Addressing the Clinical Need for a Next-Generation HRD Biomarker: Introducing Foundation Medicine’s New HRD Signature Biomarker

In recent years, there has been robust development of targeted cancer treatments called PARP inhibitors, which work by preventing cancer cells from repairing and growing. While certain genomic alterations, such as in BRCA1/2, are known to signal response to this treatment type, research indicates that the presence of these genomic alterations does not always equal a response to PARP inhibition therapy.

As a result, it continues to be challenging to identify the right patients who may benefit from PARP inhibitors. Foundation Medicine recognized the need for a tool to help healthcare providers and researchers better identify these patients, so we set out to identify a biomarker to address the gap.

Delivering innovation with impact

Homologous recombination deficiency (HRD) biomarkers are critical for identifying patients who may benefit from the use of PARP inhibitors and other DNA-damaging cancer drugs. Traditional methods of identifying genomic scarring from HRD to predict PARP inhibitor outcomes were developed using data from patients with ovarian cancer. Foundation Medicine set out to create a pan-tumor biomarker using machine learning, a type of artificial intelligence, and our wealth of genomic data to demonstrate how the presence of HRD can deliver crucial insights and inform important treatment decisions in other disease areas outside of ovarian cancer.

We developed a proprietary homologous recombination deficiency signature (HRDsig¹) biomarker on FoundationOne®CDx², our high-quality tissue-based comprehensive genomic profiling test, for use by healthcare providers in the clinic, and in an investigational capacity in trial-specific regulatory submissions or prospective and retrospective clinical trials. HRDsig is reported as a laboratory professional service that has not been reviewed or approved by the FDA.

Our HRDsig¹ is a next-generation sequencing scar-based genomic signature that does not rely on detecting homologous recombination repair (HRR) alterations in genes like BRCA or PALB2 to detect HRD, but instead on identifying genome-wide copy number (CN) features. It was built with a diverse set of more than 100 CN features and trained using our extensive pan-tumor genomic database containing data from over 500,000 patients.

From a research perspective, the goal of this new offering is to enable Foundation Medicine’s biopharmaceutical and institutional partners to refine their clinical trial enrollment, allowing the enrollment of patients with HRD from non-genomic sources but excluding patients with HRR mutations that do not lead to HRD.

HRDsig delivers better sensitivity and specificity than genomic loss of heterozygosity (gLOH), a biomarker used to identify HRD and PARP inhibitor response in ovarian cancer.³ Unlike gLOH, HRDsig is a pan-tumor biomarker that can identify more of the right patients across tumor types while avoiding the potential for false positives.⁴

Additionally, as HRDsig does not rely on HRR mutations, it can detect non-genomic mechanisms of HRD including BRCA methylation, avoiding the potential for false negatives that could occur with other testing methods.

Benefits beyond ovarian cancer 

In ovarian cancer specifically, HRDsig detects more than twice as many patients who might benefit from PARP inhibitor maintenance therapy than BRCA1/2 alterations alone, and predicts the outcome to maintenance PARP inhibitor therapy in the BRCA mutated and wild-type settings.⁵ However, the design of HRDsig means it can deliver benefits beyond ovarian cancer, offering patients across many cancer types the potential to identify additional therapies that may have otherwise been overlooked.

In prostate cancer, HRDsig was associated with improved time to treatment discontinuation and real-world overall survival on PARP inhibitors compared to patients without HRDsig detected.⁶

In breast cancer, patients with HRDsig displayed a longer median progression-free survival and significantly reduced risk of progression on PARP inhibitors compared to patients without HRDsig detected – similar trends were observed for overall survival.⁷

Looking specifically at triple-negative breast cancer (TNBC), HRDsig-positive patients were substantially more likely to have a pathological complete response (pCR) to platinum-based chemotherapy than HRDsig-negative patients. This indicates the potential of HRDsig to identify patients for whom platinum-based regimens may provide maximum benefit, but also could help to ensure that patients who are unlikely to benefit from platinum-based treatments can avoid the significant toxicity associated with treatment.⁸

Additionally, HRDsig positivity is detected in approximately 8% of pan-tumor cases in the Foundation Medicine genomic database, including over 5% of non-small cell lung cancers and gastroesophageal cancers.⁹

What’s Next for HRDsig

Foundation Medicine’s HRDsig biomarker provides value as an independent biomarker, but also in conjunction with other DNA gene alterations found using FoundationOne CDx, to predict patient outcomes to DNA-damaging therapies, like PARP inhibitors.

Additional studies assessing the clinical validity and utility of HRDsig as a biomarker for PARP inhibitors are underway, and our hope is to be able to continue to expand the utility of this novel biomarker to all patients who could benefit from the insights it delivers.