Prediction of the Effect of Renal Impairment on the Pharmacokinetics of New Drugs.

INTRODUCTION: Renal impairment may have a significant impact on the pharmacokinetics of drugs. Ad hoc studies in subjects with renal impairment are required by the regulatory authorities to propose dose adjustments in these subjects, to find a dosing regimen able to provide a systemic exposure similar to those in subjects with a normal renal function given the relevant clinical dose.
METHODS: To evaluate the main descriptors and establish a predictive model of the effect of renal impairment on the exposure of new drugs, we considered 73 marketed drugs, for which studies in subjects with different degrees of renal impairment were available in the literature. Multivariate analysis was performed using the main pharmacokinetic parameters. Other approaches, including data mining and machine learning techniques, were tested to propose models based on a categorical definition of the exposure changes.
RESULTS: Stepwise multivariate regression analyses revealed, as expected, that the fraction of dose excreted unchanged in urine and plasma protein binding were the factors primarily related to the change in exposure between subjects with normal and impaired renal function. Data mining techniques provided similar results. DISCUSSION: The pharmacokinetic predictions were however not always satisfactory, especially for drugs which, despite the negligible renal excretion, are characterized by significant increases in the systemic exposure in subjects with renal impairment. This phenomenon, interpreted considering the accumulation of endogenous metabolism inhibitors in subjects with moderate and severe renal disease (uremic toxins), cannot be fully captured and described, likely owing to an incomplete understanding of the pathophysiological phenomena and to some limitations of the available database of clinical studies.