Abstract 1353P
Background
It has been estimated that more than 30% of patients with early stage non-small cell lung cancer (NSCLC) develop recurrence despite curative intent resection. Identification of these patients may permit targeting of patients most likely to benefit from adjuvant treatment.
Methods
This was a retrospective observational study to identify predictors of recurrence following resected early-stage NSCLC using data from the Optum Enhanced Oncology database, which contains de-identified, linked enrollment, health insurance claims, and information from electronic medical records from January 1, 2007 to June 30, 2020. Patients who had stage I or II NSCLC and underwent surgery after diagnosis of NSCLC with ≥6 months of enrollment or activity prior to first evidence of lung cancer were selected. Patient characteristics were assessed as of the date of first surgery following NSCLC diagnosis (index date) and included age, sex, race, region, insurance type, stage, grade, type of surgery, smoking history, mutation status (PDL1, EGFR, KRAS, ALK), and comorbidities. Outcomes included recurrence free survival (RFS), time to recurrence (TTR), and overall survival (OS). The association between patient characteristics and outcomes were assessed using three logistic regression models adjusting for duration of follow-up.
Results
A total of 3267 patients met inclusion criteria (722 recurred, 715 died, 1134 recurred or died). Mean (SD) age was 67 (9) years (46% male, 27% current smokers, 67% stage I, 65% had lobectomy, 6.6% PDL1+, 5.1% EGFR+, 3.6% KRAS+, 1.5% ALK+). Grade information was available for only 14% of patients. Significant predictors of RFS were type of surgery, age, gender, insurance type, smoking status, KRAS+, PDL1+, and history of COPD and HIV. Predictors of TTR and OS were similar. The C-statistic was 64.4 for RFS and TTR, and 67.2 for OS suggesting relatively poor predictability. Results were robust across models.
Conclusions
It was infeasible to reliably predict the recurrence of NSCLC using data from typical claims and EHR databases. Additional prognostic information beyond routinely collected data will be beneficial to improve NSCLC outcomes.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
GRAIL LLC a subsidiary of Illumina Inc. currently held separate from Illumina Inc. under the terms of the Interim Measures Order of the European Commission dated 29 October 2021.
Funding
GRAIL LLC a subsidiary of Illumina Inc. currently held separate from Illumina Inc. under the terms of the Interim Measures Order of the European Commission dated 29 October 2021.
Disclosure
T. Delea: Financial Interests, Institutional, Sponsor/Funding: GRAIL LLC, a subsidiary of Illumina Inc. A. Moyhanan: Financial Interests, Institutional, Sponsor/Funding: GRAIL LLC, a subsidiary of Illumina Inc. R. Myers: Financial Interests, Full or part-time Employment: GRAIL LLC, a subsidiary of Illumina Inc.; Financial Interests, Stocks/Shares: GRAIL LLC, a subsidiary of Illumina Inc. K.C. Chung: Financial Interests, Full or part-time Employment: GRAIL LLC, a subsidiary of Illumina Inc.; Financial Interests, Stocks/Shares: GRAIL LLC, a subsidiary of Illumina Inc.