NEW YORK – Cancer patients with BRCA1 or BRCA2 genetic mutations appear to have strikingly different response patterns to immune checkpoint inhibitors, according to a Nature Cancer study published on Monday.
Researchers, including Timothy Chan of the Cleveland Clinic Cancer Center, reported in the published paper that mutations in BRCA1 and BRCA2 affect tumor immune microenvironments in unique ways, and crucially, that patients with BRCA2 mutations appear to respond better to immune checkpoint inhibition relative to patients with BRCA1 mutations.
Mutations in the homologous recombination repair (HRR) genes, BRCA1 and BRCA2, are well known to increase carriers' risks for hereditary breast, ovarian, pancreatic, prostate, and other cancers. Cancer patients with these mutations also respond particularly well to PARP inhibitors.
The latest findings related to the impact of these genetic mutations on immunotherapy response, however, were unexpected, according to Chan. "BRCA2 mutant patients behave very differently than BRCA1," he said. The separate mutations "differentially affect the tumor immune landscape and the gene expression levels of the T cells, and they have dramatic effects on whether patients respond or not."
Distinct differences
The Nature Cancer paper is the culmination of five years of research, part of which involved one of the largest single-cell mouse studies ever conducted, according to Chan.
Initially, Chan and colleagues retrospectively analyzed clinical data from over a thousand patients with different cancer types who had been treated with immune checkpoint inhibitors and were tested using Memorial Sloan Kettering's MSK-IMPACT multi-gene panel. They had been looking to characterize pathways in which mutated genes were associated with responses to the treatment.
It was no surprise to the researchers that truncating mutations in genes in the HRR pathway were strongly associated with improved overall survival in patients treated with immune checkpoint inhibitors, such as pembrolizumab (Merck's Keytruda) and nivolumab (Bristol Myers Squibb's Opdivo), among others. But when they looked deeper and explored which HRR genes were mutated in these patients' tumors, Chan and his colleagues were surprised.
"Our initial thought was that [tumors with HRR genetic mutations, including BRCA1 and BRCA2] will all probably behave the same," Chan said. "But they didn't behave the same."
As a next step, to ensure that what they were seeing was causal, Chan and colleagues cultivated mouse models with BRCA2- and BRCA1-mutated breast and colorectal cancer tumors. Using the mouse models, they went on to perform a number of analyses, including whole-genome sequencing, flow cytometric analysis of tumor-infiltrating lymphocytes, bulk RNA sequencing, gene set enrichment analysis, and analysis of T-cell infiltration after treatment with immune checkpoint blockade.
Altogether, their experiments revealed distinct differences between BRCA1- and BRCA2-mutated tumors. For one, pathway analysis using the Gene Ontology knowledgebase, a resource of information on the function of genes, showed that BRCA2 mutated tumors were significantly enriched for mechanisms involving T-cell cytotoxicity and antiviral responses relative to BRCA1-mutated tumors. Interferon-gamma signaling, T-cell activation, and antigen presentation were all upregulated in BRCA2-mutated tumors.
BRCA1-mutated tumors, on the other hand, were enriched for genes related to immunosuppression and immunoregulatory activity. When the mouse tumors were treated with anti-PD-1 or combined anti-PD-L1/CTLA4 checkpoint blockade, moreover, the researchers saw growth delay and increased T-cell infiltration in the BRCA2-deficient tumors but not in the BRCA1-deficient tumors.
The researchers also noted differences in immune cell types between BRCA1- and BRCA2-mutated tumors. BRCA2-mutated tumors had a higher prevalence of cytotoxic and exhausted CD4-positive and CD8-positive immune cells. BRCA1-mutated tumors were enriched for immunosuppressive tumor-associated macrophages relative to BRCA2-mutated tumors, and researchers also noted differences in terms of myeloid populations.
Moving beyond mouse models, the researchers retrospectively analyzed bulk RNA sequencing data from a cohort of patients with BRCA1- or BRCA2-mutated breast cancer in The Cancer Genome Atlas. Using the gene sets they came up with in their mouse models, the researchers performed gene set enrichment analysis and found that BRCA1-mutated breast cancer tumors were enriched for immunoregulatory genes relative to BRCA2-mutated tumors, mirroring the results seen in mice.
Chan and colleagues explained that since many patients who had their tumors sequenced using the MSK-IMPACT panel did not have germline testing, they got permission from the institutional review board to call germline variants and conduct an anonymized overall survival analysis for immunotherapy-treated patients. The analysis revealed that BRCA2 mutations were associated with improved overall survival on checkpoint inhibitors compared to BRCA1 mutations. These differential responses remained after the researchers controlled for other clinical covariates.
Together, the authors said that these findings "suggest that differences in the mutational landscape between BRCA1 and BRCA2 may partly contribute to differences in immunogenicity." They further observed that "relative to BRCA2-deficient tumors, BRCA1-deficient tumors were associated with increased levels of copy-number alterations across the genome, which have been linked to worse response to immunotherapy."
Future studies, drug combinations
Chan and his colleagues believe they may be the first group to publish on the effect of BRCA1 and BRCA2 loss on the tumor microenvironment, and to provide evidence that mutations in these two genes impact tumor immunity and patients' ability to respond to immune checkpoint inhibitors differently.
However, the researchers also acknowledged that their findings will require further validation in prospective studies.
According to Chan, the findings in the Nature Cancer paper could have particularly important implications for designing breast cancer clinical trials going forward, because the benefit of immune checkpoint inhibitor treatment in breast cancer has not been well characterized. "There's been very mixed evidence about immunotherapy helping [in] breast cancer, where BRCA mutations commonly occur," he said. "Some of the big approvals [in this setting] have been based on trials where the clinical benefit has been very small, because the patients are unselected."
For example, in 2019, the US Food and Drug Administration granted accelerated approval to atezolizumab (Genentech's Tecentriq) in combination with nab-paclitaxel for triple-negative breast cancer patients with PD-L1 positive tumors on the basis of a progression-free survival benefit. However, in a similarly designed study presented this fall, in which atezolizumab was combined with paclitaxel as opposed to nab-paclitaxel, the regimen failed to benefit the same patient population. Neither study used biomarkers other than PD-L1 to further home in on a best responder population.
Chan said that the predictive potential of BRCA1 and BRCA2 mutations described in the Nature Cancer paper could help more accurately carve out best responder populations among breast cancer patients who are considering treatment with immune checkpoint inhibitors, such as atezolizumab. Because BRCA1 mutations occur most often in TNBC, while BRCA2 mutations are more likely to occur in hormone receptor-positive breast cancer, there's a chance that selecting only TNBC patients for the atezolizumab trials wasn't the right move.
"We think the patients were selected incorrectly" for these atezolizumab studies, Chan said. "An immediate hypothesis that needs to be tested in a prospective trial is that the indication should be expanded to BRCA2 mutant tumors, which happen to be hormone receptor-positive."
The latest findings could also spur researchers to pursue immunotherapy combinations with PARP inhibitors, particularly in patients with BRCA2-mutated tumors. "PARP inhibitors enhance genetic instability of many BRCA1/2-mutant tumors leading to an anti-tumor effect," Chan said. "You could imagine based on our data that a PARP inhibitor plus immunotherapy in [BRCA2-mutant tumors] would be synergistic and much better together as a combination therapy."
Going forward, Chan said that he and fellow researchers will be working to design this type of trial. It will also be necessary to perform retrospective biomarker analyses within immune checkpoint inhibitor trials, he said, to specifically analyze patients' outcomes based on their HRR mutation status.
This research also speaks to the importance of evaluating the role of multiple biomarkers on patients' ability to respond to immune checkpoint inhibitors. For example, PD-L1 expression is often used as a predictive biomarker for immunotherapy and is indicated in a number of regulatory approvals for checkpoint inhibitors. For example, the FDA last week approved pembrolizumab (Merck's Keytruda) plus chemotherapy for triple-negative breast cancer patients who had a PD-L1 combined positive score of at least 10. The agency approved the indication based on data from a study in which only patients in this subgroup experienced improved progression-free survival on the pembrolizumab-chemo regimen compared to chemo.
However, Chan and his team were not able to consider the impact of PD-L1 expression along with BRCA1/2 mutation status into their present analyses due to the inconsistent ways in which PD-L1 expression was recorded in the datasets they used.
"That will be important to follow up on," Chan said. "It would be valuable for us to do a simple study and look at the concurrence between the PD-L1 staining levels."
Ultimately, the Nature Cancer paper could inspire a marked shift in the way researchers and drug developers stratify patients with BRCA1 and BRCA2 mutations in immunotherapy trials.
"These are deep mechanistic considerations that people have not really paid attention to," Chan said. "But these data should open the door to that. It's a pretty unequivocal finding."