A note on phase I interleaved versus parallel group ascending dose designs for concentration-QTc analyses.

Area of Science:

• Pharmacology
• Clinical Pharmacology
• Biostatistics

Background:

• Pharmacokinetic-QTc (PK-QTc) analyses are crucial in drug development, often using Phase I data.
• The impact of Phase I study designs on PK-QTc analysis precision is debated.
• Previous simulations suggested interleaved ascending dose designs offer higher power than parallel group designs.

Purpose of the Study:

• To re-evaluate the impact of interleaved versus parallel group designs on PK-QTc analysis power.
• To investigate if previous simulation conclusions hold with more realistic random effect variances and simulation setups.
• To compare the power of different study designs for PK-QTc analysis using historical data.

Main Methods:

• Re-analysis of historical QTc data to estimate random effect variances.
• Development of a simulation setup mimicking real data generation and baseline covariate correlation.
• Conducting a simulation study using revised parameters and comparing power between interleaved and parallel group designs.

Main Results:

• Estimated random effect variances from historical data were smaller than previously assumed.
• The revised simulation study showed no major power differences between interleaved and parallel group designs with equal observation numbers.
• Causal analysis supported the adequacy of the proposed simulation setup for change-from-baseline endpoints with baseline covariates.

Conclusions:

• The choice between interleaved and parallel group designs may have minimal impact on PK-QTc analysis power when sample sizes are comparable.
• Previous conclusions favoring interleaved designs may be based on overestimated random effect variances and less adequate simulation setups.
• The proposed simulation methodology provides a more realistic framework for evaluating PK-QTc study designs.