Homologous Recombination Deficiency Score Research Articles

Abstract PO1-16-03: GDP-mannose enhances the efficacy of PARP inhibitors and immunotherapy in triple-negative breast cancer

Abstract Background: Triple-negative breast cancer (TNBC) patients with high HRD scores benefit from poly (ADP-ribose) polymerase (PARP) inhibitors, while those with low HRD scores still lack therapeutic options. Homologous recombination deficiency (HRD) induction is an effective strategy to broaden the indications of PARP inhibitors. Metabolic reprogramming is a critical feature of TNBC. In this study, we aimed to explore novel metabolic biomarkers for HRD and new strategy for sensitizing PARP inhibitors. Methods: We utilized TNBC metabolomics to systematically evaluate metabolites that were correlated with HRD. A crucial metabolite, GDP-mannose (GDP-M), that impedes homologous recombination repair (HR) and sensitizes PARP inhibitors was identified and functionally validated. We further explored the detailed mechanism and proposed a potential treatment strategy utilizing GDP-M for TNBC in preclinical models. Results: Systematic metabolomic analysis revealed that GDP-mannose (GDP-M) was significantly enriched in basal-like tumors with HRD. GDP-M promoted cisplatin-induced DNA double-strand breaks by inducing HRD. Mechanistically, the low expression of the upstream enzyme GMPPA led to the endogenous upregulation of GDP-M, which further promoted the ubiquitin-mediated degradation of BRCA2 to inhibit HR. GMPPA expression and GDP-M could serve as predictive biomarkers for HRD and the response to PARP inhibitors. Therapeutically, we validated that the combination of GDP-M and PARP inhibitors synergistically inhibit tumor growth in multiple preclinical models. Moreover, GDP-M supplementation plus PARP inhibition activated STING-dependent antitumor immunity and further augmented the efficacy of anti-PD-1 antibodies. Conclusions: GDP-M promotes the degradation of BRCA2 and thus enhances the sensitivity of PARP inhibitors in TNBC. The combination of GDP-M, PARP inhibitors and anti-PD-1 immunotherapy may be a potential treatment strategy for HRD-low TNBC. Citation Format: Yi Xiao, Jiahan Ding, Yi-Zhou Jiang, Zhi-Ming Shao, Genhong Di, Fan Yang, Xiaoqing Song. GDP-mannose enhances the efficacy of PARP inhibitors and immunotherapy in triple-negative breast cancer [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO1-16-03.

Abstract PO2-08-11: The MAGIC study: Universal whole genome tumour and germline sequencing of newly diagnosed breast cancer identifies a high proportion of ER+ BRCA-like tumours

Abstract Background: Identification of germline mutations in hereditary breast cancer (HBC) genes can have profound benefits in the treatment of breast cancer (BC) and manging risk for patient and the family. In addition, sequencing of the tumours can reveal clinically relevant somatic features, such as homologous recombination deficiency (HRD) and mutational signatures, that can guide treatment which are not discernible by germline sequencing alone. We hypothesised that current standard of care where women diagnosed with BC are not routinely offered tumour sequencing is suboptimal as fails to exploit all therapeutic vulnerabilities which can be exploited in the treatment of BC. The MAGIC study is the first prospective trial in Australia of unselected invasive BCs in general oncology practice combining both germline and tumour sequencing. Methods: The MAGIC study included a total of 651 consecutive consented patients presenting with non-metastatic BC between June 2020 to March 2023. It included invasive BC, high-grade DCIS and pleomorphic LCIS. The sequencing was conducted in two phases; for phase one, 157 cases underwent whole genome sequencing (WGS) on both germline and matched BC DNA while for phase two, 494 cases underwent only germline whole exome sequencing. Germline variants were interrogated for pathogenic variants in BRCA1, BRCA2, PALB2, ATM, CHEK2, BARD1, BRIP1, RAD51B, RAD51C, RAD51D, MLH1, MSH2, MSH6, PMS2, CDH1, PTEN, STK11, TP53 and NTHL1. Tumour homologous recombination (HR) repair deficiency was calculated using HRDetect (Nat Med 2017;23:517). For BCs showing a high HRDetect score ( >0.75) all HBC gene promoter CpG islands were assessed for hypermethylation using the Twist NGS Methylation system. Mutational signatures were calculated from somatic mutations identified from whole genome sequenced BCs using the DeconstructSig package in R (Genome Biol. 2016;17:31) and referenced against the COSMIC v2 signature catalogue. Results: Actionable germline variants in any HBC gene were identified in 7.7% of cases (7.8% of invasive and 6.1% of in situ BC) with 5.2% having an actionable mutation in an HR pathway gene (BRCA1, BRCA2, PALB2 and RAD51C). Most cases with pathogenic variants showed bi-allelic inactivation in the tumour. All BRCA1, BRCA2 and PALB2 tumours showed high HRDetect and/or HRD scores consistent with loss of HR repair function, indicating them to be the underlying hereditary driver of the BC. One BC with a pathogenic PMS2 variant showed retention of the wild-type allele with no evidence of a mismatch repair mutational signature suggesting PMS2 was not the driver of this BC. Of the 157 BC in phase one that underwent WGS and did not have a germline pathogenic HBC gene variant, 16% (18/117 invasive and 3/13 DCIS) had an HRDetect score of >0.7 indicative of a homologous recombination repair defect and therefore biologically a “BRCA-like” tumour. Of these 21 BRCA-like cases, only 3 invasive BCs had clear somatic bi-allelic inactivation of an HBC gene that could explain the high HRDetect score (2 BRCA1 and 1 BRCA2). One of the 21 BRCA-like cases had a germline variant of unknown significance (VUS) in an HBC gene (a BRCA2 missense variant in an ER+ tumour. The tumour showed loss of the wild-type allele). Of the BRCA-like cases, 71% were ER positive which included 3 invasive lobular cancers. Conclusion: Tumour sequencing identified three times as many cases that might be eligible for HR repair defect targeted therapy than germline testing alone (16% versus 5.2%). Unlike a previous study (Ann Oncol., 2019 30:1071-1079), the majority of the BRCA-like BCs were ER+ with no evidence of a VUS in a known HBC gene suggesting these are driven by novel germline or somatic mutations in genes involved in DNA HR repair. Citation Format: Ian Campbell, Lisa Devereux, Kirsten Hogg, Luxi Lal, Michelle Sinclair, Lesley Stafford, Magnus Zethoven, Anita Skandarajah, Paul James, Geoffrey Lindeman, Bruce Mann, Dilanka De Silva. The MAGIC study: Universal whole genome tumour and germline sequencing of newly diagnosed breast cancer identifies a high proportion of ER+ BRCA-like tumours [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO2-08-11.

Clustering of HR + /HER2− breast cancer in an Asian cohort is driven by immune phenotypes

Breast cancer exhibits significant heterogeneity, manifesting in various subtypes that are critical in guiding treatment decisions. This study aimed to investigate the existence of distinct subtypes of breast cancer within the Asian population, by analysing the transcriptomic profiles of 934 breast cancer patients from a Malaysian cohort. Our findings reveal that the HR + /HER2− breast cancer samples display a distinct clustering pattern based on immune phenotypes, rather than conforming to the conventional luminal A-luminal B paradigm previously reported in breast cancers from women of European descent. This suggests that the activation of the immune system may play a more important role in Asian HR + /HER2− breast cancer than has been previously recognized. Analysis of somatic mutations by whole exome sequencing showed that counter-intuitively, the cluster of HR + /HER2− samples exhibiting higher immune scores was associated with lower tumour mutational burden, lower homologous recombination deficiency scores, and fewer copy number aberrations, implicating the involvement of non-canonical tumour immune pathways. Further investigations are warranted to determine the underlying mechanisms of these pathways, with the potential to develop innovative immunotherapeutic approaches tailored to this specific patient population.

Somatic mutations that affect early genetic progression and immune microenvironment in gastric carcinoma

Gastric carcinoma (GC) is a high heterogeneity and malignant tumor with a poor prognosis. The current implementation of immunotherapy in GC is limited due to the insufficient exploration of immune-related mutations and speculated early mutation events. Therefore, we performed whole-exome sequencing on 40 patients with GC to explore their genetic characteristics, shedding light on the order of genetic events, somatic mutations impacting the immune microenvironment, and potential biomarkers for immunotherapy. Regarding genetic events, TP53 disruptions were identified as frequent and early events in GC progression, often occurring alongside other gene mutations. The mutations occurring in GANS, SMAD4, and POLE were early independent events. Patients harboring CSMD3, FAT4, FLG, KMT2C, LRP1B, MUC5B, MUC16, PLEC, RNF43, SYNE1, TP53, TTN, XIRP2, and ZFHX4 mutations tended to have decreased B cells, T cells, macrophage, neutrophil, and dendritic cells infiltration, except for the ARID1A gene mutations. We also found patients with microsatellite instability-high tumors had higher homologous recombination deficiency (HRD) scores. HRD showed a positive correlation with tumor mutational burden, which might serve as indirect evidence supporting the potential of HRD as a biomarker for GC. These findings highlighted GC's high heterogeneity and complexity and provided valuable insights into the somatic mutations that affect early genetic progression and immune microenvironment.

TP53-specific mutations serve as a potential biomarker for homologous recombination deficiency in breast cancer: a clinical next-generation sequencing study

Abstract Background TP53 mutations and homologous recombination deficiency (HRD) occur frequently in breast cancer. However, the characteristics of TP53 pathogenic mutations in breast cancer patients with/without HRD are not clear. Methods Both tumor and paired blood DNA from 119 breast cancer patients were performed clinical next-generation sequencing (NGS) by a 520-gene panel. Mutations, tumor mutation burden (TMB), and genomic HRD scores were assessed from NGS data. Results All TP53 pathogenic mutations in patients were the somatic origin, which was associated with the protein expression of estrogen receptor and progestogen receptor. Compared to patients without TP53 pathologic mutations, patients with TP53 pathologic mutations had higher levels of HRD scores and different genomic alterations. The frequency of TP53 pathologic mutation was higher in the HRD-High group (HRD score≥42) relative to that in the HRD-Low group (HRD score<42). TP53 has different mutational characteristics between the HRD-Low and HRD-High groups. TP53-specific mutation subgroups had diverse genomic features and TMB. Notably, TP53 pathogenic mutations predicted the HRD status of breast cancer patients with an AUC of 0.61. TP53-specific mutations, namely HRD-Low mutation, HRD-High mutation, and HRD common mutation, predicted the HRD status of breast cancer patients with AUC values of 0.32, 0.72, and 0.58, respectively. Interestingly, TP53 HRD-High mutation and HRD common mutation combinations showed the highest AUC values (0.80) in predicting HRD status. Conclusions TP53-specific mutation combinations predict the HRD status of patients, indicating that TP53 pathogenic mutations could serve as a potential biomarker for PARP inhibitors in breast cancer patients.

Abstract LB391: Deep learning AI predicts HRD and platinum response from histologic slides

Abstract Background: Cancers with homologous recombination deficiency (HRD) can benefit from platinum salts and PARP inhibitors. Standard diagnostic tests, including FDA-approved companion diagnostics, for detecting HRD require molecular profiling, which is not universally available with global testing rates lowest among minority, rural, and other underserved populations. Methods: We trained DeepHRD, a deep-learning platform for predicting HRD from hematoxylin and eosin (H&E)-stained histopathological slides, using primary breast (n=1,008) and ovarian (n=459) cancers from The Cancer Genome Atlas (TCGA). DeepHRD was compared to four standard HRD molecular tests using breast (n=349) and ovarian (n=141) cancers from multiple external and independent datasets, including clinical cohorts with platinum complete response, progression-free survival (PFS) and overall survival (OS) endpoints. Results: DeepHRD detected HRD from held-out H&E-stained breast cancer slides in TCGA with an AUC of 0.81 ([0.77-0.85]; 95% confidence interval). This performance was confirmed in two independent primary breast cancer cohorts (AUC=0.76; [0.71-0.82]). In an external platinum-treated metastatic breast cancer cohort, samples detected as HRD had a higher complete response (AUC=0.76; [0.54-0.93]) with 3.7-fold increase in median PFS (14.4 versus 3.9 months; p-value=0.0019) and hazard ratio (HR) of 0.45 (p=0.0047) after correcting for PAM50 molecular subtype and age at diagnosis. This deep-learning classifier outperformed four genomic HRD tests used in the clinic, including standard HRD score, BRCA1/2, 26-HR gene panel and single-base substitution signature 3 (SBS3). Multiresolution spatial mapping identified morphological features utilized by DeepHRD for detecting HRD, notably enriched for neoplastic and necrotic tissues, and a higher macrophage density. Through transfer learning to high-grade serous-ovarian cancer, DeepHRD-positive samples exhibited better overall survival after TCGA first-line (HR=0.46; p=0.030) and an external neoadjuvant (HR=0.49; p=0.015) platinum-treated cohorts. Conclusion: In summary, DeepHRD exhibits consistent hazard ratios ranging from 0.45 to 0.49 across the three clinical cohorts and captures 1.8- to 3.1-fold more HRD-positive breast and ovarian cancer patients. DeepHRD-positive breast cancer patients that received platinum exhibited better complete response and PFS. Similarly, DeepHRD-positive platinum-treated ovarian cancer patients had a better OS. DeepHRD’s ability to detect HRD from digital H&E slides provides an important precision oncology tool that can be utilized in resource-constrained and underserved areas where genomic testing is generally not existent. Citation Format: Erik N. Bergstrom, Ammal Abbasi, Marcos Diaz-Gay, Loïck Galland, Sylvain Ladoire, Scott M. Lippman, Ludmil B. Alexandrov. Deep learning AI predicts HRD and platinum response from histologic slides [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr LB391.

Prediction of homologous recombination deficiency from Oncomine Comprehensive Assay Plus correlating with SOPHiA DDM HRD Solution.

Poly(ADP-ribose) polymerase (PARP) inhibitors are used for targeted therapy for ovarian cancer with homologous recombination deficiency (HRD). In this study, we aimed to develop a homologous recombination deficiency prediction model to predict the genomic integrity (GI) index of the SOPHiA DDM HRD Solution from the Oncomine Comprehensive Assay (OCA) Plus. We also tried to a find cut-off value of the genomic instability metric (GIM) of the OCA Plus that correlates with the GI index of the SOPHiA DDM HRD Solution. We included 87 cases with high-grade ovarian serous carcinoma from five tertiary referral hospitals in Republic of Korea. We developed an HRD prediction model to predict the GI index of the SOPHiA DDM HRD Solution. As predictor variables in the model, we used the HRD score, which included percent loss of heterozygosity (%LOH), percent telomeric allelic imbalance (%TAI), percent large-scale state transitions (%LST), and the genomic instability metric (GIM). To build the model, we employed a penalized logistic regression technique. The final model equation is -21.77 + 0.200 × GIM + 0.102 × %LOH + 0.037 × %TAI + 0.261 × %LST. To improve the performance of the prediction model, we added a borderline result category to the GI results. The accuracy of our HRD status prediction model was 0.958 for the test set. The accuracy of HRD status using GIM with a cut-off value of 16 was 0.911. The Oncomine Comprehensive Assay Plus provides a reliable biomarker for homologous recombination deficiency.

Abstract 2328: Next-generation sequencing-based regression algorithm to determine homologous recombination deficiency scores in a pan-cancer cohort

Abstract Background: Homologous recombination deficiency (HRD) inhibits double strand breaks from being repaired in DNA, leading to cancer cells failing to recover themselves and cell death. Previous studies found that HRD-positive ovarian cancer patients had more significant clinical benefits from poly ADP-ribose polymerase inhibitors (PARPi) treatment. For accurate detection of HRD, next-generation sequencing (whole exome, whole genome) was used to identify large-scale genomic aberrations including telomeric allelic imbalance (HRD-TAI score), loss of heterozygosity profiles (HRD-LOH score), and large-scale state transitions (HRD-LST score). So far, many studies (HRDetect, SigMA, CHORD and shallowHRD) have been conducted to accurately determine HRD in a pan-cancer cohort. As the study progressed, it has been revealed that PARPi is effective when applied to patients with HRD in not only ovarian cancer but also breast cancer, prostate cancer, and pancreatic cancer. Purpose: To develop an algorithm that accurately predict HRD score in patients with ovarian cancer, breast cancer, prostate cancer, and pancreatic cancer utilizing a machine learning algorithm. Methods: We used whole-genome sequencing (WGS) data of 710 samples from 309 patients, whole-exome sequencing (WES) data of 4,650 samples from 2,193 patients from the pan-cancer cohort of the TCGA (TCGA-OV, TCGA-BRCA, TCGA-PRAD, and TCGA-PAAD). The HRD-TAI/HRD-LST/HRD-LOH scores were calculated through structural variation analyses. After data cleaning processes, machine-learning models were trained and tested on 228 out of 2,502 samples, and validation was performed on 1,248 out of 2,502 samples from the TCGA dataset. Results: To assess the performance of the machine-learning regressor, the concordance between predictions and annotations was quantified by calculating the R squared (𝑅;;;^2). As a result of training using machine learning algorithm, we achieved a high 𝑅;;;^2 (0.904) with a RMSE (root mean squared error) score (7.649) for a pan-cancer cohort. Conclusions: Our regressor was robust and accurate when applied to 4 cancer types. Using our systematic pan-cancer analysis, we found novel insights into the mechanisms of HRD across cancer types with potential contribution to clinical practice. Citation Format: Boram Choi, Youn Jin Choi. Next-generation sequencing-based regression algorithm to determine homologous recombination deficiency scores in a pan-cancer cohort [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2328.

Abstract 1454: Expanding the therapeutic application of PARP inhibitor: AI evaluation of real-word clinico-genomic data from spontaneous canine tumors

Abstract The widespread application of Poly-ADP ribose polymerase (PARP) inhibitors in anti-cancer therapy is constantly increasing their role as a treatment strategy across various neoplasms, the majority of which are linked with BRCA deficiency. While the most significant advantages are observed in cancers with BRCA1/2 mutations, it is evident that the benefits extend beyond this specific group. Preclinical evidence encourages the exploration of PARP inhibitors in neoplasms demonstrating BRCAness or homologous recombination deficiency (HRD), both as monotherapy and in combination with chemotherapy. This study employs spontaneous canine tumors as a complementary model for cancer research, aiming to seamlessly bridge the transition from preclinical investigations to clinical application. Leveraging the FidoCure® Precision Medicine Platform, we identified biomarkers associated with prognosis and treatment prediction, specifically with the PARP inhibitor olaparib, facilitating a smoother translation from laboratory findings to practical clinical use. Analyzing real-world clinico-genomic data from 1278 dogs with cancer revealed intriguing insights of specific targeted treatments developed for humans showing a positive prognosis when applied to canine tumors with specific genomic alterations. Notably, olaparib exhibited efficacy in TP53 and BRCA1-mutated cases (OS HR 0.34, P < 0.001; HR 0.39, P = 0.004), while rapamycin (mTOR inhibitor) demonstrated promising outcomes in TP53 and RB1-mutated canine tumors (OS HR 0.73, P = 0.028; HR 0.32, P = 0.024). Further stratification by tumor types unveiled noteworthy correlations. TP53 mutant osteosarcomas exhibited improved prognosis with olaparib treatment (OS HR 0.11, P < 0.001), and soft tissue sarcomas with TP53 mutations demonstrated favorable responses to rapamycin and olaparib therapy (OS HR 0.11, P = 0.012; HR 0.07, P = 0.011). These findings underscore the potential of leveraging canine tumor models and clinico-genomic analyses to inform targeted treatment strategies, offering valuable insights for advancing precision medicine in both veterinary and human oncology. Further investigation into the association between TP53 mutation and olaparib response is warranted. The intriguing aspect lies in the role TP53 mutation has been linked in previous studies to heightened chromosomal instability and elevated HRD scores. Utilizing real-world evidence and data tools within the FidoCure dataset, we have effectively pinpointed correlations between gene mutations and survival, particularly in relation to responses to targeted therapy treatment. Canine models, mirroring human diseases with intact immune systems and comparable tumor genomic profiles, as verified by the FidoCure database, expedite clinical studies of novel treatments that face significant scalability challenges. Citation Format: Lucas Rodrigues, Kevin Wu, Garrett Harvey, Gerald Post, Abigail Hull, Aubrey Miller, Lindsay Lambert, Christina Lopes, James Zou. Expanding the therapeutic application of PARP inhibitor: AI evaluation of real-word clinico-genomic data from spontaneous canine tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1454.

Abstract B114: Effect of spatial and temporal inter-tumoral heterogeneity on homologous recombination deficiency scores in high grade serous ovarian cancer

Abstract High-grade serous ovarian cancer (HGSC) is the most common and lethal subtype of ovarian cancer. Extensive genomic instability and heterogeneity are characteristics of HGSC, with near-ubiquitous TP53 loss-of-function mutations, defects in homologous recombination (HR) repair, and extensive copy number aberrations contributing to the vast genomic heterogeneity observed. Most patients relapse and acquire resistance to platinum or PARP inhibitor (PARPi)-based therapy. Diverse mechanisms leading to therapy resistance and lack of predictive biomarkers means that matching the best treatment options to patients is difficult. Testing tumors for HR-Deficiency (HRD) status is now a clinical test to predict PARPi sensitivity in HGSC. We aimed to determine the influence of spatial and temporal heterogeneity on HRD status in HGSC. Patients (n=59) at Hammersmith Hospital (HH) underwent maximal-effort upfront cytoreduction for advanced HGSC, had their tumor dissemination mapped and tumor biopsies collected (range 4-15). Paired relapse samples were collected for 10 patients. Tumor DNA was extracted (n=5 tumors per case, plus relapse), Illumina Human OmniExpress genotyping performed for 59 cases and whole genome sequencing (WGS; 30x depth) performed for 21 cases. Allele-specific copy number (CN) was quantified using ASCAT. Multi-site CN data was also accessed from two cohorts (GSE38787, GSE40546). Tumors that failed QC or had an aberrant cell fraction <0.3 were excluded. HR scores were estimated in each cohort using different genomic instability score (GIS) or mutational signature-based algorithms and relevant cut-offs were applied to determine HRD and HR-Proficient (HRP) tumors. Variations in HR scores were observed across the cohorts with a proportion of patients presenting with a mixed HR score, displaying both HRD and HRP scores in their tumors: HH cohort (20%, 11/54 patients); GSE38787 (17%, 4/24 patients); and GSE40546 (28%, 4/14 patients) using a scarHRD algorithm. In the HH cohort (n=45), we examined if mixed HR status was associated with survival and demonstrated that patients with HR-mixed or all HRP tumors had poorer progression-free (p=0.0052) and overall survival (p=0.00092) than patients with all HRD tumors. Applying different approaches to the HH WGS data (scarHRD, CHORD, >63 score cut-off), approximately 20% of patients demonstrated HR-mixed scores in their tumors in each approach. However, depending on the approach chosen to define HRD, different cases showed HR-mixed scores, 2/21 mixed cases overlapped between the different algorithms applied with tumors changing from HRD to HRP and vice versa depending on the approach applied. Differences in HR-deficiency/proficiency scores suggests that HR status may not be consistent across advanced disseminated HGSC, indicating that a single tumor biopsy may not accurately depict spatial HGSC tumor biology. HR score variations in different methods (GIS or mutational signatures) indicates that these tests are not analogous for defining HRD and thus PARPi suitability. Citation Format: Elizabeth L. Christie, Marc Lorentzen, Maiqi Liu, Nikki Burdett, Edward Curry, Chun Hei Kwok, Ahwan Pandey, Katherine Nixon, Jennifer Ploski, James J. Clark, Iain McNeish, David D.L. Bowtell, Jonathan Krell, Christina Fotopoulou, Paula Cunnea. Effect of spatial and temporal inter-tumoral heterogeneity on homologous recombination deficiency scores in high grade serous ovarian cancer [abstract]. In: Proceedings of the AACR Special Conference on Ovarian Cancer; 2023 Oct 5-7; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(5 Suppl_2):Abstract nr B114.

Leveraging Off-Target Reads in Panel Sequencing for Homologous Recombination Repair Deficiency Screening in Tumor

Targeted tumor only sequencing has become a standard practice in cancer diagnostics. This study aims to develop an approach for robust copy number variant (CNV) calling in tumor samples using only off-target region (OTR) reads. We also established a clinical use case for homologous recombination deficiency (HRD) score estimation (HRDest) using the sum of telomeric allelic imbalance (TAI) and large-scale state transitions (LST) scores without the need for loss of heterozygosity (LOH) information. We demonstrated a strong correlation between HRD score and the sum of TAI + LST in the TCGA cohort (rho = 0.99, p < 2.2x 10-16) and in a clinical in-house cohort of 34 tumors (rho = 0.9, p = 5.1 x 10-13) comparing WES and targeted sequencing data. We compared HRDest scores from 1086 clinical cases to the TCGA data set and showed that there were no significant differences in HRD score distribution within the analyzed tumor types. As a control, commercially available HRD standards were also sequenced and the HRDest scores obtained from the OTR reads were well within the HRD reference range provided by the manufacturer. In conclusion, we showed that OTR reads of tumor-only panel sequencing can be used to determine genome-wide CNV profiles and to approximate HRD scores.

Association between Homologous Recombination Repair Defect Status and Long-Term Prognosis of Early HER2-Low Breast Cancer: A Retrospective Cohort Study.

As a newly identified subtype of HER2-negative tumors associated with a less favorable prognosis, it remains crucial to evaluate potential prognostic and predictive factors, particularly non-invasive biomarkers, for individuals with human epidermal growth factor 2 (HER2) low early-stage breast cancer (EBC). Multiple investigations have highlighted that HER2-negative patients with EBC exhibiting high homologous recombination deficiency (HRD) scores display lower rates of pathological complete response (PCR) to neoadjuvant chemotherapy (NAC). Nevertheless, no study to date has explored the correlation between HRD and the long-term prognosis in HER2-low patients with EBC. This retrospective observational study focuses on primary EBC sourced from The Cancer Genome Atlas dataset (TCGA). It reveals the gene mutation landscape in EBC with low HER2 expression and elucidates the tumor immune landscape across different HRD states. Utilizing bioinformatics analysis and Cox proportional models, along with the Kaplan-Meier method, the study assesses the correlation between HRD status and disease-specific survival (DSS), disease-free interval (DFI), and progression-free interval (PFI). Subgroup analyses were conducted to identify potential variations in the association between HRD and prognosis. In the patients with HER2-low breast cancer, patients with homologous recombination related genes (HRRGs) defects had an HRD score about twice that of those without related genes mutations, and were at higher risk of acquiring ARID1A, ATM, and BRCA2 mutations. We also found that most immune cell abundances were significantly higher in EBC tumors with high HRD than in EBC tumors with low HRD or HRD-medium, particularly plasma B-cell abundance, CD8 T-cell abundance, and M1 macrophages. In addition, these tumors with HRD-high also appear to have significantly higher tumor immune scores and lower interstitial scores. Then, we analyzed the relationship between different HRD status and prognosis. There was statistical significance (P = .036 and P = .046, respectively) in DSS and PFI between the HRD-low and HRD-high groups, and patients with HRD-high EBC showed relatively poor survival outcomes. A medium HRD score (hazard ratio, HR = 2.15, 95% CI: 1.04-4.41, P = .038) was a significant risk factor for PFI. Hormone receptor positivity is an important factor in obtaining medium-high HRD score and poor prognosis. Higher HRD scores were associated with poorer PFI outcomes, particularly in people with HR+/HER2-low. Varied HRD states exhibited distinctions in HRRGs and the tumor immune landscape. These insights have the potential to assist clinicians in promptly identifying high-risk groups and tailoring personalized treatments for patients with HER2-low EBC, aiming to enhance long-term outcomes.

Regression-based Deep-Learning predicts molecular biomarkers from pathology slides

Deep Learning (DL) can predict biomarkers from cancer histopathology. Several clinically approved applications use this technology. Most approaches, however, predict categorical labels, whereas biomarkers are often continuous measurements. We hypothesize that regression-based DL outperforms classification-based DL. Therefore, we develop and evaluate a self-supervised attention-based weakly supervised regression method that predicts continuous biomarkers directly from 11,671 images of patients across nine cancer types. We test our method for multiple clinically and biologically relevant biomarkers: homologous recombination deficiency score, a clinically used pan-cancer biomarker, as well as markers of key biological processes in the tumor microenvironment. Using regression significantly enhances the accuracy of biomarker prediction, while also improving the predictions’ correspondence to regions of known clinical relevance over classification. In a large cohort of colorectal cancer patients, regression-based prediction scores provide a higher prognostic value than classification-based scores. Our open-source regression approach offers a promising alternative for continuous biomarker analysis in computational pathology.

Homologous Recombination Deficiency Unrelated to Platinum and PARP Inhibitor Response in Cell Line Libraries

While large publicly available cancer cell line databases are invaluable for preclinical drug discovery and biomarker development, the association between homologous recombination deficiency (HRD) and drug sensitivity in these resources remains unclear. In this study, we comprehensively analyzed molecular profiles and drug screening data from the Cancer Cell Line Encyclopedia. Unexpectedly, gene alterations in BRCA1/2 or homologous recombination-related genes, HRD scores, or mutational signature 3 were not positively correlated with sensitivity to platinum agents or PARP inhibitors. Rather, higher HRD scores and mutational signature 3 were significantly associated with resistance to these agents in multiple assays. These findings were consistent when analyzing exclusively breast and ovarian cancer cell lines and when using data from the COSMIC Cell Line Project. Collectively, the existing data from established cancer cell lines do not reflect the expected association between HRD status and drug response to platinum agents and PARP inhibitors in clinical tumors. This discrepancy may extend to other tumor characteristics, highlighting the importance of recognizing potential limitations in cell line data for researchers.

Genomic mutations and homologous recombination deficiency (HRD) score in Japanese patients with metastatic hormone-sensitive prostate cancer (mHSPC): The impact of intraductal carcinoma of the prostate (IDC-P).

Genomic mutations and homologous recombination deficiency (HRD) score in Japanese patients with metastatic hormone-sensitive prostate cancer (mHSPC): The impact of intraductal carcinoma of the prostate (IDC-P).

Peri-hilar cholangiocarcinoma (phCCA): A comparative comprehensive genomic profiling study.

542 Background: CCA is a heterogenous group of malignancies that arise from epithelium of the biliary tree. The comparison of the genomic landscapes that can influence therapy selection of phCCA from other types of intra- and extra-hepatic CCA (ih/ehCCA) are not well-known. We queried differences in comprehensive genomic profiles (CGP) of CCAs and comparing genomic alterations (GA) in phCCA and other types of CCA. Methods: DNA extracted from 110 cases of phCCA, 743 cases of common bile duct CCA (cbdCCA) and 9178 cases of intrihCCA underwent hybrid capture-based CGP to evaluate GA, MSI status, TMB levels, genomic ancestry, genomic signature, HRD score and germline status. phCCA cases were confirmed at surgery and pathology as originating from either the left, right or common hepatic ducts. cbdCCA cases were confirmed either by CBD biopsies or Whipple procedures. PD-L1 expression was determined by IHC using the Dako 22C3 assay using the tumor proportional score (TPS) system. Results: Patient characteristics were similar (Table). The GA/tumor and frequencies of European ancestry were similar. Both phCCA and cbdCCA featured significantly greater frequency of ERBB2 GA than ihCCA (P=.002) and lower frequencies of FGFR2 GA (P=.001) and IDH1 GA (P=.006). ihCCA featured significantly higher frequencies of FGFR2 (11.7%) and IDH1 (13.9%; P=.006) GA than either phCCA or cbdCCA. cbdCCA had a significantly higher frequency of KRAS GA than phCCA (P=.015) and ihCCA (P&lt;.001). The KRAS G12C frequency reached 5.9% of KRAS mutated ihCCA. GA in BRAF, PIK3CA, CDKN2A and MTAP were not significantly different. TP53 GA were higher in cbdCCA than phCCA (P=0.011) and combined phCCA and cbdCCA (ehCCA) were higher than ihCCA (P=.071). MSI status, TMB levels and PD-L1 expression was similarly low in all 3 groups. Positive homologous recombination deficiency (HRD) scores based on %gLOH and high frequencies of MMR genomic signatures were similar in all 3 groups. Conclusions: phCCA appears to be a distinctive form of ehCCA featuring significantly higher frequencies of potentially targetable ERBB2 and lower frequencies of targetable FGFR2 and IDH1 mutations, when compared to ihCCA. Predictive IO markers were similarly low in both ihCCA and ehCCA.[Table: see text]

Guanosine diphosphate-mannose suppresses homologous recombination repair and potentiates antitumor immunity in triple-negative breast cancer.

Triple-negative breast cancer (TNBC) is the most aggressive subtype of breast cancer with poor prognosis. TNBCs with high homologous recombination deficiency (HRD) scores benefit from DNA-damaging agents, including platinum drugs and poly(ADP-ribose) polymerase (PARP) inhibitors, whereas those with low HRD scores still lack therapeutic options. Therefore, we sought to exploit metabolic alterations to induce HRD and sensitize DNA-damaging agents in TNBCs with low HRD scores. We systematically analyzed TNBC metabolomics and identified a metabolite, guanosine diphosphate (GDP)-mannose (GDP-M), that impeded homologous recombination repair (HRR). Mechanistically, the low expression of the upstream enzyme GDP-mannose-pyrophosphorylase-A (GMPPA) led to the endogenous up-regulation of GDP-M in TNBC. The accumulation of GDP-M in tumor cells further reduced the interaction between breast cancer susceptibility gene 2 (BRCA2) and ubiquitin-specific peptidase 21 (USP21), which promoted the ubiquitin-mediated degradation of BRCA2 to inhibit HRR. Therapeutically, we illustrated that the supplementation of GDP-M sensitized DNA-damaging agents to impair tumor growth in both in vitro (cancer cell line and patient-derived organoid) and in vivo (xenograft in immunodeficient mouse) models. Moreover, the combination of GDP-M with DNA-damaging agents activated STING-dependent antitumor immunity in immunocompetent syngeneic mouse models. Therefore, GDP-M supplementation combined with PARP inhibition augmented the efficacy of anti-PD-1 antibodies. Together, these findings suggest that GDP-M is a crucial HRD-related metabolite and propose a promising therapeutic strategy for TNBCs with low HRD scores using the combination of GDP-M, PARP inhibitors, and anti-PD-1 immunotherapy.

Assessing the Phenotype of a Homologous Recombination Deficiency Using High Resolution Array-Based Comparative Genome Hybridization in Ovarian Cancer.

Ovarian cancer (OC) cells with homologous recombination deficiency (HRD) accumulate genomic scars (LST, TAI, and LOH) over a value of 42 in sum. PARP inhibitors can treat OC with HRD. The detection of HRD can be done directly by imaging these genomic scars, or indirectly by detecting mutations in the genes involved in HR. We show that HRD detection is also possible using high-resolution aCGH. A total of 30 OCs were analyzed retrospectively with high-resolution arrays as a test set and 19 OCs prospectively as a validation set. Mutation analysis was performed by HBOC TruRisk V2 panel to detect HR-relevant mutations. CNVs were clustered with respect to the involved HR genes versus the OC cases. In prospective validation, the HRD status determined by aCGH was compared with external HRD assessments. Two BRCA mutation carriers did not have HRD. OC could approximately differentiate into two groups with characteristic CNV patterns with different survival rates. Mutation frequencies have a linear regression on the HRD score. Mutations in individual HR-relevant genes do not always indicate HRD. This may depend on the mutation frequency in tumor cells. The aCGH shows the genomic scars of an HRD inexpensively and directly.

Exploration and validation of a combined Hypoxia and m6A/m5C/m1A regulated gene signature for prognosis prediction of liver cancer

BackgroundIt is widely acknowledged that hypoxia and m6A/m5C/m1A RNA modifications promote the occurrence and development of tumors by regulating the tumor microenvironment. This study aimed to establish a novel liver cancer risk signature based on hypoxia and m6A/m5C/m1A modifications.MethodsWe collected data from The Cancer Genome Atlas (TCGA-LIHC), the National Omics Data Encyclopedia (NODE-HCC), the International Cancer Genome Consortium (ICGC), and the Gene Expression Omnibus (GEO) databases for our study (GSE59729, GSE41666). Using Cox regression and least absolute shrinkage and selection operator (LASSO) method, we developed a risk signature for liver cancer based on differentially expressed genes related to hypoxia and genes regulated by m6A/m5C/m1A modifications. We stratified patients into high- and low-risk groups and assessed differences between these groups in terms of gene mutations, copy number variations, pathway enrichment, stemness scores, immune infiltration, and predictive capabilities of the model for immunotherapy and chemotherapy efficacy.ResultsOur analysis revealed a significantly correlated between hypoxia and methylation as well as m6A/m5C/m1A RNA methylation. The three-gene prognosis signature (CEP55, DPH2, SMS) combining hypoxia and m6A/m5C/m1A regulated genes exhibited strong predictive performance in TCGA-LIHC, NODE-HCC, and ICGC-LIHC-JP cohorts. The low-risk group demonstrated a significantly better overall survival compared to the high-risk group (p < 0.0001 in TCGA, p = 0.0043 in NODE, p = 0.0015 in ICGC). The area under the curve (AUC) values for survival at 1, 2, and 3 years are all greater than 0.65 in the three cohorts. Univariate and Multivariate Cox regression analyses of the three datasets indicated that the signature could serve as an independent prognostic predictor (p < 0.001 in the three cohorts). The high-risk group exhibited more genome changes and higher homologous recombination deficiency scores and stemness scores. Analysis of immune infiltration and immune activation confirmed that the signature was associated with various immune microenvironment characteristics. Finally, patients in the high-risk group experienced a more favorable response to immunotherapy, and various common chemotherapy drugs.ConclusionOur prognostic signature which integrates hypoxia and m6A/m5C/m1A-regulated genes, provides valuable insights for clinical prediction and treatment guidance for liver cancer patients.

Homologous Recombination Deficiency (HRD) Scoring, by Means of Two Different Shallow Whole-Genome Sequencing Pipelines (sWGS), in Ovarian Cancer Patients: A Comparison with Myriad MyChoice Assay

High-grade serous ovarian cancer (HGSOC) patients carrying the BRCA1/2 mutation or deficient in the homologous recombination repair system (HRD) generally benefit from treatment with PARP inhibitors. Some international recommendations suggest that BRCA1/2 genetic testing should be offered for all newly diagnosed epithelial ovarian cancer, along with HRD assessment. Academic tests (ATs) are continuously under development, in order to break down the barriers patients encounter in accessing HRD testing. Two different methods for shallow whole-genome sequencing (sWGS) were compared to the reference assay, Myriad. All these three assays were performed on 20 retrospective HGSOC samples. Moreover, HRD results were correlated with the progression-free survival rate (PFS). Both sWGS chemistries showed good correlation with each other and a complete agreement, even when compared to the Myriad score. Our academic HRD assay categorized patients as HRD-Deficient, HRM-Mild and HRN-Negative. These three groups were matched with PFS, providing interesting findings in terms of HRD scoring and months of survival. Both our sWGS assays and the Myriad test correlated with the patient’s response to treatments. Finally, our AT confirms its capability of determining HRD status, with the advantage of being faster, cheaper, and easier to carry out. Our results showed a prognostic value for the HRD score.