Reconstructed Individual Patient Data from CheckMate 017 and 057 (5-year follow-up)

This dataset contains reconstructed pseudo-individual patient data (IPD) from the 5-year follow-up Kaplan-Meier curves of the CheckMate 017 (squamous NSCLC) and CheckMate 057 (non-squamous NSCLC) trials comparing Nivolumab vs. Docetaxel. Methods1. Digitization: KM curves for OS and PFS were extracted using WebPlotDigitizer2. Cleaning: Raw coordinates were cleaned to remove artifacts and duplicate points3. Reconstruction: The IPDfromKM R package (Liu & Lee, 2020) was used to reconstruct pseudo-IPD Contents- IPD_Master_All.csv: Final reconstructed pseudo-IPD (854 patients × 2 endpoints)- Reconstruction scripts (01_clean_raw_data.R, 02_reconstruct_ipd.R, 03_validate_risk_tables.R)- Raw WebPlotDigitizer exports and cleaned data- Risk table validation results Variables- time: Time-to-event (months)- status: Event indicator (1=event, 0=censored)- treat: Treatment (1=Nivolumab, 0=Docetaxel)- arm: Treatment arm name- analysis: Endpoint (OS or PFS) Sample Size- 427 patients per arm- 1,708 total records (OS + PFS) ValidationReconstruction validated against published risk tables:- Nivolumab OS: 0.6 patients mean error (0.7%)- Docetaxel OS: 1.0 patients mean error (2.5%)- PFS: Perfect match (0% error) Related Resources- Source publication: Hossein Borghaei et al. Five-Year Outcomes From the Randomized, Phase III Trials CheckMate 017 and 057: Nivolumab Versus Docetaxel in Previously Treated Non–Small-Cell Lung Cancer. J Clin Oncol 39, 723-733(2021). DOI:10.1200/JCO.20.01605 LicenseCC BY 4.0