Real-Time Surveillance of New Anticancer Therapies in the National Veterans Affairs System

• Published in: International journal of radiation oncology, biology, physics Vol. 111; no. 3; p. e431
• Main Authors: Bryant, A.K., Sankar, K., Ramnath, N., Green, M.
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.11.2021
• DOI: 10.1016/j.ijrobp.2021.07.1226
• ISSN: 0360-3016

The Veterans Affairs (VA) system is the largest integrated health system in the US, with one of the world's longest-running electronic medical record (EMR) systems. We describe the potential of VA EMR data for real-time surveillance of efficacy, toxicity, and uptake of new therapies in oncology. As a case study, we describe the implementation of durvalumab for adjuvant treatment of stage III non-small-cell lung cancer (NSCLC) treated with concurrent chemoradiation (CRT) after publication of the landmark PACIFIC trial in 2017. We used the Veterans Affairs Informatics and Computing Infrastructure (VINCI) to identify all intravenous administrations of durvalumab from November 2017 to January 2021 among patients with at least one inpatient or outpatient lung cancer ICD-10 diagnosis code. Cancer stage and tumor histology were algorithmically extracted from physician notes and reviewed for accuracy. Radiation therapy dates and concurrent chemotherapy delivery were ascertained through outpatient CPT codes for radiation delivery and intravenous chemotherapy infusion records. Overall survival (OS) was obtained from the VA Vital Status File (updated on a quarterly basis) and from the VA Master Patient Index (updated on a daily basis). Date of radiographic progression were obtained through physician review of extracted CT and PET scan reports. Demographic, staging, and treatment information were supplemented with VA Central Cancer Registry information where available. Statistics were performed with statistical analysis software. We identified 1106 patients with stage III NSCLC (60% stage IIIA, 35% IIIB, and 3% IIIC) treated with upfront CRT who received at least one adjuvant durvalumab infusion, of whom only 135 (12%) were represented in cancer registry data. The median age was 69 years and 95% were male. Eighty-six percent reported current or former tobacco use, and the cohort was evenly split between adenocarcinoma (48%) and squamous cell histology (48%). Patients received a median of 6 durvalumab infusions (IQR 3 to 15) over a median of 154 days (IQR 35 to 308 days). Compared to patients enrolled in the PACIFIC trial, veterans received a lower number of durvalumab infusions (6 vs. 14) and had a higher rate of systemic steroid initiation for possible immune-related toxicity (43% vs 20%). Three-year progression-free survival was 32.5% (95% CI 25% to 40%). Three-year OS was 51% (95% CI 46% to 56%) compared to 57% in PACIFIC. Veterans received fewer adjuvant durvalumab infusions than in the PACIFIC trial but maintained comparable survival. National VA administrative data, which are largely updated on a daily or weekly basis, are a unique resource to study real-time implementation of new therapies in large national cohorts. Improved case-finding algorithms are necessary to capture patients not included in the VA Central Cancer Registry.