NEW YORK (GenomeWeb) – At the annual meeting of the American Association for Cancer Research in Washington D.C. this week, representatives from Foundation Medicine presented new data on the company's adaptation of its comprehensive genomic profiling to predict tumor mutational burden and guide the use of immunotherapies.
The company also presented data offering new insights into specific molecular changes that appear to influence response to these increasingly popular drugs.
Foundation Medicine began offering tumor mutation burden (TMB) analysis using its FoundationOne panel last year, after sharing data at oncology meetings showing a high correlation between the mutation patterns in the 300-odd genes the test covers and actual genome-wide mutational load, as measured by whole-exome sequencing.
As immunotherapy has rapidly gained ground in the oncology community, research has accelerated in tandem to identify molecular markers or signatures that can best predict which patients may benefit.
"It likely hasn't escaped the attention of [attendees] that immunotherapies are transformative for some patients," Foundation Medicine CSO Phil Stephens said at the AACR meeting this week.
"There are now five approved agents hitting various points of the checkpoint cycle, but despite growing success there are many patients [who] will not respond. … Rates for the currently approved agents typically range from 10 to 30 percent across solid tumors."
Foundation Medicine is not the only entity hoping to advance methods for assessing TMB. Most recently, clinical sequencing firm Personal Genome Diagnostics announced its own plans to develop a noninvasive next-gen sequencing assay to help identify which cancer patients have the best chance at responding to treatment with immune checkpoint inhibitors.
"At a simplistic level, the explanation is that the more mutations a patient's tumor has, the more likely the tumor will express a non-self, or neo antigen, and the greater the likelihood the patient's immune system will recognize the tumor as non-self," Stephens explained at the meeting.
"We and others have demonstrated that if you sequence a megabase you can pretty much quantify mutation burden fairly accurately."
In validation data presented at the meeting, Foundation Medicine researchers reported that a FoundationOne-based TMB measure correlated strongly with whole-exome results on samples with as little as 20 percent tumor purity.
"To be completely transparent, as one gets into the lower mutation burden trying to discriminate one mutation per megabase from two or three, that's beyond the limit of the technology, but discriminating between high- intermediate- and low-TMB, sequencing one megabase does that quite effectively," Stephens said.
The gold standard in determining if this approach really works well enough, though, he added, is to look at patient outcomes or clinical response. And indeed, based on the company's early data in patients with melanoma, non-small cell lung cancer, and bladder cancer, FoundationOne-based TMB calculation correlated well with patient response to treatment, researchers reported at the meeting.
In addition to presenting a poster on the validation and clinical feasibility of FoundationOne-based TMB-analysis, the company shared new data from a study of non-small-cell lung cancer patients that pointed to a group of specific molecular alterations that also seem to play a role in patient response to immunotherapy.
Stephens said that investigators examined a series of about 1600 patients who had received FoundationOne comprehensive genomic profiling, as well as immunohistochemistry testing for PD-L1 expression.
When the company looked for biomarkers that might be enriched in patients whose tumors had high mutation-burden and low PD-L1 expression, or vice versa, they found several: The TMB-high, PD-L1 low group had significant numbers of STK11 loss of function alterations, while patients with opposite results preferentially showed enrichment for BRAF mutations and for exon-14 splice site mutations and amplifications of c-Met.
Based on literature searches and analysis of TCGA data, Stephens said the company concluded that there is encouraging evidence for how these mutations could impact immune response.
And indeed, in data from a small cohort of patients with high TMB treated with Opdivo (nivolumab), there appeared to be a statistically significant difference in response in wild-type versus STK11-mutated cancers.
This is very preliminary, Stephens said, and will have to be investigated in other, larger cohorts.
Discussion of these findings took place during a session on bridging big data and clinical practice. Since 2014, Foundation Medicine has been working with health informatics firm Flatiron Health, to integrate genomic, clinical treatment, and outcomes data
"We and others believe that marrying of clinical and genomic data will drive a new wave of discovery and improve the outcomes of patients with cancer.
At AACR, he said that the company has linked its genomic data with clinical outcomes information on over 20,000 cancer patients so far, the majority of whom were treated in the community setting.
Foundation Medicine has had luck using this database to recapitulate known findings, for example increased survival in EGFR mutant NSCLC compared to wild-type, or improved outcomes in patients treated with targeted therapy.
In that vein, researchers decided to look for evidence that would support the novel link they discovered between STK11, BRAF, and c-Met alterations and immunotherapy response in lung cancer.
According to Stephens, when the group considered multiple factors, including things like race, age, smoking, as well as these mutations, BRAF and STK11 came out at the top of the list.
There were too few patients in the database who had c-Met alterations and had also received immunotherapy for statistical significance, but there was a "clear trend in the data," for that link as well, Stephens said.