Whole Genome Landscape Analysis of Homologous Recombination Deficiency in a Pan-Cancer Cohort

I. Patient enrollment, tissue samples and clinical data acquisition

Participants were prospectively enrolled at Weill Cornell Medicine in the Institutional Review Board (IRB)-approved protocols WCM IRB #1305013903 and #1007011157. Tumor DNA for WGS was extracted from frozen tumor samples, formalin-fixed, paraffin-embedded (FFPE) archival tissue, or fluid from malignant ascites. Histopathology review was performed before DNA extraction. Germline DNA was primarily extracted from blood. When unavailable, saliva or benign tissue (frozen or FFPE) was used. We collected comprehensive clinical data, which encompassed age, sex, ethnicity, treatment-related information, radiologic findings, and pathologic data. Sequential study IDs were assigned to the samples using the format “WCM” followed by a number. These IDs are known only within the research group.

II. DNA Extraction and WGS

For DNA extraction from FFPE blocks, we used 5-micrometer-thick unstained slides macrodissected for at least 80% tumor content. For DNA extraction from frozen specimens, we employed 3mm core punches from the frozen OCT-embedded tissue. The Maxwell® 16 FFPE Plus DNA kit (Promega, Cat# AS1135) was employed, in combination with the Maxwell® 16 instrument (Promega, Madison, WI). DNA quality and quantity was assessed by using the Agilent Tapestation 4200 (Agilent Technologies) and the Qubit Fluorometer (ThermoFisher), respectively. Whole genome sequencing (WGS) was carried out at the New York Genome Center on an Illumina Novaseq6000 sequencer using 2x150 bp cycles. Libraries were generated using the KAPA Hyper Library Preparation Kit (KAPABiosystems KK8502, KK8504), targeting 500 bp fragments, in compliance with the manufacturer’s instructions. DNA fragments underwent a series of preparation steps including shearing, end-repair, adenylation, and ligation to Illumina sequencing adapters. The prepared DNA fragments were size-selected using bead-based methods and amplified ²³. Quality and quantity of the final libraries were assessed prior to sequencing.

III. WGS data processing pipeline and HRD curation

We employed the Isabl GxT analytics platform to process all WGS and RNA sequencing data and generate comprehensive reports ²⁴. Within the Isabl pipeline deployed in AWS cloud HPC environments, DNA and RNA alignment with BWAMem and STAR, quality control, somatic and germline variant calling, and annotation were performed as previously described ^25-29. Briefly, ensemble calling was done for both somatic and germline single nucleotide variants (SNV), InDels, and somatic SVs. These variant classes were then annotated and those in which at least 2 of 3 callers of each class were included for reporting. Purity, ploidy, and genome-wide copy number states were estimated with Battenberg, followed by annotation of CNA events at the gene level. Driver alterations for all variant classes and their potential treatment targets were assessed by cross-referencing protein-coding variants with COSMIC and OncoKB.

WGS-HRD scores were assigned with Isabl HRD, random forest classifier trained on a subset of the data (N=321 patients) and described in and Hadi et al. (abstract) ³⁰ to detect HRD by incorporating evidence from genome-wide SNV, indels, SV, and CNV signals (Figure S1). Briefly, individual patients with HRD were identified as the top quartile outliers for features associated with HRD (COSMIC SBS3, deletions with microhomology, and small SV duplications and deletions). Cases that had at least two outlier signatures were manually inspected for presence of HRD by visualizing circos plots and signature frequencies to identify those with HRD. 96 trinucleotide SNV contexts, 45 InDel types, 38 SV types, and LST, LOH, and TAI scores were used as features for the classifier. On a validation cohort of 556 PCAWG samples, the classifier achieved a high AUROC/AUPRC of 0.99/0.96 ³⁰. We applied the classifier on the entire cohort, which assigned a probability of presence of the HRD phenotype. The resulting score ranged from 0 to 1, with a score equal to or greater than 0.5 indicating HRD (≥0.5) and a score less than 0.5 labeled as HRR proficient (HRP <0.5). Tumor purity > 20% is required to accurately assess HRD.

IV. Fluorescence in situ Hybridization

Four-µm-thick formalin-fixed paraffin-embedded tissue sections were used for fluorescence in situ hybridization (FISH) analysis, as described in our protocols ^31-33. Bacterial artificial chromosomes were designed against loci of interest to prepare break-part dual-color FISH probes ³⁴. For BRCA2 RP11-110O22 BAC clone was labeled red and RP11-11K16 clone was labeled green, and for ATM RP11-144G7 BAC clone was labeled red and RP11-589O5 clone was labeled green. All clones were validated on normal metaphase spreads before any application on FFPE tissue. A positive break-apart was determined by one red, one green, and one yellow signal (combination of red and green signal indicating the normal chromosome homologue). At least 200 nuclei were analyzed per case using a fluorescent microscope (Olympus BX51; Olympus Optical). Cytovision 7.3.1 software was used for imaging and analysis.

Sample Characteristics and Frequency of HRD

We performed WGS analysis on 580 samples from 453 patients. Figure 1 summarizes the types of tumor samples (FFPE, frozen, or fluid) and type of lesion (primary and metastatic). These encompass 77 unique histology types (Figure S2), classified according to the MSKCC Oncotree ³⁵, and obtained from 33 unique primary sites. Sites were divided into six cancer subgroups: prostate (29%), gynecological (20%), pancreaticobiliary (15%), breast (11%), upper gastrointestinal (GI) (10%), and “others” (12%), which included smaller cancer cohorts and rare tumors (Figure 1, Figure S2).

• Download figure
• Open in new tab

Fig 1. Cohort Characteristics.

(A) Left panel, overall breakdown of cohort by specimen type (left) and whether biopsy was from a primary or metastatic tumor (right). (B) Breakdown of the Isabl HRD score and HRD status (cutoff > 0.5) across the cohort. FFPE, formalin-fixed, paraffin-embedded; GI, gastrointestinal; HRD, homologous recombination deficiency.

Among the total 580 samples, 62 samples across 53 patients exhibited HRD by WGS. The highest percentage of HRD cases were found in breast cancer (19%) followed by pancreaticobiliary (17%), gynecological (15%) and prostate cancers (8%). Upper GI cancers had the lowest percentage of HRD with only one case. In addition, 1 case of carcinoma of unknown primary in the “others” cohort harbored HRD.

Mutational landscape of HRD

While BRCAness is most commonly explained and clinically tested by alterations in BRCA1/2, dysfunction in other HRR pathway genes can also result in phenocopy. We investigated events in BRCA1, BRCA2, and other HRR pathway genes in HRD cases, accounting for loss of heterozygosity and compound hits that could result in biallelic loss of function and focusing on SVs and somatic mutations with unknown effects on HRD (Figure 2.A). Out of 62 HRD samples across 53 patients, 76% (47) harbored alterations in BRCA1 and/or BRCA2(BRCA1/2mut). In 55% (34) of samples, BRCA1/2mut had biallelic pathogenic small mutations (SNVs and InDels) and 6% (4) had homozygous deletions. Another 3% (2) harbored SVs with LOH in BRCA1/2 with predicted impact to coding sequence. Interestingly, none of those two cases with BRCA1/2 biallelic SVs had deleterious mutations in other HRR pathway-related genes (Figure 2). 11% (7) harbored BRCA1/2 small mutation VUS or SV VUS. Ultimately, the remaining 15 HRD cases (24%) were BRCAwt (Figure 2.A and Figure 2.B).

• Download figure
• Open in new tab

Fig 2. Genomic alterations and features of HRD.

(A) Landscape of BRCA1/2 and other HR-associated genes in HRD cases. (B) Top, BRCA1/2 alteration status in the HRD cohort by variant class. Bottom, further delineating alterations in genes outside of BRCA1/2 in cases that have BRCA1/2 events not considered in standard clinical practice (SSV or VUS) or are BRCAwt. (C) Genomic feature enrichment (Z-score calculated across the full HRD & HRP cohort) in HRD. Notably, the features of HRD are not distinguished by gene target or variant type. Del, deletion; Dup, duplication; GI, gastrointestinal; Hom, homozygous; HRD, homologous recombination deficiency; HRR, homologous recombination repair; MH, microhomology; Mut, mutation; SBS, single base substitution; SV, structural variant; VUS, variant of unknown significance; WGS, whole genome sequencing; WT, wild-type.

For the 15 HRD samples with BRCAwt we investigated other gene mutations that might have caused the HRD phenotype and we found that 1 had a pathogenic small mutation in FANCF, 1 had biallelic deletion in XRCC2, and 1 had biallelic FANCC deletion. The remaining BRCAwt HRD cases (n=12) either had no mutations in any of the known HRR genes (n=5; 33% of the BRCAwt HRD cases) or had a VUS small mutation or SV in one or multiple HRR pathway genes including RAD51B, PALB2, ATM (Supplementary figure 2), CHEK2, FANCD2 and RAD51 (n=7; 47% of the BRCAwt HRD cases).

Features frequently enriched in the HRD cohort consisted of genome wide small microhomology deletions(MH-dels), COSMIC signatures (V3) SBS3 or SBS40, SV deletions from 1-10kb, and SV duplications from 1-10kb (Figure 2.C). SV duplications were largely enriched in BRCA1 mutated tumors. While MH-dels were the most frequently enriched feature among HRD samples, 3 BRCAwt, RAD51B mutated samples harbored intermediate MH-dels with enrichment of SV deletions or duplications and SBS3/40 indicating a potentially distinct HRD feature profile.

It is worth noting that 9 out of 62 HRD samples represented additional biopsies. These 9 harbored similar BRCA1/2 status as the original patient sample (8 small pathogenic BRCAmut, 1 BRCAwt). In this series of patients, HRD was consistently present across all except 1 patient in which the primary tumor (HRD score 0.94) harbored an SV breakpoint in exon1 of BRCA1 coinciding with LOH while the metastasis was borderline (0.44) and did not harbor evidence of the BRCA1 SV nor LOH.

We also investigated HRD positivity using two other WGS based HRD algorithms, CHORD and HRDetect (Figure 2.A) ^3,19. Overall, 9 HRD samples were negative by CHORD, which included 3 cases that are simultaneously negative by HRDetect. Interestingly, 2 of those cases were BRCAwt with RAD51 structural variants, and one with BRCA2 monoallelic VUS somatic small mutation and BRCA1 biallelic pathogenic somatic small mutation.

Clinical and genomic characteristics of HRD cases with BRCAwt and BRCA1/2 SV

Standard clinical decisions to treat HRD cancers largely relies on the presence of small mutations or copy alteration in BRCA1/2 based on targeted panel testing. Having analyzed the mutational and feature landscape of HRD on our cohort using WGS, we then sought to examine the patient’s response to treatment in HRD cases in which clinical history was available and which presented no BRCA1/2 small mutations or copy alterations.

The first case, WCM414, is a patient diagnosed with high-grade serous fallopian tube carcinoma (HGSFC). She underwent resection then PBCT. WGS demonstrated HRD, indicated by the presence of tandem duplications and 56% contribution from SBS3. Additionally, she had BRCAwt with a biallelic small VUS mutation in ATM. The corresponding MyChoice CDx test also indicated HRD and confirmed BRCAwt status. Consequently, she was administered maintenance PARPi and has shown no evidence of disease recurrence to date, with a PFS of 20 months.

The second case, WCM209, is a patient with high-grade serous carcinoma of the ovary (HGSOC) (Figure 3.A). WGS analysis showed no mutations in BRCA1, BRCA2, or other HRR pathway genes, a finding that is consistent with the patient’s clinical genomic testing, which included FoundationOne and Invitae germline mutation assays. Notably, WGS indicated HRD, characterized by a SBS3 signature contribution of 22% and an increase in small deletions and tandem duplications (Figure 3.A). A relevant finding was a translocation resulting in a HIF1A::UIMC1 fusion involving intron 13 of UIMC1, coupled with LOH. UIMC1 (Ubiquitin Interaction Motif Containing 1) encodes a nuclear protein that collaborates with BRCA1 in recognizing and repairing DNA lesions ³⁶. This fusion was validated by RNA sequencing analysis (Figure 3.A). MyChoice CDx testing was also performed, which indicated HRD. The patient was treated with PBCT, achieving near-complete resolution of the metastatic tumor, as assessed by RECIST 1.1 criteria.

• Download figure
• Open in new tab

Fig 3. Validation of alterations in 2 WGS HRD positive cases.

(A) HGSOC (H&E, 20x) tumor harboring HRD signals as shown in the circos plot with high rate of microhomology deletions, SBS3, SV tandem duplications, and SV deletions. WGS detected an in-frame HIF1A::UIMC1 fusion that leads to a truncated, non-functional chimeric transcript that is missing exon 6, necessary for binding to BRCA1. (B) PRNE (H&E, 20x) tumor with HRD that harbored a BRCA2::TMPRSS2 fusion validated by FISH. H&E, hematoxylin and eosin; HGSOC, high grade serous ovarian carcinoma; HRD, homologous recombination deficiency; Indel, insertion deletion; LOH, loss of heterozygosity; PRNE, prostate neuroendocrine carcinoma; SBS, single base substitution; SV, structural variant.

The third case (WCM561) is a patient diagnosed with advanced prostatic carcinoma. WGS was performed on a metastatic brain lesion that exhibited a neuroendocrine histology (Figure 3.B). WGS revealed a partial homozygous loss in BRCA2 that was caused by an SV translocation at intron 21 resulting in a BRCA2::TMPRSS2 fusion accompanied by LOH in the intact portion of the gene. The presence of this fusion was validated by FISH, which demonstrated BRCA2 break-apart signals. WGS also indicated HRD in this case, with increased small deletions (Figure 3.B). Unfortunately, treatment response data for this patient was not available.

These examples illustrate how WGS can identify additional HRD cases, particularly those with BRCAwt.

Comparison of WGS HRD with commercially available assays

Current commercial tests for HRD employ only GIS or LOH scores based on allele-specific copy number segmentation, a subset of genome-wide features predictive of HRD and captured by WGS. To explore the potential clinical significance of WGS-based HRD testing, we retrospectively assessed cases with WGS, commercial HRD scores, and clinical history that included treatment response. From the 39 serous carcinoma cases in the gynecological cancer cohort, where an HRD score can inform treatment decisions, we identified 16 cases that had undergone a commercial assay that included an HRD score. This group comprised 9 cases evaluated with MyChoice CDx GIS and 7 cases with FoundationOne HRD score. The results of both commercial assays and WGS testing are summarized in Table 1. Of these cases, 3 out of 5 identified as HRD-positive by WGS were deemed HRP by commercial assays; these included 1 case tested by FoundationOne (negative) and 2 cases by MyChoice CDx, one negative and the other inconclusive for HRD. Among cases determined as HRP by WGS, 3 out of 10 were found to be positive according to FoundationOne. To delve deeper into the reasons for these discrepancies and assess their potential clinical significance, we analyzed the genomic features of all cases with divergent results. We also retrospectively correlated the scores with responses to PBCT and PARPi treatments, as per RECIST 1.1 criteria, where available.

Cases with commercial testing negative and WGS homologous recombination deficiency (HRD) status

WCM231 sample was obtained from a HGSOC peritoneal metastases. FoundationOne testing indicated HRD-negative status, although it revealed a pathogenic BRCA1 E111fs*1 mutation. WGS confirmed the same BRCA1 pathogenic mutation with LOH and showed a HRD phenotype with a high fraction of microhomology deletions (56%) with a high prevalence of COSMIC SBS3 (>6000), and small SV deletions (50) and duplications (30). The patient was treated with neoadjuvant PBCT and exhibited a partial response to treatment according to RECIST 1.1 criteria. The patient was maintained on PARPi for 10 months before experiencing recurrence.

WCM205 sample was obtained from a HGSOC splenic metastasis. MyChoice CDx showed HRP and BRCAwt. WGS did not detect BRCAmut and other HRR pathway genes mutations. However, WGS showed an HRD phenotype with a moderate high fraction of microhomology deletions (34%) and enrichment of SBS3 (>3000) with a high load of small SV deletions (49). Treatment with PBCT for recurrent peritoneal carcinomatosis showed partial response according to RECIST 1.1 criteria.

Lastly, WCM228 HGSOC sample was obtained from the ovary. MyChoice CDx testing yielded inconclusive results for HRD. Targeted NGS testing identified a BRCA2 VUS. WGS confirmed the same monoallelic BRCA2 VUS mutation and revealed a RAD51B intron 7-intron 8 deletion with loss of heterozygosity. WGS also showed an HRD phenotype with a high fraction of microhomology deletions (49%), high load of SV deletions (31) and duplications (30), and a moderate load of SBS3 (>3000). RECIST 1.1 assessment of the response to the neoadjuvant PBCT showed an overall partial response to treatment. The patient was later maintained on rucaparib and did not experience disease recurrence for 5 years following the diagnosis.

Cases with Commercial testing positive and WGS homologous recombination repair-proficient (HRP) status

WCM425 sample originated from the HGSOC omental metastasis. FoundationOne testing indicated HRD-positivity with BRCAwt, and it identified an APC VUS. WGS revealed a HRP phenotype. Rucaparib, used for maintenance therapy, was discontinued two months later due to disease progression.

WCM436 was obtained from another HGSOC omental metastasis. FoundationOne showed HRD with EP300 mutation and BRCAwt. WGS showed HRP. However, tumor purity as determined by WGS was low (0.12), and cannot derive a confident result. The RECIST 1.1 assessment showed a response to treatment with PBCT.

WCM236 sample was a HGSOC primary. FoundationOne testing indicated HRD. WGS showed an HRP phenotype, although it identified a biallelic RAD51B translocation at exon 14. The patient showed disease progression 3 months after starting olaparib maintenance (RECIST 1.1).