Introduction

Almost a quarter of the world’s population is infected with Mycobacterium tuberculosis, including approximately 13 million people in the USA [1, 2]. Most are asymptomatic and are classified as having latent TB infection (LTBI); however, understanding of the dynamics of tuberculosis (TB) infection has evolved to describe a continuum between “latent” and “active” states that includes a broad range of clinical scenarios [3,4,5,6].

People living with HIV (PWH) are 15–22 times more likely to develop active TB than people without HIV, and worldwide, TB is the leading cause of death among PWH [7]. Active TB disease can develop after recent exposure to Mycobacterium tuberculosis organisms (primary disease) or with reactivation of latent infection. The annual risk of TB disease due to reactivation of latent infection for persons with untreated HIV is approximately 3–16% per year, nearly the lifetime risk of TB (5–10%) among persons without HIV [8]. The increased risk of active TB begins in the first year after HIV infection and rises with progressive immunodeficiency [9, 10].

Furthermore, despite widespread availability of antiretroviral therapy (ART), the risk of mortality among PWH who develop tuberculosis is higher than among those with tuberculosis alone [11]. Globally, approximately 300,000 people with HIV infection died from TB in 2018 [12], even though TB is both preventable and treatable. The End TB strategy of the World Health Organization (WHO) calls for an 80% reduction in TB incidence and 90% reduction in mortality by 2030, both of which will require a dramatic increase in the number of people offered preventive therapy, and universal access to TB diagnosis and treatment for those with active disease [13].

Prevention of Tuberculosis in Persons with HIV

Antiretroviral therapy and TB-preventive therapy (TPT) are both effective interventions to prevent active TB disease in PWH. Antiretroviral therapy has been shown to reduce the risk of active tuberculosis in PWH by 67% across a wide spectrum of disease stages and CD4 cell counts [14]. TPT involves the use of one or more antituberculous drug to treat persons with LTBI who are at high risk of progression to active disease. The efficacy of TPT was first demonstrated over 60 years ago when isoniazid-preventive therapy (IPT) was used to reduce the risk of TB among Alaskan villages, household contacts, and persons living in mental health facilities [15]. Among PWH, 6 to 9 months of IPT reduces not only TB incidence but also mortality by up to 37%, regardless of CD4 cell count or treatment with antiretroviral therapy [16•, 17]. In combination with ART, TPT reduces the risk of TB disease among PWH by 76% [18].

Despite this large body of evidence, uptake of TPT for people with HIV infection is abysmally low. The often cited reasons for not providing TPT include concerns about first excluding active TB disease, poor patient adherence to long courses of therapy, drug toxicity, and unfounded fears about the emergence of drug resistance [19]. Prevention of TB disease by screening for and treating LTBI, along with ART initiation, are key components of HIV care [20•, 21•]. While access to ART has increased worldwide, screening and treatment of LTBI fall short with less than 20% of patients intended for LTBI screening completing treatment as illustrated in Fig. 1 [22]. In recent years, new shorter and better-tolerated regimens have been developed and are now recommended by the WHO, the Centers for Disease Control and Prevention (CDC), and the US Department of Health and Human Services (DHHS) [20•, 21•, 23•]. Herein, we present a summary of the current options for treating LTBI among PWH along with innovations in the field of TB-preventive therapy.

Fig. 1
figure 1

LTBI cascade of care (Adapted from Alsdurf et al. [22] with permission from Elsevier)

Full size image

Screening for LTBI in Persons with HIV

The WHO and the US DHHS recommend screening all PWH for LTBI at the time of HIV diagnosis, regardless of their epidemiologic risk factors or TB exposure history [20•, 21•]. In addition to symptom screening, the two currently available methods for the diagnosis of LTBI are the tuberculin skin test (TST) and the interferon gamma release assay (IGRA). IGRA testing is expensive and not widely available in low- and middle-income countries, and reagents for TST are often in short supply; therefore, the WHO guidelines do not require LTBI testing before initiating TPT for PWH [21•]. Moreover, PWH have been shown to benefit from TPT even with negative LTBI testing [16•, 24•, 25].

All individuals with HIV who have a new positive test for LTBI should be offered LTBI treatment after active TB disease is ruled out [20•, 21•]. Additionally, the US DHHS recommends treatment of LTBI for PWH who report close contact with anyone with infectious TB, regardless of their TB screening test results [20•]. PWH in the USA who have negative screening tests for LTBI and no recent contact with anyone with infectious TB are unlikely to benefit from treatment of LTBI [26,27,28,29].

While both the TST and IGRA help differentiate those with LTBI from those without LTBI, the diagnostic accuracy of both tests is limited. It is important to note that these tests do not distinguish between LTBI and active disease. Additionally, negative LTBI testing does not definitively rule out underlying disease.

Although the TST has been utilized for years, it has several disadvantages: patients must return for follow-up to have the test read by a provider, false-positive results can occur following exposure to other non-tuberculous mycobacteria and immunization with bacilli Calmette-Guerin (BCG), and false-negative results can occur among persons with advanced immunodeficiency and active TB disease [30, 31]. While the TST is still a recommended screening test, these limitations have led to increased use of IGRAs as a screening tool for LTBI in recent years.

Two IGRAs have been approved by the Food and Drug Administration (FDA) and are currently available in the USA: the QuantiFERON-TB Gold Plus and the T-SPOT.TB assay. The IGRAs offer some advantages over the TST including higher specificity, higher correlation with surrogate measures of exposure to TB disease, and less cross-reactivity with non-tuberculous mycobacteria and the BCG vaccine [32,33,34]. However, as with the TST, sensitivity is lower in the setting of advanced immunodeficiency and active TB disease.

Treatment Options for LTBI in PWH

There are several treatment options for PWH with LTBI; underlying comorbidities, drug–drug interactions, toxicities, and adherence should be considered when selecting a regimen.

For many years, isoniazid-preventive therapy (IPT) has been the most widely utilized and recommended regimen for the treatment of LTBI in people with HIV infection. Recent guidelines from the US National TB Controllers Association and CDC now recommend short-course rifamycin-based treatment over INH, but these recommendations do not account for the issues affecting PWH [23]. INH is effective and well tolerated, and severe toxicity is rare. Furthermore, isoniazid can be safely co-administered with any antiretroviral regimen without added risk of toxicity or drug–drug interactions. Peripheral neuropathy, hepatitis, and rash are the most common toxicities, and peripheral neuropathy can largely be prevented via supplementation with pyridoxine at a dose of 25 to 50 mg/day. Recommendations regarding the duration of therapy vary from 6 to 9 months [20•, 21•, 35]. The most significant disadvantage of the 9-month isoniazid regimen is that the majority of patients do not complete therapy [36]. Studies have shown higher completion rates with shorter regimens [36,37,38,39], and such regimens are now available (see Table 1).

Table 1 Tuberculosis-preventive therapy and compatible antiretroviral regimens
Full size table

A 4-month course of daily rifampin is recommended by the CDC and considered an alternative to IPT by the WHO and the US DHHS. A large, open-label, randomized trial comparing 4 months of rifampin (4R) with 9 months of isoniazid (9H) was recently completed, and the 4R regimen was non-inferior to 9H for the prevention of active TB despite low rates of incident active TB in both arms [50]. Treatment completion rates were significantly higher, and adverse events were less common in the 4R arm than in the 9H arm. Despite the large number of participants (> 6000 from 9 countries), only 255 had HIV, and this limits the generalizability of the findings to PWH. Furthermore, rifampin is a potent inducer of the cytochrome P450 (CYP 450) hepatic enzyme system, so ART regimens may need to be adjusted if 4R is utilized for treatment of LTBI. The US DHHS guidelines caution against starting rifampin monotherapy before ruling out active TB disease, possibly with a sputum culture or chest radiograph in addition to symptom screening, given the theoretical risk of drug resistance if rifampin monotherapy is administered in undiagnosed early stage TB disease [20•].

A 3-month course of daily isoniazid and rifampin (3HR) is also recommended by the CDC and considered an alternative regimen for the treatment of LTBI among both adults and children by the WHO [21•, 23•]. A recent systematic review showed similar efficacy rates for the 3HR regimen as compared with 6 months of IPT with comparable rates of adverse events and higher completion rates [51]. As with the 4R regimen, ART regimens may need to be adjusted because of rifampin’s potent induction of the CYP 450 enzyme system.

Rifapentine is a rifamycin with a longer half-life than rifampin and has been utilized in combination with isoniazid for the treatment of LTBI with shorter regimens that can be dosed less frequently. Rifapentine plus isoniazid once weekly for 12 weeks (3HP) has been studied in several randomized controlled trials (RCTs) and is recommended for the treatment of LTBI by the CDC and WHO and considered an alternative by the US DHHS. In two RCTs, 3HP was as effective and well tolerated as 6 to 9 months of daily IPT, including in ART-naïve PWH [52•, 53•]. Furthermore, the risk of hepatotoxicity was lower and completion rates were higher in the 3HP regimen. Although 3HP was administered via directly observed therapy (DOT) in these trials, a recently published study demonstrated non-inferior treatment completion and safety of 3HP via self-administered therapy (SAT) as compared with 3HP via DOT in persons living in the USA [54]. Although rifapentine is also a potent inducer of the CYP450 enzyme system, pharmacokinetic (PK) studies suggest that 3HP can be co-administered with efavirenz (EFV), raltegravir (RAL), and dolutegravir (DTG)-based ART regimens without dose adjustment [46,47,48].

The most recently studied regimen of an ultrashort course of therapy, 1 month of daily rifapentine plus isoniazid (1HP), is under consideration as a recommended regimen in national and international guidelines. The BRIEF-TB study (ACTG 5279) compared 1HP to 9 months of daily isoniazid in PWH and found that 1HP was non-inferior for the primary endpoint of prevention of active TB disease [24•]. Additionally, treatment completion rates were the highest ever reported in a preventive therapy trial (97%), and there was a lower incidence of adverse events in the 1HP arm. This regimen has not yet been evaluated in HIV-uninfected people, or in low TB prevalence settings. Only 23% of participants had a positive test for LTBI in the study; however, this was the largest trial of rifapentine in PWH and had the largest numbers of individuals with confirmed LTBI by TST or IGRA. Furthermore, the number of endpoints in BRIEF-TB was significant and demonstrates that the population enrolled was indeed at risk for active TB disease.

Drug Interactions

Pharmacokinetic drug–drug interactions between medications used for TPT and antiretroviral medications are common and can lead to either increased or decreased drug exposure. Isoniazid results in few drug–drug interactions with most antiretrovirals, and no dose adjustment is needed, but rifampin and rifapentine are both potent inducers of the CYP450 enzyme system and the uridine diphosphate glucuronosyltransferase (UGT) 1A1 enzyme, and thus result in numerous drug–drug interactions. The CYP450 system is responsible for the metabolism of many commonly prescribed antiretrovirals including nucleoside reverse transcriptase inhibitors (NRTIs), protease inhibitors (PIs), elvitegravir, and maraviroc [20•]. Raltegravir is primarily metabolized via UGT1A1, and bictegravir and dolutegravir are metabolized by both CYP3A4 and UGT1A1. The ART regimens that are compatible with each TPT regimen are outlined in Table 1 along with the evidence to support their use. Importantly, bictegravir is not recommended in combination with either rifampin or rifapentine due to significantly reduced exposure [55].

Although most NRTIs are compatible with rifampin-containing LTBI regimens, the effect of co-administration of tenofovir alafenamide (TAF) and rifamycins is unclear. Although co-administration of TAF and rifampin results in decreased plasma exposure of TAF, a study of 23 healthy volunteers results in higher intracellular levels of tenofovir when TAF was administered with rifampin as compared with tenofovir disoproxil fumarate [40].

Special Circumstances

Children

Although the majority of the evidence regarding TPT in children is based on a 6-month course of IPT, 3HR has been shown to be as effective as 6 months of IPT and was associated with a lower risk of adverse events and a higher adherence rate [56]. 3HP is recommended for children over the age of 2, and 1HP has been studied in children over the age of 13 years. A child-friendly dispersible formulation of rifapentine is under study (TBTC Study 35, NCT03730181), and the safety and pharmacokinetics of 1HP in children will be studied by the NIH-funded IMPAACT network (C5019).

Pregnancy

Pregnant women living with HIV and LTBI are at high risk of progression to active TB disease, but the safety, efficacy, and optimal timing of TPT in pregnant women with HIV remain unclear. An RCT of IPT during pregnancy (immediate) or after delivery (deferred) in 956 women with HIV on ART revealed higher than expected rates of treatment-related adverse events as well as higher adverse pregnancy outcomes in the immediate IPT arm [57]. Post hoc analysis of 126 pregnancies among women who received either 3HP or IPT in the PREVENT TB and iADhere trials did not reveal worse pregnancy outcomes, but only 1 participant in this analysis had HIV [58]. The US DHHS Guidelines for Opportunistic Infections recommend against the use of rifapentine in pregnancy because of fetal loss and malformations in animal studies [20•]. A study of the safety and pharmacokinetics of 1HP versus 3HP in early versus late pregnancy is in development by the IMPAACT network (CS5021).

Drug-Resistant TB

Recent modeling data estimates that, globally, three in every 1000 people carry latent multidrug-resistant (MDR) tuberculosis infection, and the ideal treatment strategy for MDR LTBI is unknown [59]. The WHO recommends selecting a regimen based on the drug resistance profile of the source case, but additional data are needed [21•]. Three large randomized clinical trials are underway to answer this important question, as effective treatment of MDR LTBI is a critical component of efforts to eliminate TB disease (NCT03568383, ISRCTN92634082, and ACTRN12616000215426).

Research Gaps/Future Directions

Although effective treatment and prevention for TB are available, the impact of these on the global epidemic has been limited and progress is very slow. Recent studies in TB vaccines show some promise, although the number of people with HIV infection included in any of the trials is very limited [60, 61]. In the absence of an effective vaccine, we must rely on provision of preventive therapy to all PWH at risk and prompt diagnosis and treatment of those with active disease.

While significant progress has been made in the development of TPT regimens that are shorter and more easily tolerated, novel strategies are needed to improve LTBI treatment completion rates. Access to rifapentine is limited globally, and the drug is much more expensive than either INH or rifampin [62]. Child-friendly drug formulations are needed, as well as more investigation of TPT in pregnancy. Even shorter courses of TB-preventive therapy that include some newer TB drugs are effective in a mouse model and could be evaluated in human trials [63]. Long-acting drug formulations have proven to be successful in the treatment of HIV, and the application of this technology to the field of TB prevention could have a substantial impact [64, 65]. A long-acting/extended release tuberculosis working group was established in 2015 under the umbrella of the Long-Acting/Extended Release Antiretroviral Assistance Resource Program (LEAP) and recently received funding from Unitaid to develop long-acting drugs for the prevention of TB [66, 67].

National guidelines recommend screening for LTBI in all PWH, but current testing rates are well below target [22, 68]. Although transition from the TST to IGRA-based screening has improved adherence to screening guidelines in many areas where IGRA testing is available, further work is needed to address this first step off in the LTBI treatment cascade.

Additionally, better diagnostic tools are needed as the current tools have limited sensitivity and specificity and do not distinguish between those with subclinical progressing infection and those with nonprogressing latent infection. Without a better biomarker to identify those at high risk of disease progression, the number needed to treat in order to prevent TB is very large, especially in low-prevalence settings [69]. Although many tools ranging from 18F-fluorodexoyglucose positron emission tomography imaging [70] to quantitative blood plasma-based cytokine and chemokine studies [71] to gene expression markers [72] have been explored, results are inconsistent and further research is needed.

An important knowledge gap that remains is the utility of repeat courses of TPT in PWH. Continuous isoniazid has been shown to provide more durable protection in high-burden settings, but uptake of such a regimen is negligible and not recommended outside of a few high-burden countries [73]. Repeat administration of shorter regimens such as 3HP may be associated with a more durable protection with higher completion rates, and one study of periodic 3HP versus a single course of 3HP is currently under way [74].

Conclusions

Improvements must be made in the cascade of LTBI treatment if we are to make progress in reducing the global burden of TB disease. ART and TPT are an essential component of LTBI treatment, and both ART and TPT regimens can be optimized to avoid drug–drug interactions, decrease pill burden and frequency, shorten the duration of therapy, and minimize side effects in order to increase adherence and treatment completion rates.