Can non-renal clearance and/or bioavailability of drugs be assessed reliably in renal impairment studies with oral drug administration?

Abstract

The causes for the variable susceptibility of renal clearance (CLr) and bioavailability (F) of drugs in renal impairment are still unknown. We investigated whether the impact of chronic kidney disease (CKD) on non-renal clearance (CLnr) or F can be appraised when drug administration is by the oral route only in dedicated renal impairment studies (DRIS), as is routinely done when developing drugs intended for oral use. A literature search on DRIS administering drugs orally only or orally and intravenously was conducted. Seven drugs administered orally only with notable CLnr and 2 drugs administered by the oral and intravenous routes with negligible CLnr were identified. Regressions of oral clearance (CL/F), normalized by absolute bioavailability in healthy subjects (F1), on CLr were performed for the drugs with notable non-renal elimination to determine the impact of CKD on CLnr. Regressions of CL/F and CL on CLr were conducted for the drugs with negligible CLnr to determine F. Excessive variability in CL/F and CLr precluded evaluation of CLnr for 1 drug with notable CLnr and F1 < 0.01. A categorization based on the susceptibility of CLnr to CKD appeared possible for the 6 drugs with notable non-renal elimination if the parameters of the F1 normalized regressions of CL/F on CLr are taken at face value, i.e., if equality of F and F1 is assumed. However, the true relationship between F and F1 in subjects with varying renal function is unknowable for drugs with significant CLnr when administered orally only. F of drugs with significant CLnr may be altered by a reduced activity of uptake-transporters and/or enzymes so that in renal impaired subjects both absorption and first pass metabolism of intact drug may be reduced relative to healthy subjects, making it impossible to predict whether F in the former or latter population is greater. Bioavailability of drugs with negligible CLnr may depend primarily on the integrity of uptake-transporters so that F in healthy subjects is expected to be greater than in renal impaired subjects. Apparently accurate estimates of F for drugs with negligible CLnr may be obtained from DRIS with oral administration by using the reciprocal of the slope of the regressions. A reliable assessment of the impact of CKD on CLnr for drugs with significant non-renal elimination requires information after oral and intravenous administration in the same DRIS study. However, apparently accurate estimates of F for drugs with negligible non-renal elimination may be obtained in DRIS with oral drug administration only, but validation of the proposed method with other drugs exhibiting negligible non-renal elimination and variable F1 is required.