Reexamining transaminase elevation in Phase I clinical trials: the importance of baseline and change from baseline.

PURPOSE: The aims of the present analysis were to determine prevalence of transaminase elevation in placebo-treated healthy volunteers in our historical phase I clinical trials and to assess which factors were associated with it.
METHODS: We performed a retrospective analysis of alanine transaminase (ALT) levels in 481 placebo-treated healthy volunteers from 20 phase I trials by examining ALT elevation rates using the upper limit of normal values (ULN) as the cutoff as well as changes from baseline in actual ALT values.
RESULTS: The ULN for ALT ranged from 32 to 72 IU/L across the studies. Although the overall ALT elevation rate (4.4%) from pooled datasets was low, the elevation rates were higher in more recent studies than in earlier ones. While elevation rates at baseline and during placebo treatment did not differ significantly, ALT maximal levels during placebo treatment were significantly higher than baseline levels. Moreover, baseline ALT levels were found to be more important in predicting ALT elevation during placebo treatment than demographic and study design factors.
CONCLUSION: Baseline level and changes from baseline in transaminase are important variables to examine in addition to elevation above ULN for more reliably interpreting liver signals in Phase I clinical trials.