Personalized Tuberculosis Treatment Through Model-Informed Dosing of Rifampicin

Abstract

Background and objective This study proposes a model-informed approach for therapeutic drug monitoring (TDM) of rifampicin to improve tuberculosis (TB) treatment. Methods Two datasets from pulmonary TB patients were used: a pharmacokinetic study (34 patients, 373 samples), and TDM data (96 patients, 391 samples) collected at Radboud University Medical Center, The Netherlands. Nine suitable population pharmacokinetic models of rifampicin were identified in the literature and evaluated on the datasets. A model developed by Svensson et al. was found to be the most suitable based on graphical goodness of fit, residual diagnostics, and predictive performance. Prediction of individual area under the concentration-time curve from time zero to 24 h (AUC 24) and maximum concentration (C max) employing various sampling strategies was compared with a previously established linear regression TDM strategy, using sampling at 2, 4, and 6 h, in terms of bias and precision (mean error [ME] and root mean square error [RMSE]). Results A sampling strategy using 2-and 4-h blood collection was selected to be the most suitable. The bias and precision of the two strategies were comparable, except that the linear regression strategy was more biased in prediction of the AUC 24 than the model-informed approach (ME of 9.9% and 1.5%, respectively). A comparison of resulting dose advice, using predictions on a simulated dataset, showed no significant difference in sensitivity or specificity between the two methods. The model was successfully implemented in the InsightRX precision dosing platform. Conclusion Blood sampling at 2 and 4 h, combined with model-based prediction, can be used instead of the currently used linear regression strategy, shortening the sampling by 2 h and one sampling point without performance loss while simultaneously offering flexibility in sampling times. Key Points We propose a new model-informed therapeutic drug monitoring method for rifampicin. The proposed method uses only two blood samples at 2 and 4 h after dosing. In contrast to linear regression-based methods, the proposed method allows for flexible sampling times and prediction of optimal dose adjustment.

Individualized Dosing With High Inter-Occasion Variability Is Correctly Handled With Model-Informed Precision Dosing—Using Rifampicin as an Example

“…As EBEs and shrinkage are dependent on the sampling design, different EBEs for each individual would have been achieved with a different sampling design. However, the sampling PK design used in this work is supported by the work by van Beek et al (2019) who identified this sampling design as most informative for deriving EBEs of rifampicin among those evaluated. If a different design for any reason would be applied, the absolute dose prediction error would most likely be higher.…”

“…Besides high accuracy in dose predictions, another advantage of a MIPD approach is the limited amount of samples and sampling occasions needed to forecast the next dose. It has previously been shown, that taking merely two blood samples (2 and 4 h post-dose) is sufficient to characterize individual PK parameters (van Beek et al, 2019) when using a model-based approach. In addition, the results of this simulation study demonstrate that two sampling occasions are sufficient to capture the IOV and individual exposure, since the decrease in imprecision and bias in dose predictions was statistically significant when information from two occasions were used to estimate EBEs, compared to only using information from a single occasion (Tables 5 and 6).…”

“…Rifampicin plasma concentrations were simulated (given covariates and doses) for 1,000 virtual patients per dose level (10,20,25,30, and 35 mg/kg), following a sparse sampling schedule including samples pre-dose (5 min before dose) and at 2 and 4 h post-dose relating to the suggested sampling scheme by van Beek et al (2019). Samples were taken at three sampling occasions (days 1, 7, and 14).…”

“…Svensson et al developed a MIPD approach for dose individualization of high-dose rifampicin, able to handle all the above mentioned complexities in PK (Svensson et al, 2019), based on a POPPK model (Svensson et al, 2018a) that has been shown to be best suitable for MIPD of rifampicin (van Beek et al, 2019). In this approach, an average exposure (AUC 0-24h at steady state: 235 h mg/L) corresponding to a high dose of 35 mg/kg in the PanACEA HIGHRIF1 trial, an open-label phase II multiple dose-rising trial registered at www.clinicaltrials.gov (NCT01392911) (Boeree et al, 2015;Svensson et al, 2018a;Svensson et al, 2019) is targeted, a dose that was found to be safe, while still resulting in high efficacy.…”

Model-Based Biomarker Selection for Dose Individualization of Tyrosine-Kinase Inhibitors

“…Limitations related to the use of model-based dose individualization for TKIs mainly involve the complexity and computational power related to the model building process, and consequently the availability of such models and software to have a direct implementation at the point of care (van Beek et al, 2019). In addition, since the current models are based upon rich, but often more selected, datasets derived from clinical trials they might require further adjustments to represent the parameter distributions of the entire patient population (Keizer et al, 2018).…”