Introductions: Tuberculosis is the most common comorbidity among people living with HIV (PLWH) and remains a leading cause of HIV-related mortality worldwide. Early preventive treatment for latent tuberculosis infection (LTBI) is essential to prevent disease progression and reduce transmission. Recently, the one-month daily combination of rifapentine and isoniazid (1HP) regimen has demonstrated high efficacy and good tolerability in PLWH. However, reports have indicated that the bioavailability of rifapentine is reduced in HIV-positive individuals. As a potent inducer of CYP3A4, rifapentine may also affect the pharmacokinetics (PK) of ritonavir, a commonly co-administered antiretroviral agent primarily metabolized by CYP3A4 and widely used as a PK enhancer in HIV therapy. Despite these concerns, clinical data on rifapentine use in PLWH, particularly among the Chinese population, remain scarce. Model-informed strategies can be used to integrate existing knowledge to support evidence-based medication practice of rifapentine.
Aims: To 1) investigate ethnic differences in rifapentine PK via both “bottom-up” and “top-down” approaches; 2) integrate rifapentine PK from non-Chinese populations with HIV and LTBI to inform the development of a physiologically based pharmacokinetic (PBPK) model of Chinese patients with the same comorbidities; 3) assess the potential interaction between rifapentine and ritonavir in this population through PBPK simulations.
Methods: Rifapentine concentrations from Chinese adults with LTBI receiving the 1HP regimen (300 mg isoniazid plus 600 mg rifapentine once daily) were used to externally validate a published non-Chinese population pharmacokinetic (PopPK) model. Physicochemical properties, ADME parameters, and clinical data of rifapentine were collected from the literature to support stepwise development of PBPK models for both non-Chinese and Chinese adults. Both approaches were used to assess whether ethnic differences exist in the PK of rifapentine. Subsequently, assuming similar HIV effects on Chinese/non-Chinese patients, rifapentine PK profiles in Chinese with HIV were simulated from the published PopPK model to inform the development of a PBPK model for Chinese patients with HIV and LTBI. Finally, the potential interaction between rifapentine (750 mg once daily) and ritonavir (100 mg twice daily) was assessed through clinical trial simulations.
Results: Both the GoF and VPC plots showed that the non-Chinese rifapentine PopPK model adequately described the PK from Chinese LTBI patients receiving the 1HP regimen. Meanwhile, the developed PBPK models of rifapentine for both non-Chinese and Chinese adults performed similarly with comparable Cmin,ss (6.25 vs 8.09 mg/L), Cmax,ss (20.33 vs 19.44 mg/L) and AUCss (287.61 vs 325.03 h*mg/L) under 600 mg once daily. Furthermore, the simulation of Chinese PBPK also aligned with the prediction from non-Chinese PopPK model. All of these suggested that ethnic differences in rifapentine PK were negligible. Subsequently, the PBPK model for Chinese patients with HIV and LTBI was developed via calibrating the small intestinal transit time in the non-HIV PBPK model to describe the simulated rifapentine PK profiles from Chinese with HIV. This adjustment reflected the reduced bioavailability of rifapentine in HIV-positive individuals due to gastrointestinal reaction. Finally, the clinical trial simulations of PBPK models revealed that rifapentine reduced the Cmax,ss of ritonavir by approximately 30% under the given dosing regimen. Increasing the ritonavir dose to 150 mg twice daily was predicted to restore exposure levels comparable to those observed without rifapentine coadministration.
Conclusion: The integrated modeling strategy suggested that no ethnic differences existed in rifapentine PK between Chinese and non-Chinese populations and the ritonavir dose should be increased to 150 mg twice daily when co-administered with rifapentine. This study demonstrated the use of clinical pharmacometrics in evidence-based medication practice when limited data is available. In future, this model-informed medication practice needs to be re-assessed when more clinical information is collected.