Treatment Optimization in Patients Co-Infected with HIV and Mycobacterium tuberculosis Infections: Focus on Drug–Drug Interactions with Rifamycins
- First Online:
- Cite this article as:
- Regazzi, M., Carvalho, A.C., Villani, P. et al. Clin Pharmacokinet (2014) 53: 489. doi:10.1007/s40262-014-0144-3
- 443 Views
Tuberculosis (TB) and HIV continue to be two of the major causes of morbidity and mortality in the world, and together are responsible for the death of millions of people every year. There is overwhelming evidence to recommend that patients with TB and HIV co-infection should receive concomitant therapy of both conditions regardless of the CD4 cell count level. The principles for treatment of active TB disease in HIV-infected patients are the same as in HIV-uninfected patients. However, concomitant treatment of both conditions is complex, mainly due to significant drug–drug interactions between TB and HIV drugs. Rifamycins are potent inducers of the cytochrome P450 (CYP) pathway, leading to reduced (frequently sub-therapeutic) plasma concentrations of some classes of antiretrovirals. Rifampicin is also an inducer of the uridine diphosphate glucuronosyltransferase (UGT) 1A1 enzymes and interferes with drugs, such as integrase inhibitors, that are metabolized by this metabolic pathway. Rifampicin is also an inducer of the adenosine triphosphate (ATP) binding cassette transporter P-glycoprotein, which may also lead to decreased bioavailability of concomitantly administered antiretrovirals. On the other side, rifabutin concentrations are affected by the antiretrovirals that induce or inhibit CYP enzymes. In this review, the pharmacokinetic interactions, and the relevant clinical consequences, of the rifamycins—rifampicin, rifabutin, and rifapentine—with antiretroviral drugs are reviewed and discussed. A rifampicin-based antitubercular regimen and an efavirenz-based antiretroviral regimen is the first choice for treatment of TB/HIV co-infected patients. Rifabutin is the preferred rifamycin to use in HIV-infected patients on a protease inhibitor-based regimen; however, the dose of rifabutin needs to be reduced to 150 mg daily. More information is required to select optimal treatment regimens for TB/HIV co-infected patients whenever efavirenz cannot be used and rifabutin is not available. Despite significant pharmacokinetic interactions between antiretrovirals and antitubercular drugs, adequate clinical response of both infections can be achieved with an acceptable safety profile when the pharmacological characteristics of drugs are known, and appropriate combination regimens, dosing, and timing of initiation are used. However, more clinical research is needed for newer drugs, such as rifapentine and the recently introduced integrase inhibitor antiretrovirals, and for specific population groups, such as children, pregnant women, and patients affected by multidrug-resistant TB.
Tuberculosis (TB) and HIV are major causes of morbidity and mortality worldwide, and together are responsible for the death of millions of people every year.
In 2012, the World Health Organization (WHO) estimated that there were 8.6 (range 8.3–9.0) million new cases of TB worldwide, including 1.1 (range 1.0–1.2) million cases of TB/HIV co-infection . Of the 2.8 million new TB cases that were screened for HIV in 2012, 20 % tested HIV positive . Globally, an estimated 35.3 (range 32.2–38.8) million people were living with HIV in 2012, a continuously rising number as a result of the increasing number of people who are receiving life-saving antiretroviral therapy (ART) .
TB remains the leading cause of death among people living with HIV. Since 2004, TB-related deaths among people living with HIV have declined by 36 % worldwide, but in 2012 there was an unacceptable toll of 320,000 lives lost due to HIV-associated TB .
TB is a curable condition in virtually all affected people. The objective of TB treatment is to cure the patient, restoring quality of life and labor productivity, and to prevent relapse, long-term physical sequelae, and death. Furthermore, efficient TB treatment reduces the risk of transmission of Mycobacterium tuberculosis to others and prevents the emergence of resistant strains . The WHO recommends that any newly diagnosed TB patient be treated with a standard regimen consisting, for the initial intensive phase of 2 months, of four drugs (namely rifampicin, isoniazid, pyrazinamide, and ethambutol), followed by a continuation phase of two drugs (rifampicin and isoniazid) for a period of at least 4 months . The use of fixed-dose combinations of antituberculars is strongly recommended, with the aim of simplifying prescription on the health sector side and adherence on the patient’s side [3, 4].
ART is highly effective in reducing the risk of progression of HIV infection, preventing the development of AIDS and death, as well as reducing viral transmission. According to the US Department of Health and Human Services, the regimens of choice for antiretroviral-naïve patients include  (a) efavirenz/tenofovir disoproxil fumarate/emtricitabine; (b) ritonavir-boosted atazanavir + tenofovir disoproxil fumarate/emtricitabine; (c) ritonavir-boosted darunavir + tenofovir disoproxil fumarate/emtricitabine; and (d) raltegravir + tenofovir disoproxil fumarate/emtricitabine. The selection of a regimen should be individualized on the basis of virologic efficacy, toxicity, pill burden, dosing frequency, drug–drug interaction potential, resistance testing results, and co-morbid conditions . The WHO recommends a public health approach for ART, and specifically targets resource-constrained countries and settings with limitations in human resources, where the majority of people needing ART live. In its recent consolidated antiretroviral guidelines, once-daily regimens comprising a non-thymidine nucleoside reverse transcriptase inhibitor (NRTI) backbone (tenofovir disoproxil fumaratein association with emtricitabine or lamivudine) and one non-NRTI (NNRTI) (usually efavirenz) are the preferred choices in adults, adolescents, and children older than 3 years . For children younger than 3 years, a protease inhibitor (PI)-based regimen is the preferred approach .
The most relevant clinical consequence of the interaction between the HIV virus and M. tuberculosis is the HIV-driven increase in the risk of progression from M. tuberculosis infection to disease. Hence, it is not surprising that ART lowers the incidence of TB . In addition, HIV infection increases the risk of a poor TB treatment outcome . Ideally, TB treatment in TB/HIV co-infected patients could achieve success rates similar to those of HIV-uninfected patients, provided that treatment is started early, and that HIV infection is concomitantly treated. Randomized clinical trials show that adding ART to TB treatment in co-infected patients improve treatment outcome and reduce mortality , regardless of the level of the CD4 cell count [10–12]. The benefits of ART extend to patients with multidrug-resistant (MDR)-TB and HIV co-infection .
Although the number of HIV-positive TB patients on ART has grown from a very low level in 2004 to reach 0.3 million in 2012, the target of 100 % co-infected patients receiving concomitant treatment for both infections  is still out of reach: among TB patients notified in 2012 who had a documented HIV-positive test result, 57 % were on ART globally . These figures witness the challenges of combining TB and HIV treatment.
In the present manuscript we review and synthesize the literature on the interactions between rifamycins and antiretroviral drugs in order to give practical recommendations on how to optimize antiretroviral and antitubercular therapies in patients co-infected with HIV and M. tuberculosis infections.
2 Search Strategy and Selection Criteria
We initially searched for original articles or reviews published in the last 5 years through MEDLINE under the medical subject heading major topics “Antitubercular Agents” and “Anti-Retroviral Agents” with the following subheadings: “metabolism,” “pharmacokinetics,” “pharmacology,” and “therapy.” An initial list of 153 articles was retrieved, we read the abstracts, and chose the manuscripts related to drug interactions between any rifamycin and any antiretroviral drug. Other articles were selected from the references of initially selected works. There were no restrictions on language of publication a priori.
Manuals and guidelines on TB and HIV therapy were searched for on the official websites of the WHO, the International Union Against Tuberculosis and Lung Diseases (The Union), EuroTB and European Centre for Disease Prevention and Control (ECDC), Centers for Disease Prevention and Control (CDC), and AIDSinfo.
The search was updated on 31 January 2014.
3 Treatment Principles in Patients with HIV and Mycobacterium tuberculosis Co-Infection
Concomitant treatment of both the conditions of TB and HIV is complex due to significant drug–drug interactions between antiretrovirals and antitubercular drugs, and the overlapping toxicity [15, 16]. Despite these challenges, adequate clinical response to both infections can be achieved with an acceptable safety and efficacy profile when the pharmacological characteristics of available drugs are known and used to guide treatment choice, dosing, and timing of initiation [5, 6, 14].
The principles of TB treatment in HIV-infected patients are the same as in HIV-uninfected ones. All identified patients should start a standard TB regimen as soon as the diagnosis is made. There is limited evidence that a treatment duration of ≥8 months may improve outcomes in co-infected patients , but 6 months of therapy is recommended by international guidelines [3, 5].
Although all currently available rifamycins (rifampicin, rifabutin, and rifapentine) have clinically significant interactions with a number of antiretroviral drugs, a rifamycin should be included in the TB regimen for patients receiving ART, because rifamycin-lacking regimens have significantly higher failure/relapse rates (11.7 vs. 6 %; p = 0.002) , and would require a substantially longer treatment duration (12–18 months).
All HIV-infected patients with active TB should receive ART during TB treatment, as ART reduces mortality in HIV-infected TB patients . The timing of antiretroviral initiation during antitubercular treatment varies in different recommendations. The US guidelines recommend an early start (within 2 weeks of starting antitubercular treatment) in patients with CD4 counts <50 cells/mm3 . In patients with CD4 counts above that threshold, ART should be initiated within 2–12 weeks of starting antitubercular treatment . WHO recommendations do not discriminate on the basis of the CD4 cell count: ART should be started as soon as possible within the first 8 weeks of antitubercular treatment [6, 14]. However, it is specified that TB/HIV co-infected patients with profound immunosuppression (e.g., CD4 counts <50 cells/mm3) should receive ART immediately within the first 2 weeks of antitubercular treatment [6, 14].
Mortality rates can be exceedingly high in HIV-infected individuals with MDR-TB . Among HIV-infected patients using second-line TB drugs, there is evidence of a lower risk of death and a higher likelihood of TB cure in patients who received concomitant ART than in those who did not . In these patients, ART should be initiated as soon as possible (within the first 8 weeks) following antitubercular treatment initiation, irrespective of CD4 cell count .
Whenever TB develops among people living with HIV who are already taking ART, ART should be continued (if there is no evidence of failure), and possibly modified if required according to anticipated drug interactions with antituberculars.
4 Rifamycins: Pharmacokinetics, Metabolism, and Bases of Drug Interactions with Antiretrovirals
Rifampicin is the most commonly used rifamycin. The recommended dose of rifampicin for treatment of active TB is 10 mg/kg/day (usual dose 600 mg) or 10 mg/kg three times a week (usual dose 600 mg) . The drug is readily absorbed from the gastrointestinal tract; within 2–4 h after ingestion of a dose of 600 mg, maximum (peak) plasma concentrations (Cmax) may reach 7–10 mg/L . The absorption of rifampicin is reduced by about 30 % when the drug is ingested with food. Rifampicin is widely distributed throughout the body. It is present in effective concentrations in many organs and body fluids, including cerebrospinal fluid. Rifampicin is about 80 % protein bound and most of the unbound fraction is not ionized and therefore is diffused freely in tissues. Its half-life ranges from 2 to 5 h. Approximately 85 % of the drug is metabolized in the liver; the drug and its metabolites are mainly excreted in bile (60–65 %) and eliminated in stools. About 6–15 % of the drug is excreted in non-metabolized form and is reabsorbed in the intestine (entero-hepatic recirculation), progressively increasing the serum concentrations of the drug .
Rifampicin is a potent inducer of the cytochrome P450 (CYP) system, and can therefore increase the metabolism of numerous drugs that are partially or completely metabolized by CYP when these drugs are administered concomitantly with rifampicin [24, 25]. The inductive effect of rifampicin is most marked on the CYP3A and CYP2C subfamilies, which account for more than 80 % of the CYP isoenzymes [24, 25]. An in vitro study found that rifampicin induced the expression of CYP2B6 by 8.8-fold and CYP3A4 by 55.1-fold .
The induction effect of rifampicin on the CYP system is concentration dependent . The most frequently used dose of rifampicin in clinical interaction studies is 600 mg, although CYP3A4 seems to be strongly induced already at 450 mg daily . The duration of the rifampicin treatment is of great importance. From studies on the effect of rifampicin on pharmacokinetics of verapamil, which is metabolized by CYP3A4 and CYP2C9, it was estimated that full induction of these enzymes was reached after 1 week of treatment with rifampicin . In a study by Magnusson et al. [28, 29], the half-life of induction (50 % of the maximal induction) of CYP3A4 was found to be 70 h and 90 % of the fully induced enzyme level was attained after about 10–14 days.
Rifampicin is also an inducer of the uridine diphosphate glucuronosyltransferase (UGT) 1A1 enzymes and interferes with drugs, such as the integrase inhibitors, that are metabolized by glucuronidation via the UGT1A1 isozymes .
Finally, rifampicin induces the adenosine triphosphate (ATP) binding cassette transporter P-glycoprotein, which may lead to decreased bioavailability of concomitantly administered drugs . The co-localization of P-glycoprotein and CYP enzymes in enterocytes, hepatocytes, and renal tubular cells may enhance the effects of rifampicin on common substrates, causing more extensive pre-systemic metabolism and accelerated drug elimination .
Several studies have shown a lack of effect of drugs that are either inducers (e.g., efavirenz) or inhibitors (e.g., ritonavir) of CYP3A on rifampicin, in agreement with the notion that rifampicin is not a substrate of CYP3A.
Rifabutin is structurally similar to rifampicin, but there are important pharmacokinetic differences between the two drugs. The physicochemical property that contributes most to the pharmacokinetic divergence of these drugs is the difference in their lipid solubility. Other factors that distinguish the pharmacodynamics of rifabutin from those of rifampicin include the antimicrobial activities of its two major metabolites, 25-O-desacetyl rifabutin (with activity almost equivalent to that of rifabutin) and 31-OH rifabutin (with an activity equivalent to approximately 10 % of that of rifabutin) .
In HIV-infected subjects who received oral capsules of rifabutin, the bioavailability of a single dose was 20 % . Rifabutin plasma concentrations are best described by a two-compartment open model, with a terminal elimination half-life of 36–45 h. The low bioavailability of rifabutin is probably due to high biliary excretion and substantial presystemic metabolism. Food decreases the rate, but not the extent, of absorption of rifabutin in the capsule formulation .
Rifabutin is a moderate inducer of the CYP3A4 enzymes and therefore may reduce the plasma concentrations of drugs that are principally metabolized by those enzymes. As an inducer, rifabutin is 40 % as potent as rifampicin .
Unlike rifampicin, rifabutin is metabolized by the CYP3A. Drugs that induce or inhibit CYP enzymes may, therefore, decrease or increase, respectively, concentrations of concurrently administered rifabutin . Bidirectional drug–drug interaction is anticipated in the case of rifabutin and antiretroviral coadministration.
The length of time that a rifamycin is above a certain threshold concentration is the most important index associated with M. tuberculosis killing and the prevention of resistance. Thus, one of the strategic goals of the WHO, the pharmaceutical industry, and research scientists has been to develop rifampicin congeners with long half-lives that would make intermittent dosing more effective. Rifapentine is characterized by a longer terminal elimination half-life of 13–14 h, compared with 2–3 h for rifampcin. The rationale for the administration of rifapentine at doses of 600 mg twice weekly or even once weekly in the treatment of TB is provided by the “in vitro” pharmacodynamics model of TB demonstrating that, despite the short half-life of rifampicin, its long post-antibiotic effect prevents regrowth during the entire 1-week dosing interval, so that cycles of killing and regrowth are not encountered. Moreover, rifapentine has a greater ability to penetrate macrophages, achieving a four- to fivefold greater ratio of intracellular accumulation than rifampicin, resulting in marked reduction of the mycobacterial burden that is maintained over a 4-week period after once-weekly exposures of infected macrophages in an experimental model of intracellular infection .
After oral administration of a single 600 mg oral dose, rifapentine is absorbed slowly from the gastrointestinal tract, reaching Cmax values in plasma of about 15 μg/mL within 4–5 h . Ingestion of the dose with a high-fat meal increases the Cmax and area under the plasma concentration–time curve (AUC) by 43–50 % over fasting values [36, 37]. Rifapentine is metabolized mostly by the liver and is excreted predominantly (70 %) in feces. The drug is metabolized by hydrolysis and deacetylation to 25-O-desacetylrifapentine, which is microbiologically active, contributing 38 % of the drug’s overall activity [36, 38]. The metabolite reaches Cmax in 14.4–17.8 h, and the mean elimination half-life is 13.3–24.3 h.
Rifapentine, like other rifamycins, induces the CYP enzymes—specifically, the CYP3A4, CYP2C8, and CYP2C9 isoenzymes . The relative potency as inducers is as follows: rifampicin, 1; rifapentine, 0.85; and rifabutin, 0.40. Therefore, interaction of rifapentine and drugs that are metabolized by these enzymes are anticipated to be clinically significant. One study suggests that the maximal induction of these enzymes occurs within 4 days after receipt of the first dose and returns to the baseline level within 14 days after rifapentine is discontinued ; however, despite limited data, there is no reason to expect rifapentine to induce enzymes faster than rifampicin (which requires at least 7 days) . Any drug known to have interactions with rifampicin (e.g., oral contraceptives, methadone, β-blockers, benzodiazepines, and oral anticoagulants) should be considered to have similar interactions with rifapentine, unless proven otherwise.
Rifapentine is the most recently approved of the rifamycins, licensed only in the USA and Canada, and its use in clinical settings is still limited. An increased risk of developing rifamycin resistance was demonstrated while treating TB/HIV co-infected patients with an intermittent (once weekly) rifapentine regimen; this effect was independent from the use of antiretroviral drugs. Acquired rifamycin resistance had not been demonstrated in HIV-uninfected patients . It has been speculated that the long half-life of rifapentine and the very short half-life of isoniazid likely leads to exposure to a single drug and to the development of resistance . On the basis of these findings, a drug with a duration of activity longer than that of isoniazid may be needed to ensure effective once-weekly therapy with rifapentine in HIV-infected patients .
Currently, rifapentine is not recommended in HIV-infected patients receiving ART for treatment of latent TB infection (LTBI) or active TB, unless in the context of a clinical trial . This has probably discouraged any systematic attempt to investigate the magnitude of the enzyme induction effect of rifapentine on the metabolism of antiretroviral drugs. A manufacturer-sponsored study evaluating the effects of both once-weekly and daily rifapentine at the dose recommended for the treatment of LTBI on efavirenz is underway .
Due to the paucity of available data on drug–drug interactions, rifapentine is not further considered in this review.
5 Rifamycins and Antiretroviral Interactions, and Relevant Clinical Implications
5.1 Nucleoside Reverse Transcriptase Inhibitors
When considering drug interactions, the parent NRTI concentration in plasma may be important for toxicity, but the pharmacokinetic parameter most closely associated with the activity of the nucleoside analogs is the intracellular concentration of the active form, the triphosphate derivative, and there is no close relationship between serum concentrations of the parent drug and the intracellular concentration of the triphosphate metabolite . Formal drug interaction studies have not been performed between rifamycins and all of the currently approved nucleoside analogs, but significant interactions are unlikely because NRTIs are not metabolized by the liver and their concentrations are not affected by medications that induce or inhibit hepatic CYP enzymes. One exception is zidovudine, a drug metabolized by glucuronidation via the UGT1A1 isozymes, which is induced by rifampicin. The interaction between zidovudine and rifampicin results in a 47 % decrease in the AUC of zidovudine . Nevertheless, no dose adjustment is recommended, presumably because the correlation between serum and intracellular concentrations of zidovudine remains poorly defined.
5.2 Non-Nucleoside Reverse Transcriptase Inhibitors
Drug interactions between rifamycins and non-nucleoside reverse transcriptase inhibitors, and recommendations for use
Interactions in pharmacokinetic studies
Clinical consequences of the interaction
Efavirenz median Cmax, Ctrough, and AUC ↓ by 24, 18, and 10 %
Efavirenz efficacy is similar in HIV patients with or without concomitant rifampicin treatment
An efavirenz-based regimen is superior to a nevirapine-based regimen
Efavirenz AUC reduction reversed by efavirenz dose increase to 800 mg
Nevirapine Cmax ↓ 36–39 %, AUC12 ↓ 31–36 %, and Cmin ↓ 21–39 %
Nevirapine efficacy is similar in HIV patients with or without concomitant rifampicin treatment
Alternative regimen in resource-constrained countries in which PI-based ART cannot be used due to the unavailability of rifabutin, and whenever efavirenz cannot be used 
Contraindicated ; if it must be used, no lead-in dosing recommended and close monitoring of adherence and plasma HIV RNA warranted
Consider therapeutic drug monitoring, if available 
Sub-therapeutic concentrations of nevirapine reported in 21–38 % of treated patients
Inferior rate of virologic success compared with efavirenz
No change in rifampin exposure
Efficacy of nevirapine compared with efavirenz:
Rilpivirine AUC24, Cmax, and Cmin ↓ by 80, 69, and 89 %, respectively
No change in efavirenz exposure
Association possible, increasing rifabutin dose at 450–600 mg od. Clinical utility questionable
Rifabutin Cmax ↓ by 29 % and the AUC24 ↓ by 37 %
Rifabutin reduction reverted by increasing rifabutin dose to 600 mg twice weekly
nevirapine Cmin ↓ 16 %. Rifabutin AUC ↑ 17 %, 25-O-desacetyl rifabutin AUC ↑ 24 %
Association possible, at standard drug doses, but clinical utility questionable
Rilpivirine AUC24, Cmax, and Cmin ↓ by 46, 35, and 49 %, respectively
Etravirine AUC ↓ 37 %. Rifabutin AUC ↓ 17 %
Association possible, at standard drug doses, provided that no additional potent cytochrome P450 inducers are concomitantly used
The range of acceptable efavirenz plasma concentrations after 12 h (C12) is currently proposed to be 1–4 μg/mL . Concentrations below the level of 1.0 μg/mL have been associated with a more than twofold increase in the risk of treatment failure [47, 48]. Efavirenz is largely cleared by CYP2B6 and, to a lesser extent, by CYP3A4. However, efavirenz concentrations in the presence of rifampicin vary widely, partly due to the effect of pharmacogenetic variability in CYP2B6 alleles . One study showed that polymorphisms of CYP2B6 genotypes determine variations in efavirenz and nevirapine plasma concentrations that are much higher than those due to the concomitant administration of rifampicin in TB/HIV co-infected Thai adults: the mean 12-h post-dose plasma efavirenz concentration in patients with a TT genotype at weeks 6 and 12 of ART and 1 month after rifampicin discontinuation (10.97 ± 2.32, 13.62 ± 4.21, and 8.48 ± 1.30 mg/L, respectively) were significantly higher than those with GT (3.43 ± 0.29, 3.35 ± 0.27, and 3.21 ± 0.22 mg/L, respectively) (p < 0.0001) or GG genotypes (2.88 ± 0.33, 2.45 ± 0.26, and 2.08 ± 0.16 mg/L, respectively) (p < 0.0001) . Patients with a T/T CYP2B6 polymorphism, which is more common in sub-Saharan African and African American populations than in Europeans and European Americans, who receive rifampicin and efavirenz treatment are more than 30 times more likely to achieve efavirenz Cmax values above 4 μg/mL than patients with G/G CYP2B6 polymorphism . Efavirenz Cmax values above 4 μg/mL have been associated with a higher risk of toxicity of the central nervous system in some studies [47, 48, 51].
Earlier studies showed that rifampicin reduces the median Cmax, the measured concentration at the end of a dosing interval (trough plasma concentration; Ctrough), and the integral of the AUC of efavirenz by 24, 18, and 10 %, respectively [52, 53]. More recent pharmacokinetic studies in Europe , India , and Africa [48, 52] demonstrated no significant decrease in efavirenz concentrations in the presence of rifampicin-based TB treatment.
There is substantial evidence from clinical studies on the efficacy and safety of efavirenz in combination with currently recommended rifampicin-based antitubercular regimens. Randomized clinical studies have shown that efavirenz efficacy is similar in HIV patients with or without concomitant rifampicin treatment , and is equivalent or superior to nevirapine (see below); however, no adequately powered randomized trial has been published comparing an efavirenz- and a PI-based ART regimen in TB/HIV patients .
There is evidence that the approximate 22 % reduction in efavirenz AUC when coadministered with rifampicin can be converted with an efavirenz dose increase to 800 mg in patients weighing >60 kg [53, 54]. Other studies have shown that due to genetic ethnic variability and differences in average body mass index, no dose adjustment of efavirenz is needed in high TB/HIV burden areas in Africa [47, 58] and Asia . The higher efavirenz dose may be considered on an individual basis (with a weight threshold) in Europe and the USA.
Nevirapine is extensively metabolized via CYP oxidative metabolism, mainly via the CYP3A family. It is also an inducer of CYP enzymes and may result in lower plasma concentrations of other drugs that are extensively metabolized by CYP3A or CYP2B. Because of the auto-induction effect of nevirapine, there is an approximately 1.5- to 2-fold increase in the apparent oral clearance of the drug after 2–4 weeks of dosing with 200–400 mg daily.
Exposure to rifampicin did not significantly differ between patients receiving and not receiving nevirapine . However, pharmacokinetic studies in TB/HIV co-infected patients showed 31–36 % reduction in serum nevirapine AUC from time zero to 12 h (AUC12), 36–39 % reduction in Cmax, and 21–32 % decrease in the nevirapine minimum steady-state plasma concentration (Cmin) before and after rifampicin [60, 61] (Table 1). Other studies confirmed a reduction of 37 %  in the mean concentrations and of 39 % in the Cmin of nevirapine  attributable to rifampicin. Sub-therapeutic concentrations of nevirapine were reported in 21–38 % of co-infected patients on rifampicin-based antitubercular treatment in Thailand  and South Africa , although the extent of the reduction of AUC of nevirapine during concomitant rifampicin treatment substantially decreases over time, possibly limiting its clinical impact . The clinical consequences of the interaction are still debated. No risk of increased virologic failure was observed in uncontrolled studies in Malawi [66, 67] and Thailand . However, nevirapine achieved an inferior rate of virologic success compared with efavirenz in another large South African study: the proportion of persons with virologic failure doubled when comparing nevirapine initiated during rifampicin-containing antitubercular therapy and nevirapine initiated in persons who did not have TB . In the first large randomized trial comparing an efavirenz- versus a nevirapine-based regimen for TB/HIV co-infected patients receiving rifampicin, C12 values below the recommended threshold were observed in 3.1 and 21.3 % of the cases on efavirenz and nevirapine treatment, respectively. Despite the fact that multivariate analysis revealed that patients with low C12 values were 3.6 times more likely to develop all-cause treatment failure, the proportion of patients with virologic failure was similar in the two groups . The rate of persistent viral suppression was also similar in a 4-year follow-up study conducted in two groups of 70 TB/HIV co-infected patients treated with rifampicin and nevirapine, and compared with 70 HIV monoinfected patients treated with nevirapine . A second pivotal open-label, non-inferiority, randomized controlled trial conducted at three sites in southern India compared efavirenz-based ART with nevirapine 400 mg once daily-based ART in TB/HIV co-infected patients under rifampicin treatment . The trial was halted by the data and safety monitoring board at the second interim analysis because the nevirapine arm was inferior and associated with more frequent virologic failure and death. These data were not confirmed in a recent randomized open-label trial conducted in India to compare efavirenz and nevirapine (200 mg twice daily) in TB/HIV patients who were followed-up for 2 years . No significant difference was found between the groups in clinical, immunological, or virologic failure rates. The overall mortality was 17 % with no significant difference between the two groups. Except for the lead-in period of 14 days, the mean nevirapine concentration remained above 3 mg/L. Finally, a multicenter, open-label, randomized, non-inferiority trial comparing efavirenz- and nevirapine-based regimens in association with rifampicin was conducted at three health centers in Maputo, Mozambique. Although non-inferiority of the nevirapine-regimen could not be demonstrated, the intention-to-treat analysis for virologic suppression at week 48 showed that this outcome was reached by a similar proportion of patients (64.6 and 69.8 % for nevirapine and efavirenz, respectively—one-sided 95 % CI of the difference of efficacy 11.7 %) . There is evidence against the use of nevirapine 600 mg/day in patients receiving rifampicin due to an increased risk of hypersensitivity reactions .
Rifabutin has fewer interactions with NNRTI drugs, but is affected by the induction effect of these drugs on its own metabolic pathway. In a study of healthy volunteers, rifabutin had little or no effect on efavirenz concentrations at a standard dose of 300 mg daily . In the same study, the mean Cmax was decreased by 29 % by efavirenz, and the mean AUC decreased by 37 % . A small pharmacokinetic study of 15 patients with TB/HIV co-infection who were treated with an increased dose of rifabutin 600 mg twice weekly and efavirenz 600 mg daily showed that the mean efavirenz AUC from time zero to 24 h (AUC24) was 10 % higher than that in 35 historical subjects with HIV infection who were not taking rifabutin . The serum concentration of nevirapine decreased by 16 % by the concomitant administration of rifabutin [15, 75].
Data on drug interactions of rifampicin and rifabutin with the newer NNRTIs rilpivirine or etravirine are limited. Rilpivirine interaction with rifamycins was investigated in two pharmacokinetic studies. Rifampicin dosed at 600 mg once daily together with rilpivirine 150 mg once daily was found to reduce rilpivirine AUC24, Cmax, and Cmin by 80, 69, and 89 %, respectively, when given to 16 HIV-negative volunteers for 7 days . In an 11-day study of rilpivirine 150 mg once daily and rifabutin 300 mg once daily in 18 HIV-negative volunteers, the AUC24, Cmax, and Cmin of rilpivirine were reduced by 46, 35, and 49 %, respectively . No clinical study investigated the association of rifampicin or rifabutin and rilpivirine .
Studies of the association between rifampicin and etravirine are unavailable, but pharmacokinetic interactions are clearly anticipated . Drug–drug interactions between etravirine and rifabutin were examined in one randomized, open-label, two-period, crossover trial in HIV-negative, healthy volunteers. Rifabutin decreased etravirine exposure by 37 %. This study concluded that etravirine can be coadministered with 300 mg of rifabutin once daily in the absence of an additional potent CYP inducer .
5.3 Protease Inhibitors
Drug interactions between rifamycins and protease inhibitors, and recommendations for use
Interactions in pharmacokinetic studies
Clinical consequences of the interaction
LPV/r AUC12 and Ctrough ↓ 75 and 99 %, respectively
High-dose LPV/r (800/200 mg bid or 400/400 mg bid) determined high rate of AEs and treatment discontinuation
High-dose LPV/r (800/200 mg bid or 400/400 mg bid) could be used , but requires close clinical and laboratory monitoring for possible hepatotoxicity, and treatment-limiting AEs very common
Increasing LPV/r to 800/200 mg bid or 400/400 mg bid reverts the concentration ↓
ATZ/r Cmin ↓ by 97 %
Increasing atazanavir to 400 mg bid determined high rate of AEs and discontinuation
Rifabutin 150 mg qod: 90–100 % with sub-therapeutic Cmax values of rifabutin, and Cmax and AUC ↓ by 36 and 26 %, respectively
Preferred alternative for combination therapy whenever efavirenz cannot be used
LPV/r dose unchanged. Rifabutin dose 150 mg od, with close monitoring for rifabutin-associated AEs
Rifabutin 150 mg qod steady-state concentration ↓ by 44 %, and rifabutin 150 mg od ↑ by 32 % compared with rifabutin 300 mg alone
LPV/r concentrations unaffected
Rifabutin 150 mg twice a week: rifabutin Cmax, AUC, and Cmin ↑ by 149, 48, and 40 %, respectively; 25-O-desacteyl rifabutin Cmax, AUC, and Cmin ↑ by 6.77-, 9.90-, and 10.45-fold, respectively
Association possible. ATZ/r dose unchanged
Rifabutin dose 150 mg od, with close monitoring for rifabutin-associated AEs
Rifabutin 150 mg qod: 44 % of subjects below Cmax threshold for efficacy
Rifabutin 150 mg qod: rifabutin Cmin ↑ by 64 % and Cmax ↓ by 28 %, while the AUC of 25-O-desacetylrifabutin ↑ by 881 %
Association possible. DRV/r dose unchanged
Rifabutin dose 150 mg od, with close monitoring for rifabutin-associated AEs
Rifabutin 150 mg qod: rifabutin AUC48 unchanged, Cmax ↓ by 14 %, while AUC48 and Cmax of 25-O-desacetylrifabutin ↑ by 11- and 6-fold, respectively
Association possible. FPV/r dose unchanged
Rifabutin dose 150 mg od, with close monitoring for rifabutin-associated AEs
Rifabutin 150 mg single dose: rifabutin AUC, Cmax, and C12 ↑ by 190, 70, and 114 %, respectively
25-O-desacetyl-rifabutin AUC, Cmax, and C12 ↑ by 333, 86, and 176 %, respectively
Association possible. TPV/r dose unchanged
Rifabutin dose 150 mg od, with close monitoring for rifabutin-associated AEs
A suggested strategy to overcome the dramatic reduction of serum concentrations of PIs determined by the concomitant administration of rifampicin is to increase the doses of the PI. There is very little evidence from clinical studies concerning this strategy. Pharmacokinetic data for 19 healthy subjects  and 21 HIV-infected individuals who did not have TB  showed that doubling the dose of lopinavir/ritonavir or adding additional ritonavir (so that lopinavir:ritonavir = 1:1) can overcome the inducing effect of rifampicin. However, subsequent clinical studies in volunteers receiving rifampicin with high-dose saquinavir , atazanavir , and lopinavir  were prematurely interrupted due to the high rate of liver toxicity. One recent study conducted in 18 adult patients on a high-dose lopinavir/ritonavir-based antiretroviral regimen (half on a 400/400 mg twice-daily dose and half on a 800/200 mg twice-daily dose) and rifampicin-based antitubercular therapy confirmed that the majority (77 %) reached lopinavir Ctrough above the recommended concentration of 1 mg/L and showed that none developed grade 3 or 4 liver toxicity . The authors suggested that the toxicity and tolerability of the association may be improved in TB/HIV co-infected patients compared with healthy subjects, especially if antitubercular treatment is given to patients who are already established on HIV therapy. However, this was not confirmed by a retrospective study of 15 patients receiving the lopinavir/ritonavir 400/400 mg twice-daily super-boosted regimen in South Africa, which showed a significantly higher probability of liver toxicity and treatment discontinuation than the standard lopinavir/ritonavir regimen .
Rifabutin is the preferred rifamycin to use in HIV-infected patients with active TB on a PI-based regimen because the risk of substantial drug interactions with PIs is lower with rifabutin than with rifampicin [44, 93], and a rifabutin-based regimen is equally as effective as a rifampicin-based regimen [94–96]. No dose adjustment is required for the PI as concentrations in the blood are unaffected by rifabutin . However, rifabutin own blood concentrations are affected by PIs, leading to increased serum concentrations and an increased risk of adverse effects if standard doses of this rifamycin are used. Based on data from healthy volunteers, and aiming at preventing rifabutin adverse effects such as anterior uveitis, in 2004 the CDC recommended a dose of rifabutin 150 mg three times a week or every other day when combined with PIs .. Since then, a few pharmacokinetic studies have reported on serum rifabutin concentrations when the drug was used at lower dose in association with lopinavir/ritonavir. In particular, in two small studies almost all subjects with combined treatment had a sub-therapeutic Cmax of rifabutin (therapeutic threshold at 0.5 μg/mL) [99, 100]. The Cmax and AUC were reduced by 36 and 26 %, respectively, compared with the patients receiving rifabutin alone . The risk associated with low serum concentrations of rifamycins is represented by a higher likelihood of treatment failure and emergence of resistance; the Tuberculosis Trials Consortium (TBTC)/US Public Health Service Study 23 defined an AUC24 of 4.5 mg/mL as the cut-off value for risk of emergence of resistance to rifamycins . Available evidence demonstrates that the danger of fostering resistance associated with sub-therapeutic concentrations of rifabutin is real [102, 103]. The largest pharmacokinetic study to identify the most appropriate rifabutin dose conducted to date is a randomized, open-label, multi-dose, two-arm, crossover trial conducted in Vietnamese adults with HIV-associated TB to compare daily with three times weekly rifabutin in association with lopinavir/ritonavir. Rifabutin 150 mg daily with lopinavir/ritonavir was associated with a 32 % mean increase in the rifabutin average steady-state concentration compared with rifabutin 300 mg alone. In contrast, the rifabutin average steady-state concentration decreased by 44 % when rifabutin was given at 150 mg three times per week with lopinavir/ritonavir. With both dosing regimens, two- to five-fold increases of the 25-O-desacetyl- rifabutin metabolite were observed when rifabutin was given with lopinavir/ritonavir compared with rifabutin alone .
As far as other PI drugs are concerned, a randomized, multiple-dose, parallel-group study conducted in healthy subjects compared the association of boosted atazanavir plus rifabutin 150 mg twice weekly with rifabutin alone . The rifabutin Cmax, AUC, and Cmin increased by 149, 48, and 40 %, respectively, as well as 25-O-desacteyl rifabutin (6.77-, 9.90-, and 10.45-fold increases, respectively) when rifabutin was coadministered with atazanavir/ritonavir. A post hoc simulation analysis suggested that the 150 mg every other day dose of rifabutin was appropriate in association with boosted atazanavir. Remarkably, this trial was terminated early due to a high proportion of subjects developing neutropenia. However, similarly to what was extensively observed for lopinavir/ritonavir, when shifting the study population from healthy volunteers to TB/HIV co-infected patients, these results were not confirmed by the only available observational study conducted in 16 adult HIV-infected TB patients being treated for TB with a 150 mg three times weekly rifabutin regimen associated with standard boosted atazanavir. In this study, the rifabutin Cmax was below the lower therapeutic limit (<0.3 μg/mL) in seven patients, and ten patients had a Ctrough below the minimal inhibitory concentration against M. tuberculosis (0.06 μg/mL) . For darunavir, a single study on healthy volunteers is available, which compared ritonavir-boosted darunavir at a standard dose (600/100 mg twice daily) alone, rifabutin 300 mg once daily, and darunavir standard dose with rifabutin 150 mg every other day. The rifabutin Cmin increased by 64 % and Cmax decreased by 28 %, while the AUC of 25-O-desacetylrifabutin was increased by 881 % during combined treatment . This study again demonstrated a high rate of adverse events during combined treatment, leading the authors to contraindicate further similar studies in healthy volunteers. A randomized, open-label, two-period, two-sequence, balanced, crossover drug interaction study was conducted in 22 healthy adult subjects to assess rifabutin pharmacokinetic parameters during concomitant administration of boosted fosamprenavir (700/100 mg twice daily) with rifabutin 150 mg every other day compared with the standard rifabutin 150 mg once daily. The rifabutin AUC from time zero to 48 h (AUC48) was unchanged; Cmax was decreased by 14 %, while the AUC48 and Cmax of 25-O-desacetylrifabutin were increased by 11- and 6-fold, respectively . The effect of steady-state ritonavir-boosted tipranavir 500/200 mg on the pharmacokinetic parameters of a single dose of rifabutin and 25-O-desacetyl-rifabutin were determined in a controlled, two-period study with healthy volunteers. The AUC, Cmax, and C12 of rifabutin were increased by 190, 70, and 114 %, respectively, and the metabolite concentrations were increased by 333, 86, and 176 %, respectively .
5.4 Integrase Inhibitors
Drug interactions between rifamycins and integrase inhibitors, and recommendations for use
Interactions in pharmacokinetic studies
Clinical consequences of the interaction
Raltegravir (400 mg bid)
Pharmacokinetics of a single dose of raltegravir 400 mg: raltegravir C12, AUC, and Cmax ↓ by 61, 40, and 38 %, respectively
AUC12 ↓ by 40 %; doubling the dose of raltegravir (800 mg bid): AUC ↑ by 27 %
Association possible, increasing the dose of raltegravir to 800 mg bid
Doubling raltegravir to 800 mg bid: raltegravir AUC12 ↑ 27 %, Ctrough ↓ 53 % and Cmax ↑ 62 % compared with standard raltegravir and no rifampicin
Raltegravir 800 mg bid: AEs or virological failure not reported
Rates of HIV suppression similar in TB/HIV patients treated with raltegravir 400 mg bid, raltegravir 800 mg bid, or efavirenz
Dolutegravir 50 mg bid: AUC24 and Ctrough↑ by 33 and 22 %, respectively, compared with dolutegravir 50 mg od alone
Not included in current recommendations, but association is theoretically possible
Raltegravir AUC12 ↑ 19 %, Cmax ↑ 39 %, Cmin ↓ 20 %
Association possible, with both raltegravir and rifabutin at standard doses
Elvitegravir AUC ↓ 21 %, Cmin ↓ 67 % by rifabutin 150 mg qod
No significant change in rifabutin AUC
Dolutegravir 50 mg od
Dolutegravir geometric mean ration of AUC24 and Ctrough ↓ by 5 and 30 %, respectively, by rifabutin 300 mg od
Not included in current recommendations, but association is theoretically possible
No published clinical study investigated the association of raltegravir and rifabutin. However, the “in vitro” characterization of the induction potency of rifabutin on UGT1A1 showed that the geometric mean ratio of raltegravir AUC12, C12, and Cmax were, respectively, 1.19, 0.80, and 1.39 for raltegravir and rifabutin versus raltegravir alone .
Elvitegravir undergoes metabolism primarily by the CYP3A4/5 and partly by glucuronidation through the UGT1A1 and 1A3 isozymes . The interaction between elvitegravir, either boosted by ritonavir or cobicistat, and rifampicin has not been investigated in pharmacokinetic studies. A pharmacokinetic study among 19 healthy volunteers assessed the potential interaction between elvitegravir/ritonavir (300/100 mg once daily) and rifabutin 150 mg every other day. Exposures of elvitegravir and rifabutin were not altered, but the rifabutin metabolite was increased five- to 20-fold. Three subjects terminated the study prematurely due to a grade 4 adverse event (neutropenia) . In a pharmacokinetic study of the interaction between cobicistat-boosted elvitegravir and rifabutin 150 mg every other day, the AUC, Cmax, and Ctrough concentrations of elvitegravir were reduced by 20, 10, and 67 %, respectively .
Similarly to raltegravir, dolutegravir is mainly metabolized by UGT1A1, with some contribution from CYP3A. In a phase I pharmacokinetic drug interaction study healthy subjects received dolutegravir at standard dose (50 mg once daily) or at double dose (50 mg twice daily) together with either rifampicin 600 mg or rifabutin 300 mg. When administered at a double dose and associated with rifampicin, the geometric mean ratios of dolutegravir AUC24 and Ctrough were 1.33 and 1.22, respectively, compared with dolutegravir alone. Rifabutin, at the standard 300 mg once-daily dose, determined a slight decrease in the dolutegravir geometric mean ratio of AUC24 and Ctrough of 0.95 and 0.70, respectively . It is unclear whether such a decrease is dose dependent.
5.5 CCR5 Antagonists
Drug interactions between rifamycins and CCR5 antagonists, and recommendations for use
Interactions in pharmacokinetic studies
Clinical consequences of the interaction
Maraviroc 100 mg bid: maraviroc AUC12 ↓ 67 %, Cmax ↓ 70 % by rifampicin 600 mg od
Maraviroc 200 mg bid with rifampicin 600 mg od: AUC and Cmax unchanged compared with maraviroc 100 mg bid alone
Not recommended due to lack of evidence
Not recommended due to lack of evidence
5.6 Fusion Inhibitors
Enfuvirtide, the first fusion inhibitor approved for the treatment of HIV-1 infection, is a synthetic peptide that binds to HIV-1 glycoprotein 41, blocking the fusion of viral and cellular membranes. Enfuvirtide does not influence concentrations of drugs metabolized by CYP3A4, CYP2D6, or N-acetyltransferase, and has only minimal effects on those metabolized by CYP1A2, CYP2E1, or CYP2C19 . In vitro and in vivo studies indicate that enfuvirtide has a low potential to interact with concomitantly administered drugs .
6 Current Recommendations for the Treatment of TB/HIV Co-Infected Patients
In agreement with the evidence described above, international guidelines consistently recommend an efavirenz-based ART regimen in combination with a standard rifampicin-based anti-TB regimen for the treatment of TB/HIV co-infected patients [5, 6, 14]. NRTI drugs can be used as the backbone of any HIV regimen, and no dose modification is required . The dose of efavirenz is 600 mg once daily, although efavirenz 800 mg may also be used in persons weighing >50 kg  or >60 kg . Remarkably, more than 95 % of adults and children living with HIV who are on ART worldwide are on a recommended first-line regimen, consisting of two NRTIs and one NNRTI, which, in most cases, is efavirenz . However, this proportion reflects a global average and is susceptible to rapid changes in the near future. As a matter of fact, people living with HIV in northern countries are much more likely to be on second-line HIV therapy; moreover, the 2013 WHO guidelines for consolidated antiretroviral treatment, which promotes extensive virologic monitoring of treated patients, will likely expand the number of people living with HIV on second-line HIV therapy in resource-constrained settings.
Combined treatment of TB/HIV co-infected patients is challenging whenever efavirenz cannot be used, due to intolerance or resistance. Regimens composed entirely of NRTIs are less effective than combinations of two classes of antiretroviral drugs (e.g., NNRTI + NRTI) [124–126]. This option is included in WHO recommendations for the treatment of TB/HIV co-infected patients who cannot take efavirenz, but its efficacy has never been compared with standard initial ART (e.g., efavirenz + two NRTIs) among patients receiving rifampicin.
Nevirapine is a logical substitute for efavirenz as it is used as the first-line antiretroviral for a large proportion of persons living with HIV. Clinical studies on nevirapine-based regimens are available, but evidence is inconsistent; it shows that nevirapine is probably a weaker alternative to efavirenz for the treatment of TB/HIV co-infected patients. Inconsistencies exist among international guidelines as well: while the WHO recommends nevirapine as an alternative regimen whenever efavirenz cannot be used , the CDC recommends against the association of rifampicin and nevirapine . Our opinion is that the association of nevirapine and rifampicin, without lead-in dosing, may be an acceptable choice in resource-constrained countries in which PI-based ART cannot be used due to the unavailability of rifabutin, and whenever efavirenz cannot be used.
Rifabutin could be used with standard doses of efavirenz or nevirapine [5, 6]. Based on pharmacokinetic simulations and very limited pharmacokinetic data, the dose of rifabutin should be increased from 300 to 450 or 600 mg when combined with efavirenz [5, 44]. However, there is little clinical rationale for the combined use of efavirenz or nevirapine and rifabutin.
Based on the anticipated pharmacokinetic interactions and the limited available clinical data, the use of rifampicin with the new NNRTIs rilpivirine and etravirine is contraindicated [5, 78, 79]. For the same reasons, rifabutin should not be used with rilpivirine. However, based on very limited data, rifabutin 300 mg once daily could be used with standard-dose etravirine provided that no additional potent CYP inducers are concomitantly used .
Commonly, a second-line antiretroviral regimen is based on drugs belonging to non-NNRTI classes. Globally, the vast majority of people living with HIV who are receiving a second-line antiretroviral are on a PI-based regimen, mainly lopinavir/ritonavir; in more affluent countries, however, a second-line antiretroviral may include any other available class of antiretrovirals.
Rifampicin should not be administered with PIs, whether boosted or unboosted, at standard doses. According to WHO recommendations , rifampicin can be administered with super-boosted PIs, e.g., lopinavir/ritonavir 800/200 mg twice daily or 400/400 mg twice daily, but this option does not appear in the US Department of Human Services (DHHS) guidelines [5, 44]. Due to the substantial evidence of poor tolerability of super-boosted PI regimens, larger studies would be necessary before this approach can be widely used, taking in consideration that it would be poorly suitable for a large-scale public health approach to ART.
Rifabutin is the best rifamycin to be used in association with PIs. No dose adjustment of the PI is necessary, but the optimal dose of rifabutin still needs to be conclusively assessed. Based on the available evidence, 150 mg daily is the most rational rifabutin dose when coadministered with lopinavir/ritonavir. The same probably applies to boosted atazanavir, and, possibly, to boosted darunavir. However, because limited safety data are available with this dose, patients taking rifabutin 150 mg in combination with PIs should be monitored for rifabutin-related toxicity .
Based on the very limited available clinical data, rifampicin could be used with raltegravir, doubling the dose of raltegravir to 800 mg twice daily. Raltegravir could be used with rifabutin as well, with both drugs administered at standard doses, based on the evidence from in vitro studies that rifabutin does not alter raltegravir pharmacokinetics to a clinically meaningful degree.
In contrast, the use of cobicistat-boosted elvitegravir in association with any rifamycin is not recommended. Dolutegravir is not mentioned in current international recommendations; however, the scarce available information on interactions between this drug and rifampicin and rifabutin show a profile similar to that of raltegravir and the association could have clinical applications.
Although the coadministration of maraviroc and rifabutin is not contraindicated by the DHHS , given the virtual absence of pharmacokinetic and clinical data, coadministration of maraviroc and rifamycins should probably be avoided.
There are no limitations for the use of enfuvirtide in association with any antitubercular drug.
7 Conclusions and Future Directions
Decisions on the treatment of TB/HIV co-infection should be made taking into consideration the potential pharmacokinetic and/or pharmacodynamic drug–drug interactions, in order to avoid drug toxicity or reduced efficacy. Most antiretroviral medications are eliminated or metabolized via the CYP enzyme system, or the UGT1A1 enzymes, or are extruded by the P-glycoprotein transporters, all of which are induced by the rifamycins. Moreover, rifabutin is metabolized via the CYP enzyme system, which makes interactions bidirectional. Adequate clinical response of both infections can be achieved with an acceptable safety profile when the pharmacological characteristics of drugs are known, and appropriate combination regimens, dosing, and timing of initiation are used.
Pharmacokinetic studies are important to pave the way of knowledge and predict interactions with potential clinical significance. However, it is generally impossible to identify threshold levels of the serum drug concentration, which makes interactions clinically significant. This is because of the existence of a number of potential co-factors and variables that are often poorly known, such as, for instance, the minimum effective concentration of some antiretrovirals. Clinical studies would therefore be essential for the development of guidelines; however, these are scant, except for combination studies of rifamycins and NNRTIs. Clinical studies on combination treatment for TB/HIV co-infection should be encouraged, but barriers exist; for example, the number of patients with TB/HIV co-infection who could enter clinical studies is small in countries with low TB endemicity, where HIV drugs belonging to the new classes (such as integrase inhibitors or co-receptor antagonists) are largely available. Conversely, the use of second-line HIV drugs is still very limited in less affluent countries, where the majority of TB/HIV co-infections occurs.
The role of therapeutic drug monitoring (TDM) is poorly defined, though it is currently being recommended for some clinical situations, such as monitoring the rifabutin concentration when the drug is used at 150 mg once daily in combination with PIs. Though extremely useful for the management of individual cases, TDM would hardly fit the public health approach to the management of antiretroviral therapy in resource-limited settings, where treatment regimens that do not require TDM would definitely be preferable.
More data are required for specific drugs: for example, rifapentine is a very attractive rifamycin for the treatment of active TB and, more so, of LTBI. However, the drug is currently banned in HIV-infected persons receiving antiretroviral therapy, and recommendations will not be updated unless new evidence on safety and efficacy is produced. Similarly, information on drug interaction of the newer integrase inhibitors is still very limited, preventing wider use of combinations between these drugs and rifamycins.
More data are also needed to steer recommendations for special populations. Pediatric formulations and dosing guidelines for rifabutin are not available for children and NNRTI-based therapy is not recommended as a preferred therapy for perinatally infected infants under the age of 1 year . Evidence on optimal dosing for ritonavir-boosted lopinavir in children receiving rifampicin is expanding. Super-boosted lopinavir treatment in association with rifampicin (ritonavir in addition to doses of co-formulated lopinavir/ritonavir to achieve mg to mg parity of ritonavir and lopinavir) ensures serum concentrations of lopinavir comparable with those of children treated with standard-dose lopinavir/ritonavir alone [127, 128]. Moreover, virologic response appears to be similar among children receiving super-boosted lopinavir and rifampicin and those receiving standard-dose lopinavir/ritonavir alone . Remarkably, using a similar approach by doubling the dose of lopinavir in association with rifampicin appears to achieve inferior results . No published data are available on the combined effects of pregnancy and rifampicin on antiretroviral drug concentrations and HIV treatment efficacy, making it difficult to formulate recommendations for the treatment of HIV-related TB among pregnant women. Finally, patients affected by MDR-TB deserve more research as well, although they are supposed to have fewer problems from drug–drug interactions as rifamycins are not part of the TB regimen. Knowledge of the metabolic pathways of some second-line drugs is incomplete because many of these drugs were developed and licensed decades ago. Conversely, new TB drugs, such as bedaquiline and delamanid, will soon become the cornerstone of MDR-TB treatment, but their use in TB/HIV co-infected patients is uncertain in the absence of data from clinical trials.
Conflict of interest
No sources of funding were used in the preparation of this review. Mario Regazzi, Anna Cristina Carvalho, Paola Villani and Alberto Matteelli have no conflicts of interest that are directly relevant to the content of this review.