RESOLVE-IPD: High-Fidelity Individual Patient Data Reconstruction and Uncertainty-Aware Subgroup Meta-Analysis

TL;DR

RESOLVE-IPD enables accurate IPD reconstruction and robust, uncertainty-aware subgroup meta-analyses, strengthening the reliability and transparency of secondary evidence synthesis in precision oncology.

Abstract

Individual patient data (IPD) from oncology trials are essential for reliable evidence synthesis but are rarely publicly available, necessitating reconstruction from published Kaplan-Meier (KM) curves. Existing reconstruction methods suffer from digitization errors, unrealistic uniform censoring assumptions, and the inability to recover subgroup-level IPD when only aggregate statistics are available. We developed RESOLVE-IPD, a unified computational framework that enables high-fidelity IPD reconstruction and uncertainty-aware subgroup meta-analysis to address these limitations. RESOLVE-IPD comprises two components. The first component, High-Fidelity IPD Reconstruction, integrates the VEC-KM and CEN-KM modules: VEC-KM extracts precise KM coordinates and explicit censoring marks from vectorized figures, minimizing digitization error, while CEN-KM corrects overlapping censor symbols and eliminates the uniform censoring assumption. The second component, Uncertainty-Aware Subgroup Recovery, employs the MAPLE (Marginal Assignment of Plausible Labels and Evidence Propagation) algorithm to infer patient-level subgroup labels consistent with published summary statistics (e.g., hazard ratio, median overall survival) when subgroup KM curves are unavailable. MAPLE generates ensembles of mathematically valid labelings, facilitating a propagating meta-analysis that quantifies and reflects uncertainty from subgroup reconstruction. RESOLVE-IPD was validated through a subgroup meta-analysis of four trials in advanced esophageal squamous cell carcinoma, focusing on the programmed death ligand 1 (PD-L1)-low population. RESOLVE-IPD enables accurate IPD reconstruction and robust, uncertainty-aware subgroup meta-analyses, strengthening the reliability and transparency of secondary evidence synthesis in precision oncology.

Figures (16)

• Figure 1: Schematic overview of the RESOLVE-IPD framework.
• Figure 2: KM plots derived from reconstructed subgroup-level IPD using the VEC-KM + CEN-KM pipeline, overlaid with the ground-truth KM curves. (A) Overall population. (B) Biomarker-high subgroup. (C) Biomarker-low subgroup reconstructed directly using VEC-KM + CEN-KM. (D) Biomarker-low subgroup reconstructed via the KM subtraction method.
• Figure 3: True subgroup KM curves overlaid with (A) reconstructed KM curves from all initial candidate labelings $\mathcal{G}_{\mathrm{MAPLE}}$, and (B) the refined set of plausible labelings $\mathcal{G}_{\mathrm{MAPLE}}^*$, showing excluded (red) and retained (green) labelings.
• Figure 4: Reconstructed KM curves overlaid on the original publication curve for RATIONALE-302 and on KM-GPT–derived curves extracted from low-fidelity raster images for the remaining studies. (A) RATIONALE-302 PD-L1–low subgroup reconstructed via the High-Fidelity IPD Reconstruction procedure. (B)-(D) KEYNOTE-181, ESCORT and ATTRACTION-3 PD-L1–low subgroups reconstructed by MAPLE. IO = Immunotherapy, Chemo = Chemotherapy.
• Figure 5: Uncertainty-propagating meta-analysis of immunotherapy (IO) versus chemotherapy (Chemo) in the PD-L1-Low subgroup. (A) Ensemble of OS curves from 500 Monte Carlo simulations of the propagating meta-analysis, demonstrating a consistent separation between IO and Chemo across all plausible subgroup data configurations. (B) Time-varying HR with 95% CI from each run. Intervals excluding the null value (HR=1) are colored blue, indicating statistical significance.