Patterns of Genomic Instability in Ovarian Cancer Patients

Obstetrics and Gynecology Oncology Biochemistry Medicine Genetics Life Sciences
ovarian cancer genomic instability PARP inhibitors homologous recombination deficiency clinical trials BRCA mutations precision medicine

Background and Research Question

Ovarian cancer, as a fatal gynecological malignancy, has become the focus of recent studies, particularly on expanding the feasibility of targeted therapies. The introduction of poly(ADP-ribose) polymerase (PARP) inhibitors has marked a groundbreaking treatment option for ovarian cancer, especially for patients with homologous recombination deficiency (HRD). However, aside from the classic BRCA1 and BRCA2 mutations (BRCAM), the additional genes or variations involved in HRD remain unclear. Furthermore, understanding the patterns of non-BRCA genomic instability and their implications on HRD testing and precision treatment remains insufficiently explored.

To address these gaps, Alan Barnicle and colleagues conducted a pioneering analysis. This investigation used data from six clinical trials (SOLO1, PAOLA-1, STUDY 19, SOLO2, OPINION, and LIGHT) involving nearly 2,000 ovarian cancer patients. The study aimed to systematically explore the patterns of genomic instability driven by BRCA and non-BRCA genetic factors in HRD, providing critical insights into optimizing PARP inhibitors and other targeted therapies.

Source and Methods

Published in 2024 in Genome Medicine, this study was conducted by researchers from AstraZeneca, the University of Lyon, Gustave Roussy, and other institutions. They analyzed 2,147 tumor samples from high-grade ovarian cancer patients across six Phase II/III multicenter clinical trials including SOLO1 (NCT01844986), PAOLA-1 (NCT02477644), and others. Using the Myriad Genetics Myriad myChoice® CDx test, this study assessed genomic instability scores (GIS) to evaluate HRD across BRCA and non-BRCA genetic mutations.

Study Design and Workflow

Sample Collection and Genetic Testing

The study focused on tumor samples with mutations in BRCA1/2, non-BRCA homologous recombination repair (HRR) genes, and non-HRR genes (e.g., PIK3CA, CCNE1). Key workflow steps included:

  1. Patient and Sample Recruitment:

    • Tumor tissue samples were collected from 2,147 high-grade ovarian cancer patients, predominantly from diagnostic or post-chemotherapy biopsies.
    • Both formalin-fixed paraffin-embedded (FFPE) tissues and blood DNA samples were assessed.

  2. Genomic Testing and GIS Calculation:

    • HRD was evaluated using the Myriad myChoice® method, incorporating genomic loss of heterozygosity (LOH), telomeric allelic imbalance (TAI), and large-scale state transitions (LST) to calculate scores and evaluate tumor HRD positivity.

  3. Statistical Analysis:

    • BRCAM (n = 1,021), non-BRCA HRRM (n = 121), and non-HRRM (n = 1,005) cohorts were compared. Statistical methods analyzed GIS patterns, HRD positivity (GIS ≥ 42), and biallelic loss (inactivation of both alleles of a gene).

Innovative Contributions

This study is the first to combine GIS with a broad panel of over 100 genes, exploring potential drivers of HRD and strategies for PARP inhibitor application beyond BRCA-related factors.

Research Results

GIS Characteristics and HRD Patterns in BRCAM

BRCAM, contributing to 47.6% of the cases, exhibited high GIS and robust HRD positivity (93.9% achieved HRD-positive status). Specific findings include:

  1. BRCA1 vs. BRCA2 Distinctions:

    • Tumors with BRCA1 mutations had significantly higher median GIS (64, IQR 55–71) compared to those with BRCA2 mutations (median GIS: 59; p < 0.001).
    • Over 90% of BRCAM samples displayed biallelic loss, regardless of whether the mutations were germline or somatic.

  2. LOH and GIS Associations:

    • Biallelic BRCA loss was linked to significantly higher GIS (median 62) compared to monoallelic loss (median 39; p < 0.001).

GIS Analysis of Non-BRCA HRRM and Non-HRRM Tumors

Non-BRCA HRRM Patterns

Non-BRCA HRRM tumors (e.g., RAD51C, RAD51D, BRIP1, and PALB2) accounted for 121 cases and were characterized by:

  1. Highly Variable GIS: Median GIS was 42 (IQR 29–58), higher than non-HRRM but lower than BRCAM.
  2. Gene-Specific Drivers of HRD:
    • Mutations in RAD51C, RAD51D, and BRIP1 resulted in notably higher GIS (RAD51D median GIS: 62).
    • These mutations also exhibited high rates of biallelic gene inactivation.

Non-HRRM Population

Non-HRRM tumors (n = 1,005, of which 868 had evaluable GIS) primarily exhibited low GIS (median: 32; IQR: 20–55), though select genes like NF1 and RB1 correlated with high GIS.

  1. NF1 and RB1 vs. CCNE1 and PIK3CA:
    • NF1 and RB1: These alterations were associated with elevated GIS (49 and 55, respectively) and substantial biallelic inactivation.
    • CCNE1 and PIK3CA: These alterations correlated with significantly lower GIS levels (24 and 32, respectively), suggesting distinct resistance mechanisms to DNA damage.

Clinical Implications and Heterogeneity Across Studies

  • Inconsistencies Between Data

    • For example, while PAOLA-1 did not show strong HRD enrichment among non-BRCA HRRM tumors, OPINION demonstrated significantly higher GIS in patients with RAD51C and RAD51D mutations.

  • Significance of Patient Heterogeneity:

    • The findings underscore the complexity of HRD-positive and HRD-negative populations, emphasizing the importance of multi-omics approaches for stratification.

Conclusions and Scientific Significance

This study provides a comprehensive genomic perspective on HRD and its drivers in ovarian cancer, with several key takeaways:

  1. HRD-Positive Patients as Targets for PARP Inhibitors:

    • Aside from BRCAM, patients with non-BRCA HRRM (e.g., RAD51C, RAD51D) could also benefit from PARP inhibitors or alternative DNA repair-targeted therapies.

  2. Insights Into Non-HRRM Tumors:

    • Alterations in NF1 or RB1 were associated with high genomic instability, highlighting potential therapeutic targets. Conversely, CCNE1 and PIK3CA mutations exhibited resistance phenotypes, potentially requiring tailored treatment strategies.

  3. Broader Implications for Precision Medicine:

    • Though ovarian cancer is a model for HRD, findings such as high GIS in RB1-altered tumors could provide rationale for expanding PARP inhibitor indications to other cancers.

This research represents a significant advance in understanding the drivers of genomic instability in ovarian cancer and offers a foundation for developing next-generation targeted therapies.