Treatment of tuberculosis infection (latent tuberculosis) in nonpregnant adults with HIV infection

INTRODUCTION — 

Treatment of individuals with active tuberculosis (TB) is the first priority for TB control; an important second priority is identification and treatment of individuals with TB infection (TBI) [1-6]. Issues related to treatment of TBI in nonpregnant patients with human immunodeficiency virus (HIV) infection will be reviewed in more detail here.

The diagnosis of TBI is discussed separately. (See "Tuberculosis infection (latent tuberculosis) in adults: Approach to diagnosis (screening)" and "Use of the tuberculin skin test for diagnosis of tuberculosis infection (tuberculosis screening) in adults" and "Use of interferon-gamma release assays for diagnosis of tuberculosis infection (tuberculosis screening) in adults".)

Issues related to treatment of TBI in individuals without HIV are discussed separately. (See "Treatment of tuberculosis infection (latent tuberculosis) in nonpregnant adults without HIV infection" and "Tuberculosis disease (active tuberculosis) in pregnancy" and "Tuberculosis infection (latent tuberculosis) in children".)

TERMINOLOGY AND DEFINITIONS — 

TB terminology is inconsistent in the literature [7]. Relevant terms are defined in the table (table 1).

Definitions for drug-resistant TB are summarized in the table (table 2).

RATIONALE FOR TBI TREATMENT — 

Tuberculosis infection (TBI) is caused by inhalation of viable tuberculous bacilli. These organisms usually persist in an inactive state, known as TBI; in some cases, there is rapid progression to active TB disease.

Individuals with TBI are asymptomatic and not infectious. Latent TB bacilli remain viable and may reactivate years later, causing active symptomatic and transmissible TB disease [8]. The natural history of TB is discussed further separately. (See "Tuberculosis: Microbiology, pathogenesis, and immunology".)

Following primary TB infection, patients with HIV infection are at greater risk of progressing to active TB disease than individuals without HIV [8,9]; the interval between exposure and disease can be less than two months [9,10]. In addition, patients with untreated HIV and TBI are 30 to 100 times more likely to reactivate with TB disease than individuals without HIV [11-14]. Antiretroviral therapy reduces risk of developing active TB disease among individuals with HIV and TBI [15-18]. (See 'Initiation of antiretroviral therapy' below.)

Treatment of TBI in patients with HIV is associated with two important benefits:

●Treatment of TBI reduces the risk of progression to active TB disease [19]. Among patients in Africa with HIV and TBI, the risk of developing active TB disease is 5 to 10 percent per year [20,21]. Overall, treatment of TBI in individuals with HIV reduces active TB disease by 62 percent and mortality by 26 percent; most studies have demonstrated the benefit is greatest among those with a positive test for TBI (via tuberculin skin test or interferon-gamma release assay) [20].

●Treatment of TBI reduces TB transmission [15,20,22]. Mathematical models have predicted that provision of TBI therapy to at least half of individuals with HIV and TBI would be cost effective and would result in a substantial reduction in TB disease [22,23].

For these reasons, treatment of TBI in people with HIV has been favored by expert panels for many years [24]. However, despite consistent and clear recommendations from authoritative agencies (including the World Health Organization) [5,6], it is estimated that in 2018 and 2019 less than 50 percent of patients with HIV newly accessing HIV care received TBI treatment worldwide [25,26].

WHOM TO TREAT — 

TBI treatment is warranted for individuals with HIV infection in the following circumstances (in some cases, it may not be possible to definitively establish the presence of TBI):

●Individuals with recent contact with a person with active TB disease

●Individuals with clinical suspicion for prior TB (eg, fibrotic disease on chest radiograph consistent with healed TB) and no documented history of adequate TB treatment

●Individuals with positive tuberculin skin test (TST) or interferon-gamma release assay (IGRA) in the absence of active TB

●If TBI testing is not available − All individuals with HIV infection and risk of TB exposure (particularly if the CD4 cell count is <200 cells/microL, which confers greater risk for development of TB disease), once active TB has been excluded

Testing for TBI via TST or IGRA is warranted prior to initiation of therapy; multiple studies suggest that the benefits of treatment are greatest among patients with HIV infection and positive TST or IGRA [15,17,20,27,28]. In one systematic review of longitudinal studies, the incidence of active TB was 11 times higher among individuals with HIV infection and a positive IGRA or TST than those with a negative test [29]. Similarly, in two systematic reviews of 12 controlled trials evaluating treatment regimens administered for 2 to 12 months [20], and three trials of extended duration isoniazid (INH) [30], TBI treatment was associated with reduction in TB incidence [20,30] and death [30] only among patients with HIV who had positive TST at enrollment.

For patients with HIV infection in whom there is strong likelihood of TB exposure, but TBI testing is not available (despite all efforts of the provider), TBI treatment is warranted. This approach is supported by a randomized trial in South Africa including 1329 individuals with HIV treated with INH or placebo; the incidence of TB over a three-year period was lower among those who received treatment than those who did not (2.3 versus 3.6 per 100 person-years; hazard ratio 0.63, 95% CI 0.41-0.94) [31].

Issues related to diagnosis of TBI are discussed further separately. (See "Tuberculosis infection (latent tuberculosis) in adults: Approach to diagnosis (screening)".)

PRIOR TO TREATMENT

Exclude active TB disease — Prior to initiation of treatment for TBI, all patients must be evaluated for active TB disease, to avoid inadvertent undertreatment of active disease leading to development of drug resistance [32]. The evaluation includes clinical history, physical examination, and chest radiograph; it also may require microbiologic testing of respiratory samples. (See "Tuberculosis infection (latent tuberculosis) in adults: Approach to diagnosis (screening)", section on 'Excluding TB disease'.)

The increased risk of active TB begins as soon as HIV infection is acquired and is especially high among patients with a CD4 cell count <200 cells/mm³ [33]. Patients with advanced immunosuppression may present with atypical clinical manifestations including extrapulmonary TB disease and/or mediastinal lymphadenopathy or with subclinical infection and a normal chest radiograph.

Excluding active TB in individuals with HIV may be more difficult than in individuals without HIV. Among individuals with HIV, symptoms may be minimal or nonspecific, and the yield of diagnostic procedures may be limited because of many factors; these include presence of extrapulmonary disease (with absence of lung involvement in some cases) and normal chest radiography (pulmonary disease present but relative lack of inflammation related to immune suppression).

The World Health Organization (WHO) suggests symptoms screening is adequate to exclude active TB [5,6]; however, in a systematic review of 21 studies including more than 15,000 people with HIV, the sensitivity of symptoms screen alone was only 51 percent in those on antiretroviral therapy (ART). Based on data from two studies, the addition of chest radiography improved the sensitivity of active TB detection to 89 percent [34]. Hence, we strongly advise chest radiography to exclude active TB before initiation of TBI treatment.

The investigation for active TB should include at least two respiratory samples for acid-fast bacilli (AFB) smear and TB culture; sputum induction may be warranted. If sputum is nondiagnostic, bronchoscopy may be useful to obtain additional respiratory samples for diagnosis of TB as well as to facilitate evaluation for other causes of pulmonary infiltrates and respiratory symptoms.

For individuals with HIV in whom it is difficult to exclude active TB (for example, individuals with relevant clinical manifestations [cough >2 weeks' duration, fevers, night sweats, and/or weight loss] but negative sputum AFB smear), initiation of treatment for active disease (pending microbiology data) is reasonable, particularly in patients with advanced immunosuppression or in resource-limited settings. If there is no improvement in symptoms or radiographic findings with empiric treatment for active TB and microbiologic studies are negative, additional evaluation for an alternative cause should be pursued, and the patient may be switched to a regimen for treatment of TBI. In such cases, if the patient has taken two months of standard therapy for active TB (ie, isoniazid [INH], rifampin, and pyrazinamide), the patient may be considered to have completed adequate treatment for TBI [1].

The diagnosis of active TB disease is discussed in detail separately. (See "Diagnosis of pulmonary tuberculosis in adults".)

Assess comorbidities — Comorbidities that should be considered in planning treatment of TBI include underlying liver disease, neuropathy, and concomitant medications.

Regimens for treatment of TBI are associated with hepatotoxicity; the risk of hepatotoxicity is greatest with INH and less so for the rifamycin-based regimens. Baseline liver enzyme testing is appropriate for the following individuals [24,35]:

●Patients receiving ART. In earlier guidelines, HIV infection was considered a risk factor for drug-induced liver injury [24]. Subsequently, an extensive review concluded that the likelihood of developing INH-related hepatotoxicity among individuals with HIV is comparable with that of individuals without HIV [36].

●Patients with underlying liver disease (eg, hepatitis B or C, alcoholic hepatitis, or cirrhosis) or risk of underlying liver disease

●Pregnant and postpartum patients within three months of delivery

●Patients who consume alcohol regularly

●Patients on other medication(s) with potential hepatotoxicity

●Advanced age (especially >50 years)

Baseline liver enzyme testing may also be considered on an individual basis for other patients but need not be performed routinely. Active hepatitis and end-stage liver disease are relative contraindications to the use of hepatotoxic agents for TBI treatment. For patients with baseline serum transaminase levels greater than three times the upper limit of normal, TBI treatment decisions should be deferred pending evaluation of the underlying cause.

The risks and benefits of initiating treatment of TBI in patients with abnormal baseline liver function (alanine aminotransferase [ALT] >3 times the upper limit of normal) should be considered carefully. Clinical considerations include baseline ALT level, alcohol use, age, presence of viral hepatitis, and the risk of TB progression [36]. If therapy is required, serum transaminases, bilirubin, and international normalized ratio should be monitored every two to four weeks [36].

The potential for drug-drug interactions with rifamycins is high for individuals taking the following drugs: warfarin, oral or other hormonal contraceptives, corticosteroids, some antihypertensives, some antiarrhythmics, some antidepressants, some anticonvulsants, methadone, and some antiretroviral drugs (including tenofovir alafenamide, some non-nucleoside reverse transcriptase inhibitors, all protease inhibitors, and all integrase inhibitors). Details about specific interactions may be obtained by using the drug interactions program included within UpToDate. (See "Rifamycins (rifampin, rifabutin, rifapentine)".)

SELECTING A REGIMEN — 

The approach to treatment of TBI in individuals with HIV depends on a number of factors, including possible drug interactions in individuals taking antiretroviral therapy (ART), known or suspected susceptibility of the presumed infecting organism, likelihood of patient adherence with longer regimens, and local TB transmission rates.

Approach in most clinical settings

Regimen selection and ART considerations — For treatment of TBI in patients with HIV infection on ART, we favor either 3HP (isoniazid [INH] and rifapentine [RPT] administered weekly for three months via directly observed therapy) or 4R (rifampin [RIF] self-administered daily for four months) over INH monotherapy (table 3).

We favor using a rifamycin-based regimen (either 3HP or 4R) over INH monotherapy in view of the similar efficacy, improved treatment completion rates, and significantly lower rates of hepatotoxicity of 3HP or 4R compared with INH, in patients with HIV infection [37,38] as well as patients without HIV infection [39,40].

We do not favor use of 3HR (daily INH and rifampin [RIF] for three months). Use of 1HP (daily INH and RPT for one months) may be feasible; more data on safety and tolerability of this regimen is needed.

The above approach differs from that of the World Health Organization, which recommends one the following regimens: 3HR, 3HP, or INH daily for 6 or 9 months, and lists the following as alternative regimens: 4R or 1HP [5,6].

The above approach also differs somewhat from the United States Department of Health and Human Services (DHHS) Opportunistic Infection Guidelines, in which the preferred therapies are 3HP (or 1HP) and INH monotherapy [41].

Use of a rifamycin-based regimen requires careful attention to drug interactions [1]. A discussion of the available data on the efficacy for these regimens is presented below. (See 'Regimen efficacy and toxicity' below.)

Drug interactions related to rifamycins are a very important consideration for management of patients with HIV infection on ART. Treatment decisions must be individualized, given that (1) integrase strand transfer inhibitor (INSTIs)-based regimens are preferred for treatment of HIV infection, (2) rifamycin-based regimens (3HP and 4R) are preferred for treatment of TBI, and (3) coadministration of INSTIs and rifamycins is associated with significant risk of drug interactions that may reduce the efficacy of the INSTIs. (See 'Drug interactions: ART and rifamycins' below.)

Treatment decisions should be guided by the patient's current ART regimen; in some circumstances, we consider switching to an alternative ART regimen with minimal rifamycin-based drug interactions for the duration of TBI treatment.

●Use of 3HP – The regimen 3HP (INH and RPT administered weekly for three months via directly observed therapy [DOT]) is preferred for treatment of TBI in patients with HIV infection, given greatest efficacy and safety data (see 'Isoniazid and rifapentine (3HP)' below).

RPT induces cytochrome P450 isoenzymes and can cause significant interactions [42]. The 3HP regimen may be used in the following circumstances:

•Patients on a dolutegravir [DTG]-based regimen (administered once daily) who are virologically suppressed [43].

•Patients taking (or willing to switch to) a raltegravir (RAL)- or efavirenz (EFV)-based regimen [43,44].

•For patients on TAF, the need to switch to alternate HIV therapy for the duration of TBI treatment is unclear. DHHS guidelines suggest that TAF and RPT can be co-administered with close follow-up monitoring for efficacy [41]. (See 'Drug interactions: ART and rifamycins' below.)

RPT significantly reduces other integrase inhibitor concentrations (bictegravir, cabotegravir, elvitegravir-cobicistat) [41], and coadministration is contraindicated.

●Use of 4R – The regimen 4R (RIF self-administered daily for four months) may be used if DOT (for administration of 3HP) is not feasible. This regimen may be used for patients taking (or willing to change to) a dolutegravir-based or raltegravir-based regimen; in such cases the dose of DTG or RAL should be doubled [45]. Alternatively, patients can switch to an EFV-based regimen. Daily rifampin may not be used for patients on bictegravir, TAF, and a number of other antiretroviral agents.

4R regimen is safe and efficacious in patients without HIV, but data among patients with HIV are limited (see 'Rifampin (4R)' below).

●Use of 1HP – The regimen 1HP (INH and RPT self-administered daily for one month) is a potential option for treatment of TBI in patients with HIV infection; more data on safety and tolerability of this regimen is needed. (See 'Isoniazid and rifapentine (1HP)' below.)  

This regimen may be used for patients on a dolutegravir-based regimen who are virologically suppressed; in such cases the dose of DTG should be doubled [46].

●Use of INH – For patients who are unable or unwilling to switch to a rifamycin-compatible regimen, we use INH monotherapy (daily for nine months) for treatment of TBI. The WHO favors daily INH for six or nine months [6,47]; completion rates may be higher with shorter duration, but shorter duration carries risk of lower efficacy. The United States DHHS Opportunistic Infection Guidelines also list INH as the preferred therapy. (See 'Isoniazid monotherapy' below.)

●Use of 6LFX – We favor use of 6LFX in people with HIV infection and exposure to individuals with MDR-TB or RR-TB (table 2), in keeping with WHO 2024 guidance [6]. In addition, 6LFX may be used for individuals in whom rifamycins cannot be given, as an alternative to INH monotherapy (such as in the context of liver disease). Further safety and tolerability data of this regimen are needed. (See 'Levofloxacin monotherapy (6LFX)' below.)

Efavirenz should not be administered with a fluoroquinolone.

●Regimens not recommended

•3HR – The 3HR regimen has been studied in three randomized trials among adults with HIV infection [21,48,49]. In these studies, completion rates with 3HR were not significantly better than with 6H [48]. In two trials including a total of 1254 participants, adverse events requiring discontinuation of therapy were more common with 3HR than with 6H (28 of 636 receiving 3HR versus 9 of 618 receiving 6H) [21,48] although less than with 12 months of INH [50]. The efficacy of 3HR and isoniazid monotherapy regimens was similar in the three studies [21,48,50].

Hence, 3HR confers no benefit over INH monotherapy in terms of completion or efficacy, has similar toxicity, and also confers risk of rifampin-induced drug interactions (see 'Drug interactions: ART and rifamycins' below). Our approach is similar to Canadian recommendations [51]. The Centers for Disease Control and Prevention (CDC) have conditionally recommended 3HR [1], while WHO recommends use of 3HR in settings where rifamycins are available as fixed dose combinations of INH and RIF [5,6].

•RIF plus pyrazinamide – RIF plus pyrazinamide should not be used for treatment of TBI due to hepatotoxicity [52].

•INH for 36 months – Previously, this regimen was recommended in settings with very high TB transmission; it is no longer recommended due to toxicity, poor completion rates, low certainty of supporting evidence, and expanded availability of ART in these settings [6].

●Nitrosamine impurities – Given detection of nitrosamine impurities in samples of RIF and RPT announced by the US Food and Drug Administration (FDA) in August 2020 [53], INH monotherapy may be considered as an alternative regimen for adults with newly diagnosed TBI. However, for adults at highest risk of new TB infection or TB reactivation (such as close contacts, recent converters, immunocompromised individuals, and those who will be receiving tumor necrosis factor-alpha inhibitors), and particularly for those at increased risk of liver toxicity (age >50 years, hepatitis C coinfection, or other liver disease), use of rifamycin-containing regimens may be considered after individual patient discussion of potential risks and benefits. To date, there is no information on the clinical significance of these nitrosamine impurities, nor is there information on possible impurities in rifabutin. (See "Rifamycins (rifampin, rifabutin, rifapentine)", section on 'Nitrosamine impurities'.)

●Additional considerations – Important considerations with respect to regimen selection and administration include:

•Rifamycin drug interactions - RIF and RPT have significant interactions with many other medications including the following: warfarin, oral and other hormonal contraceptives, corticosteroids, some antihypertensives, some antiarrhythmics, some antidepressants, some anticonvulsants, methadone, buprenorphine. There are many interactions with antiretroviral drugs (including TAF, all non-nucleoside reverse transcriptase inhibitors [NNRTIs], all protease inhibitors, and all INSTIs). Specific interactions can be determined by use of the drug interactions program included within UpToDate.

•Risk for hepatotoxicity - Regimens for treatment of TBI are associated with hepatotoxicity. The risk of hepatotoxicity is greatest with INH and significantly less with 4R or 3HP. The risk appears to be somewhat less with 3HR than with 6 to 12 months of INH [21,48,50], although in these studies overall rates of drug discontinuation due to adverse events was similar with 3HR and the longer isoniazid monotherapy regimens. Additional side effects associated with rifamycins include rash and thrombocytopenia. (See 'Assess comorbidities' above and 'Monitoring and adherence' below.)

•Risk for INH-induced neuropathy - Peripheral neuropathy occurs in up to 2 percent of patients taking INH due to interference with metabolism of pyridoxine and can be prevented with pyridoxine supplementation [54]. For patients with HIV infection on an TBI regimen containing INH, we coadminister pyridoxine (dosing for daily INH regimen: 25 to 50 mg/day; dosing for weekly isoniazid/rifapentine regimen: 50 mg once weekly) [37]. We also administer pyridoxine to infants of lactating patients receiving INH.

•Adverse effects of 3HP - Important side effects of 3HP include hypersensitivity or flu-like symptoms (eg, lightheadedness, dizziness, headache, nausea or vomiting, syncope, rash, or angioedema); these occurred in 3.8 percent of patients treated with 3HP in one trial [39]. For this reason, 3HP usually is administered via DOT, to facilitate side effect review and treatment withholding if needed. However, self-administration may be acceptable for patients who can reliably take their medications on schedule and inform their providers promptly of side effects (while withholding the next dose pending provider review) [39,42,55]. (See "Treatment of tuberculosis infection (latent tuberculosis) in nonpregnant adults without HIV infection", section on 'Isoniazid and rifapentine (3HP)'.)

•Adherence - Self-administered INH is frequently associated with poor adherence; for patients with poor adherence and contraindications to rifamycins, administration of INH via DOT may be the optimal approach.

Regimen efficacy and toxicity

Rifamycin-based regimens — Use of a rifamycin-based regimen over INH monotherapy for treatment of TBI in patients with HIV infection on antiretroviral therapy is supported by a network meta-analysis including 20 studies (including 16 randomized trials) with median sample size 616; administration of INH monotherapy (6 to 12 months) was no more efficacious for TB prevention than rifamycin-containing regimens (incidence rate ratio [IRR] 1.0, 95% CI 0.8-1.4); however, INH monotherapy was associated with higher all-cause mortality (IRR 1.6, 95% CI 1.2-2.0) and higher risk of severe hepatotoxicity (risk difference 8.9 per 100 persons treated, 95% CI 2.8-14.9) [56].

The FDA announced detection of nitrosamine impurities in samples of RIF and RPT in August 2020. This is discussed further above. (See 'Regimen selection and ART considerations' above.)

If RIF or RPT is used to treat TBI, clinicians should assess the potential for drug-drug interactions among the rifamycins and different ART regimens. (See 'Drug interactions: ART and rifamycins' below.)

Isoniazid and rifapentine (3HP) — Thus far, the greatest cumulative clinical and trial experience with INH and RPT, although limited, has been a weekly regimen for three months (3HP) in which doses are directly observed. A regimen of INH and RPT administered daily for one month (1HP) has been studied in patients with HIV, as described below. However, given limited data for this approach, combination therapy with INH and RPT should be administered for three months as 3HP pending further study (table 3). (See "Treatment of tuberculosis infection (latent tuberculosis) in nonpregnant adults without HIV infection", section on 'Isoniazid and rifapentine (3HP)'.)

●Use of ART with 3HP – ART regimens that may be used for patients treated with 3HP are discussed above. (See 'Regimen selection and ART considerations' above.)

Use of 3HP among patients on a dolutegravir (DTG)-based regimen (administered once daily) who are virologically suppressed is supported by pharmacokinetic data; in one study including 61 participants treated with DTG-based ARV and 3HP, trough concentrations of DTG remain adequate without dose adjustment despite a 26 percent reduction in DTG exposure [43].

●Dosing – Weight-based RPT dosing may result in underdosing among the lower weight categories; in addition, RPT levels are approximately 30 percent lower in people with HIV [57]. Pending further data and consideration by authoritative agencies, we continue to dose RPT as summarized in the table (table 3).

●Side effects – Important side effects of 3HP include hypersensitivity or flu-like symptoms (eg, lightheadedness, dizziness, headache, nausea or vomiting, syncope, rash, or angioedema); these occurred in 3.8 percent of patients treated with 3HP in one trial [39]. For this reason, 3HP usually is administered via DOT, to facilitate side effect review and treatment withholding if needed. In one trial including more than 1000 patients with TBI comparing self-administration of 3HP with DOT, the rate of treatment completion was lower in the self-administered group [55]. However, self-administration may be acceptable for patients who can reliably take their medications on schedule and inform their providers promptly of side effects (while withholding the next dose pending provider review) [39,42,55].

●Efficacy data for 3HP – Use of 3HP in patients for treatment of TBI in patients with HIV infection is supported by the following trial data:

In a trial including more than 4000 patients with HIV infection on ART (largely efavirenz-based regimens; median CD4 count 440 cells/mcL) in South Africa, Ethiopia, and Mozambique, patients were randomly assigned to treatment with 3HP given for one course, 3HP given one course annually for two years, or INH daily for six months (6H) [58]. Approximately 60 percent of participants had negative QuantiFERON TB Gold Plus test results. The incidence of TB among the two groups treated with 3HP was similar (1.21 versus 1.26 per 100 person-years; hazard ratio 0.96, 95% CI 0.61-1.50), suggesting no benefit of repeat 3HP treatment. The incidence of TB was comparable among those treated with 3HP or 6H (hazard ratio 1.60, 95% CI 0.58-4.42), despite more favorable treatment completion rates with 3HP (90.4 versus 50.5 percent; risk ratio 1.78, 95% CI 1.61-1.95).

In a trial including nearly 400 patients with HIV infection (median CD4 count approximately 500 cells/mcL) in the United States, Brazil, Spain, Peru, Canada, and Hong Kong, patients were randomly assigned to treatment with 3HP or INH daily for 9 months (9H) [37]. Enrolled participants were not taking ART or planning to initiate ART in the next 90 days. The TB case rate was lower among those treated with 3HP (0.39 versus 1.25 per person-year). The treatment completion rate was higher with 3HP than 9H (89 versus 64 percent); the rate of adverse drug reactions was comparable between the groups.

Rifampin (4R) — Thus far, safety and efficacy data available from randomized trials for use of 4R in patients with HIV are limited.

In a secondary analysis of two randomized trials including 270 patients with HIV infection treated with 4R or 9 months of INH, drug-related adverse events were less common among patients treated with 4R than those treated with 9H (risk difference -2.1 percent, 95% CI -5.9 to 1.6) [38]. Rates of active TB were also nonsignificantly lower with 4R than 9H (rate difference per 100 person-years -0.4: 95% CI -2.2 to 1.3). These findings are reassuring, given that among people without HIV in the same two trials, 4R had a significantly higher completion rate (compared with 9H), a significantly lower rate of severe adverse events (particularly hepato-toxicity), with non-inferior efficacy [40,59].

ART regimens that may be used for patients treated with 4R are discussed above. (See 'Regimen selection and ART considerations' above.)

Among patients without HIV, of the available TBI regimens, 4R is associated with the lowest rate of hepatotoxicity [1,60,61]. (See "Treatment of tuberculosis infection (latent tuberculosis) in nonpregnant adults without HIV infection".)

Isoniazid and rifapentine (1HP) — Use of 1HP in patients for treatment of TBI in patients with HIV infection is supported by a randomized trial including nearly 3000 patients with HIV infection (of whom 43 percent were receiving EFV-based ART, 7 percent were receiving nevirapine-based ART, and 50 percent were not on ART) in which one month of daily self-administered INH and RPT was compared with nine months of daily INH (9H) [62]. Among the study participants, 23 percent had a positive test for TBI; 60 percent had a negative test; the remainder were not tested but resided in a region with TB prevalence 60 to 100 cases per 100,000 population. The incidence rates for TB disease or death during follow-up (median 3.3 years) were comparable between the groups (0.65 versus 0.67 per 100 person-years); the rates of serious adverse events were also comparable (6 versus 7 percent). The treatment completion rates were high in both arms, although higher among those who received the one-month regimen (97 versus 90 percent).

It may be difficult to extrapolate the trial results to routine clinical settings. The trial include broad eligibility criteria and most patients were tuberculin skin test (TST) negative; only 13 percent of patients had CD4 <250 cells/microL. In addition, the safety of 1HP requires further evaluation in other settings and populations; the toxicity was similar between the groups, yet toxicity is a major limitation of 9H.

Isoniazid and rifampin (3HR) — We do not favor use of 3HR for treatment of TBI in patients with HIV infection; this regimen confers no benefit over INH monotherapy in terms of efficacy, toxicity is similar as with INH alone (ie unacceptable), and it adds the risks associated with rifampin, particularly drug-drug interactions. (See 'Regimen selection and ART considerations' above.)

The efficacy and hepatotoxicity of 3HR is similar to that of INH administered for six months. In a randomized trial among more than 2000 individuals with HIV evaluating 3HR and INH for six months, the relative risk of TB with INH alone (compared with placebo) was 0.33 (95% CI 0.14-0.77) and with INH and RIF was 0.40 (95% CI 0.18-0.86) [21].

Some trials supporting recommendations for use of 3HR actually used four months of HR [63]. In a meta-analysis of five trials among patients with and without HIV infection, the toxicity of 3HR or 4HR was the same as that of INH administered for 6 to 12 months [63].

Isoniazid monotherapy — Of the available TBI regimens, INH is associated with the highest rate of hepatotoxicity in adults [60,61,64,65]. This can result in hospitalization [64,66] and fatal hepatoxicity [64,65]. In addition, the risk of mortality due to INH hepatotoxicity in adults with HIV may be under-recognized in high TB incidence settings; in one cohort including more than 4900 patients with HIV in Eritrea treated with 6H, life-threatening liver toxicity was observed in 15 cases and 4 died [65].

Prior to the availability of clinical trial data supporting use of rifamycin-based regimens, INH monotherapy was favored for treatment of TBI in patients with HIV, given its efficacy and compatibility with ART [17,18,20]. This approach is based on inference from clinical trials in patients without HIV and subgroup analyses of data from trials using treatment interventions of different durations [24,41,67]. INH monotherapy remains an acceptable alternative agent for patients with contraindications to use of rifamycins (such as unmanageable drug-drug interactions or hypersensitivity). (See "Treatment of tuberculosis infection (latent tuberculosis) in nonpregnant adults without HIV infection", section on 'Isoniazid monotherapy'.)

We favor daily INH dosing (300 mg) for nine months; the DHHS and the WHO recommend six to nine months [3,6]. Based on limited data in patients without HIV infection, the efficacy of six months is lower than the efficacy of nine months (risk reduction 69 versus 93 percent) [68], but completion rates are better [24]. In the setting of difficulty with adherence, providers may prefer to concentrate efforts in ensuring six months of therapy.

Any ART regimen may be used for patients on INH monotherapy, as there are no clinically significant drug interactions [41]. For patients who cannot tolerate INH and who are on an ART regimen containing drugs that interact with rifamycins and cannot be managed with dose adjustments and/or substitutions, the approach requires careful individualized review of medications and potential drug interactions. In such cases, consultation with an expert is warranted prior to selecting a regimen for treatment of TBI.

Extended duration of INH monotherapy (36 months) is no longer recommended, even in settings with very high transmission rates [6].

Levofloxacin monotherapy (6LFX) — We favor use of levofloxacin for six months among people with HIV infection and exposure to individuals with rifampin-resistant or multidrug-resistant tuberculosis [5]. We also favor use of this regimen for individuals in whom rifamycins cannot be given, as an alternative to INH monotherapy. Dosing is provided in the table (table 3).

Evidence for use of 6LFX is discussed separately. (See "Treatment of tuberculosis infection (latent tuberculosis) in nonpregnant adults without HIV infection", section on 'Drug-resistant tuberculosis infection (TBI)'.)

Drug interactions: ART and rifamycins — The potential for drug-drug interactions with rifamycins in individuals taking ART is high. These interactions often are bi-directional and may be complex. Antiretroviral drugs should be reviewed carefully for compatibility with rifamycins [41,69].

Rifamycins accelerate metabolism of TAF, all NNRTIs, all protease inhibitors, all integrase strand transfer inhibitors (INSTIs; bictegravir cabotegravir, dolutegravir, elvitegravir, and raltegravir), and the CCR5 antagonist maraviroc. RPT appears to cause less severe drug interactions than RIF [2]. (See "Rifamycins (rifampin, rifabutin, rifapentine)" and "Treatment of drug-susceptible pulmonary tuberculosis in nonpregnant adults with HIV infection: Initiation of therapy", section on 'Drug interactions' and "Selecting antiretroviral regimens for treatment-naïve persons with HIV-1: General approach".)

●Integrase strand transfer inhibitors (INSTIs) - RIF causes decreases in concentrations of INSTIs; this can be managed by doubling the dose of dolutegravir or raltegravir; In one study DTG exposure was adequate when co-administered with Rifampin if the dose was doubled to 50mg BID [45]. This dosing was also adequate if the dose of RIF was 10mg/kg/day (standard) or 35mg/kg/day in the same study [45]. However, other INSTIs (bictegravir, cabotegravir, elvitegravir-cobicistat) may not be coadministered with RIF [70,71].

●Efavirenz (EFV) – RIF only leads to modest reduction in EFV concentrations, and standard doses of EFV can still achieve good clinical outcomes [41]. Use of nevirapine with RIF is not recommended due to low levels with RIF [72]. RIF significantly decreases concentrations of the other NNRTIs (doravirine, etravirine, nevirapine, and rilpivirine) and they should not be used together [73-76]. RIF cannot be used with any HIV protease inhibitor, due to significant decreases in protease inhibitor concentrations [41,77].

●Tenofovir alafenamide (TAF) – TAF is a potent nucleotide reverse transcription inhibitor that is used commonly in HIV treatment regimens. Strong P-glycoprotein inducers like RIF, rifabutin, and RPT likely inhibit the absorption of TAF, and the United States FDA Prescribing Information recommends against administration of these agents with TAF because of risk of treatment failure [78]. However, while administration of TAF with RIF to healthy volunteers reduced both plasma and intracellular TAF exposure by 36 percent relative to administration of TAF alone, intracellular concentrations of the active drug metabolite were more than four times higher than levels following administration of tenofovir disoproxil fumarate (TDF) alone; these findings suggest that coadministration of TAF and rifampin may be acceptable [79].

Settings with extremely high TB transmission rates — In a few settings, such as the South African gold mines, where close quarters, poor ventilation, and frequent HIV coinfection produced some of the highest TB rates in the world, the optimal approach to TBI treatment is uncertain. In one cluster-randomized study in this setting, mass screening and treatment for TBI had no significant effect on TB control, despite successful use of INH in preventing TB during TBI treatment [80]. In such settings, there are no data on the impact of any short rifamycin-containing regimen; further study is needed.

Empiric initiation of treatment for active TB disease (four-drug therapy) does not appear superior to preventive therapy for reduction of mortality. (See "Treatment of drug-susceptible pulmonary tuberculosis in nonpregnant adults with HIV infection: Initiation of therapy", section on 'Role of empiric TB therapy'.)

MONITORING AND ADHERENCE — 

The optimal approach to liver enzyme testing for patients with HIV receiving treatment for TBI (in the absence of other risk factors) is uncertain. The risk of drug-induced hepatotoxicity associated with isoniazid (INH) increases with age [81] and alcohol use [81]. There is no consistent evidence of increased toxicity of TBI therapy in individuals with HIV [82,83]. (See "Treatment of pulmonary tuberculosis in adults with HIV infection: Follow-up after initiation of therapy", section on 'Drug-related side effects'.)

Issues related to baseline laboratory testing are discussed above. (See 'Assess comorbidities' above.)

For patients with abnormal baseline aminotransferases and for patients with additional risk factors for hepatotoxicity (eg, pre-existing liver disease), we favor subsequent laboratory monitoring of serum transaminases, bilirubin, and international normalized ratio every two to four weeks [24,36].

For patients with normal baseline aminotransferases who are initiating treatment for TBI, we favor a single aminotransferase assay after approximately one month of therapy. Otherwise, laboratory monitoring is warranted only in the setting of clinical symptoms or other concern for hepatotoxicity.

We are in agreement with some experts who favor withholding TBI treatment in patients with symptomatic hepatitis and an alanine aminotransferase level (ALT) greater than three times the upper limit of normal, as well as among those with an ALT greater than five times the upper limit of normal (whether symptomatic hepatitis is present or not) [24,36].

Patients should be counseled on the symptoms and signs of drug-induced hepatitis (eg, right upper quadrant pain, jaundice, nausea, vomiting, loss of appetite, dark urine) and be instructed to stop therapy and report these symptoms promptly should they be experienced. Patients should be monitored monthly for signs and symptoms of adverse drug events and development of active TB ; regular patient-doctor interactions also provide an opportunity to reinforce the importance of drug adherence.

Issues related to clinical monitoring for patients on INH are discussed further separately. (See "Isoniazid: An overview".)

INITIATION OF ANTIRETROVIRAL THERAPY — 

Treatment for TBI may be started at any time, regardless of antiretroviral therapy (ART) status. We prefer to stagger initiation of ART and TBI treatment, to allow for assessment of side effects should any occur. For patients with a new diagnosis of HIV and TBI, we begin ART promptly and initiate TBI treatment a month later. (See 'Selecting a regimen' above.)

Early initiation of ART has been associated with lower incidence of active TB among patients with HIV in TB-endemic areas [15,17,27,84-86]:

●In a randomized trial including more than 2000 patients with HIV in the Ivory Coast with CD4 <800 cells/mm³, both ART and six months of isoniazid preventive therapy were independently associated with lower rates of active TB [27]. There was a significant benefit of early ART even among participants with CD4 of 500 to 800 cells/mm³ at enrollment. (See "Use and impact of antiretroviral therapy for HIV infection in resource-limited settings", section on 'Tuberculosis'.)

●In a randomized trial including more than 4600 patients with HIV around the world with CD4 counts >500 cells/mm³, active TB developed less frequently among those who received immediate ART than among those who started ART only when their CD4 count dropped to 350 cells/mm³ or when they developed a clinical condition that required ART (14 versus 20 percent) [87].

SPECIAL CIRCUMSTANCES

Drug-resistant TBI — Issues related to drug-resistant TB are discussed above and separately. (See "Treatment of tuberculosis infection (latent tuberculosis) in nonpregnant adults without HIV infection", section on 'Drug-resistant tuberculosis infection (TBI)'.)

Pregnancy — Issues related to TBI during pregnancy among patients with HIV are discussed separately. (See "Tuberculosis disease (active tuberculosis) in pregnancy".)

SOCIETY GUIDELINE LINKS — 

Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Diagnosis and treatment of tuberculosis".)

SUMMARY AND RECOMMENDATIONS

●Rationale – Treatment of tuberculosis infection (TBI) in patients with HIV is associated with two important benefits: reducing the risk for progression to active tuberculosis (TB) disease (which is greater among individuals with HIV those without HIV) and reducing TB transmission. (See 'Rationale for TBI treatment' above.)

●Whom to treat – We recommend TBI treatment for nonpregnant adults with HIV in the following circumstances (see 'Whom to treat' above):

•Individuals with recent contact with a person with active TB disease (Grade 1A)

•Individuals with no signs or symptoms of active TB and clinical suspicion for prior TB (eg, fibrotic disease on chest radiograph consistent with healed TB) with no documented history of adequate TB treatment (Grade 1A)

•Individuals with positive tuberculin skin test (TST) or interferon-gamma release assay (IGRA) in the absence of active TB (Grade 1A)

•If TBI testing is not available − All individuals with HIV infection (particularly if CD4 cell count <200 cells/microL) and risk of TB exposure, once active TB has been excluded (Grade 1B)

●Exclude active TB – Patients must be evaluated for active TB disease prior to initiation of TBI treatment, to avoid monotherapy and risk of TB drug resistance. We strongly advise chest radiography; respiratory samples for acid-fast bacilli smear and TB culture should also be obtained. (See 'Exclude active TB disease' above.)

●Approach in most clinical settings - For treatment of TBI in patients with HIV infection on antiretroviral therapy (ART), we suggest a rifamycin-based regimen over isoniazid (INH) monotherapy (table 3) (Grade 2B), given the efficacy, favorable completion rates, and relatively low hepatotoxicity rates.

However, use of such regimens requires careful attention to drug interactions. Treatment decisions should be guided by the patient's current ART regimen; in some circumstances, we consider switching to an alternative regimen with minimal rifamycin-based drug interactions for the duration of TBI treatment. (See 'Regimen selection and ART considerations' above.)

•Rifamycin-based regimens

-Use of 3HP – The regimen 3HP (INH and rifapentine [RPT] administered weekly for 3 months via directly observed therapy [DOT]) is preferred for treatment of TBI in patients with HIV infection. 3HP may be given to patients on a raltegravir- or efavirenz- (EFV) based regimen; it may also be given to patients on a dolutegravir-based regimen (without dose adjustment, administered once daily), if virologically suppressed. Patients on tenofovir alafenamide would need to switch for tenofovir disoproxil fumarate for the duration of TBI treatment.

-Use of 4R – The regimen 4R (rifampin [RIF] self-administered daily for four months) may be used if administration of 3HP via DOT is not feasible. 4R may be given to patients on an EFV-based regimen (without dose adjustment). Patients on a dolutegravir or raltegravir-based regimen require a doubling of the dose. 4R may not be used for patients on bictegravir, TAF, and a number of other antiretroviral agents.

-Use of 1HP – The regimen 1HP (INH and RPT self-administered daily for one month) is a potential option for treatment of TBI in patients with HIV infection; more data on safety and tolerability of this regimen is needed.

•Alternative regimens

-Use of INH – For patients who are unable or unwilling to switch to a rifamycin-compatible ART regimen, we suggest INH monotherapy (daily for nine months) for treatment of TBI (Grade 2B). The World Health Organization (WHO) favors daily INH for six months because of better completion rates, but this carries risk of lower efficacy. The United States Department of Health and Human Services Opportunistic Infection Guidelines also list INH as preferred therapy.

-Use of 6LFX – The regimen 6LFX (levofloxacin administered daily for six months) may be used for individuals in whom rifamycin-based therapy cannot be given, as an alternative to INH monotherapy.

•No role for 3HR − We do not favor use of 3HR for treatment of TBI in patients with HIV infection; this regimen confers no benefit over INH monotherapy in terms of efficacy or toxicity, but it does confer risk of rifampin-induced drug interactions. This approach differs from the WHO, which suggests use of 3HR in settings where rifamycins are only available as fixed dose combinations of INH and RIF.

●Monitoring and adherence (See 'Monitoring and adherence' above.)

•Symptoms − Patients receiving TBI treatment should have monthly monitoring for symptoms of hepatitis. In addition, patients should be educated to stop their medication and seek prompt evaluation if they develop such symptoms.

•Laboratory testing − We perform baseline liver enzyme testing for patients on ART. For patients with abnormal baseline liver enzymes and for patients with risk for hepatotoxicity, we perform subsequent laboratory monitoring every two to four weeks.

•Treatment discontinuation − TBI treatment should be discontinued for symptomatic patients with serum transaminase levels greater than three times the upper limit of normal and for asymptomatic patients with serum transaminase levels greater than five times the upper limit of normal. TBI treatment should not be resumed even after liver function tests have improved.

●Initiating ART – Treatment for TBI may be started at any time, regardless of ART status. For patients with a new diagnosis of HIV and TBI, we begin ART promptly and initiate TBI treatment a month later. (See 'Initiation of antiretroviral therapy' above.)