig. 4B). In particular, those with a mutation in BRIP1 (median GIS 46; IQR 41–58), RAD51C (median GIS 58; IQR 48–66), RAD51D (median GIS 62; IQR 54–69), and PALB2 (median GIS 64; IQR 58–74) had a relatively higher median GIS than the other non-BRCA HRR genes (Fig. 4B) such as CDK12 (median GIS 33; IQR 23–39) (Fig. 4B). The GIS distribution for individual genes was broadly consistent across studies (Additional file 1: Fig. S7); however, mutation frequencies for individual genes were not always consistent between studies and small patient numbers mean that gene-by-gene–level data for individual studies should be interpreted with caution.Fig. 4A Gene-specific zygosity and B GIS distribution in non-BRCA HRRm tumors. Panel A shows results for 98 patients with a non-BRCA HRRm and evaluable zygosity. Excludes tumors with co-occurring HRR, as gene-specific zygosity on the patient level cannot be assessed. Cases where gene-specific zygosity could not be assessed are not shown. In panel B, the box plot shows median (IQR) and whiskers indicate 1.5 times the IQR above Q3 and below Q1. The dashed horizontal line denotes the GIS cutoff of 42. Excludes tumors with co-occurring HRR, as gene-specific HRD and zygosity on the patient level cannot be assessed, and samples are only included where a GIS was calculated (103 patient samples). An additional gene (CHEK1) is not shown in panels A or B, as no individual mutations were detected. GIS, genomic instability score; HRD, homologous recombination deficiency; HRR, homologous recombination repair; HRRm, HRR mutation; IQR, interquartile range; Q, quartile A Gene-specific zygosity and B GIS distribution in non-BRCA HRRm tumors. Panel A shows results for 98 patients with a non-BRCA HRRm and evaluable zygosity. Excludes tumors with co-occurring HRR, as gene-specific zygosity on the patient level cannot be assessed. Cases where gene-specific zygosity could not be assessed are not shown. In panel B, the box plot shows median (IQR) and whiskers indicate 1.5 times the IQR above Q3 and below Q1. The dashed horizontal line denotes the GIS cutoff of 42. Excludes tumors with co-occurring HRR, as gene-specific HRD and zygosity on the patient level cannot be assessed, and samples are only included where a GIS was calculated (103 patient samples). An additional gene (CHEK1) is not shown in panels A or B, as no individual mutations were detected. GIS, genomic instability score; HRD, homologous recombination deficiency; HRR, homologous recombination repair; HRRm, HRR mutation; IQR, interquartile range; Q, quartile Genomic instability patterns with non-HRRm (gene mutations other than BRCA or non-BRCA HRRm) Patterns of genomic instability were assessed in patients without BRCAm or non-BRCA HRRm (i.e., the non-HRRm population, n = 1005, of whom 868 had evaluable GIS) relative to the BRCAm population. Overall, non-HRRm tumors had a significantly lower median GIS (median GIS 32 [IQR 20–55]) than BRCAm tumors (P < 0.001; Fig. 5). In non-HRRm tumors, some individual genes had a high (≥ 42) median GIS (i.e., MAP3K1, MYC, RB1, KMT2D, CSMD3, SOX2, NF1, MAP2K4, NTHL1) and some had a low (< 42) median GIS (i.e., PTEN, STAG2, PTPRD, FANCM, ERBB2, TSC1, MCL1, BLM, PIK3CA, MYH, NBN, CCNE1, ARID1A, NRAS, KRAS). On a gene-by-gene level, patterns of GIS were generally stochastic and/or limited in sample size. However, some select genes showed consistent patterns.Fig. 5GIS distribution in tumors with tBRCAm and non-HRRm (gene alterations other than tBRCAm or non-BRCA HRRm) in PAOLA-1, OPINION, LIGHT, and Study 19. The box plot shows median (IQR) and whiskers indicate 1.5 times the IQR above Q3 and below Q1. The dashed horizontal line denotes the GIS cutoff of 42. The figure shows results for 1295 patient samples (427 tBRCAm with an evaluable GIS and 868 non-HRRm). Only mutations that qualified as deleterious/suspected deleterious were included in the analysis. ‘Other’ represents all genes that are mutated < 5 times across the datasets. Study 19 data were derived from the FoundationOne® assay results (Foundation Medicine, Inc., Cambridge, MA, USA), rather than Myriad MyChoice® CDx test results. Overall GIS for tBRCAm and non-HRRm are shown for completeness. The figure does not consider co-occurrence with TP53. BRCAm, BRCA1 and/or BRCA2 mutation; GIS, genomic instability score; HRRm, homologous recombination repair mutation; IQR, interquartile range; Q, quartile; tBRCAm, tumor BRCAm GIS distribution in tumors with tBRCAm and non-HRRm (gene alterations other than tBRCAm or non-BRCA HRRm) in PAOLA-1, OPINION, LIGHT, and Study 19. The box plot shows median (IQR) and whiskers indicate 1.5 times the IQR above Q3 and below Q1. The dashed horizontal line denotes the GIS cutoff of 42. The figure shows results for 1295 patient samples (427 tBRCAm with an evaluable GIS and 868 non-HRRm). Only mutations that qualified as deleterious/suspected deleterious were included in the analysis. ‘Other’ represents all genes that are mutated < 5 times across the datasets. Study 19 data were derived from the FoundationOne® assay results (Foundation Medicine, Inc., Cambridge, MA, USA), rather than Myriad MyChoice® CDx test results. Overall GIS for tBRCAm and non-HRRm are shown for completeness. The figure does not consider co-occurrence with TP53. BRCAm, BRCA1 and/or BRCA2 mutation; GIS, genomic instability score; HRRm, homologous recombination repair mutation; IQR, interquartile range; Q, quartile; tBRCAm, tumor BRCAm In non-HRRm tumors, NF1- (median GIS 49; IQR 25–60; n = 90) and RB1- (median GIS 55; IQR 30–71; n = 45) altered tumors had a high GIS (Fig. 6A). In all patients assessed, NF1 and RB1 mutations co-occurred with BRCAm in 33.5% (55/164) and 44.3% (39/88) of cases, respectively. Although both NF1 and RB1 mutations are known to co-occur with BRCAm (where GIS is predominantly high) [24], median GIS was high in tumors with NF1 and RB1 mutations in both BRCAm (RB1, median GIS 65; IQR 51–71; NF1, median GIS 59; IQR 48–66) and non-BRCAm (RB1, median GIS 55; IQR 33–69; NF1, median GIS 50; IQR 28–60) states, across gene alteration subtypes. GIS was significantly higher for both RB1-mutated and NF1-mutated tumors relative to non-NF1-/non-RB1-mutated tumors (P < 0.001 in both cases) in non-BRCAm states (Fig. 6A). RB1 and NF1 alterations also co-occurred with each other in these cohorts at a rate of 1.4% (n = 7) and 1.3% (n = 13) in BRCAm and non-BRCAm tumors, respectively.Fig. 6GIS in tumors with A NF1 and RB1 alterations, B CCNE1 alterations, and C PIK3CA alterations according to tBRCAm and non-tBRCAm status. The box plot shows median (IQR) and whiskers indicate 1.5 times the IQR above Q3 and below Q1. The dashed horizontal line denotes the GIS cutoff of 42. Figures do not consider co-occurrence with TP53. BRCAm, BRCA1 and/or BRCA2 mutation; GIS, genomic instability score; IQR, interquartile range; Q, quartile; tBRCA, tumor BRCAm GIS in tumors with A NF1 and RB1 alterations, B CCNE1 alterations, and C PIK3CA alterations according to tBRCAm and non-tBRCAm status. The box plot shows median (IQR) and whiskers indicate 1.5 times the IQR above Q3 and below Q1. The dashed horizontal line denotes the GIS cutoff of 42. Figures do not consider co-occurrence with TP53. BRCAm, BRCA1 and/or BRCA2 mutation; GIS, genomic instability score; IQR, interquartile range; Q, quartile; tBRCA, tumor BRCAm Conversely, in all patients assessed, tumors with CCNE1 alterations (n = 49) had a low GIS (median 24; IQR 18–29), were predominantly non-BRCAm (43/49; 87.8%), and had a significantly lower GIS than non-CCNE1-mutated tumors (P < 0.001) in non-BRCAm states (Fig. 6B). A similar pattern was also observed for PIK3CA-mutated tumors (n = 43) (median 32, IQR 14–52), which were predominantly non-BRCAm (38 of 43; 88.4%), where a significantly lower GIS was seen with PIK3CA-mutated versus non-PIK3CA-mutated tumors (P ~ 0.02) in non-BRCAm states (Fig. 6C). The pattern of gene alterations reported in PAOLA-1, OPINION, LIGHT, and Study 19 in the context of BRCAm, HRD, and non-BRCA HRRm is shown in Additional file 1: Fig. S8. Discussion We analyzed genomic instability data from six phase II or III studies evaluating olaparib as maintenance therapy (SOLO1 [8, 13], PAOLA-1 [9, 22], Study 19 [14–16], SOLO2 [18, 21], OPINION [19, 20]) or as treatment (LIGHT [12, 17]). Our study comprehensively describes the relationship between GIS and a 108-gene panel in > 2000 patients with newly diagnosed [8, 9, 13, 22] or platinum-sensitive relapsed [12, 14–21] high-grade serous or high-grade endometrioid ovarian cancer, and to our knowledge represents the largest combined HRD analysis of high-grade epithelial ovarian cancer using a commercially available HRD assay. To our knowledge, for the first time, our study shows the relationship between GIS and a broad panel of genes in carefully curated patients without BRCAm or HRRm. We demonstrated that GIS has an overall bimodal distribution in high-grade serous or high-grade endometrioid ovarian cancer, with high GIS in BRCAm tumors, variable GIS in non-BRCA HRRm tumors (possibly reflecting differences in the role of these genes in the HRR pathway), and predominantly low GIS in non-HRRm tumors. This pattern was consistent across individual studies and in patients with newly diagnosed ovarian cancer or PSROC. In our analysis and consistent with the individual studies [7, 25, 26], BRCAm was associated with high GIS and high rates of biallelic loss, irrespective of BRCAm being germline or somatic in origin. Our analysis also demonstrated that GIS was high regardless of mutation subtype. Although both BRCA1m and BRCA2m tumors had a very high GIS (with a similar IQR), a significantly higher score was observed for those with BRCA1m. Clinical trial data have demonstrated significant benefit of PARP inhibitor treatment in ovarian cancer patients with BRCAm; however, some studies suggest that patients with BRCA2m have relatively greater sensitivity to PARP inhibition than those with BRCA1m [27, 28]. In the SOLO1 trial, benefit from olaparib was demonstrated in patients with a BRCAm regardless of whether genome-wide LOH scores were high or low [25]. These apparently discordant observations may reflect the fact that GIS is only a surrogate marker for sensitivity to PARP inhibition and not the sole determinant. It is possible that once the threshold for GIS has been passed, other factors, such as drug resistance mechanisms, play a more prominent role in tumor sensitivity to PARP inhibition. Further work is needed to understand the relationship between BRCAm and GIS in BRCAm tumors with low genomic instability. As shown in phase III trials [9–11], HRD, as determined by genomic instability testing, is critical in the newly diagnosed ovarian cancer setting to identify which patients may experience the greatest benefit from PARP inhibitor maintenance therapy [29, 30]. HRD is a measure of global genomic instability induced by defects in the HRR pathway, while HRRm reflects the presence of mutations in specific genes involved in the HRR pathway. Our data demonstrate that HRD and HRRm analyses are not interchangeable and identify different sub-populations of patients. The finding that there was no statistically significant difference in median GIS across the non-BRCA HRRm population in the newly diagnosed ovarian cancer and PSROC settings was unexpected as it suggests that a difference in GIS does not explain why HRRm was predictive of PARP inhibitor benefit in Study 19 [7], but not in PAOLA-1 [31]. Some selection for patients with an overall higher GIS might have been expected in the platinum-sensitive relapsed setting given these studies only included patients known to have platinum-sensitive disease, whereas patients with no evidence of disease following cytoreductive surgery and whose sensitivity to platinum was unknown were eligible for inclusion in the first-line studies. Non-BRCA HRRm were low in prevalence, not enriched in HRD-positive tumors, and heterogeneous with regards to biallelic loss and GIS. RAD51C, RAD51D, BRIP1, and PALB2 are genes known to be associated with a hereditary risk of ovarian cancer [32–34]. The small number of PALB2 mutations limits interpretation; however, RAD51C, RAD51D, and BRIP1 mutations were associated with both high rates of biallelic loss and high GIS, suggesting that these genes might be true drivers of HRD in ovarian cancer. Mutations in these genes accounted for the majority of non-BRCA HRRm in OPINION, compared with PAOLA-1, LIGHT, and Study 19 where other non-BRCA HRRm genes were also prevalent. This could explain why we observed a higher degree of GIS in OPINION relative to the other studies. In terms of other non-BRCA HRR genes, studies suggest that loss-of-function mutations in CDK12 confer sensitivity to PARP inhibition [35]. A possible explanation for the relatively low GIS associated with CDK12 in this analysis is that CDK12 mutations have a distinct genomic instability pattern characterized by focal tandem repeats [36] and the Myriad MyChoice® CDx assay does not detect this particular genomic instability signature. Our knowledge of what constitutes an HRRm gene continues to evolve. Only two patients with variants in PPP2R2A were included in this analysis; low GIS was observed in both cases. After the studies included in this analysis were initiated, data from the PROfound study demonstrated that no benefit of olaparib over control therapy was observed in patients with prostate cancer and PPP2R2A mutations and was not predictive of benefit from PARP inhibitor therapy [37]. In terms of alterations outside of BRCAm or HRRm, those in MAP3K1, MYC, RB1, KMT2D, CSMD3, SOX2, NF1, MAP2K4, and NTHL1 had a high median GIS and those in PTEN, STAG2, PTPRD, FANCM, ERBB2, TSC1, MCL1, BLM, PIK3CA, MYH, NBN, CCNE1, ARID1A, NRAS, and KRAS had a low median GIS. It is now understood that RB1 loss is not mutually exclusive with HRRm, and improved outcomes have been reported when RB1 loss co-occurs with HRRm [38, 39]. This analysis demonstrates that even in tumors without BRCAm or HRRm, NF1 and RB1 mutations were associated with relatively high genomic instability; further work is needed to understand the mechanistic relationship between NF1 and RB1 and GIS. By contrast, mutations in CCNE1 and PIK3CA were associated with relatively low genomic instability that was found to be significantly lower than that seen in non-CCNE1-altered and non-PIK3CA-mutated tumors, respectively, in non-BRCAm states. To our knowledge, PIK3CA mutations have not previously been shown to be associated with low GIS and an association between CCNE1 amplification and low GIS has previously only been observed in smaller datasets [40]. Increased cyclin E expression, either by CCNE1 gene amplification, copy-number gain, or elevated protein expression, is associated with poor clinical outcomes and resistance to DNA-damaging drugs in ovarian cancer [24, 41]. For example, CCNE1 amplification was correlated with shorter relapse-free survival in patients with ovarian carcinomas treated with platinum-based chemotherapy [41]. The results of our analysis and others [42, 43] suggest that patients with CCNE1 amplifications (but without BRCAm) may potentially benefit from alternative targeted therapies. PIK3CA mutations are also associated with poor response to platinum-based chemotherapy and platinum resistance in patients with ovarian cancer [44]. Patients with alterations in CCNE1 or PIK3CA therefore represent a population with very high unmet medical needs. The optimal treatment of CCNE1-amplified and PIK3CA-mutated tumors warrants further investigation; the PI3K/AKT/mTOR pathway is one of the potential therapeutic targets for CCNE1-mutated [45] and PIK3CA-mutated [44] tumors. Ovarian cancer is the archetypal homologous recombination deficient tumor and is the only tumor type to date where genomic instability status is predictive of benefit from PARP inhibitor therapy. Although correlating genomic instability with clinical benefit was beyond the scope of the current analysis, previous analyses of PAOLA-1 [9, 22], Study 19 [7], OPINION [19, 20], and LIGHT [12, 17] have evaluated clinical outcome according to HRD status. Further work is needed to understand the clinical implications of these data and to evaluate potential links between genomic instability and clinical benefit from PARP inhibitor therapy for individual non-BRCA HRR and non-HRR genes. To date, post hoc exploratory analysis found that non-BRCA HRR gene panels were not predictive of the efficacy of maintenance olaparib plus bevacizumab in PAOLA-1 [31]. By contrast, subgroup analyses based on non-BRCA HRRm demonstrated the benefit of olaparib treatment in Study 19 [7], olaparib activity comparable with that seen in BRCAm in the ORZORA trial [46], and longer PFS benefit in patients with non-BRCA HRRm relative to patients with non-HRRm in the OPINION trial [47]. One explanation for these apparent discrepancies might be a difference in signal between the first-line [31] and relapsed disease [7] settings, with the relapsed disease population more likely to be enriched for platinum sensitivity. These studies included small numbers of individual non-BRCA HRR genes and differences between the studies in gene-to-gene prevalence may be another potential explanation for apparent discrepancies. For example, of the non-BRCA HRRm subgroup, CDK12, BRIP1, and RAD51C accounted for 24%, 13%, and 17%, respectively, in PAOLA-1 [31]; 36%, 15%, and 18%, respectively, in ORZORA [46]; and 9%, 21%, and 24%, respectively, in OPINION [47]. Furthermore, these studies were not individually powered to assess the predictive power of these non-BRCA HRR genes, making comparisons difficult. In terms of other PARP inhibitors, RAD51C and RAD51D mutations predicted response to treatment with rucaparib patients with relapsed ovarian cancer in a post hoc exploratory analysis of ARIEL2 [48]. Although we found BRIP1 mutations had a high GIS, they are yet to be established as predictive of PARP inhibitor response [31]. Better understanding of drivers of genomic instability in ovarian cancer may open up opportunities for PARP inhibitor use in other indications where GIS alone has not been shown to be predictive. Our study identified a further nine genes (MAP3K1, MYC, RB1, KMT2D, CSMD3, SOX2, NF1, MAP2K4, and NTHL1) which, when mutated, were associated with high median GIS and an additional 15 genes (PTEN, STAG2, PTPRD, FANCM, ERBB2, TSC1, MCL1, BLM, PIK3CA, MYH, NBN, CCNE1, ARID1A, NRAS, and KRAS) which, when mutated, were associated with low median GIS; some of these genes (e.g., CCNE1, ARID1A, NRAS, and KRAS) are linked with replication stress. Thus, our data highlight the heterogeneity of HRD-positive and HRD-negative populations, which appear to have different biology, and the need to better personalize treatment options that target alternative pathways. Further clinical studies would be required to validate these targets as biomarkers for PARP inhibitor benefit in ovarian cancer, which is beyond the scope of this study. These data also signal the need to explore PARP inhibitor benefit in other tumor types harboring mutations shown here to associate with high GIS and, potentially, an HRD phenotype. This analysis is associated with several limitations. The first is pooling data from six different trials—with variations in study design, treatment setting, and patient selection criteria as well as diagnostic tests applied—which may have contributed to differences in the results between studies. For example, these data do not represent the patterns of genomic instability in an all-comer population, as some of the trials selected for biomarkers prospectively (OPINION [19, 20], SOLO1 [8, 13], SOLO2 [18, 21], LIGHT [12, 17]), whereas others did not (PAOLA-1 [9, 22], Study 19 [14–16]). Hence, the prevalence of BRCAm, HRRm, and non-HRRm is not reflective of all patients with ovarian cancer. In addition, LIGHT was conducted in the later-line treatment setting in patients who had received one or more prior lines of platinum-based chemotherapy [12, 17], whereas the other studies evaluated olaparib maintenance therapy in patients who responded to platinum-based chemotherapy (in combination with bevacizumab in PAOLA-1 [9, 22]). It is not clear to what extent this heterogeneity may have impacted the pooled dataset used in this analysis. Secondly, while BRCA1 methylation data are available from Study 19 [7] and PAOLA-1 [49], methylation data were not available from the other studies included in this analysis. Hypermethylation may partly explain the genomic instability seen in some non-HRRm tumors. BRCA1 hypermethylation is found in approximately 10% of ovarian tumors [50] and is associated with high levels of genomic instability [7]. In PAOLA-1, of the 72 BRCA1 or RAD51C methylated tumors with a valid GIS, 92% had a GIS ≥ 42 [49]. Similarly, a recent study in high-grade serous ovarian carcinoma showed that homozygous methylation of the RAD51C promoter is predictive of sensitivity to PARP inhibition [51]. Lack of methylation data was a limitation when assessing genomic instability in the non-BRCAm non-HRRm subgroup. A further limitation is the possibility that some large rearrangements in BRCA (germline or somatic) and some CCNE1 amplifications may not have been detected because of the assays used, causing tumors with these alterations to be underrepresented in the analysis. In addition, data regarding whole-genome duplication were not available for this analysis; however, minor allele frequency was assessed to ensure it was at zero when evaluating biallelic loss. Finally, this analysis only includes clinical trials of olaparib and may not be representative of all PARP inhibitors. In summary, this analysis of genomic data from six olaparib studies [8, 9, 12–22], including > 2000 patients with ovarian cancer, reveals distinct patterns of genomic instability in this patient population. We demonstrate that GIS has an overall bimodal distribution in ovarian cancer, with germline and somatic BRCAm tumors typically having a high GIS and high biallelic loss. We show that non-BRCA HRRm are low in prevalence and heterogeneous with regards to biallelic loss and GIS. Non-BRCA HRR genes that were associated with high GIS included BRIP1, RAD51C, and RAD51D. In terms of mutations outside of BRCAm or HRRm, NF1 and RB1 mutations had a relatively high GIS, and CCNE1-amplified and PIK3CA-mutated tumors had a low GIS and were predominantly non-BRCAm. Conclusion This analysis of data from six studies evaluating olaparib as maintenance therapy or treatment provides valuable insight into patterns of genomic instability and potential drivers of HRD, other than BRCAm, among patients with ovarian cancer. These data will help guide future research into the potential clinical effectiveness of anti-cancer treatments in ovarian cancer, including PARP inhibitors as well as other precision oncology agents. Supplementary Information Additional file 1: Table S1. Summary of ovarian cancer trials included in this analysis. Table S2. Assessment of mutation status and GIS in the ovarian cancer trials used in this analysis. Table S3. Full gene panel included in the Myriad tumor tissue testa (Myriad Genetic Laboratories, Inc.). Figure S1. GIS distribution by patient race in (A) all patients and (B) patients with a tBRCAm, by tumor histology in (C) all patients and (D) patients with a tBRCAm, and by primary tumor location in (E) all patients and (F) patients with a tBRCAm. Figure S2. GIS distribution in tumors with and without tBRCAm by individual study in PAOLA-1, OPINION, LIGHT, Study 19, SOLO1, and SOLO2. Figure S3. GIS distribution in tumors from patients (A) in response after first-line platinum-based chemotherapy and (B) with platinum-sensitive relapsed disease, and (C) GIS distribution by germline and somatic tumor BRCAm status and in non-tBRCAm tumors. Figure S4. (A) Gene-specific zygosity in tumors with BRCA1m or BRCA2m, (B) gene-specific zygosity and the rate of biallelic loss in tumors with a BRCAm, and (C) gene-specific zygosity and the rate of biallelic loss in tumors with germline BRCA1m or BRCA2m and somatic BRCA1m or BRCA2m. Figure S5. Gene-specific zygosity in patients with BRCA1m or BRCA2m by individual study. Figure S6. GIS distribution in (A) non-BRCA HRRm tumors from patients in response after first-line platinum-based chemotherapy and with platinum-sensitive relapsed disease and (B) in tumors with non-BRCA HRRm by individual study in PAOLA-1, OPINION, LIGHT, and Study 19. Figure S7. GIS distribution in patients with non-BRCA HRRm by individual study (PAOLA-1, OPINION, LIGHT, and Study 19). Figure S8. Genomic alterations detected in PAOLA-1, OPINION, LIGHT, and Study 19. Additional file 1: Table S1. Summary of ovarian cancer trials included in this analysis. Table S2. Assessment of mutation status and GIS in the ovarian cancer trials used in this analysis. Table S3. Full gene panel included in the Myriad tumor tissue testa (Myriad Genetic Laboratories, Inc.). Figure S1. GIS distribution by patient race in (A) all patients and (B) patients with a tBRCAm, by tumor histology in (C) all patients and (D) patients with a tBRCAm, and by primary tumor location in (E) all patients and (F) patients with a tBRCAm. Figure S2. GIS distribution in tumors with and without tBRCAm by individual study in PAOLA-1, OPINION, LIGHT, Study 19, SOLO1, and SOLO2. Figure S3. GIS distribution in tumors from patients (A) in response after first-line platinum-based chemotherapy and (B) with platinum-sensitive relapsed disease, and (C) GIS distribution by germline and somatic tumor BRCAm status and in non-tBRCAm tumors. Figure S4. (A) Gene-specific zygosity in tumors with BRCA1m or BRCA2m, (B) gene-specific zygosity and the rate of biallelic loss in tumors with a BRCAm, and (C) gene-specific zygosity and the rate of biallelic loss in tumors with germline BRCA1m or BRCA2m and somatic BRCA1m or BRCA2m. Figure S5. Gene-specific zygosity in patients with BRCA1m or BRCA2m by individual study. Figure S6. GIS distribution in (A) non-BRCA HRRm tumors from patients in response after first-line platinum-based chemotherapy and with platinum-sensitive relapsed disease and (B) in tumors with non-BRCA HRRm by individual study in PAOLA-1, OPINION, LIGHT, and Study 19. Figure S7. GIS distribution in patients with non-BRCA HRRm by individual study (PAOLA-1, OPINION, LIGHT, and Study 19). Figure S8. Genomic alterations detected in PAOLA-1, OPINION, LIGHT, and Study 19.