ATLANTA (GenomeWeb) – A clinico-genomic real-world database can be useful to study the impact of rare genetic mutations in cancer and the efficacy of drugs in molecularly defined patient subsets, researchers demonstrated this week at the American Association for Cancer Research annual meeting.
AACR Project GENIE (Genomics Evidence Neoplasia Information Exchange), launched in 2015, is a registry of genomic and clinical data of thousands of cancer patients treated at 19 participating institutions in North America and Europe. The goal of the project is to collect this data and share it to improve decision making in cancer care and fuel research.
The project publicly releases data twice a year after a period of exclusivity for the participating centers. At the first data release in 2017, the registry included information on around 18,900 sequenced tumors and as of the latest release in January this year, the registry contains data from 59,400 sequenced tumors spanning more than 80 major cancer types. In a few months, when there is another public release, the registry is expected to grow to around 70,000 sequenced tumors, said Charles Sawyers, chair for the AACR Project GENIE Steering Committee.
"It's continuing to ramp up," said Sawyers, who also chairs the human oncology and pathogenesis program at Memorial Sloan Kettering Cancer Center, and who provided an update on GENIE at the AACR annual meeting. "People are using the data."
AACR has published two papers to date, one in 2017 in Cancer Discovery describing the mutational landscape from tumors sequenced as part of the first data release and another in JCO Clinical Cancer Informatics last year that details how the participating institutions, which all use different sequencing platforms and collect and store data in different ways, harmonize the information into one registry.
"The real challenge in establishing this [registry] and the holy grail is to get the longitudinal follow-up [on patient outcomes,]" Sawyers said. Institutions that belong to the consortium have agreed to provide longitudinal outcomes data for specific projects when queried by GENIE headquarters.
At the meeting this week, researchers presented data from two such projects where longitudinal data within GENIE was used. The projects highlight the utility of the registry in creating synthetic control arms within real-world studies, something the US Food and Drug Administration is interested in using to support regulatory decisions on drugs.
In December, the FDA released its "Framework for Real-World Evidence Program," in which the agency discussed the possibility of evaluating the efficacy of products within nonrandomized, single-arm trials and comparing to controls gleaned from real-world data. Rajeshwari Sridhara from the FDA highlighted that the agency has already used real-world data in its supplemental approval of Amgen's B-cell acute lymphoblastic leukemia drug blinatumomab (Blincyto).
In this vein, David Hyman presented data gleaned from GENIE to conduct a case control study comparing estrogen receptor-positive HER2-negative metastatic breast cancer patients with AKT1 E17K mutations and controls without these mutations; while Monica Arnedos from Gustave Roussy Cancer Campus in France did a similar comparison of hormone receptor positive, HER2-negative breast cancer patients with and without ERBB2 mutations. Both of these markers are rare in the population, but the scale of GENIE offers a chance to study their impact on cancer patients.
In the study on AKT1 E17K mutations, researchers matched cases and controls at the institutional level, according to if they were born within a 10-year range, if they were sequenced within a four-year range, and breast cancer histologic subtype. The comparison showed that advanced breast cancer patients with rare AKT1 E17K mutations and those who are wildtype have similar disease trajectories in terms of sites of metastases and overall survival.
Treatment outcomes with endocrine therapy, fulvestrant, and chemotherapy were also similar between the groups, though the data suggested that patients with mutations may have a longer duration on treatment with an mTOR inhibitor compared to the wildtype population.
GENIE similarly enabled evaluation of ERBB2 mutations, which occur in 1 percent to 2 percent of breast cancers. Patients with these rare mutations have been reported to be resistant to some HER2 targeted therapies but possibly responsive to Puma Biotechnology's neratinib (Nerlynx). The FDA has approved the drug for early-stage HER2-overexpressed/amplified breast cancer patients after adjuvant treatment with a trastuzumab (Herceptin)-based therapy.
Researchers identified 45 individuals with metastatic, HER2 expression/amplification-negative breast cancer with these ERBB2 (also called HER2) mutations among 8,700 breast cancer patients in the registry. They then matched these cases to 90 ERBB2-wildtype controls based on race, gender, birth year, and age at sequencing. There were no major differences in the clinicopathological features between the two groups, and overall response rate and time to progression didn't differ.
In this study, researchers identified 19 patients with ERBB2 mutations who received a HER2-inhibitor — 14 received neratinib, three received lapatinib, and two received an undisclosed drug. Analysis of outcomes in the mutation-positive group showed a non-significant trend toward overall survival with neratinib treatment, though this study wasn't designed to look at this outcome specifically.
Puma is conducting the Phase II SUMMIT basket trial involving patients with solid tumors and activating, somatic HER2 mutations, that will enroll breast cancer patients and evaluate their response to neratinib. The company recently reported data from this study on 11 patients with HER2-mutant cervical cancers. Puma was involved in the analysis of the Project GENIE data, but did not reply to questions about how the insights from the registry fits in with the company's broader development plan for neratinib.
"As you see we get deeper phenomics through these various projects," Sawyers said. Currently within GENIE, the genomic data in the registry is all somatic, but in the future, the project has ambitions to collect data on germline DNA, cell-free DNA, RNA sequences, and epigenetics.
In the near term, however, the project plans to focus on expanding phenomics capabilities. The phenotypic data parameters in the registry currently include information on tumor type, histology, vital stats, and demographics. However, the consortium members and AACR leaders have agreed to expand the phenotypic parameters for all patients in GENIE to include information on medications and outcomes, which are difficult parameters to grab from electronic medical records.
"They require intensive curation efforts," said Sawyers. "That's a project we'll launch soon and take this to a whole another level and be unique among data registries."