e15557 Background: Oncology has greatly benefited from the use of clinical genomics to aid diagnostic and treatment decisions. To inform targeted therapy, standard practice dictates molecular testing of metastatic colorectal cancer (mCRC) tumors through genomic testing (GT), including molecular biomarker, single gene, or next-generation sequencing (NGS). While persisting disparities in CRC screening, treatment, and outcomes have been well described, little is known about differences in GT for mCRC. Using all-payer electronic health record-derived de-identified data from a real-world database generated from routine clinical care across the U.S., we identified combinations of demographic and clinical characteristics, rather than individual factors, associated with GT. Methods: This study used the nationwide Flatiron Health electronic health record (EHR)-derived de-identified database. During the study period, the de-identified data originated from approximately 280 US cancer clinics (̃800 sites of care). Our study population included 26,524 patients with mCRC during the years 2013-2020. We evaluated documentation of receipt of GT (individual biomarker or NGS) measured within one year of diagnosis of metastatic disease, by demographic (age, sex, and race/ethnicity), payer type, de novo metastatic cancer (vs. progression to metastatic disease), tumor site (colon vs. rectum), Eastern Cooperative Oncology Group (ECOG) performance status, and year of diagnosis. We identified Elixhauser comorbid conditions from ICD codes, and grouped them by count (0-4, 5+). We conducted Classification and Regression Tree (CART) analysis, a machine learning approach to identify combinations of demographic and clinical characteristics associated with receipt of GT. Results: Nearly 83% had documented GT, 90% of whom did so within 6 months of diagnosis. The largest group of patients with GT at 90% consisted of individuals with colon cancer, 64 years of age or younger, 0-4 comorbidities, and an ECOG score of 0 or 1 (n = 2,004). Conversely, the lowest rates of GT ( < 60%) were observed among women 75 years of age or older, despite having 0-4 comorbidities (n = 1,314). In this group of patients, GT was higher among those with ECOG score of 1 than among those with high ECOG scores (66% and 54%, respectively). On the other hand, 71% of patients 75 years of age or older with 5+ comorbidities and high ECOG scores received GT. Conclusions: Considerable variations exist in GT across subgroups of the population. Additional analysis is warranted to characterize young and healthy patients who did not receive GT, and those who did get tested despite their older age and compromised health status. Future analyses will investigate whether documentation of receipt of GT is associated with treatment decisions and outcomes.
Validation of Colorectal Cancer Screening in the Electronic Health Record for Identifying Patients Due for Screening in a Pragmatic Trial
