Penalized regression for left‐truncated and right‐censored survival data

Abstract

High-dimensional data are becoming increasingly common in the medical field as large volumes of patient information are collected and processed by high-throughput screening, electronic health records, and comprehensive genomic testing. Statistical models that attempt to study the effects of many predictors on survival typically implement feature selection or penalized methods to mitigate the undesirable consequences of overfitting. In some cases survival data are also left-truncated which can give rise to an immortal time bias, but penalized survival methods that adjust for left truncation are not commonly implemented. To address these challenges, we apply a penalized Cox proportional hazards model for left-truncated and right-censored survival data and assess implications of left truncation adjustment on bias and interpretation.We use simulation studies and a high-dimensional, real-world clinico-genomic database to highlight the pitfalls of failing to account for left truncation in survival modeling.

Prognostic value of plasma circulating tumor DNA fraction across four common cancer types: a real-world outcomes study

“…To account for this, risk-set adjustment was carried out, including only patients who have met all inclusion criteria at each time point as at risk in Kaplan–Meier and Cox model analyses. 25 The assumption of independent left truncation was verified by univariable modeling of the effect of delayed entry time on the survival outcome. R version 4.2.1 software (R Foundation for Statistical Computing, Vienna, Austria) was used for all statistical analyses.…”

Comparative Effectiveness of Immune Checkpoint Inhibitors vs Chemotherapy by Tumor Mutational Burden in Metastatic Castration-Resistant Prostate Cancer

“…Electronic health record OS data derivation has been previously described . OS risk intervals were instead initiated at time of CGP report (left truncation) if the CGP report was received after treatment initiation, as patients who die before a CGP report is delivered are not included in the database . Differences in time-to-event outcomes were assessed with the log-rank test and Cox proportional hazard models.…”

Mismatch repair deficiency, next-generation sequencing-based microsatellite instability, and tumor mutational burden as predictive biomarkers for immune checkpoint inhibitor effectiveness in frontline treatment of advanced stage endometrial cancer

“… 17 Because patients cannot enter the database until a comprehensive genomic profiling report is delivered, overall survival risk intervals were left truncated to the date of report to account for immortal time. 24 25 Flatiron Health database mortality information is a composite derived from three sources: the electronic health record, Social Security Death Index, and a commercial death dataset from obituaries and funeral homes. The mortality information has been externally validated in comparison to the National Death Index with >90% accuracy.…”