Skip to main content
Premium Trial:

Request an Annual Quote

MSKCC Cancer Sequencing Team Shares First Data on Risk Variants From Clinical Germline Test


NEW YORK (GenomeWeb) – Having reached a milestone of 10,000 cases sequenced so far, and running about 150 new cases per week, the team behind Memorial Sloan Kettering Cancer Center's MSK-IMPACT test is beginning to collect important data on the ability of the broad matched tumor and normal next-generation sequencing test to better guide patients to existing therapies and clinical trials.

The center has also begun to analyze results from its early efforts to offer patients clinical germline cancer susceptibility information from their MSK-IMPACT sequencing data, a service MSKCC launched with approval from New York State and CLIA in March of last year.

MSK-IMPACT, which stands for Memorial Sloan Kettering-Integrated Mutation Profiling of Actionable Cancer Targets, is a 410-gene matched tumor/normal NGS panel that detects a range of DNA alterations including point mutations, small insertions and deletions, copy number variations, and structural rearrangements, and is available to patients treated at MSKCC — mostly those with advanced metastatic disease.

At last month's annual meeting of the American Association for Cancer Research, Mike Berger, associate director of MSKCC's Marie-Josée and Henry R. Kravis Center for Molecular Oncology and a co-developer of MSK-IMPACT, presented early results from the new clinical germline testing program, an update on the test in general, and some future directions such as liquid biopsy.

Before launching the germline interpretation service, Berger and colleagues conducted an anonymized retrospective study of about 1,600 patients who had been tested for somatic alterations through MSK-IMPACT between March and October 2014.

That research, published this January in JAMA Oncology, revealed that about 15 percent of patients harbored at least one presumed pathogenic germline variant in some known Mendelian disease-associated gene, and 13 percent had variants that were associated with cancer susceptibility.

One particularly interesting finding from the retrospective analysis was that for almost half of the subjects who did have a germline variant of some kind, it was in the context of a tumor type not normally associated with that particular variant. For example BRCA1 and BRCA2 mutations occurred frequently in patients with tumors other than the expected breast, ovarian, prostate, and pancreatic cancers.

Based on these findings, the MSKCC team decided to start reporting not only somatic tumor mutations but also germline results from MSK-IMPACT, putting in place a consent process for sequenced patients that allows them to opt in to added germline analysis in addition to the standard somatic test reports.

At the AACR meeting, Berger said that among 648 patients who received the clinical germline analysis since it launched last year, 20 percent have had a purportedly pathogenic variant associated with inherited cancer risk.

The higher rate in this clinically tested group versus the team's retrospective cohort probably rests on the fact that the clinically analyzed patients represent a biased sample — they were referred specifically for germline testing by their physicians based on various clinical factors, and, at least in the early months of the new service, mostly had cancer types where the presence of germline cancer susceptibility mutations is known to be higher.

Meanwhile, with 10,000 total sequenced cases on the somatic interpretation side of things, Berger also said that the MSK-IMPACT cohort is becoming large and mature enough for the team to start to tease out information about the clinical utility of their endeavor.

For one, Berger said, the implementation of the test appears to have positively affected clinical trial enrollment at MSKCC. In a slide, he showed how accrual to basket trials — in which patients are recruited across different tumor types based on the presence a specific molecular feature — showed a sharp upswing after MSK-IMPACT was implemented two years ago.

Time will tell how this in turn may affect the success of these trials, Berger said. "We are taking a prospective approach to use this data as best we can to drive patients toward the right studies so hopefully so-called exceptional responses are less exceptional and more common in these trials," he said at the AACR meeting.

The group is also working on its first analyses of how many patients tested actually have an actionable alteration in their somatic sequencing data, and amongst those, how many are then treated based on that information.

At the AACR meeting, he highlighted one early analysis of lung cancer patients that revealed that 77 percent of them appear to harbor some sort of clinically relevant alteration. Moreover, the vast majority have received relevant therapy either on label or in clinical trial, he said.

Speaking with GenomeWeb this week, Berger added that there are subgroups of MSKCC researchers who are doing the same type of analysis for other specific cancer types, as well as a team, which he leads, that is doing broad overview of the full IMPACT-tested cohort.

"Each disease team is going back and looking at the rate that patient care was altered," he said. "We are tabulating that … and we have a lot of meetings where we review different diseases and try to figure out, tumor type by tumor type, which areas this is making the biggest difference."

Looking to the future, Berger and colleagues are currently working on something they hope will be the next expansion of MSK-IMPACT — liquid biopsy.

In his presentation at AACR, he said that he and colleagues have run the IMPACT panel on blood samples from a number of cases, and have shown that it works on cell-free DNA, at least in some cancer types.

For others, the fraction of tumor-derived DNA appears to be too low for IMPACT. For those cases, the group is working on developing a complementary assay using a more targeted panel and hybridization capture-based deep sequencing, down to 10,000x, using duplicate reads to suppress background errors.

"It's not that different from what other leading groups have published. We are following all of it closely and making some tweaks of our own," Berger said.

This is currently all on the research side. Liquid biopsy results are not being returned to patients. But according to Berger, the group hopes to be able to offer some sort of clinical blood-based targeted NGS within the next year.