NEW YORK – Treating HER2-negative metastatic breast cancer patients with targeted therapies matched to genomic alterations driving their tumors can benefit them, but only when treatment decisions are guided by a validated system ranking the evidence underlying those biomarker-treatment associations.
At the 2021 San Antonio Breast Cancer Symposium on Tuesday, Fabrice André, head of research at the Institute Gustave Roussy in France, presented data showing that patients whose metastatic breast cancers harbor genomic alterations with top-tier ranking (tiers I and II) according to the ESMO Scale for Clinical Actionability of Molecular Targets (ESCAT) derive a progression-free survival (PFS) benefit with matched targeted treatment versus standard-of-care chemo. Meanwhile, the study showed that patients with alterations in lower tiers, denoting less robust evidence or more limited clinical actionability, are no better off with matched targeted treatment than standard chemo.
"What this trial is saying is that it's absolutely mandatory, when we do genomics-selected targeted therapy, to rank the genomic alteration," André said in a discussion of the data held Monday for members of the media. "Only the [tiers] I or II should be taken into account."
The findings come from a pooled analysis of the Phase II randomized SAFIR02-BREAST and SAFIR-PI3K clinical trials. Both trials enrolled patients with HER2-negative metastatic breast cancer who were eligible for first- or second-line chemotherapy treatment. Before beginning treatment, patients underwent pretreatment biopsies of the metastatic tumor tissue, which researchers from the National Federation of French Cancer Centers, including Gustave Roussy's André, used for next-generation sequencing and copy number analysis of germline testing.
After conducting the NGS genomic analyses, BRCA1/2 germline testing, and comparative genomic hybridization on more than 1,450 patients' samples, the investigators randomly assigned 238 of those with stable disease after six to eight initial chemotherapy cycles to receive either targeted therapies matched to their genomic alterations or maintenance chemo.
Of the randomized patients, 157 received biomarker-matched targeted therapies and 81 received maintenance chemo. Patients enrolled in the SAFIR-PI3K trial who had PI3K mutations were specifically assigned to treatment with alpelisib (Novartis' Piqray) plus fulvestrant or to maintenance chemotherapy. Patients in SAFIR02-BREAST received either maintenance chemo or one of nine different therapies based on their genomic targets.
Before assessing the PFS benefit of matched targeted treatment versus maintenance chemotherapy — the study's primary aim — André and colleagues classified each patient's genomic alteration using the ESCAT ranking system.
ESCAT was established to help oncologists interpret genomic sequencing results and is similar in aim to Memorial Sloan Kettering's Oncology Knowledge Base, or OncoKB, that is widely accepted among US oncologists and regulators for guiding therapy decisions based on molecular profiling. These resources recognize that detected alterations vary in terms of clinical utility and targeted treatment benefit and rank the underlying evidence to help oncologists make treatment decisions most likely to benefit their patients.
Some alterations, for instance, may match to drugs that regulators have approved for that particular biomarker-defined indication or that may be supported by data from large, randomized trials. The therapeutic impact of other alterations, meanwhile, could be less well studied and understood. Still others may only have preclinical evidence suggesting an impact on targeted treatment, or no evidence at all.
Researchers have shown previously that the ESCAT system can be implemented in routine clinical practice, and some hope the resource will become widely used in precision oncology and even influence policy and reimbursement decisions.
According to the scale, tier I alterations are those with matched drugs that clearly improve patient outcomes in clinical trials. Tier II alterations, meanwhile, match to drugs that have shown antitumor activity in clinical trials, but the magnitude of benefit is still unclear. Tier III alterations have drug matches that are suspected to improve outcomes based on clinical trial data in other tumor types or with other similar molecular alterations. Tier IV alterations have only preclinical evidence supporting a matched drug; tier V alterations have matches associated with objective response, but no clinically meaningful benefit; and tier X alterations have no evidence at all supporting matched treatment.
For their analyses, André and colleagues sorted outcomes according to ESCAT tiers. First, they considered only the 115 patients who had ESCAT I or ESCAT II alterations and found a clear improvement in outcomes with targeted therapy. Specifically, among these patients, the median PFS with matched targeted therapy was 9.1 months versus 2.8 months with maintenance chemotherapy.
However, when the investigators repeated the analysis in the full group of 238 patients in the intention-to-treat population, which included those with alterations in tiers III through X, there was no statistically significant benefit observed with targeted therapy versus chemo. Taking a closer look at these outcomes in a subgroup analysis, the researchers found that ESCAT level classification was highly predictive of whether biomarker matched treatment would benefit patients.
"For patients who do not present the level I or II genomic alterations, there is absolutely no benefit of using the genomic report," Andre said, acknowledging that many oncologists, himself included, tend to overinterpret alterations identified in patients and may prescribe a targeted therapy based on a drug-biomarker association that isn't fully validated.
"This study clearly showed that we as oncologists should not use genomic information when the target has no previously validated trials," Andre added.
Beyond investigating the primary endpoint and determining that matched treatment is beneficial specifically for patients with ESCAT tier I or II mutations, André and colleagues used the SAFIR02-BREAST and SAFIR-PI3K trial data to conduct several exploratory analyses. Using their SNP array analyses, for example, they found that patients who were resistant to prior treatment with CDK4/6 inhibitors presented amplified CDK4 in their metastatic sites far more frequently than in their primary tumors.
Finally, among patients harboring germline BRCA1/2 mutations, those who had loss of heterozygosity or had homologous recombination deficiency experienced improved PFS with the PARP inhibitor olaparib (AstraZeneca/Merck's Lynparza).
During a discussion at SABCS following his presentation of the data on Tuesday, André took a step back and underscored that his research clearly showed that clinical decision support systems ought to be built into every tumor profiling report.
He clarified that the point of the study wasn't to determine whether oncologists should use genomics to guide breast cancer treatment — in fact, he said that they absolutely should — but rather to shed light on the value of using a scoring system to determine when, exactly, to take action based on genomic findings.
Furthermore, the aim of the study wasn't necessarily to promote the specific ESCAT system used in this case for ranking therapeutic actionability, but to stress the need for similar decision support resources to help oncologists practice precision oncology as new drugs and biomarkers are identified.
"It's a very simple message," he continued. "We need to integrate the level of evidence for genomic alterations in each [test] report that we deliver."