Exposure Response Analysis of the Effect of Loperamide on the Cardiac Repolarization Interval in Healthy Adult Subjects
Belén Valenzuela (1), Per Olsson Gisleskog (2), Iolanda Cirillo (3), Jay Ariyawansa (4), Saberi Rana Ali (5), Juan José Pérez-Ruixo (1), Oliver Ackaert (1)
(1) Janssen Clinical Pharmacology and Pharmacometrics, Beerse, Belgium. (2) POG Pharmacometrics, London, UK. (3) Janssen Clinical Pharmacology and Pharmacometrics, Raritan, US. (4) Janssen Statistical and Decision Sciences, Raritan, US. (5) Janssen Established Products, Titusville, US
Introduction and Objectives: Loperamide is indicated for the symptomatic control of acute and chronic diarrhea via selective binding and activation of μ-opioid receptors in the gut wall. The objective of this analysis was to assess the relationship between the plasma concentrations of loperamide and its N-desmethyl loperamide metabolite (M1) and the potential QT interval prolongation, using therapeutic and supratherapeutic dose levels of loperamide.
Methods: Data from a randomized, double-blind, placebo- and positive-controlled, single-dose, 4-way (placebo, loperamide 8 mg, loperamide 48 mg and moxifloxacin 400 mg) crossover study in healthy adult subjects [1] was used for the concentratrion-QTc analysis. Triplicate electrocardiograms (ECGs) were extracted from 12-lead digital Holter recordings at scheduled timepoints on Day 1, 2 and 14 of each treatment period. For each subject and treatment, the average of 3 sets of triplicate predose ECG taken prior to the dose administration was used as the baseline. ECG measurements were obtained at the same time as plasma concentrations of loperamide and M1 metabolite, resulting in paired PK-ECG measurements. The PK and PD data were analysed using non-linear mixed effect modelling approach (NONMEM® version 7.4.3.) [2]. The data processing, statistical analysis, diagnostic plots, and post-processing of results were carried out in R, version 3.6.2 [3].
Results: A total of 53 subjects with 1408 time-matched PK and QT measurements were available for the analysis. The Fridericia correction method was found appropriate to correct for any HR effect on QT. The placebo-corrected change from baseline QTcF (∆∆QTcF) was selected as primary response variable. Hysteresis between both loperamide and M1 plasma concentrations, and ΔΔQTcF was observed in the supratherapeutic (48 mg) dose group. A population PK-PD analysis was performed using an effect compartment model to collapse the hysteresis. To this end, fit-for-purpose population PK models were developed enabling the characterization of the individual plasma concentration profiles of loperamide and M1. Subsequently, a univariate analysis, characterizing the relationship between ΔΔQTcF and loperamide exposure, as well as ΔΔQTcF and M1 exposure separately, showed a slightly better model fit for the M1 metabolite. When combining both loperamide and M1 in one model, the collinearity between both variables resulted in high variance inflating factor, high relative standard errors of parameter estimate (>100%) and a very high condition number that suggested the model was ill-conditioned. This finding prevented to have a stable model that reliably estimate the relative contribution of loperamide and M1 on ΔΔQTcF.
The model best describing the linear concentration-ΔΔQTcF relationship was the pre-specified model where M1 concentrations in the effect-compartment were characterizing the exposure-dependent effect of loperamide in ΔΔQTcF. An intercept of -1.66 (95% confidence interval [CI]: -2.84 to -0.48) msec was estimated and the ΔΔQTcF increases with a slope 0.544 (95%CI: 0.364 to 0.724) msec per ng/mL of M1 concentration in the effect compartment. The model-predicted mean ΔΔQTcF at the observed geometric mean of the maximum concentration in the effect compartment following a single dose of 8 mg (2.1 ng/mL) and 48 mg (14.2 ng/mL) was -0.526 msec (90% CI: -1.51 to 0.462 msec) and 6.06 msec (90% CI: 3.86 to 8.27 msec), respectively, with the upper limit of the two-sided 90% CI below 10 msec for both doses. A sensitivity analysis considering the loperamide concentrations in the effect compartment, instead of M1, led to the same conclusion. The exposure-response (ER) analysis confirms the results of the intersection-union test, that concluded the absence of an effect of loperamide or M1 on cardiac repolarization following single dose administration of an 8 mg therapeutic (maximum approved over-the counter (OTC) daily dose for adults) and a 48 mg supra-therapeutic dose.
Conclusions: ER analysis based on QTc Holter data supports that within the investigated loperamide and M1 concentration range, there is no evidence of an effect of loperamide or M1 on cardiac repolarization.
References:
[1] A Study to Evaluate the Effects of Loperamide (JNJ-289679) on Electrocardiogram Intervals in Healthy Adult Participants. ClinicalTrials.gov identifier: NCT04225078. Accessed January 23, 2023. https://clinicaltrials.gov/ct2/show/NCT04225078?term=R018553NAP1001&draw=2&rank=1
[2] Beal SL et al. 1989-2011. NONMEM Users Guides. Icon Development Solutions, Ellicott City, Maryland, USA.
[3] R Development Core Team (2012). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org/