Pharmacokinetic and pharmacodynamic considerations of rifamycin antibiotics for the treatment of tuberculosis

KEYWORDS:

• Rifamycins
• pharmacokinetics
• pharmacodynamics
• tuberculosis

1. Introduction

Tuberculosis (TB) treatment was revolutionized in 1965 with the discovery of rifamycins by Sensi et al. [1]. With the introduction of rifamycins, treatment duration drastically decreased from 18 to 6 months [2]. Rifamycins, which include rifampicin, rifabutin, and rifapentine, are bactericidal DNA-dependent RNA polymerase inhibitors, that inhibit RNA synthesis of several microorganisms including Mycobacterium tuberculosis. Rifamycins contribute bactericidal activity toward the persisters, which are metabolically less active – up to now, this sterilizing activity is unparalleled [3]. Rifamycins remain the cornerstone of TB treatment, however, are limited by their low barrier for resistance, particularly when its concentrations or concentrations of companion drugs are low at the site of infection [4]. In this review, we discuss the pharmacokinetic and pharmacodynamic characteristics of rifamycins and their implications on clinical use in patients with TB.

2. General pharmacokinetic considerations

With repeated dosing, rifamycins induce their own clearance through the liver (autoinduction) via their effects on metabolizing enzymes. This leads to a decrease in rifamycin exposure during the first 6–14 days of oral administration. This effect is seen with all three rifamycins where autoinduction results in up to 45% decrease in exposure. Rifampicin and rifabutin are about 80% protein bound to albumin (rifapentine is >95% protein bound) and distribution occurs intra- and extracellularly due to their polar nature with 60–90% excretion in feces and urine (Table 1).

Table 1. Summary of pharmacokinetic and pharmacodynamic parameters of rifamycins.

Drug	Cmax (mg/mL)	T1/2 (hours)	Protein Binding %	CYP450 Enzyme induction	MIC* mg/mL
Rifampicin	8–24	3–4	80	+++	0.125–0.25
Rifapentine	8–30	13–15	>95	++	0.01–0.05
Rifabutin	0.3–0.9	25–62	85	+	0.03–0.06

CYP450: cytochrome P450, T1/2: half-life, C_max: maximum concentration

*Minimum Inhibitory Concentration for susceptible strains of Mycobacteria.

2.1. Rifampicin

Rifampicin is absorbed at its maximum on an empty stomach with a C_max of 8–24 µg/mL usually achieved within 2–4 h using the standard 10mg/kg dose; the C_max of >8 µg/mL is used clinically as a target for therapeutic drug monitoring, though it is not evidence-based. Despite the short half-life of rifampicin, it has a long post-antibiotic effect through its active metabolites. Rifampicin PK is highly variable, and the drug displays Michaelis-Menten PK. Patients with HIV infection may have reduced serum concentrations of rifampicin, particularly in the setting of severe immunocompromised [5]. Patients who are underweight also have up to 30% reduction in rifampicin exposures, related to malnutrition coupled with low dosing when mg/kg-based dosing algorithms are applied [6,7]. In our opinion, higher rifampicin dosing should routinely be applied in low-weight patients or those with advanced HIV. Of note, across patient populations, rifampicin concentrations are commonly ‘subtherapeutic’. While outcomes are still favorable in the majority of patients, there is growing knowledge about those harder-to-treat patients who do not have favorable outcomes. Rifampicin, in particular, is not robust to even small adherence lapses [8].

Several studies are evaluating higher doses of rifampicin as a strategy for TB treatment shortening. Rifampicin used at a dose of 20mg/kg did not demonstrate any difference in treatment outcomes compared to standard dosing; however, at 35mg/kg, faster sputum conversion in liquid medium was observed with no increased risk of toxicities compared to the standard dose used in the controls [9,10]. This suggests that bold, rather than modest, increases in rifampicin dosing will be needed for treatment shortening.

2.2. Rifapentine

Rifapentine has a longer half-life than rifampicin and is a slightly more potent antimycobacterial agent with a daily rifapentine dose of 5–10 mg/kg being equipotent to 30–40 mg/kg of rifampicin in murine models [11]. Unlike rifampicin and rifabutin, absorption is maximal when ingested following a fatty meal. Rifapentine demonstrates reduced exposure in HIV infected patients, blacks and men [12].

Its highly protein-bound nature (up to 97%) and modest penetration into caseum may contribute to the lower microbiologic activity seen in patients with large pulmonary cavities in phase 2 trials [13,14]. However, in murine models, although lower concentrations of rifapentine have been found in caseum, it sterilizes the lesions at concentrations similar to or lower than rifampicin, intimating its efficacy in cavitary disease [15].

Use of rifapentine has been evaluated as a means of treatment shortening when used with quinolones [16]. Though treatment shortening was not achieved, a six-month regimen that had once-weekly dosing of rifapentine with moxifloxacin during the continuation phase was effective. Higher daily doses of rifapentine of 1200 mg daily compared to the standard 600 mg have also been shown to lead to faster sputum clearance [12]. Daily 1200–1800 mg doses are under evaluation in 4-month arms in Phase 3 trials.

2.3. Rifabutin

Rifabutin achieves its peak concentrations after 3–4 h, and it has a large volume of distribution, with C_max values only reaching 0.3–0.9µg/ml with 300 mg daily dosing. It also has a longer half-life than rifampicin (25 h). Rifabutin’s role is generally as a substitute for rifampicin in patients on protease inhibitors and those with some rifampicin toxicities.

Initial recommendations of 150 mg thrice weekly when given with a boosted protease inhibitor were demonstrated to be insufficient, and now the recommended dose is 150 mg daily. Rifabutin is 50% renally cleared, and so a dose reduction of up to 50% is required in patients with renal impairment. Rifabutin also has a unique toxicity – uveitis – and cytopenias appear to be more common with this drug. Thus, rifabutin has a narrow therapeutic margin than the other rifamycins, and increasing the dose is not a viable treatment-shortening strategy.

3. Drug–drug interactions

All rifamycins are potent inducers of cytochrome (CYP) P450 enzymes as well as Phase 2 enzymes and transporters (including P-glycoprotein), creating challenges for patients taking concomitant medications. Rifabutin, however, is a less potent inducer of cytochrome P450 than rifampin. However, unlike rifampicin and rifapentine, it is a substrate of CYP3A4 and so sometimes is a victim of drug–drug interactions through this pathway.

Persons living with HIV bear a disproportionate share of TB disease and TB-related morbidity and mortality. Aside from pill burden and overlapping toxicities, drug–drug interactions make HIV-TB co-treatment a challenge. Rifampicin, the standard rifamycin used in first-line TB treatment, can be used with NRTI (including, in our opinion, tenofovir alafenamide [17]), without dose adjustments. It can also be used with efavirenz without dose modifications and with raltegravir or dolutegravir, provided the integrase inhibitor dose is doubled to twice daily. Rifampicin reduces cabotegravir and bictegravir concentrations significantly, more work is needed to determine whether or not these DDI can be overcome with dose modifications [18]. Rifampicin and rifapentine reduce exposure to almost all known protease inhibitors and co-administration is not recommended, with the exception of lopinavir/ritonavir, where the use of double doses can be considered, with close monitoring, in settings where rifabutin is not available [19,20].

Given that rifamycins have such potent and unique sterilizing activity, one might like to combine them with new drugs being developed for TB. However, this class of drugs’ induction effects makes this difficult to do. For example, the AUC of bedaquiline, which is being considered for the treatment of drug-sensitive TB, was markedly lowered in healthy volunteers receiving rifapentine or rifampicin [21]. The same is true of pretomanid [22]. Rifabutin, on the other hand, has less effect on bedaquiline pharmacokinetics and when used in combination, demonstrated higher potency in ex vivo experiments [21,23].

4. Pharmacodynamic considerations

Rifamycins demonstrate early bactericidal activity, achieving high kill-rates within the first hours – weeks on treatment and demonstrate concentration-dependent killing that correlates with C_max/minimum inhibitory concentration (MIC) ratio and area under the concentration-time (AUC): MIC ratio AUC/MIC. The C_max/MIC and AUC/MIC ratios are useful pharmacokinetic/pharmacodynamic (PK/PD) biomarkers of treatment response. Rifampicin has higher MICs than rifabutin and rifapentine for susceptible mycobacteria (Table 1). Interestingly, even with ‘susceptible’ strains (having MIC below the 1.0 mcg/mL breakpoint), higher MIC values were associated with increased risk of relapse [24]. Also, with the current standard dosing, even small lapses in adherence (skipping doses on weekends, for example) have a significant impact on the risk of unfavorable treatment outcomes [8]. These recent studies provide further clinical evidence that we are on the steep part of the dose–response curve with current rifampicin dosing.

5. Future considerations for dosing strategies

Given how unforgiving the current dose of rifampicin is – exposures are inadequate in the setting of slightly higher MIC, imperfect adherence, low weight or co-infection – and the fact that we are at the lower end of the dose–response curve, there is a strong rationale to increase the dose for all patients, not just in treatment-shortening scenarios. While the maximal effective and tolerated dose of rifampicin remains unknown, there still seems to be room for dose escalation to safely achieve maximum effect, especially in light of recent trials showing doses of up to 35 mg/kg daily appear safe. The relationship between dose and enzyme induction also remains uncharacterized, though recent studies suggest only a modest increase in auto-induction as rifampicin doses increase beyond 10 mg/kg daily [25]. Similarly, further work is needed to understand the relationship among dose, site-of-action PK, and emergence of resistance for the different rifamycins.

6. Expert opinion

Rifamycins have been around over half a century. Yet we still do not know how to use them optimally. While failure, relapse, and on-treatment acquired resistance remain uncommon, several emerging lines of evidence suggest that current drug dosing is too low for many patients. Target concentrations are ill-defined and may differ by the patient group. Rifamycins, specifically rifampicin and rifapentine, hold promise as treatment-shortening agents, but significant dose increases above currently licensed doses will be required to achieve that several enrolling or upcoming trials are exploring high-dose rifamycin-based regimens. Rifabutin has too narrow a therapeutic index to be tested at higher doses, but high-dose rifampicin and rifapentine carry the liability of high drug interaction potential. However, with care, these promising treatments can be evaluated in patients requiring concomitant medicines for co-morbidities (e.g. HIV) or TB itself (e.g. new chemical entities). Future studies need to continue to delve further into the maximum-tolerated dose of rifamycins and drug–drug interactions at higher doses.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

A reviewer of this manuscript has disclosed working with a group that does clinical trials in TB and work with rifapentine. Their group has received funding and material support from Sanofi and other manufacturers. This has never benefited any individuals in the group. Sanofi has not influenced choice of trials or presentation of results.

Peer reviewers on this manuscript have no other relevant financial or other relationships to disclose.

Acknowledgments

C Sekaggya-Wiltshire receives research support from NURTURE under NIH grant D43TW010132.

Additional information

Funding

This paper was not funded.