Tuberculosis disease among people with HIV: therapeutic advances

Abstract

Over the past 80 years, tuberculosis treatment has evolved with the development of all-oral treatments, which are now given for 4–6 months for drug-sensitive tuberculosis and 6–9 months for drug-resistant tuberculosis. Treatment success is often reduced among people with HIV due to an interplay of factors, including immune dysregulation, lower drug concentrations, complexities of cotreatment (eg, high pill burden and overlapping toxicities), and social factors. Recent clinical trials have shown that among adults and adolescents, treatment duration can be decreased to 4 months with repurposed therapeutics for drug-sensitive tuberculosis, and a four-drug regimen of isoniazid, rifapentine, moxifloxacin, and pyrazinamide has become part of WHO recommendations. Among children with drug-sensitive, non-severe tuberculosis disease, a 4-month regimen of standard tuberculosis drugs (eg, isoniazid, rifampicin, pyrazinamide, and ethambutol) is non-inferior to a 6-month regimen. Following recent research advances for drug-resistant tuberculosis, a 6-month regimen containing a potent combination of bedaquiline, pretomanid, linezolid, and moxifloxacin is a new standard for people with and without HIV. The tuberculosis drug development pipeline contains promising new therapeutics in various stages of development. To accelerate tuberculosis elimination, future research should focus on shortened treatment duration, and safer and effective therapeutics for tuberculosis-affected populations globally, including people with HIV, children, and pregnant people, and should assess newer modalities of treatment delivery.

Introduction

Tuberculosis has remained the most prominent cause of death due to an infectious disease in the past two decades, which was only briefly exceeded by SARS-CoV-2 during the COVID-19 pandemic.¹ The overall global burden of tuberculosis has declined only slightly over the past two decades and the absolute burden remains substantial in many low-income and middle-income countries (LMICs). In 2023, the total burden of tuberculosis was 10·8 million cases, with people with HIV accounting for 6% of cases, and with less than 5% of cases due to multidrug-resistant and rifampicin-resistant tuberculosis.² Globally, 1·25 million people with tuberculosis died in 2023, of which 161 000 were people with HIV, highlighting the substantial population-level effect of active tuberculosis disease among people with HIV, accounting for 25% of HIV mortality globally.¹

Among people with HIV, unfavourable tuberculosis treatment outcome rates are higher than for people without HIV.^1,3 Furthermore, treatment outcomes for people with HIV and drug-resistant tuberculosis have been historically poor, with low long-term survival rates.² Overall, the high number of adverse tuberculosis treatment outcomes among people with HIV has been attributed in part to delayed treatment of HIV, high pill burden with cotreatment, drug–drug interactions, overlapping toxicities between antituberculosis therapy and antiretroviral therapy (ART), other intercurrent infections, and the disseminated nature of tuberculosis among people with HIV.^4,5 Although the pill burden for HIV treatment has reduced substantially with the single-pill fixed-dose combinations in many countries with high tuberculosis burden,⁶ the tuberculosis treatment pill burden still remains a concern, with the current standard of rifamycin requiring 6 months of treatment.

WHO and the UN set ambitious targets for ending the tuberculosis epidemic globally, including treating 40 million people with tuberculosis: 3·5 million children and 1·3 million people with multidrug-resistant tuberculosis (including 115 000 children); between 2018 and 2022.⁷ Overall, 84% of treatment targets (71% for children) were achieved by 2022. However, only 55% of treatment targets for multidrug-resistant tuberculosis were met, and only 19% of treatment targets were achieved for children with multidrug-resistant tuberculosis.¹ In September, 2023, a second High-Level UN meeting set a new and more ambitious target of 97% coverage of tuberculosis treatment by 2027.⁸ To meet these targets, improved regimens with shorter duration, lower pill burdens, fewer drug–drug interactions, and minimal toxicity need to be prioritised.

Because tuberculosis tends to occur in people with moderate to severe immunodeficiency, combining antituberculosis therapy and ART is essential for optimal outcomes, although for many years clinicians treated tuberculosis first and deferred ART for long periods. A series of clinical trials in the early 2000s showed that initiating ART early in the course of treating people with HIV and tuberculosis—2–8 weeks after starting tuberculosis therapy—resulted in better survival, particularly as early as 2 weeks for those with advanced HIV.^9,10 Subsequent research has focused on managing drug–drug interactions between antituberculosis and ART regimens to facilitate concomitant treatment of both diseases.

ART for people with HIV and tuberculosis

Rifamycin antibiotics (eg, rifampicin, rifapentine, and rifabutin) are the mainstay of tuberculosis treatment due to the sterilising activity of this class of drugs, but they complicate HIV–tuberculosis cotreatment because they are potent inducers of hepatic cytochrome P450 enzymes, as well as P-gp and UGT1A1: enzymes and transporters involved in antiretroviral drug disposition (table 1). Dolutegravir (a WHO-recommended first-line integrase strand inhibiter used globally) is predominantly metabolised by UGT1A1 with a minor contribution of CYP3A4.²⁶ The activity of dolutegravir is most closely correlated with trough concentrations. Co-administration of rifampicin and dolutegravir (50 mg twice daily) results in decreased dolutegravir area under the curve (AUC), maximum concentration (C_max), and trough concentration (C_trough) compared with twice-daily dolutegravir alone, but increased these parameters compared with daily dolutegravir 50 mg alone. The INSPIRING study showed that receiving dolutegravir 50 mg twice daily was effective in ART-naive people with HIV and tuberculosis as virological success was achieved by 81% of participants at week 24 and 75% of participants at week 48.¹³ Although these rates of suppression are lower than the virological success typically achieved in trials among patients with HIV alone, these rates are typical for patients with HIV-associated tuberculosis. Global guidelines now recommend dolutegravir 50 mg twice daily for people with HIV when co-administered with rifampicin.²⁷ In 2023, the RADIANT-TB study showed similar rates of HIV viral suppression in people with HIV and tuberculosis, ART-naive patients concomitantly treated with rifampicin, and dolutegravir 50 mg taken once daily versus twice daily, suggesting that once-daily dosing might be adequate for most patients, perhaps because of the prolonged binding time of dolutegravir to its target.²⁸ Real-world data similarly support the notion that once-daily dosing is adequate.²⁹

Table 1:

Considerations for antiretroviral and rifamycin cotreatment in people with HIV

	Metabolism	Effect of rifamycin on antiretroviral concentrations			Comments
		Rifampin	Rifapentine	Rifabutin
Integrase strand transfer inhibitor
Dolutegravir	UGT1A1 (nearly 80%); CYP3A1; P-gp	Rifampin with dolutegravir (50 mg) twice daily vs dolutegravir (50 mg) alone once daily in healthy volunteers:11 dolutegravir AUC increases by 33%; dolutegravir Cmin increases by 22%; rifampin with dolutegravir (50 mg) once daily vs dolutegravir (50 mg) alone twice daily in healthy volunteers:11 dolutegravir AUC decreases by 54%; dolutegravir Cmin decreases by 72%	Rifapentine (900 mg) once weekly (with 3-month isoniazid and rifapentine) with dolutegravir (50 mg) once daily: dolutegravir AUC reduces by 26%; dolutegravir Cmin reduces by 47%; rifapentine (600 mg) once daily (with 1-month isoniazid and rifapentine) with dolutegravir (50 mg) twice daily:12 dolutegravir AUC and Cmin increases by 5%	Rifabutin (300 mg) once daily: dolutegravir AUC remained the same; dolutegravir Cmin reduced by 30%	Double recommended dolutegravir dose to 50 mg twice daily with rifampin: tenofovir disoproxil fumarate (300 mg), lamivudine (300 mg), and dolutegravir (50 mg) combination one-pill morning dose plus evening dose of dolutegravir (50 mg) for people with HIV >30 kg;13,14 evidence for dolutegravir double dose in children >6 years;15 consider alternative to rifampin (eg, rifabutin) in people with HIV with integrase strand transfer inhibitor-associated resistance mutation; evidence from RADIANT-TB trial suggests that once-daily dose of dolutegravir is sufficient in people receiving rifampin-based therapy: no dolutegravir dose adjustment needed if co-administered with rifabutin; once-weekly rifapentine (with 3-month isoniazid and rifapentine) and once-daily rifapentine (with 1-month isoniazid and rifapentine) for tuberculosis preventive therapy: no dolutegravir dose adjustment is needed
Bictegravir	UGT1A1; CYP3A4; P-gp	In HIV-negative volunteers:16bictegravir AUC reduces by 60%; Ctrough reduces by 80%	Significant decrease in bictegravir concentrations	Rifabutin (300 mg) once daily:17 bictegravir AUC reduced by 38%; bictegravir Cmin reduced by 56%	Current guidelines do not recommend co-administering with rifampin, rifapentine, and rifabutin; INSIGHT-TB trial (CAPRISA 093) showed high efficacy of bictegravir, emtricitabine, and tenofovir alafenamide combination twice daily in people with HIV and tuberculosis receiving rifampin as 94% (75 of 80) of participants had a viral load <50 copies/mL at week 24
Raltegravir	UGT1A1	Raltegravir (400 mg) twice daily:18 raltegravir AUC reduces by 40%; raltegravir Cmin reduces by 61%	Rifapentine (600 mg) once daily: raltegravir Cmin reduces by 41%	Raltegravir AUC increases by 19%; raltegravir Cmin reduces by 20%	Double raltegravir dose to 800 mg twice daily; REFLATE TB showed raltegravir (400 mg) twice daily might be an alternative to efavirenz for the treatment of HIV and tuberculosis;18,19 double weight-based raltegravir dose in children older than 4 weeks with rifampin;20 do not co-administer with once-daily rifapentine; no raltegravir dose adjustment with rifabutin21
Cabotegravir	UGT1A1 (main); CYP3A1; P-gp	Oral cabotegravir (30 mg): AUC reduces by 59%; Cmin decreases by 50%; intramuscular cabotegravir: no data; significant decrease in cabotegravir expected	No data for rifapentine; significant descrease in cabotegravir expected	Oral cabotegravir: cabotegravir AUC reduces by 23%; cabotegravir Cmin reduces by 26%; intramuscular cabotegravir: reduced intramuscular cabotegravir and rilpivirine concentration expected	Contraindicated with rifampin and rifapentine for both oral and intramuscular cabotegravir; no cabotegravir dose adjustment for oral cabotegravir with rifabutin; contraindicated for rifabutin with intramuscular cabotegravir due to significant rilpivirine concentration reduction
Non-nucleoside reverse transcriptase inhibitor
Efavirenz	CYP 3A4	Efavirenz AUC reduces by 26%	No significant effect on efavirens	Rifabutin AUC reduces by 38%	Efavirenz (600 mg) once daily with rifampin or rifapentine; increased rifabutin dose to 450–600 mg/day
Doravirine	CYP 3A4	Doravirine AUC reduces by 88%	Doravirine AUC reduces by 29%; Cmin reduces by 31%	Doravirine AUC reduces by 50%	Contraindicated with rifampin; no data for doravirine and rifapentine on tuberculosis treatment; double doravirine dose to 100 mg twice daily if cotreatment with rifabutin
Rilpivirine	CYP 3A4	Oral rilpivirine (25 mg) once daily: rilpivirine AUC reduces by 80%; intramuscular rilpivirine: decreased rilpivirine concentration expected	Oral and intramuscular rilpivirine: decreased rilpivirine concentration expected	Intramuscular rilpivirine: decreased rilpivirine concentration expected; oral rilpivirine: no effect on rilpivirine concentration if double dose of rilpivirine (50 mg) is used	Contraindicated for both oral and intramuscular rilpivirine with rifampin or rifapentine; contraindicated for intramuscular rilpivirine with rifabutin; double rilpivirine oral dose to 50 mg once daily if cotreatment with rifabutin
Etravirine	CYP 3A4	Significant decrease in etravirine concentrations possible	Significant decrease in etravirine concentrations possible	Rifabutin metabolite AUC reduces by 37%	Do not co-administer with rifampin or rifapentine; use rifabutin (300 mg) once daily if etravirine is administered without boosted protease inhibitor; do not use etravirine with rifabutin plus boosted protease inhibitor
Protease inhibitor
All protease inhibitors	CYP3A4; P-gp	Decreased protease inhibitor concentration by >75%	Daily (with 1-month isoniazid and rifapentine) and weekly (with 3-month isoniazid and rifapentine): decreased protease inhibitor concentration expected	Unboosted atazanavir: increased rifabutin AUC; ritonavir-boosted atazanavir with rifabutin (150 mg) once daily vs rifabutin (300 mg) alone: rifabutin AUC increased 110%; metabolite AUC increased by 210%; ritonavir-boosted darunavir with rifabutin (150 mg) every other day vs rifabutin (300 mg) alone once daily: rifabutin metabolite AUC increased 881%; ritonavir-boosted lopinavir with rifabutin (150 mg) once daily vs rifabutin (300 mg) alone: rifabutin AUC increased 203%; metabolite increased by 375%; protease inhibitors and cobicistat: reduction in cobicistat expected; increase in rifabutin expected	Contraindicated with rifampin; do not co-administer with rifapentine; rifabutin (150 mg) once daily if cotreatment with unboosted atazanavir; rifabutin (150 mg) once daily if cotreatment with ritonavir-boosted atazanavir, ritonavir-boosted darunavir, or ritonavir-boosted lopinavir; monitor rifabutin toxicity (particularly uveitis and neutropenia) if cotreatment with ritonavir-boosted atazanavir, ritonavir-boosted darunavir, or ritonavir-boosted lopinavir; do not co-administer protease inhibitors and cobicistat with rifabutin
Capsid inhibitor
Lenacapavir	Moderate CYP3A4 inhibitor UGT1A1	Lenacapavir AUC reduces by 85%22	Reduced lenacapavir expected	Reduced lenacapavir expected	Administration of lenacapavir with rifampin is contraindicated;23 concomitant administration of lenacapavir with rifapentine and rifabutin is not recommended
Nucleoside reverse transcriptase inhibitor
Tenofovir alafenamide	P-gp	Tenofovir alafenamide with rifampin vs tenofovir alafenamide alone: plasma tenofovir alafenamide AUC reduces by 55%; plasma tenofovir AUC reduces by 36%; tenofovir alafenamide with rifampin vs tenofovir disoproxil fumarate alone: intracellular tenofovir diphosphate AUC increases 4·2-fold	Decreased tenofovir alafenamide concentration expected	Decreased tenofovir alafenamide concentration possible	Based on intracellular tenofovir diphosphate, standard dose tenofovir alafenamide can be used with rifampin;24,25 ongoing study (NCT04424264; investigating tenofovir alafenamide 25 mg in HIV–tuberculosis receiving rifampin); use with caution with rifapentine and rifabutin and monitor virological response

AUC=area under the curve. C_min=minimal concentration. C_trough=trough concentration.

The integrase strand transfer inhibitor bictegravir is metabolised mainly by CYP3A4 and UGT1A1. Rifampicin reduces bictegravir concentrations, and even twice-daily dosing results in a 60% reduction in trough concentrations (table 1).²⁴ Recent data from the CAPRISA 093 INSIGHT study (n=120),³⁰ in which bictegravir, emtricitabine, and tenofovir alafenamide was given twice daily versus a dolutegravir-based twice-daily regimen showed that both regimens were effective in ART-naive adults with tuberculosis and HIV and CD4 cell count of more than 50 cells per μL, as 94% of participants achieved virological success at week 24. These findings support the use of a bictegravir, emtricitabine, and tenofovir alafenamide twice-daily dose in HIV infection during cotreatment with rifampicin, although this recommendation has not yet been evaluated for inclusion in treatment guidelines.

Tenofovir alafenamide is a P-gp substrate that has been contraindicated with rifampicin, and tenofovir disoproxil fumarate has been the preferred agent (table 1). In a study among healthy volunteers, plasma tenofovir diphosphate AUC was reduced by 54% and C_trough was reduced by 55%, following administration of tenofovir alafenamide and emtricitabine plus rifampicin. However, intracellular tenofovir diphosphate was still 4·2-fold higher compared with tenofovir disoproxil fumarate without rifampicin,²⁴ suggesting that cotreatment of tenofovir alafenamide at standard doses and rifampicin might be acceptable. Although a clinical study of concurrent use of tenofovir alafenamide and rifampicin in people with HIV and tuberculosis is ongoing (NCT04424264), European AIDS Clinical Society guidelines in 2023 recommended tenofovir disoproxil fumarate or tenofovir alafenamide-based ART during cotreatment for HIV and tuberculosis.²⁵ Tenofovir alafenamide has better bone and renal safety than tenofovir disoproxil fumarate and is now a key component of first-line, generic, fixed-dose, three-drug combinations containing tenofovir alafenamide plus dolutegravir and emtricitabine, or bictegravir plus emtricitabine. Tenofovir alafenamide roll-out in many LMICs is feasible, but the risk–benefit balance is still under study in several contexts. Information regarding co-administration of additional ART with first-line tuberculosis therapy is provided in table 1.

Bedaquiline exposure is modulated by CYP3A4 isoenzymes and is affected by both inducers and inhibitors. Efavirenz, a CYP3A inducer, reduces exposure to bedaquiline and its M2 metabolite by up to 50%.³¹ Integrase strand transfer inhibitors, such as dolutegravir, allow the use of bedaquiline without meaningful drug–drug interactions expected. Protease inhibitors increase the AUC of bedaquiline and should be used with caution.³²

Tuberculosis treatment in adults and adolescents with HIV

Although earlier guidelines recommended continuing tuberculosis therapy for 9–12 months, data from the era of effective ART show that a standard 6-month course of isoniazid and rifampicin with pyrazinamide and ethambutol for the first 8 weeks is highly effective in people with HIV. Efforts to develop simpler or shorter regimens to treat drug-susceptible tuberculosis have been promising after initial setbacks.^33,34 A 1999 study of a simplified continuation phase of tuberculosis treatment with once-weekly rifapentine and isoniazid for the last 4 months of treatment found high rates of recurrence with emergence of rifamycin resistance in people with HIV.³³ However, the RIFAQUIN trial found that a 6-month regimen—with moxifloxacin, rifampicin, pyrazinamide, and ethambutol for the first 2 months followed by weekly treatment with rifapentine and moxifloxacin in the continuation phase—successfully treated tuberculosis in almost all patients with and without HIV. However, a 4-month regimen using twice-weekly rifapentine and moxifloxacin in the continuation phase was not effective.³⁴

The TBTC Study 31/ACTG A5349 assessed two 4-month regimens that substituted rifampicin with rifapentine, and moxifloxacin with ethambutol. Treatment outcomes for people with HIV were better in the rifapentine and moxifloxacin group than in the control group, with 85·5% of participants having successful treatment versus 78·1%.³⁵ Outcomes of participants with HIV treated in the rifapentine-only substitution group were inferior, with only 70·6% having successful treatment.36 Most of the participants were on ART, and when outcomes were assessed by CD4 strata, in those with a CD4 cell count higher than 500 cells per μL the odds of unfavourable outcomes were 1·9-fold higher in the experimental 4-month regimen group compared with those with a CD4 cell count of lower than 200 cells per μL.³⁶ This 4-month treatment is now endorsed as an alternative regimen for people with HIV and those with drug-susceptible tuberculosis by the US Centers for Disease Control and Prevention and WHO.^37,38 Another trial of a novel regimen (STAND), with pretomanid, moxifloxacin, and pyrazinamide given for 6 months or 4 months, was stopped early because of high rates of toxicity and inferior outcomes in people with HIV, with a third having unsuccessful treatment compared with 14% in those receiving the standard of care.³⁹ A novel, treatment-shortening study (TRUNCATE-TB) examining 2-month regimens for the primary composite outcome of death, ongoing treatment, or active disease at week 96, showed that an 8-week bedaquiline, linezolid, isoniazid, pyrazinamide, and ethambutol regimen was non-inferior to the standard treatment, although more participants receiving the 8-week regimen required re-treatment. However, this trial did not include people with HIV (table 2).⁴⁰

Table 2:

Global therapeutic clinical trials for drug-sensitive and drug-resistant tuberculosis in adults, adolescents, and children with and without HIV in the past decade

	Trial name	Trial period	Study groups	Region	Age	Sample size (n)		Results*		Summary of findings
						Overall	People with HIV	Overall	People with HIV
Adult drug-sensitive tuberculosis trials
Paton et al (2023)40	TRUNCATE-TB	2018–20	Group 1: standard treatment (2-month HRZE + 4-month HR); group 2: one of the 8-week regimens (rifampicin–linezolid, rifampicin–clofazimine, pretomanid–linezolid, or bedaquiline–linezolid) followed by extension for persistent disease and re-treatment for relapse	India, Indonesia, Philippines, Singapore, Thailand, and Uganda	18–65	675	0	adjRD 7·4 (95%CI 1·7 to 13·2)	NA	A strategy involving initial treatment with an 8-week bedaquiline–linezolid regimen was non-inferior to standard treatment for tuberculosis with respect to the outcome; the strategy was associated with a shorter total duration of treatment and with no evident safety concerns
Metcalfe et al (2024)41	ACTG 5362/CLO-FAST	2021–24	Group 1: PHZE + clofazimine 300 mg for 2 weeks; then PHZE + clofazimine 100 mg for 6 weeks; then pretomanid–isoniazid–pyrazinamide + clofazimine 100 mg for 5 weeks; group 2 (standard of care): 2-month HRZE + 4-month HR; group 3: then PHZE + clofazimine 100 mg for 4 weeks; then treatment per standard of care (2-month HRZE + 4-month HR)	Haiti, India, Malawi, South Africa, and Zimbabwe	≥18	104	30	aHR 1·17 (0·79 to 1·73)	NA	13-week regimen containing clofazimine and rifapentine was not as effective as hoped to allow treatment shortening to 3 months
Pettit et al (2023)36	TBTC Study 31/ACTG 5349	2016–21	Group 1: 2-month HRZE + 4-month HR; group 2: 2-month PHZE + 2-month pretomanid–isoniazid; group 3: 2-month PHZM + 2-month pretomanid–isoniazid–moxifloxazine	Brazil, China, Haiti, India, Kenya, Malawi, Peru, South Africa, Thailand, Uganda, USA, Viet Nam, and Zimbabwe	≥12	2516	214	absRD −7·4% (−20·6 to 6·0)	absRD −7·4% (−20·8 to 6·0)	Group 3 regimen was non-inferior to group 1; group 2 regimen was not non-inferior to group 1; CD4 subgroup analysis: odds ratio (OR) 0·61 (0·12–3·00) for 200–499 cells/μL group, and OR 1·90 (0·60–6·03) for ≥500 cells/μL group, with 100–199 cells/μL being the reference category
Tweed et al (2021)39	STAND trial	2015–18	200 mg pretomanid daily, 400 mg moxifloxacin, and 1500 mg pyrazinamide for 6 months (6Pa200MZ) or 4 months (4Pa200MZ); 100 mg pretomanid daily for 4 months in the same combination (4Pa100MZ); 2-month HRZE + 4-month HR	Georgia, Kenya, Malaysia, Philippines, South Africa, Tanzania, Uganda, and Zambia	≥18	284	58	absRD 6·6% (2·2 to 15·4)	NA	Pretomanid–moxifloxacin–pyrazinamide regimens did not achieve non-inferiority in this underpowered trial
Adult drug-resistant tuberculosis trials
Conradie et al (2020)42	Nix-TB	2015–20	6-month BPaL	South Africa	≥14	109	56 (CD4 >50 cells/μL)	Nine (8%) unfavourable outcomes	No effect on the outcome; three participants (8·6%) had unfavourable outcomes	BPaL led to a favourable outcome in high percentage of cases; linezolid toxicity was commonly observed: peripheral neuropathy (81%) and myelosuppression (48%); CD4 subgroup analysis not available
Conradie et al (2022)43	ZeNix	2017–22	6-month bedaquiline–pretomanid with: group 1: 1200 mg linezolid for 26 weeks; group 2: 1200 mg linezolid for 9 weeks; group 3: 600 mg linezolid for 26 weeks; group 4: 600 mg linezolid for 9 weeks	Georgia, Moldova, Russia, and South Africa	≥14	181	36 (CD4 >100 cells/μL)	84% to 93% across all BPaL groups had favourable outcomes	No effect on outcome	Risk-benefit ratio favoured 600 mg linezolid for 26 weeks
Nyang’wa et al (2022)44	TB-PRACTECAL	2017–22	Stage 1: group 1: 6-month BPaLM; group 2: 6-month BPaLC; group 3: 6-month BPaL; group 4: standard of care; stage 2: group 1: 6-month BPaLM; group 2: 9–20 months of standard of care	Belarus, South Africa, and Uzbekistan	≥15	552	153 (any CD4 cell count)	absRD −37 (−53 to −22)	No effect on outcome; absRD: 11·4% (−48·5% to 25·6%)	24-week all-oral regimen was non-inferior and safer than control; CD4 strata outcomes not available
Guglielmetti et al (2024)45	End TB	2016–23	Group 1: 9-month BLMZ; group 2: 9-month BCLLfxZ; group 3: 9-month BDLLfxZ; group 4: 9-month DCLLfxZ; group 5: 9-month DCMZ group 6: standard of care per WHO guidelines	Georgia, India, Kazakhstan, Lesotho, Pakistan, Peru, and South Africa	≥15	754	98 (any CD4 cell count)	BCLLfxZ was superior; BLMZ, BDLLfxZ, and DCMZ were non-inferior	No effect on outcome	Three shortened, all-oral regimens were non-inferior to current, well performing standard of care
Conradie (2024; NCT04062201; unpublished)46	Beat TB	2019–24	Group 1: 6-month BDLLfxC; group 2: South African standard of care (ie, 9-month BHZELLfxC)	South Africa	≥6	400	200 (any CD4 cell count)	Unpublished	No effect on outcome	Non-inferiority was shown
Paediatric drug-sensitive tuberculosis trial
Turkova et al (2022)47	SHINE	2015–20	Group 1: 2-month HRZE + 2-month HR; group 2: 2-month HRZE + 4-month HR	India, South Africa, Uganda, and Zambia	<16	1204	127 (any CD4 cell count)	adjRD −0·4 (−2·2 to 1·5)	No effect on outcome; adjRD −4·3 (−14·9 to 6·2)	4-month regimen was non-inferior to 6-month regimen among children with non-severe tuberculosis; CD4 strata outcomes not available

HRZE=Isoniazid, rifampicin, pyrazinamide, and ethambutol. HR=Isoniazid and rifampicin. adjRD=adjusted risk difference. NA=not applicable. PHZE=pretomanid, isoniazid, pyrazinamide, and ethambutol. aHR=adjusted hazard ratio. PHZM=pretomanid, isoniazid, pyrazinamide, and moxifloxacin. absRD=absolute risk difference. BPaL=bedaquiline, pretomanid, and linezolid. BPaLM=BPaL and moxifloxacin. BPaLC=BPaL and clofazimine. BLMZ=bedaquiline, linezolid, moxifloxacin, and pyrazinamide. BCLLfxZ=bedaquiline, clofazimine, linezolid, levofloxacin, and pyrazinamide. BDLLfxZ=bedaquiline, delamanid, linezolid, levofloxacin, and pyrazinamide. DCLLfxZ=delamanid, clofazimine, linezolid, levofloxacin, and pyrazinamide. DCMZ=delamanid, clofazimine, moxifloxacin, and pyrazinamide. BDLLfxC=BDLLfx and clofazimine. BHZELLfxC=bedaquiline, isoniazid, pyrazinamide, ethambutol, linezolid, levofloxacin, and clofazimine.

^*Results are reported in line with the outcome measure in each paper.

Optimal treatment of tuberculosis and HIV co-infection requires effective combinations of agents for both infections. The strategy of delaying ART until tuberculosis treatment is completed results in high mortality.⁹ Initiating ART within 2–8 weeks of the onset of tuberculosis treatment reduces mortality and improves tuberculosis and HIV outcomes.^9,10 For patients with CD4 cell counts lower than 50 cells per μL, starting ART within 2 weeks results in better survival, whereas for those with higher CD4 counts, survival is similar whether treatment begins within 2 weeks or 8 weeks.⁴⁸ However, rates of adverse reactions, notably immune reconstitution inflammatory syndrome (IRIS), increase when treatment is started earlier. For people with CD4 cell counts lower than 50 cells per μL, the increase in IRIS is offset by an improvement in mortality.⁴⁹ For those beginning ART early, adjuvant treatment with prednisolone reduces the risk of IRIS-related complications in people with CD4 cell counts lower than 100 cells per μL.

WHO recommends that all people with HIV and tuberculosis receive ART within 2 weeks of starting tuberculosis treatment.⁵⁰ This guideline is pragmatic and intends to ensure programmatic implementation of ART uniformly. US and UK guidelines recommend starting ART within 2 weeks for those with CD4 cell counts lower than 50 cells per μL and no later than 8 weeks otherwise.^51,52

Tuberculosis treatment in children with HIV

HIV increases the incidence of tuberculosis in children by approximately eight-fold versus children without HIV. Although ART reduces tuberculosis risk by around 70%, it does not eliminate tuberculosis.⁵³ Children with HIV remain at increased risk for tuberculosis,⁵⁴ have poorer tuberculosis treatment outcomes than HIV-negative children, and face high mortality risk.^55,56 A study in Uganda showed that HIV-positive status more than doubled the risk of mortality in children admitted to hospital with presumed tuberculosis (adjusted hazard ratio [aHR] 2·45; 95% CI 1·37–4·38).⁵⁷ Younger age⁵⁵ and undernutrition are also substantial risk factors for tuberculosis-associated mortality.

The SHINE trial—a multicentre, open-label, randomised controlled trial—compared 4-month versus 6-month treatment regimens with standard drugs in children (aged 3 months to younger than 16 years) with smear-negative, non-severe tuberculosis, using paediatric-dispersible fixed-dose formulations and WHO weight-band dosing recommendations. The median age of participants was 3·5 years, and 127 (11%) had HIV. The 4-month regimen was found to be non-inferior to the standard 6-month regimen, with consistent findings for children with HIV (table 2).^56,47 Children with HIV were more likely to be underweight and have anaemia than children without HIV (median weight-for-age −2·3 kg [IQR −3·3 to −0·8] vs −1·0 kg [−1·8 to −0·2]; p<0·01; haemoglobin 9·5 g/dL [8·7–10·9] vs 11·5 g/dL [10·4–12·3]; p<0·01). Regardless of tuberculosis treatment duration, children with HIV had higher rates of mortality (aHR 2·6; 95% CI 1·2–5·8), and admission to hospital (adjusted odds ratio 2·4; 95% CI 1·3–4·6) than children without HIV. Most children on ART received either efavirenz-based or ritonavir-boosted lopinavir-based regimens, but only 61% achieved a viral load lower than 1000 copies per mL at 48-week follow-up,⁵⁶ highlighting the need for more effective and tolerable ART options. The SHINE findings led to the rapid revision of WHO paediatric tuberculosis treatment guidelines in 2021, strongly recommending short treatment regimen for non-severe tuberculosis among children with or without HIV, and with implementation guidance based on the availability of chest radiography and Xpert MTB/RIF (Xpert; Cepheid, Sunnyvale, CA, USA) or Xpert MTB/RIF Ultra (Xpert-Ultra) assays.⁵⁸ This approach has been adopted by national programmes in many countries, with outcomes of the real-life roll-out awaited.

Adolescents aged 12–17 years with and without HIV (n=63 [3%]) were included in the TBTC Study 31/ACTG A5349 (table 2).³⁶ The 4-month rifapentine-based regimen with moxifloxacin was non-inferior to the standard 6-month regimen, with similar findings in adolescents.³⁵

A nested, pharmacokinetic substudy in the SHINE trial was the first to evaluate exposures of first-line tuberculosis drugs in children with and without HIV using paediatric-dispersible formulations of rifampicin, isoniazid, and pyrazinamide for children under 25 kg.^46,57 Area under the concentration–time curve from 0 h to 24 h with dispersible tablets were low for ethambutol (all weight bands), rifampicin (<12 kg), isoniazid (<8 kg), and pyrazinamide (<8 kg). Children weighing more than 25 kg on adult formulations had low exposures to rifampicin, isoniazid, and ethambutol. There were no significant differences in the exposure by HIV status.⁵⁹

Systematic reviews produced conflicting results on the effect of HIV status on antituberculosis drug exposures in children. Although these reviews confirmed low AUCs for most first-line antituberculosis drugs with WHO-recommended dosing (particularly in children at the lowest and highest weight bands), two reviews concluded that HIV status was associated with lower antituberculosis drug exposures (although one review did not have data on ART).²⁰ However, a model-based, individual data-pooled meta-analysis suggested that reduced antituberculosis drug exposures were associated with concomitant ART, particularly ritonavir-boosted lopinavir, rather than HIV status alone.⁶⁰

Integrase strand transfer inhibitor-based ART is currently recommended as the preferred HIV treatment option for children,^25,61,62 with dolutegravir being the most commonly used agent. HIV and tuberculosis substudies in the ODYSSEY and EMPIRICAL multicentre, randomised controlled trials, showed that twice-daily dolutegravir administration effectively overcomes rifampicin’s enzyme-inducing effect.^63,64 In ODYSSEY, children with HIV-associated tuberculosis receiving double-dose dolutegravir had virological suppression (<400 HIV-1 RNA copies per mL) in 97% of participants at the end of tuberculosis treatment or at 24 weeks after ART initiation, whichever occurred later.⁶³

Boosted darunavir and atazanavir cannot be co-administered with rifampicin, and efavirenz is either contraindicated in children with previous non-nucleoside reverse transcriptase inhibitor exposure (eg, failed vertical transmission prophylaxis or treatment failure) or unavailable for children younger than 3 years. Lopinavir-based treatment with ritonavir super-boosting (lopinavir–ritonavir ratio 4:1) remains the only alternative.⁶⁵ However, single-entity, paediatric ritonavir remains unavailable in most countries, hindering the implementation of this approach. Various strategies, including double-dose ritonavir-boosted lopinavir, semi super-ritonavir-boosted lopinavir (with additional once-daily crushed ritonavir 100 mg), and increased ritonavir-boosted lopinavir dosing and frequency have been explored, but none have shown adequate lopinavir exposures.^66,67 Therefore, a gap remains for alternative ART options for children with HIV and tuberculosis co-infection with dolutegravir resistance or intolerance.

Rifabutin, with a less potent effect on metabolising (UGT and CYP34A) and transporter (P-gp) enzymes, has the potential to replace rifampicin and be co-administered with boosted protease inhibitors,⁶⁸ but its limited availability, particularly in LMICs, precludes its use (table 1).

Drug-resistant tuberculosis in people with HIV

The treatment of multidrug-resistant and rifampicin-resistant tuberculosis has seen remarkable improvements with the advent of new medications and improved diagnostic techniques. In the past decade, three new drugs were approved for treating multi-drug-resistant and rifampicin-resistant tuberculosis: bedaquiline, delamanid, and pretomanid, including the repurposed linezolid as part of the bedaquiline, pretomanid, and linezolid regimen. Shorter, less toxic regimens have replaced the lengthy and arduous treatments of the past, offering new hope for patients and health-care providers. In December, 2022, WHO endorsed the bedaquiline, pretomanid, and linezolid regimen with or without the addition of moxifloxacin and with a daily linezolid dose of 600 mg regimen for the treatment of multidrug-resistant and rifampicin-resistant tuberculosis.³⁷

The groundbreaking Nix-TB trial used the bedaquiline, pretomanid, and linezolid regimen in patients with fluoroquinolone and injectable-resistant multidrug-resistant tuberculosis, and those with multidrug-resistant tuberculosis treatment intolerance or failure.⁴² This single-arm trial conducted in South Africa enrolled 109 participants, 56 (51%) of whom were people with HIV on ART with a baseline CD4 cell count above 50 cells per μL. 98 (90%) participants had a favourable outcome at 6 months post treatment with no difference between patients with and without HIV (table 2). Seven deaths were reported during the trial, with four deaths occurring in people with HIV. A high linezolid dose of 1200 mg daily led to 81% of participants to develop peripheral neuropathy. Overall, 85% of participants had an adverse event requiring interruption or linezolid dose reduction.⁴²

The ZeNix trial sought to optimise the dose and duration of linezolid for the bedaquiline, pretomanid, and linezolid regimen.⁴³ In this partially blinded study, participants were randomly allocated to different doses and durations of linezolid (600 mg for 2 months, 600 mg for 6 months, 1200 mg for 2 months, or 1200 mg for 6 months) along with bedaquiline and pretomanid. 39 (20%) of 181 participants with HIV were on ART. 84–93% of participants across all four bedaquiline, pretomanid, and linezolid treatment groups had a favourable outcome. In the planned subgroup analyses, HIV status did not influence outcomes (table 2).⁴³

The TB-PRACTECAL trial was an adaptive, phase 2–3, randomised controlled trial that evaluated the efficacy and safety of three 24-week all-oral regimens for the treatment of rifampicin-resistant tuberculosis in people with or without HIV.⁴⁴ In stage 2 of the trial, a 24-week regimen of bedaquiline, pretomanid, linezolid, and moxifloxacin was compared with a 9–20-month standard-of-care regimen. In the modified intention-to-treat analysis, 11% of patients in the bedaquiline, pretomanid, linezolid, and moxifloxacin group and 48% of those in the standard-of-care group had an unfavourable outcome, largely due to treatment interruption (table 2).⁴⁴ These pivotal trials show similar safety and efficacy in people with HIV compared with people without HIV, leading to the recommendation of the same dose and duration of treatment.

BEAT TB was a pragmatic randomised controlled trial conducted in South Africa. A six-month regimen containing bedaquiline, delamanid, linezolid, levofloxacin, and clofazimine was compared with the South African standard of care for rifampicin-resistant tuberculosis and pre-extensively drug-resistant tuberculosis (9-month regimen containing bedaquiline, isoniazid, pyrazinamide, ethambutol, linezolid, levofloxacin, and clofazimine).⁴⁶ Over half of participants enrolled in this study were people with HIV. Non-inferiority was shown (table 2). In June, 2024, this evidence was reviewed by the WHO Guidelines Development Group and was suggested as an option for eligible patients regardless of their HIV status. The BEAT-TB trial evidence included children, adolescents, and pregnant and breastfeeding people, flagging the possible use of the regimen in these population groups.⁴⁶

The end TB trial (table 2) was a randomised controlled trial evaluating five different 9-month regimens that found that three regimens—bedaquiline, linezolid, moxifloxacin, and pyrazinamide; bedaquiline, linezolid, levofloxacin, clofazimine, and pyrazinamide; and bedaquiline, delamanid, linezolid, levofloxacin, and pyrazinamide—were non-inferior to controls in patients with drug-resistant tuberculosis in whom fluroquinolone resistance had been excluded. WHO suggested using 9-month all-oral regimens in place of an 18-month regimen, with preference given to bedaquiline, linezolid, moxifloxacin, and pyrazinamide.⁴⁶

Therapeutic considerations among special populations

Pregnant and lactating people

Despite the case burden in pregnancy and postpartum, there remains a dearth of high-quality evidence to inform the treatment of tuberculosis in this setting as pregnant and lactating people have been uniformly excluded from tuberculosis therapeutic trials.⁶⁹ The lack of data on safety of tuberculosis regimens and individual drugs has resulted in inconsistent guidelines for tuberculosis treatment in pregnancy. US guidelines recommend a pyrazinamide-sparing 9-month regimen for drug-susceptible tuberculosis for pregnant people with and without HIV.⁷⁰ In contrast, WHO guidelines recommend a 6-month regimen including pyrazinamide for pregnant people with and without HIV.^37,71

A pharmacokinetic study of 27 pregnant women with tuberculosis found decreases of 25% in isoniazid concentrations and 39% in ethambutol concentrations in the third trimester, but the significance of these decreases is unclear and no dose adjustment during pregnancy is recommended.⁷² Shorter regimens, such as 4-month isoniazid, rifapentine, pyrazinamide, and moxifloxacin are not currently recommended in pregnancy or lactation regardless of HIV status as there are no pregnancy-specific data about the safety or outcomes of this regimen.

Concerns about drug–drug interactions between HIV ART and rifamycins (eg, rifampicin and rifapentine) also appear in pregnancy and lactation with no specific dosing changes recommended, although data on this interaction are scarce.

Treatment of multidrug-resistant and rifampicin-resistant tuberculosis in pregnancy is more complex as certain drugs are avoided due to fetal toxicity concerns.⁷³ Reproductive toxicity studies in animals have observed increased congenital anomalies and CNS toxicities with ethionamide; therefore, the use of this drug is generally avoided. Aminoglycosides cross the placenta and can result in fetal hearing loss, but their use has been described in pregnancy; these injectables, however, should be avoided now that newer all-oral regimens are available. Current WHO guidance recommends a 9-month all-oral regimen with linezolid instead of ethionamide for pregnant people in whom resistance to fluoroquinolones has been excluded.⁴⁶ A systematic review of 275 cases from ten observational studies of multidrug-resistant tuberculosis treatment outcomes during pregnancy observed 72·5% treatment success, 6·8% mortality, 18·4% loss to follow-up, and 0·4% treatment failure.⁷⁴ Reports from case series with newer agents, such as bedaquiline and delamanid, suggest that these agents can be safely used with favourable outcomes, but these data are from small, non-comparative case series.⁷⁵ Short-course regimens of bedaquiline, pretomanid, and linezolid with or without moxifloxacin are not currently recommended during pregnancy or lactation as the safety of pretomanid remains unclear.¹ The BEAT-TB regimen (ie, bedaquiline, delamanid, linezolid, levofloxacin, and clofazimine) and the three End TB trial regimens (ie, bedaquiline, linezolid, moxifloxacin, and pyrazinamide; bedaquiline, linezolid, levofloxacin, clofazimine, and pyrazinamide; and bedaquiline, delamanid, linezolid, levofloxacin, and pyrazinamide) can be used in pregnant people as per the WHO rapid communication in June, 2024.⁴⁶

Few data on tuberculosis drugs are available regarding transplacental transfer, concentrations in breastmilk, and what concentrations pass onto infants. Based on the available data, most tuberculosis drugs result in low concentrations in breastmilk.⁷⁶ However, bedaquiline has very high concentrations in breastmilk and concentrations of the drug might be high in infants as well.⁷⁷

Breastmilk is often an important form of nutrition and there is no contraindication to breastfeeding while on tuberculosis drugs. However, tuberculosis preventive therapy for infants who are exposed to tuberculosis should be considered. For infants born to mothers with HIV who are receiving the prophylactic nevirapine, there is no concerning interaction between maternal rifampicin and infant nevirapine.⁷⁸

Immune reconstitution inflammatory syndrome

After starting ART there is a rapid decline in plasma HIV viral load, accompanied by initial restoration of immune function. During this period, in patients who have recently started tuberculosis therapy, the recovering immune system might drive inflammatory reactions targeted at Mycobacterium tuberculosis antigens at sites of tuberculosis disease. These reactions are termed paradoxical tuberculosis-associated IRIS (TB-IRIS).⁷⁹ Due to the cyclical nature of ART disengagement and re-engagement,⁸⁰ paradoxical TB-IRIS is now frequently also seen in patients who interrupt ART, return to care with active tuberculosis, and those who start tuberculosis treatment and then restart ART. Major risk factors for paradoxical TB-IRIS include low CD4 cell count, high HIV viral load, disseminated or extrapulmonary tuberculosis, and short duration between starting tuberculosis treatment and ART.⁸¹

Paradoxical TB-IRIS symptoms typically commence 1–2 weeks after starting or restarting ART. Symptoms include recurrence of night sweats, fever, cough, malaise, and weight loss, and might be accompanied by new or worsening radiological pulmonary infiltrates. Enlargement of tuberculosis lymph nodes is seen in almost 40% of cases,⁸² and might complicate with suppuration and formation of draining sinuses. Abdominal manifestations include spleen and liver enlargement, peritoneal inflammation, and abdominal lymph node enlargement. Renal granulomatous inflammation with renal impairment might occur.⁸³ Pleural and pericardial effusions might accumulate. Neurological involvement is the most life-threatening form and features include meningitis, enlarging tuberculomas, tubercular abscesses, and cerebral oedema.⁸⁴ In a meta-analysis, 25% of patients who developed paradoxical TB-IRIS required hospitalisation and the condition caused death in 2% of patients.⁸¹

The diagnosis of paradoxical TB-IRIS is clinical and relies on the typical features, temporal relation to ART initiation, and exclusion of plausible differential diagnoses, such as drug-resistant tuberculosis and other opportunistic infections. A consensus case definition for use in research has been published.⁸⁵ Patients with milder forms do not require specific treatment, but can be provided with information and reassurance. For patients with substantial symptoms, corticosteroids provide symptom relief. A small, placebo-controlled trial showed that taking prednisone for 4 weeks (dosed at 1·5 mg/kg per day for 2 weeks followed by 0·75 mg/kg per day for 2 weeks) resulted in significantly reduced duration of hospital stay and outpatient therapeutic procedures, and more rapid improvement in symptoms and radiological features.⁸⁶ Non-steroidal anti-inflammatory drugs have been used to treat paradoxical TB-IRIS although no evidence from trials are available.⁸¹ In patients with refractory paradoxical TB-IRIS despite corticosteroid therapy, a range of immunomodulatory therapies have been used with anecdotal reports of response, including thalidomide,⁸⁷ infliximab,⁸⁸ and anakinra.⁸⁹ In patients with rapidly enlarging effusions due to paradoxical TB-IRIS, needle aspiration might be indicated. Interruption of ART for paradoxical TB-IRIS is rarely necessary, but is considered in life-threatening neurological forms.

Another placebo-controlled, randomised controlled trial showed that moderate-dose prednisone (40 mg/day for 2 weeks followed by 20 mg/day for 2 weeks) was well tolerated and effective in reducing the risk of paradoxical TB-IRIS by 30% in patients treated for HIV-associated tuberculosis who were starting ART and at high risk for paradoxical TB-IRIS (CD4 ≤100 cells per μL and starting ART within 30 days of tuberculosis treatment).⁹⁰

The other form of TB-IRIS is unmasking tuberculosis-associated IRIS. This form occurs in patients with undiagnosed active tuberculosis at the time of ART initiation who present with marked inflammatory features of tuberculosis within the first 3 months of ART. The mechanisms underlying unmasking TB-IRIS—similar to paradoxical TB-IRIS—are thought to be driven by the recovering immune system. Patients should be started on anti-tuberculosis therapy. The role of anti-inflammatory therapy in unmasking TB-IRIS has not been systematically studied.⁷⁹

Tuberculous meningitis

In settings with high HIV prevalence, most adult tuberculous meningitis diagnosed are in people with HIV. Tuberculous meningitis mortality in adults with HIV (around 40%) is 2·5-fold higher than in adults without HIV.⁹¹

International guidelines for tuberculous meningitis treatment in adults recommend rifampicin, isoniazid, pyrazinamide, and ethambutol for 2 months followed by a continuation phase of rifampicin and isoniazid for 7 months. Due to the poor penetration of first-line drugs into the CNS, there are now several ongoing trials of novel antimicrobial regimens for tuberculous meningitis, with many participants being people with HIV. Strategies being evaluated include higher-dose rifampicin, higher-dose isoniazid, or the addition of linezolid. Other trials are evaluating aspirin as a host-directed therapy strategy given that strokes frequently complicate tuberculous meningitis, and that aspirin has both anti-inflammatory and antithrombotic properties.⁹²

Corticosteroids reduce mortality in patients with tuberculous meningitis overall,⁹³ but in a trial evaluating dexamethasone versus placebo that enrolled exclusively people with HIV, there was no significant reduction in mortality or disability (HR 0·85; 95% CI 0·66–1·10).⁹⁴ ART initiation could be complicated by paradoxical TB-IRIS, which carries a high mortality risk.⁸⁴ ART initiation should be deferred in patients with tuberculous meningitis: WHO advises deferring ART 4–8 weeks from the initiation of tuberculosis treatment.⁶¹

Disseminated tuberculosis

Disseminated tuberculosis refers to the spread of M tuberculosis via the bloodstream to multiple organs. The risk of dissemination increases as CD4 cell counts decline. In a meta-analysis, the predicted probability of having a blood culture positive for M tuberculosis was 45% among people with HIV and active tuberculosis treated at hospital.⁹⁵ In postmortem studies among people dying with HIV-associated tuberculosis, mainly in hospitals, 88% had evidence of dissemination to multiple organs, including in the spleen, liver, and lymph nodes.⁹⁶ Dissemination is associated with higher mortality: blood culture positivity and higher M tuberculosis burden in the bloodstream increased mortality risk significantly.⁹⁵ Tuberculosis bacteraemia is among the most common causes of severe sepsis in people with HIV in tuberculosis-endemic settings. In a Ugandan study, 23% of adults with HIV who were hospitalised with severe sepsis had a positive M tuberculosis blood culture.⁹⁷

Disseminated tuberculosis might be associated with granulomatous hepatitis, renal impairment, coagulopathy, and other features of sepsis, including lactate elevation.⁹⁸ An immune profile characterised by elevation of innate immune cell cytokines and chemokines was associated with both increased disseminated M tuberculosis load and mortality risk.⁹⁸

Two trials are currently being conducted evaluating specific treatment strategies for this patient population, including higher-dose tuberculosis drugs (NCT04738812) and corticosteroids (NCT04951986).

Research in progress and future priorities

Despite its burden spanning centuries, advances in tuberculosis therapeutics were emerging at a slow pace until improvements in the past decade. There are now over 41 new tuberculosis-targeted chemical entities in various stages of development (figure).^99,100 The research and development landscape is now poised to make substantial contributions to international guidelines and clinical practice, should they show promising results.

Figure: Current landscape of tuberculosis treatment trials with new chemical entities in adult and paediatric populations.

Drug-sensitive, drug-resistant, and pulmonary disease trials; tuberculous meningitis trials; and disseminated disease trials are included. OBR=Optimised Background Regimen.

Although many trials are assessing new chemical entities, representative inclusion of people with HIV is crucial. A major limitation, however, is that many trials restrict entry of people with lower CD4 cell counts because even a single serious adverse event of often unexplained cause, which occurs more often in people living with HIV, poses a much higher risk to the development of the drug. Furthermore, many previously conducted efficacy trials did not include pregnant people or children. With expanding experience and advocacy, many efficacy trials are now including adolescents and pregnant people whenever feasible.

Future tuberculosis therapeutics require research that identifies effective regimens that are shorter, safer, and have fewer drug–drug interactions than current treatments, which can be effectively scaled globally and include people with HIV, children and pregnant people (panel). To achieve this goal, a multipronged approach is needed that includes facilitation of research, inclusion of target populations, and a strategy that optimises care and treatment (panel). Although optimising treatment for achieving a cure is the overarching goal, drug delivery strategies that optimise adherence should be a priority, specifically when polypharmacy is the norm in people with HIV. New early-phase research on long-acting injectables and nanotechnology developments are prospects for fewer health-care visits and thus improved adherence to care and treatment.

Panel: Future research priorities for tuberculosis therapeutics in the context of HIV.

Facilitating research for tuberculosis

• Ensuring geographical representation, particularly among countries with high HIV–tuberculosis burden

• Enhancing clinical trial capacity and infrastructure across diverse settings (ie, from geographically representative, high-income, middle-income, and low-income settings)

• Increasing investments in tuberculosis research globally

• Expediting and streamlining regulatory processes worldwide

Inclusive target population in tuberculosis therapeutic trials

• Including sufficient numbers of people with HIV to assess safety and efficacy

• Performing trials across all CD4 cell count ranges, including people with advanced HIV disease

• Including children and pregnant people with and without HIV

• Facilitating research for the full range of people with HIV, such as those with substance use, heavy alcohol use, undernutrition, or other opportunistic infections

Focus of tuberculosis therapeutic research

• Optimising the duration and dosing of existing drugs (including drug-resistant cases) among people with HIV

• Evaluating new therapies for efficacy, safety, dosing, and drug–drug interactions with existing and newer antiretroviral drugs

• Optimising adherence by developing newer drug delivery strategies (eg, long-acting injectable agents)

• Developing fixed-dose combinations for suitable antiretroviral and antituberculosis drug candidates

• Evaluating tuberculosis treatment options for treatment failure and relapses among people with HIV

• Evaluating the role of host-directed therapies, such as steroids for tuberculous meningitis or disseminated tuberculosis, in people with HIV

• Evaluating the role of host-directed therapies for advanced HIV disease (CD4 <200 cells per μL) for prevention of tuberculosis-associated immune reconstitution inflammatory syndrome

Search strategy and selection criteria.

References for this Review were identified through searches of PubMed with the search terms “tuberculosis”, “therapeutics”, “people with HIV”, “children with HIV”, “antiretrovirals”, and “TB-HIV drug interactions” from Jan 1, 2000, to Oct 31, 2024. Articles were also identified through searches of the authors’ own files. Only papers published in English were reviewed. The final reference list was generated on the basis of originality and relevance to the broad scope of this Review.