DDPGPSurv

DDPGPSurv estimates optimal intravenous busulfan AUC intervals to support precision dosing during allogeneic stem cell transplantation for patients with acute leukemia.

Key Features:

• Bayesian Nonparametric Regression: Integrates the dependent Dirichlet process with Gaussian processes (DDP-GP) for flexible Bayesian nonparametric survival regression.
• Personalized Therapeutic Intervals: Tailors optimal AUC intervals based on individual covariates including age and disease status at transplant (complete remission versus active disease) to balance toxicity and relapse risk.
• Data-Driven Insights: Analyzes a dataset of 151 patients to identify substantial variation in optimal therapeutic AUC intervals across demographic and clinical subgroups.
• Comparative Performance: Uses extensive simulation studies to validate the DDP-GP model and compare its performance to alternative methods in similar settings.

Scientific Applications:

• Precision dosing in allo-SCT for acute leukemia: Supports determination of busulfan dosing regimens by estimating covariate-specific AUC targets for allogeneic stem cell transplantation in acute leukemia.
• Pharmacokinetic/pharmacodynamic variability analysis: Applies to contexts where PK/PD variability influences therapeutic efficacy and toxicity and AUC-guided dosing is relevant.
• Evaluation of conditioning regimens: Provides a statistical framework to assess how conditioning regimen AUC relates to clinical outcomes and to inform regimen optimization.

Methodology:

Implements a dependent Dirichlet process with Gaussian processes (DDP-GP) Bayesian nonparametric survival regression to estimate covariate-specific AUC intervals, applied to a 151-patient dataset and evaluated via simulation studies.