TL;DR
This study reveals that high expression of the DNA repair protein RAD51 in ovarian cancers is linked to resistance to platinum-based chemotherapy and an "immune-excluded" tumor microenvironment. This finding identifies a specific patient subgroup that may not benefit from standard treatments and could guide the development of new therapeutic strategies.
What Problem Does This Paper Address?
Ovarian cancer, particularly high-grade serous ovarian carcinoma (HGSOC), is often treated with platinum-based chemotherapy. However, a significant challenge is primary platinum resistance, where tumors don't respond to initial treatment. Homologous recombination deficiency (HRD) is known to sensitize tumors to platinum and PARP inhibitors. While HRD is a key factor, understanding the role of hyperactive homologous recombination (HR) – the opposite of HRD – is less clear. RAD51 is a crucial mediator of HR, and its overexpression has been observed in cancers, but its clinical significance, especially in relation to platinum resistance and the tumor's immune environment, has been difficult to quantify and remains unclear due to limitations in measurement techniques. This paper aims to address this by quantitatively assessing RAD51 expression and its impact on platinum response and immune infiltration in ovarian cancer.
Methods
The researchers employed advanced techniques to quantify RAD51 protein expression. They used fluorescent immunohistochemistry (fIHC) combined with multispectral imaging and automated analysis (Vectra 2 System) to measure RAD51 protein levels in formalin-fixed, paraffin-embedded (FFPE) tissue samples. This quantitative approach, termed the "RAD51 Vectra score" (RAD51vs), overcomes the limitations of traditional, less scalable methods like foci counting. They also performed gene expression analysis (RNA sequencing) and in-situ multiplexed imaging to assess immune cell infiltration and the expression of immune-regulatory genes. Statistical analyses included Kaplan-Meier survival analysis, Cox proportional hazards models, and one-way ANOVA to correlate RAD51 expression with clinical outcomes and immune profiles.
Datasets
The study utilized several independent cohorts of ovarian cancer patients:
- British Columbia Cancer (BCC) cohort: (n=284) used for initial fIHC and survival analysis.
- SCOTROC4 clinical trial cohort: (n=268) a phase III trial of carboplatin monotherapy, used for independent validation of RAD51 expression and outcomes.
- TCGA (The Cancer Genome Atlas) cohort: (n=566) used for mRNA expression and immune infiltration analyses.
- GSE9891 (Australian Ovarian Cancer Study - AOCS): (n=267) used for mRNA expression and immune infiltration analyses.
- GSE26712 (Massachusetts General Hospital - MGH): (n=185) used for mRNA expression and immune infiltration analyses.
- GSE3149 (Duke University Hospital - Duke): (n=146) used for mRNA expression and immune infiltration analyses.
The study also used in-vitro models of RAD51 overexpression in HGSOC cell lines.
Key Insights and Ideas
- RAD51 as a marker of platinum resistance: Contrary to the expectation that HRD (which sensitizes to platinum) is the sole driver, this study highlights that hyperactive HR, indicated by high RAD51, is associated with resistance to platinum chemotherapy.
- Immune exclusion as a mechanism: High RAD51 expression doesn't just affect tumor cells directly; it actively shapes the tumor microenvironment by leading to an "immune-excluded" phenotype, where cytotoxic T-cells are kept out of the tumor core. This likely contributes to poor treatment outcomes.
- HRD negativity is key: The prognostic significance of high RAD51 is most pronounced in tumors classified as "HRD negative" by genomic scar assays. This suggests a synergistic approach using both HRD status and RAD51 expression could better stratify patients.
- In vitro vs. in vivo disconnect: Interestingly, while high RAD51 expression in cell lines didn't directly confer platinum resistance, it significantly altered immune-regulatory gene expression, pointing to the crucial role of the in vivo tumor microenvironment.
Key Results
- Survival Impact: RAD51-High tumors were associated with significantly shorter progression-free survival (PFS) and overall survival (OS) in both the BCC and SCOTROC4 cohorts. For example, in the BCC cohort, median PFS was 1.1 years for RAD51-High vs. 1.7 years for RAD51-Low (p<0.001).
- Platinum Resistance in SCOTROC4: In the SCOTROC4 trial (carboplatin monotherapy), RAD51-High cases showed poorer PFS and OS (median PFS 0.99 vs. 1.2 years, p=0.013). They were also more likely to relapse within 12 months (57% vs. 38%).
- HRD Negative Stratification: In HRD-negative patients from the SCOTROC4 cohort, RAD51-High cases had significantly worse PFS (median 0.92 vs. 2.3 years, p=0.012) and OS (median 1.8 vs. not reached, p=0.001).
- Immune Exclusion: Multiplexed imaging confirmed that RAD51-High tumors exhibited an exclusion of CD3+/CD8+ cytotoxic T-cells, particularly in BRCA wild-type tumors.
- Gene Expression Changes: RAD51-High tumors consistently showed altered expression of immune-related genes across multiple cohorts, with enrichment in T-cell and B-cell mediated immunity pathways.
Limitations and Future Work
The authors acknowledge that this is a retrospective analysis of a subset of patients, and findings need to be further validated in prospective cohorts from clinical trials. They also note that the predictive value of RAD51 for platinum resistance was restricted to HRD-negative tumors, and the reasons for this are unclear, possibly due to the small sample size of HRD-positive cases or that RAD51 upregulation alone doesn't overcome marked HR deficiency in vivo.
Future work should focus on:
- Exploring the exact relationship between high RAD51 and negative immune regulators in cancer.
- Investigating whether high RAD51 protein levels define a subset of HR-proficient EOC patients who will not benefit from PARPi maintenance therapy.
- Developing better therapeutic approaches for cancers with this RAD51-high, immune-excluded phenotype, potentially including anti-angiogenic agents like bevacizumab.
- Comparing RAD51 protein and mRNA expression in large cohorts to identify the most clinically applicable predictive biomarker.
- Evaluating anti-CD47 antibodies as a potential therapeutic strategy, given CD47's consistent high expression in RAD51-High tumors.
Summary updated