Minisymposium explores real-world examples of AACR Project GENIE applications
Launched in 2015, the AACR Project GENIE registry has emerged as a powerful resource to not only facilitate clinical decision making but also impact translational research, said Trevor J. Pugh, PhD, UHN Princess Margaret Cancer Centre, during the Minisymposium Advancing Cancer Research Through an International Cancer Registry: AACR Project GENIE Use Cases.
“Since the first release, nine data sets have now been released, the most recent containing data from more than 112,000 patients,” said Pugh, presenting on behalf of the AACR GENIE Genomics and Analysis Working Group. “We’ve logged over 7,600 users, over 280 citations, and the top 3 cancer types are represented by data from over 10,000 patients.”
The data set continues to grow, he said, through ongoing contributions and collaborations, additional institutions, and an increased breadth of profiling.
“We also have seen a general trend towards increasing gene content on clinical panels, and the clinical data continues to get richer,” Pugh said. “In general, GENIE is really an ideal framework for a global precision medicine strategy, led first and foremost by very clear and strong governance, real commitment and operationalization of data standards, and real impact on both clinical care and cancer research.”
Evolution of OncoKB, a precision oncology knowledge base
Sarah P. Suehnholz, PhD, Kravis Center for Molecular Oncology, Memorial Sloan Kettering Cancer Center, talked about how they have used AACR Project GENIE in conjunction with OncoKB, Memorial Sloan Kettering’s precision oncology knowledge base.
“The goal of this project was to use the AACR Project GENIE cohort annotated with OncoKB versions from 2018 and 2021 to evaluate changes in the clinical actionability landscape,” Suehnholz said. “We used the AACR GENIE cohort version 9, which consists of almost 100,000 samples across 19 institutions and 51 solid tumor types.”
A key utility of knowledge bases such as OncoKB, Suehnholz said, are their ability to assign clinical actionability to investigational biomarkers based on compelling clinical data prior to FDA approvals.
“The field of precision medicine is rapidly evolving, and this is evident by the dramatic shift in the clinical actionability landscape over the past five years,” Suehnholz said. “This includes three tumor type agnostic FDA drug approvals—two for the anti-PD1 antibody pembrolizumab for MSI-H solid tumors and TMB-H solid tumors, as well as the FDA approval of the NTRK inhibitor larotrecinib for NTRK fusion-positive solid tumors.”
Suehnholz noted that between 2016 and today there were 49 biomarker-based and tumor-type specific FDA drug approvals, including 28 that occurred in 2020 amid the COVID-19 pandemic.
Real-world clinical outcomes for patients with advanced NSCLC
Beilei Cai, PhD, Novartis, presented the first reported analysis of real-world data for advanced non-small cell lung cancer (aNSCLC) using the AACR Project GENIE database.
“Although patients with advanced non-small cell lung cancer without common actionable oncogenic drivers, such as EGFR mutations or ALK rearrangements, may receive immunotherapy and/or chemotherapy and targeted therapies, there is still high unmet medical need,” Cai said.
The study aims to explore overall survival and progression-free survival based on both radiographic imaging and medical oncologist assessment among NSCLC patients with EGFR wild type and ALK wild type treated in the real-world academic setting using AACR Project GENIE data, with a focus on first-line treatments.
A total of 263 patients satisfied criteria for inclusion in the study, and 175 (67 percent), 31 (12 percent), 14 (5 percent), and 38 (14 percent) initiated chemotherapy, IO-monotherapy, TT, and bevacizumab + chemotherapy respectively.
Among their findings, Cai reported that the median progression-free survival in patients with stage IV NSCLC was 5.0 (95 percent CI, 3.9-5.7) months, and median overall survival was 15.1 (95 percent CI, 11.6-20.6) months
“As biomarker testing has become more commonly used in recent years and more treatments become available, we look forward to repeating the analysis as the updated NSCLC Project GENIE datasets become available,” Cai said.
Racial differences among patients with early-onset colorectal cancer
Andreana N. Holowatyj, PhD, MS, Vanderbilt University Medical Center, presented the results of a study looking at the possible molecular features that may underlie colorectal cancer (CRC) disparities among young adults.
“Evidence is accumulating from clinical and molecular studies to suggest that early-onset colorectal cancer is a distinct malignancy from late-onset disease,” she said. “However, as the burden of early-onset colorectal cancer sharply varies across diverse populations, our understanding of the molecular characteristics of early-onset colorectal cancer across these groups remain quite limited.”
In their study, Holowatyj and her colleagues leveraged data from the AACR Project GENIE consortium that included 6,120 individuals agse 18-89 years at cancer sequencing with a pathologically confirmed CRC diagnosis across 12 institutions worldwide.
“We sought to characterize racial differences in tumor mutation burden specifically within the population of early-onset microsatellite-stable (MSS) colorectal cancer cases,” she said. “Strikingly, we observed that non-Hispanic Blacks, but not Asians or Pacific Islanders, had a significantly higher tumor mutational burden when compared with non-Hispanic Whites.”
Holowatyj said these patterns were consistent across both early-onset and late-onset groups in the study.
“This study provides the first-of-its-kind evidence that molecular features of early-onset CRC may differ by race, such that the development of therapeutic modalities that target these distinct molecular features may yield clinical implications for patients with early-onset colorectal cancer,” Holowatyj said. “Moving forward upon validation of these findings, this work emphasizes a timely need to integrate mechanistic laboratory-based approaches with human studies to unravel molecular underpinnings of early-onset colorectal cancer, as well as other gastrointestinal cancer disparities.”
Racial and ethnic differences in genomic profiling of early-onset colorectal cancer
The rising incidence of CRC in younger adults was the focus of a study presented by David M. Hein, MS, University of Texas Southwestern Medical Center.
“Between 1992 and 2017, colorectal cancer rates in young adults increased across all races and ethnicities and was fastest in non-Hispanic whites. Racial and ethnic differences in incidence and outcome are documented, but etiologies are understudied,” Hein said. “We wished to study genomic differences among young adult colorectal cancer patients by combining our population, which is enriched with Hispanic patients, with the GENIE database.”
In the study, Hein said they assessed the frequency of somatic mutations in 22 genes commonly mutated and clinically important in CRC. They used Fisher’s Exact Test to look for enrichment in mutations among these four racial and ethnic groups—non-Hispanic Asian, non-Hispanic Black, White Hispanic, and non-Hispanic White.
Among their findings, Hein reported that Hispanic patients with early-onset colorectal cancer appear to have enrichment of somatic mutations in Lynch/DNA repair genes and that KRAS mutations are more common among Black patients.
“Our findings highlight the need for more effort in ensuring greater diversity in genetic databases so that we are adequately representing our diverse young colorectal cancer population,” he said.
Defining real-world recurrence in the Biopharma Collaborative data
Jessica A. Lavery, MS, Memorial Sloan Kettering Cancer Center, reported the results of a study that suggests that the GENIE Biopharma Collaborative (BPC) phenomic data can be used to reliably estimate the date of recurrent disease among stage 1-3 patients who underwent surgery.
“This is an important first step in working towards a structured, consistent, methodologic standard for defining recurrence from EHR data, and it enables identification of metastatic cohorts to estimate progression-free survival,” Lavery said. “We note that multiple sources documenting cancer recurrence are needed, and that an intuitive landmark approach is not sufficient for determining recurrent events and estimating the timing of recurrence.”
Lavery said that estimates of recurrence-free survival (RFS) in this cohort will not be representative of RFS in the population of patients with NSCLC because of the inclusion criteria of genomic sequencing to be included in the GENIE BPC data.
“Our goal here is really to estimate the timing of recurrence to determine when a patient is eligible to progress, so that we have a larger cohort of patient for progression-free survival,” Lavery said. “Future work includes validating this algorithm against patients who did not undergo surgery, as well as expanding the algorithm to multiple cancer sites.”
A replay of this session is available to registrants through June 21, 2021.