Provisional CDC Guidance for the Use of Pretomanid as part of a Regimen [Bedaquiline, Pretomanid, and Linezolid (BPaL)] to Treat Drug-Resistant Tuberculosis Disease

Tuberculosis (TB) is caused by bacteria of the Mycobacterium tuberculosis complex, most commonly M. tuberculosis (MTB). TB infection is usually transmitted from one person to another by airborne droplet nuclei containing the bacteria. Multidrug-resistant tuberculosis (MDR TB), caused by MTB that is resistant to at least isoniazid (INH) and rifampin (RIF), has traditionally required longer, more intensive treatment than drug-susceptible disease, including 15–21 months of treatment after culture conversion with 4–7 drugs that are less effective, more toxic, and more costly in treating MDR TB than is a standard first-line regimen in treating drug-susceptible TB [1]. Rifamycin-monoresistant (RMR) TB, in the absence of INH resistance, although rare, has also been challenging to treat and associated with poor outcomes [2].

Even as efficacy of drug regimens has improved, side effects and length of treatment have continued to be challenges in treating patients with drug-resistant TB [3, 4]. During 2014–2018, 524 new cases of MDR TB were reported in the United States and U.S.-affiliated areas (territories and freely associated states), including 443 with resistance to isoniazid and rifampin alone, 72 with additional resistance to a fluoroquinolone or an injectable (assessed as amikacin, kanamycin, capreomycin), and 9 with additional resistance to both a fluoroquinolone and an injectable (CDC, unpublished data, 2021). Of 518 MDR TB patients alive at diagnosis, 63% were reported as completing treatment within 24 months, and 8% died before treatment completion (CDC, unpublished data, 2021).

On August 14, 2019, the Food and Drug Administration (FDA) approved the use of pretomanid as part of an all-oral combination (with bedaquiline and linezolid, together [BPaL]) administered by direct observation to adults with a diagnosis of pulmonary extensively drug-resistant or treatment-intolerant or nonresponsive (TI/ NR) MDR TB, based on limited clinical safety and efficacy data [5]. Pretomanid, known as molecule PA-824 while in development, is a novel bicyclic nitroimidazole with antimycobacterial activity comparable to INH [6]); it was developed by TB Alliance under license from Novartis and is available from Viatris (formerly Mylan). Pretomanid kills actively replicating MTB by inhibiting mycolic acid biosynthesis that is needed for cell wall production and kills nonreplicating MTB by nitric oxide release [7].

Bedaquiline is available from Janssen, with a patient assistance program in place for uninsured patients (https://www.jjpaf.org), and linezolid is available from multiple manufacturers. In murine models, the 3-drug regimen was found to reduce bacterial counts in the lungs and to be associated with fewer relapses at 2 and 3 months than any 2-drug combination of the respective three drugs.

FDA approves drug products for lawful marketing for specific intended uses based on data that establish safety and efficacy and FDA approves labeling specific to those uses. FDA approved pretomanid under FDA’s Limited Population Pathway for Antibacterial and Antifungal Drugs. This pathway allows for FDA and industry to have a route to approval for development of antibacterial and antifungal drugs targeting infections that lack effective therapies [8].

Based largely on the results of the Nix-TB trial [9], FDA approved pretomanid in the context of BPaL, i.e., to be administered as part of a combination regimen with bedaquiline and linezolid. Pretomanid was conditionally approved with the following warnings and precautions:

• Hepatic adverse reactions
• Myelosuppression
• Peripheral and optic neuropathy
• QT prolongation
• Reproductive effects
• Lactic acidosis

Subsequently, the results of the ZeNix trial were published and support the inclusion of these warning and precautions [10].

Scientific literature search of peer-reviewed manuscripts and presented abstracts was conducted with support from the CDC library. The more recent literature search included all studies Jan 1, 2020 through Dec 31, 2022, updated from the prior literature search Jan 1, 2020 through October 2021, with “pretomanid” as the keyword, using Medline, Embase, Cochrane, and Scopus databases and excluding only duplicate citations. Due to the limited number of available studies, a formal evidence evaluation framework was not used; therefore, this guidance is based on the available evidence and expert opinion.

Primary data supporting use of BPaL originated from the Nix-TB therapeutic trial, a small, single-arm, open-label South African study with specific inclusion and exclusion criteria for a high-mortality condition with limited treatment options that included 65% of patients (n=71) with MDR TB additionally resistant to at least one fluoroquinolone and injectable medication. The remaining 35% (n=38) of patients had TI/NR MDR TB [9], i.e., they did not respond to treatment with an available regimen for 6 months or more prior to enrollment or were unable to continue a second-line drug regimen because of drug intolerance. Over 50% of the participants in the Nix-TB study were persons with HIV and the results regarding favorable outcome were consistent when stratified by category of drug-resistance and regardless of HIV status.

In August 2020, a draft of initial proposed guidance was provided to seven non-CDC subject matter experts (SMEs) with drug-resistant TB expertise in epidemiology, clinical research, diagnosis, treatment, laboratory testing, and public health programs. Experts provided: 1) individual perspectives on the review; 2) experience with use of various regimens in treating drug-resistant TB; and 3) individual viewpoints on the proposed guidance. The initial draft guidance was also shared for public comment during the Advisory Council for the Elimination of Tuberculosis meeting on December 11, 2020. As clinicians have provided feedback on the guidance and data from new studies have been published, updates have been incorporated into this revision.

Management decisions in the care of patients with drug-resistant TB may be modified based on what is clinically indicated by unique patient circumstances.

Candidates for the BPaL regimen

General Use

In the Nix-TB trial, drug resistance was documented by phenotypic or genotypic tests [9]. In the United States, rapid molecular testing for resistance in patients being evaluated for drug-resistant TB should be obtained with confirmatory sequencing, available at many local or state public health laboratories or through CDC’s Molecular Detection of Drug Resistance (MDDR) service, followed by phenotypic testing. For more information about MDDR, see CDC’s Laboratory User Guide for U.S. Public Health Laboratories: Molecular Detection of Drug Resistance (MDDR) in Mycobacterium tuberculosis Complex by DNA Sequencing (https://www.cdc.gov/tb/topic/laboratory/mddr-user-guide.htm). To submit a sample for the MDDR service, complete CDC’s MDDR Request Form.

Persons with HIV

Based largely on the results of the Nix-TB trial in which the majority of participants were persons with HIV [9], the BPaL regimen can be used in patients living with HIV. Although pharmacokinetic data are not available to describe the impact of pretomanid within the BPaL regimen, this drug is metabolized through the CYP3A pathway, suggesting levels or drug concentrations in blood and tissue may be reduced with use of efavirenz (EFV) [11]. EFV should therefore not be used with pretomanid [11,12]. Additionally, given known drug interactions with bedaquiline, both EFV and cobicistat (COBI) should be avoided in the antiretroviral therapy (ART) regimen for a patient receiving BPaL [12]. Persons with HIV with a CD4+ count >50 cells/μL receiving an EFV- and COBI-sparing antiretroviral regimen without additional known drug-drug interactions with medications in the BPaL regimen are therefore candidates for this regimen.

Pregnant Persons and Children

Pretomanid 200mg within the BPaL regimen does not have an approved indication for use in pregnant persons or in children. Both these patient populations were excluded from enrollment in the Nix-TB trial [9], although participants >14 years of age were enrolled in the ZeNix trial [10] and >15 years were enrolled in the TB PRACTECAL trial [13]. Providers experienced in treating patients with drug-resistant TB disease need to weigh the risks and benefits of treating pregnant patients and children with BPaL on a case-by-case basis.

Other Populations

The following TB patients were among those excluded from the Nix-TB trial:

• those with baseline peripheral neuropathy of grade 3 or 4
• patients who had already received over 2 weeks of linezolid or bedaquiline
• patients with significant cardiac arrhythmia requiring medication, QTcF¹ >500 ms, history of risk factors for arrhythmias, concomitant use of any drug known to prolong QTc interval
• patients with baseline liver function tests over 5 times the upper limit of normal (ULN)
• patients with serum hemoglobin less than 8 gm/dL or concomitant use of any drug known to induce myelosuppression
• patients with serum potassium below the limits of normal
• patients with concomitant use of any drug known to induce myelosuppression
• patients with TB infection with suspected resistance to bedaquiline, pretomanid, or linezolid.

There are insufficient data to make a recommendation on whether the above patients are candidates for the BPaL regimen at this time. Patients who have had prior linezolid or bedaquiline may or may not still be eligible for the BPaL regimen but have been excluded in the relevant clinical trials; results of resistance testing conducted after receiving these agents should be taken into consideration. In addition, interpretation of results of the ZeNix and TB PRACTECAL trials needs to take into consideration that certain other patient populations were excluded from the ZeNix trial:

• persons with HIV infection and CD4+ cell count <100 per mm³
• patients with a risk of arrhythmia
• patients with an alanine aminotransferase level and an aspartate aminotransferase level >3 times ULN
• patients with peripheral neuropathy of grade 3 or higher at baseline
• patients who had previously received any of the three trial drugs or delamanid for 2 weeks or more before enrollment

and from the TB PRACTECAL trial:

• patients who were pregnant
• patients with an alanine aminotransferase level or an aspartate aminotransferase level >3 times ULN
• patients with a QTcF >450 msec or with structural heart disease
• patients with TB infection with suspected resistance to bedaquiline, pretomanid, or linezolid.

Results from additional studies utilizing pretomanid in short-course treatment regimens for patients with all forms of MDR TB have been and are actively being published. In particular, results of the TB PRACTECAL trial showed noninferiority of a regimen of pretomanid, bedaquiline, linezolid, and moxifloxacin for the treatment of MDR TB for 6 months to locally accepted standard of care, leading to early study termination [13].

¹The QT interval is dependent on heart rate and may be corrected by calculation to improve detection of patients at increased risk of ventricular arrhythmia. One correction formula focuses on QT interval divided by cube-root of RR (QTcF), where RR is the interval from the onset of one QRS complex (graphical deflections seen on an electrocardiogram [ECG] that correspond to the depolarization of the right and left ventricle with each heartbeat) to the onset of the next QRS complex, measured in milliseconds.

Initiating the BPaL Regimen

Any patient who is under consideration for or initiated on the BPaL regimen should be reported in a timely manner to local and state TB public health authorities (http://www.tbcontrollers.org/community/statecityterritory/). A physician with expertise in drug-resistant TB treatment should be involved in the patient’s treatment plan [1, 14]. Physicians with this expertise may be available through the local or state public health TB control program; CDC funds TB programs and TB Centers of Excellence (TB COEs) to facilitate timely medical consultations at no charge for patients with complex or drug-resistant TB disease (https://www.cdc.gov/tb/education/tb_coe).

Prior to initiation, providers should ensure the ability to complete safety and adherence monitoring of the BPaL regimen (see Warnings and Precautions and Adverse Event Monitoring).

Pharmacovigilance is the science and programmatic practice of the detection, understanding, assessment, and prevention of adverse effects from pharmaceuticals [50]. Monitoring, risk identification, quantification, and assessment of the use of medicines can minimize harm to patients and facilitate the identification of hazards and the need for necessary preventive measures, including potential regulatory actions. Pharmacovigilance should include a thorough, initial history and physical exam in addition to an initial measurement of specific laboratory and diagnostic testing to establish the patient’s baseline. This includes measuring a patient’s temperature, weight, and blood pressure as well as performing a complete blood count (CBC), urinalysis (UA), and other organ-specific labs (ALT, AST, GGT, amylase, lipase, blood creatinine phosphokinase), as well as serum potassium, calcium, and magnesium, with correction if abnormal. Performing peripheral nerve assessments and electrocardiograms (ECG) are also important in light of adverse event reporting from two key clinical trials [9, 14].

Monitoring while on treatment is an imperative, including inquiring about side effects since last visit, monitoring blood pressure and temperature, and ordering appropriate laboratory testing, QTcF interval monitoring, sputum collection, adverse event (AE) monitoring, and monitoring of clinical response (e.g., chest radiograph, symptoms, weight). Therapeutic drug monitoring (TDM) for linezolid with dose adjustment may be helpful [1, 13]. TB control programs should train providers in the process of monitoring and managing adverse events given the limited data available for pretomanid and minimal data on adverse events with use of BPaL. More on this subject is noted below in the section entitled Patient Monitoring.

Patient-centered care is a critical component of the treatment plan for patients receiving the BPaL regimen. When patients are educated about their diagnosis and given the ability to understand and participate in their treatment, potential barriers to treatment can be addressed and outcomes can be optimized.

Dosing

CDC-recommended doses in the BPaL regimen in adults are as follows:

• Pretomanid 200 mg administered orally once daily for 26 weeks.
• Bedaquiline 400 mg administered orally once daily for 2 weeks, followed by 200 mg administered orally 3 times weekly, with at least 48 hours between doses, for 24 weeks for a total treatment duration of 26 weeks.
• Linezolid starting at 600 mg daily and continuation at 600 mg or reduction to 300 mg daily or interruption of dosing as necessary for known linezolid adverse reactions of myelosuppression, peripheral neuropathy, and optic neuropathy. There may be circumstances where a 1200 mg initiation dose might be considered by some expert clinicians.

In the Nix-TB trial, high rates of neuropathy and myelosuppression occurred during treatment with a linezolid 1200 mg dose (81% and 48%, respectively); of the 109 patients treated with BPaL, only 16 (15%) completed treatment using an initial linezolid dose of 1200 mg [9]. Of note, in the Nix-TB trial, TDM was not available at any point for linezolid dose adjustments to minimize serum troughs. In other treatment regimens for MDR TB in the United States, the recommended initiation dose of linezolid for MDR treatment regimens is 600 mg daily and TDM is suggested to minimize toxicity [1].

In the ZeNix study, published in September 2022, patients were treated for 6 months with bedaquiline, pretomanid, and were equally randomized, dose-blinded, to daily linezolid starting at 1200 mg for 6 months (1200L6m), 1200 mg for 2 months (1200L2m), 600 mg for 6 months (600L6m), or 600 mg for 2 months (600L2m) [10]. In that study, 181 participants were enrolled, including 20% who were HIV positive. A high success rate at the primary endpoint, similar to Nix-TB, was observed: 93% in 1200L6m, 89% in 1200L2m, 91% in 600L6m and 84% in 600L2m. Notably fewer adverse events in the 600L6m and 600L2m arms were reported, and participants with lower linezolid dosing also had fewer treatment interruptions or discontinuations. In the early period after FDA approval, the majority in a cohort of 20 patients in the United States who received BPaL initiated linezolid at a 600 mg daily dose and achieved good outcomes [16].

To maximize successful outcomes while also minimizing adverse events, initiating BPaL with a reduced dose of linezolid of 600 mg daily is therefore preferred to initiation with linezolid 1200 mg daily. In December 2022, the World Health Organization (WHO) also endorsed 600 mg rather than 1200 mg starting dose of linezolid [17].

Administration

BPaL is taken with food. Pretomanid is available in 200 mg tablets and is swallowed whole with water (https://www.drugs.com/pro/pretomanid-tablets.html). Providers should carefully review all medications the patient is taking to evaluate for any drug-drug interactions with components of BPaL.

At present, pretomanid currently has approved use in combination with bedaquiline and linezolid in the treatment of TB [18]; additional published data from clinical studies examining other pretomanid-based regimens are forthcoming. The BPaL regimen should be administered by directly observed therapy and with case management strategies. Such strategies could include the use of incentives and enablers (e.g., food certificates, bus passes, cash, housing) to ensure adherence to the treatment regimen. Patients should be advised that nonadherence to a treatment regimen could result in treatment failure, relapse, or acquired resistance. An evaluation for the development of resistance to the BPaL regimen, including repeat drug susceptibility testing, is recommended for patients with treatment failure or relapse (see Microbiologic Monitoring).

Clinicians are encouraged to contact their state or local drug-resistant TB experts, CDC’s TB program, or their designated TB Center of Excellence for technical assistance, including advice for safety monitoring or referral for TDM of linezolid.

For linezolid, some experts use serum drug levels to adjust the linezolid dose or dosing interval to minimize the trough level below 2 μg/mL, which has been associated with lower risk of known toxicities (e.g., myelosuppression, peripheral neuropathy, and optic neuritis), and to achieve a peak serum linezolid level of 12–26 μg/mL [19–21]. Some experts also adjust linezolid dose to ensure the patient’s peak serum linezolid concentration is at least 4–16 times over the MTB organism’s Minimum Inhibitory Concentration (MIC) (or Critical Concentration) [22, 23]. In the ZeNix trial, only participants on high dose (1200 mg) linezolid had treatment interruptions; events precipitating these interruptions included anemia, neutropenia, fatigue, increased lipase or lactic acid, dizziness or peripheral sensory neuropathy [10]). In the same study, events precipitating dose reduction included most frequently anemia, neutropenia or peripheral sensory neuropathy. Similar to other second-line TB medications, linezolid should be provided to patients with close clinical and laboratory monitoring, and in the setting of severe toxicity, discontinuation should be considered.

Treatment with the BPaL regimen can be extended beyond 26 weeks up to 9 months (39 weeks) based on delayed treatment response within the first 8 weeks as assessed by time to culture conversion, persistent culture positivity, clinical response to treatment, and other underlying clinical factors, or modified based on adverse events.

Persons receiving the BPaL regimen should be monitored weekly for nausea, headache, hemoptysis, chest pain, arthralgia, rash, liver toxicity, signs of optic and peripheral neuropathy; treatment should be modified as clinically indicated. Complete blood count, comprehensive metabolic panel (to include potassium, bicarbonate, creatinine, serum aminotransferases, bilirubin, amylase, lipase, calcium, magnesium), and thyroid studies to rule out hypothyroidism (thyroid stimulating hormone level) should be conducted at baseline. A complete blood count and listed components of the comprehensive metabolic panel should be repeated monthly. Patients should also be monitored for other suspected adverse reactions tailored to side effects specific to other drugs in the regimen and the potential for drug interactions with bedaquiline and pretomanid; specific examples of adverse reactions observed in the Nix-TB trial using pretomanid, bedaquiline, and linezolid are detailed below [9].

Clinically significant safety risks associated with the BPaL regimen in the Nix-TB trial were peripheral neuropathy, optic neuritis, myelosuppression, hepatotoxicity, and pancreatitis [9]. Most adverse events were managed by treatment interruptions, dose reductions, or discontinuations of linezolid alone. In 6 participants, adverse events led to death (gastrointestinal hemorrhage, pancreatitis with hemorrhage, sepsis). As per study design, pretomanid and bedaquiline doses were not adjusted during the Nix-TB trial. Below, both the safety risks and adverse event data from studies are described, as set out in the Full Prescribing Information for Pretomanid [5].

Patients receiving BPaL must be monitored closely for adverse effects. Initial data submitted to FDA in 2019 from the Nix-TB trial warned of hepatotoxicity, myelosuppression, peripheral and optic neuropathy, QT prolongation, pancreatitis, reproductive effects (in rats) and lactic acidosis. In the ZeNix trial, adverse effects were noted involving the nervous system (peripheral neuropathy, headache), gastrointestinal tract (nausea, vomiting, dyspepsia, abdominal pain, constipation), skin (acne, pruritus), blood cells (anemia, neutropenia, thrombocytopenia), heart (QT prolongation), and a variety of lab abnormalities.

Evaluate all patients for signs and symptoms of tuberculosis disease during treatment as an indication of nonresponse to the regimen and evaluation for acquired resistance. Microbiologic surveillance through sputum specimens for acid-fast bacilli smear and culture is recommended monthly throughout and at end of treatment.

Gastrointestinal, including liver

• At baseline, aspartate aminotransferase, alanine aminotransferase, bilirubin, and alkaline phosphatase should be assessed.
• Once regimen initiated, liver-specific labs should be monitored at 2 weeks, then monthly, and, if symptomatic, more frequent monitoring should be considered, especially if administered with other hepatotoxic medications or in those with underlying liver disease.
  • Evidence of new or worsening liver dysfunction include fatigue, anorexia, nausea, jaundice, dark urine, liver tenderness, or hepatomegaly.
  • Such symptoms should prompt additional evaluation by the clinician.
• An increase of serum aminotransferases to more than 3 times upper limit of normal (ULN) should be followed by repeat testing within 48 hours.
• Discontinuation of BPaL should be considered for any of the following
  • serum aminotransferase levels are more than 8 times ULN
  • serum aminotransferase levels are greater than 5 times ULN and serum aminotransferase elevations persist beyond 2 weeks
  • serum aminotransferase levels are accompanied by total bilirubin level >2 times ULN
• National guidelines for management of hepatotoxicity should be followed [24]. Testing for HIV and viral hepatitis should be performed, and other hepatotoxic medications discontinued.
• To minimize the potential for hepatotoxicity among persons receiving BPaL, patients should be advised to avoid alcohol and other hepatotoxic drugs, including herbal supplements and should be monitored closely as noted below.
• BPaL can be administered to patients with mild to moderate hepatic impairment (Child-Pugh A or B) but should be avoided in patients with severe hepatic impairment (Child-Pugh C).
• Patients who develop recurrent nausea, vomiting, or abdominal pain should receive immediate medical evaluation, including evaluation of bicarbonate and lactic acid levels, transaminases, lipase/amylase levels, and the interruption or dose adjustment of linezolid or interruption of BPaL.

Summary of gastroenterology data from clinical trials, including hepatotoxicity

• Hepatotoxicity was associated with BPaL treatment in the Nix-TB trial and was managed by temporary discontinuation of the regimen [9]. Specifically, one patient had levels of ALT >3 times ULN, total bilirubin >2 times ULN, and alkaline phosphatase <2 times ULN [9]. One patient had ALT >3 times ULN, total bilirubin >2 times ULN, but alkaline phosphatase >2 times ULN. Both had interruption in the treatment, and the regimen was successfully restarted and completed. Hepatotoxicity is included in the warnings and precautions section of the bedaquiline packet insert. Hepatic transaminases were increased, and drug-induced liver injury was reported in phase III trials of pretomanid administered with other antimycobacterial drugs. Studies NC-005 and NC-006 treatment regimens included pretomanid: in NC-005, bedaquiline, pretomanid, and pyrazinamide, with or without moxifloxacin (BPaMZ and BPaZ) [25]; in NC-006, also known as the “STAND” trial, moxifloxacin, pretomanid, and pyrazinamide (MPaZ) [26]. In study NC-006, three patients died around week 4 of treatment due to acute liver failure; these 3 deaths were considered related to the MPaZ regimen.
• In the Nix-TB trial, persons with HIV experienced elevations in hepatic transaminases and hemopoietic cytopenia (i.e., anemia, neutropenia, and thrombocytopenia) more often than did HIV-negative subjects [9], which would be expected because of higher likelihood of underlying comorbidity and concomitant treatment with antiretrovirals and other medications.
• In the ZeNix trial, 47/181 (26%) of participants overall had one or more liver-related adverse events, with similar numbers across the different treatment groups [10]. In the Nix-TB trial, for related gastrointestinal symptoms, 39 of 109 (36%) participants experienced nausea, 31/108 (28%) vomiting, 19/108 (17%) dyspepsia, 6/109 (6%) diarrhea, and 7/109 (6%) abdominal pain [9]. Similar related symptoms were reported in ZeNix trial: diarrhea 12/181 (7%), 15/181 (8%) nausea, 5/181 (3%) constipation [10].
• Acute pancreatitis was reported in 3 of 109 (2.8%) patients in the Nix-TB trial (9). It is not clear whether this was associated with the study drugs or with potential risk factors for pancreatitis (e.g., alcohol use, HIV, and antiretroviral therapy such as lopinavir/ritonavir). Pancreatitis was reported in animal studies of bedaquiline [27, 28].

Renal and Other Metabolic Considerations

Pretomanid is minimally excreted by the kidneys, does not require dosage adjustment in patients with mild to moderate renal impairment (not requiring dialysis), and should be administered with caution in patients with severe renal impairment requiring dialysis.

Summary of data from clinical trials on renal issues

• Patients with creatinine >2 times ULN were excluded from the Nix-TB trial per protocol, and a safety signal for renal toxicity was not evident in the post baseline creatinine levels. Five of 109 (5%) patients had creatinine elevations > ULN and ≤2 times ULN at baseline [9]. Based on these limited data and clinical trial exclusions, safety of the BPaL regimen or pretomanid in patients with renal impairment has not been assessed.
• In the Nix-TB trial, there were no signals of renal impairment [9]. However, in the ZeNix trial, proteinuria was reported in 6/181 (3%) and hematuria in 3/181 (2%) of all trial participants [10].

Summary of data from clinical trials on lactic acidosis

Lactic acidosis occurred in 3/109 (2.8%) patients in the Nix-TB trial, in which all participants initiated BPaL with a linezolid dose of 1200 mg daily, and was associated with the use of linezolid [9]. Lactic acidosis is listed in warnings in the linezolid US package insert [29]. Lactic acidosis was not reported in other trials of pretomanid-containing regimens [25, 26, 27].

Cardiovascular

Persons receiving the BPaL regimen should be closely monitored for signs of cardiac toxicity with repeated ECG and other measures as follows:

• At baseline, prior to BPaL initiation, persons should undergo blood pressure measurement and an ECG to screen for prolonged QT interval.
• Persons receiving the BPaL regimen should then be monitored closely for signs of cardiac toxicity (evidence of QT prolongation) with ECGs at least 2, 12, and 24 weeks after starting treatment [30].
• Discontinuation of bedaquiline and all other QTcF-prolonging drugs should be considered if the patient develops:
  • clinically significant ventricular arrhythmia, or
  • a QTcF of >500 ms (confirmed by repeat ECG).
• Serum potassium, calcium, magnesium, and thyroid function tests should be obtained at baseline and monthly whenever clinically indicated, especially if QT prolongation is detected. Any abnormalities that are detected should be addressed, and electrolytes should be monitored until fully corrected.
• The concurrent use of other drugs known also to cause QTcF prolongation could be associated with increased risk of cardiotoxicity when patients are receiving bedaquiline. Persons who receive BPaL should be monitored weekly with an ECG if they:
  • receive other QTcF prolonging drugs;
  • have a history of torsade de pointes, congenital long QT syndrome, hypothyroidism and bradyarrhythmias, or uncompensated heart failure; or
  • have serum calcium, magnesium, or potassium levels below the lower limits of normal.
• If pretomanid is discontinued for cardiotoxicity, ECGs should be monitored frequently to confirm that QTcF has returned to baseline. If syncope occurs, an ECG should be obtained to evaluate for QTcF prolongation.

Summary of cardiac data from clinical trials, including QT prolongation:

In the ZeNix trial, 6 of 181 (3%) of participants experienced non-cardiac chest pain, and 13 of 181 (7%) reported hypertension [10].

Mild QT prolongation was reported in Nix-TB trial in 6 of 109 (5.5%) of participants while no prolongation events were reported in the ZeNix trial [9, 10]. QT prolongation is associated with bedaquiline. Pretomanid had no clinically significant effect on the QT interval in a Thorough QT (TQT) study [31].

Neurologic, including Optic Neuritis and Neuropathy

Optic Neuritis: Visual acuity and color vision assessment with Ishihara testing should be monitored in all patients receiving linezolid as part of the BPaL regimen. If symptoms of visual impairment are identified, interrupt linezolid dosing and obtain prompt ophthalmologic evaluation.

Neuropathy: While receiving linezolid as part of the BPaL regimen, patients should be monitored for symptoms (e.g., tingling, pain, or numbness of hands or feet) and signs of peripheral neuropathy with assessments of neurologic and visual function, and linezolid dosage should be adjusted as needed. Clinical evaluation of peripheral neuropathy should include questions regarding tingling, burning, freezing, numbness, stinging, or itching of the hands and feet, as well as a physical examination including a monofilament test. For more information about the monofilament test please visit: https://www.annfammed.org/content/7/6/555.long

Summary of clinical data on neurologic issues

• In the Nix-TB trial, optic neuritis developed in 4 of 109 (4%) participants; in the ZeNix trial, optic neuritis developed in 2 of 181 (1%) [9, 10]. In both trials, the optic neuritis resolved.
• Across both Nix-TB and ZeNix trials, neuropathy was the most common AE reported with use of the BPaL regimen.
• Peripheral neuropathy associated with linezolid was the most common AE, reported in 87/109 (80%) of patients in the Nix-TB trial, in which participants initiated BPaL with a linezolid dose of 1200 mg daily for at least 4 weeks [9]. Over 50% of the AEs were moderate to severe, including some that were not reversible by end of study. Peripheral neuropathy increased with the duration of BPaL treatment. Prior studies have documented irreversibility of some linezolid-associated neuropathies [32, 33].
• In the Nix-TB trial, headache was reported in 22 of 109 (20%) participants and in 13 of 181 (7%) participants in the ZeNix trial.
• In the NIX-TB and ZeNix trials, the known adverse reaction of optic neuritis associated with linezolid was fully reversible when linezolid was discontinued [9, 10]. Ophthalmological examinations and Age-Related Eye Disease Study 2 (AREDS2) scores showed no evidence that pretomanid induces cataract formation in humans [8, 34].

Hematologic, including Myelosuppression

At baseline, a CBC should be done and thereafter, CBC should be monitored regularly, such as weekly, during the first 6–8 weeks, then monthly as needed based on symptoms. Interrupting linezolid or adjusting the dose or dosing interval, aiming for serum drug level <2 μg/mL, should be done in patients who develop or have worsening myelosuppression. Importantly, other etiologies for myelosuppression should also be investigated while making any treatment adjustments.

Summary of clinical data from clinical trials on hematologic issues

In the Nix-TB trial, 38 of 109 (35%) of participants experienced anemia, 8 of 109 (7%) neutropenia, and 6 of 109 (6%) thrombocytopenia. In this trial in which all participants initiated BPaL with a linezolid dose of 1200 mg daily, myelosuppression associated with linezolid was reversible except for in 3 patients who had ongoing events at the time of data cutoff [9]. Myelosuppression increased with the duration of BPaL treatment.

In the ZeNix trial, 14 of 181 (8%) participants experienced anemia, and 16 of 181 (9%) experienced neutropenia [10].

Additional adverse events of special interest/safety signals for pretomanid from animal/preclinical studies

• Neurotoxicity: Convulsions and ataxia were noted in cynomolgus monkeys receiving 5 times the maximum recommended human dose (450 mg/kg)[14]. Neither ataxia, convulsions, nor other CNS-related clinical signs were seen in monkeys dosed for 39 weeks at exposures (based on area under the curve over 24 hours) similar to the predicted exposures in patients at the maximum recommended human dose. Two participants in Nix-TB had convulsions but had reported a medical history of convulsions [9].
• Lens Disorders/Cataract: Cataracts were seen in rat studies with 2 times the human exposure at maximum recommended human dose over 26 weeks [14]. Cataracts did not develop in monkeys given daily doses of pretomanid 100 mg/kg for 39 weeks. This dose produced an average area under the curve over 24 hours (AUC_0-24) value of more than twice the exposure patients would experience at the maximum recommended human dose of pretomanid. No cataracts were observed in the following human studies of pretomanid: NC-002, NC-005, NC-006, and Nix-TB [9, 25, 26].
• Testicular Toxicity: In animal toxicology studies of pretomanid, testicular toxicity resulting in infertility was noted in male rats (U.S. Food and Drug Administration. Drug prescribing information: Pretomanid https://www.accessdata.fda.gov/drugsatfda_docs/label/2019/212862s000lbl.pdf). Reduced fertility was observed in male rats given daily oral pretomanid for 13 weeks at 30 mg/kg. Spermatocyte degeneration was noted in rats at 1.5 times the human exposure. The no observed adverse effect level for testicular toxicity was 50 mg/kg, which was associated with an AUC of 70 μg*h/m. This is 1.2 times the maximum recommended human dose. At 100 mg/kg (3.5 times the clinical dose), pretomanid was associated with reduced body weight, complete irreversible infertility, testicular atrophy, lower sperm counts, reduced sperm motility, lower serum inhibin B concentration, and higher serum FSH concentration [18]. These effects were irreversible at 100 mg/kg. Reduced fertility appears to be a result of the effects on spermatogenesis in males. However, male hormones were normal in the following human phase 2 and 3 studies: NC-002, NC-005, NC-006. Other data support lack of adverse effects on male reproductive hormones with pretomanid-containing regimens: Analysis of serum hormone levels in males from four clinical TB trials found that pretomanid-containing treatment was not associated with testicular toxicity and led to improvement in underlying hypogonadism [36].
• Female Reproductive and Maternal Toxicity: Female rats dosed daily with oral pretomanid for 2 weeks showed body weight loss, reduced feed consumption, reduced number of estrous stages per 14 days, and a greater incidence of prolonged diestrus at 100 mg/kg/day [34]. The number of live fetuses and fetal body weight were lower, and skeletal development was slowed. Rat pups from dams that experienced maternal toxicity showed a slight delay in the age at which the air-drop righting reflex developed and an increase in basic and fine movement as well as total distance travelled. Reproductive and developmental toxicity have not been evaluated in human studies.

Additional serious adverse events observed in Phase III and Phase III human studies

• In study NC-002, serious adverse events in patients treated with pretomanid, moxifloxacin, and pyrazinamide included agranulocytosis, elevated hepatic transaminases, pneumothorax, atrioventricular block, seizure, and dyspnea [26].

BPaL use should be accompanied by microbiologic monitoring. At least one sputum specimen should be submitted for culture monthly throughout and at end of treatment, even after conversion to negative cultures. This is consistent with United States and World Health Organization guidance for care of patients with drug-resistant TB [1]. Laboratory results, including those from phenotypic drug susceptibility testing (DST), rapid molecular testing, or a combination of both should be obtained to document drug-resistant TB status. Ideally, molecular DST would be performed to evaluate the presence of mutations known to be associated with first- and second line antituberculosis drug resistance as well as with newer drugs like bedaquiline [37]. These results, in combination with phenotypic results, are preferrable for informing clinical decision making.

• With use of the BPaL regimen, isolates of MTB grown from the initial and any subsequent monthly specimen, including those taken before BPaL treatment initiation, should be evaluated for resistance or changes in susceptibility to regimen drugs through use of phenotypic methods, and when available, molecular methods. Among bedaquiline, pretomanid, and linezolid, the Clinical and Laboratory Standards Institute (CLSI) has defined a critical concentration for susceptibility testing using the fluorescence-based, commercial, shorter-incubation liquid media system (i.e., BD BACTEC Mycobacteria Growth Indicator Tube System or MGIT) for linezolid only [38]. CLSI has not yet defined a critical concentration or breakpoint for testing of bedaquiline but has published a minimum inhibitory concentration (MIC) quality control range for broth microdilution testing of MTB H37Rv for a custom MIC plate that includes both linezolid and bedaquiline [38, 39]. Additionally, WHO has recommended critical concentrations for testing of linezolid, in 7H10 Middlebrook, 7H11 Middlebrook, and MGIT, and bedaquiline in 7H11 Middlebrook and MGIT [40]. There are currently no recommended critical concentrations for testing of pretomanid, although a provisional critical concentration has been proposed at 1 µg/ml in MGIT (European Medicines Agency, Dovpela (Pretomanid) Drug Information Sheets, Summary of Product Characteristics, Annex 1, Page 10; available at: https://www.ema.europa.eu/en/documents/product-information/dovprela-epar-product-information_en.pdf), and the drug is not yet widely available for laboratory testing. In the Nix-TB trial, phenotypic susceptibility testing for pretomanid was performed by evaluation of MIC using MGIT for baseline and interim evaluation. In the Nix-TB trial, all baseline isolates had pretomanid MIC values of ≤1 μg/ml in MGIT [9]. Phenotypic DST for BPaL is likely to be limited to larger reference laboratories (see below for further information on accessing DST). Minimum inhibitory testing by the broth microdilution method may be appropriate for monitoring changes in susceptibility. CDC can assist in identifying a laboratory that can perform susceptibility testing for this purpose [41].
• Regarding molecular testing, mutations in Rv0678, atpE, and pepQ have been reported as associated with resistance to bedaquiline as have mutations in rplC and rrl for linezolid [42–48]. Pretomanid resistance is thought to be associated with mutations in the deazaflavin (cofactor F420) dependent nitroreductase known as Ddn or mutations in the genes fgd1, fbiA, fbiB, and fbiC involved in cofactor 420 synthesis needed for conversion of the prodrug into its active form [49]. There are existing knowledge gaps regarding which mutations reliably predict resistance for all three drugs of the BPaL regimen. However, even with these limitations, evaluation of isolates by both phenotypic and molecular methods from patients treated with this regimen is critical for resistance monitoring and detection of changes in susceptibility over time. This is particularly important for bedaquiline given that in the Nix-TB trial, one of three patients experiencing relapse was reported to have an increased bedaquiline MIC that was associated with a mutation detected in Rv0678 [9]. Some reports indicate detection of resistance to bedaquiline even in the absence of previous exposure to the drug [50, 51]. CDC’s laboratory is working toward implementation of molecular and growth-based testing for bedaquiline and linezolid. In addition, CDC is conducting surveillance of pretomanid resistance by MIC testing as part of a global post-marketing surveillance study. Susceptibility testing for drugs in this regimen may also be available in other laboratories. CDC can assist in identifying a source for this purpose.

Mutations associated with resistance to medications in the BPaL regimen

Patients treated with the BPaL regimen should be followed for 2 years after treatment completion to evaluate for any evidence of recurrence. This recommendation is not based on clinical trials but rather on reasonable expert opinion in treatment of patients with bedaquiline-containing regimens outside of clinical trials [52].

Any patient with drug-resistant TB must be reported in a timely manner to the local and state TB public health authorities (https://www.tbcontrollers.org/community/statecityterritory/). Serious drug side effects, product quality problems, and therapeutic failures should be reported to FDA’s MedWatch program at https://www.accessdata.fda.gov/scripts/medwatch/index.cfm?action=reporting.home or by calling 1-800-FDA-1088. Any BPaL-treatment-associated adverse event leading to hospital admission or death should also be reported by clinicians to local or state health departments.