TUTORIAL

The Treatment of Tuberculosis

Charles A. Peloquin1,* and Geraint R. Davies2,3

Tuberculosis (TB) remains a leading cause of infectious death worldwide, and poverty is a major driver. Clinically, TB
presents as “latent” TB and active TB disease, and the treatment for each is different. TB drugs can display “early
bactericidal activity (EBA)” and / or “sterilizing activity” (clearing persisters). Isoniazid is excellent at the former, and
rifampin is excellent at the latter. Pyrazinamide and ethambutol complete the first-­line regimen for drug-­susceptible
TB, each playing a specific role. Drug-­resistant TB is an increasing concern, being met, in part, with repurposed drugs
(including moxifloxacin, levofloxacin, linezolid, clofazimine, and beta-­lactams) and new drugs (including bedaquiline,
pretomanid, and delamanid). One challenge is to select drugs without overlapping adverse drug reaction profiles.
QTc interval prolongation is one such concern, but to date, it has been manageable. Drug penetration into organism
sanctuaries, such as the central nervous system, bone, and pulmonary TB cavities remain important challenges. The
pharmacodynamics of most TB drugs can be described by the area under the curve (AUC) divided by the minimal
inhibitory concentration (MIC). The hollow fiber infection model (HFIM) and various animal models (especially mouse
and macaque) allow for sophisticated pharmacokinetic/pharmacodynamic experiments. These experiments may
hasten the selection of the most potent, shortest possible regimens to treat even extremely drug resistant TB. These
findings can be translated to humans by optimizing drug exposure in each patient, using therapeutic drug monitoring
and dose individualization.

Tuberculosis (TB) remains a leading cause of infectious death for pyrazinamide (US clinicians) or rifaPentine (researchers and
worldwide, despite the availability of effective and inexpensive non-­US clinicians). Let the reader beware.
treatments.1 Poverty is a major driver in sustaining TB globally.
Some may consider it a social disease with medical consequences.2 Tuberculosis (TB)
Because of its unique cellular construction and life cycle, the treat- Clinically, TB can be described in two general forms –­“latent” TB
ment of the disease caused by Mycobacterium tuberculosis (Mtb) and active TB disease.3 This represents an oversimplification of a
is long, which in turn leads to nonadherence.3 Incentives for the continuum from exposed but not infected, exposed and infected
pharmaceutical industry to develop new drugs are few, and many but cleared, infected but not clinically apparent (i.e. “latent”), to
companies choose not to develop TB drugs. When they do, con- active TB disease with clinical manifestations.3
siderable public support likely is involved.4 Fortunately, new drugs
have emerged recently for the treatment of drug-­resistant TB, and Latent TB infection (LTBI)
research is focused on shortening the duration of treatment for all So called “latent” TB is a condition where skin or blood tests
forms of TB.5 demonstrate that the immune system recognizes Mtb, but there
Since the vast majority of humans who become infected with are no signs or symptoms of active disease. Under these condi-
Mtb are able to eliminate or contain it, clearly most humans already tions, it is estimated that fewer than 1000 live bacilli reside within
have the immune capacity to control Mtb. However, approximately the host, largely or completely contained by the immune system.
10% of the human population lacks this immunological capability, Patients with latent TB infection (LTBI) are not infectious be-
and because the immunology of TB is incompletely understood, cause they do not have active pulmonary lesions.1,3 LTBI pres-
transferring this immune capability to vulnerable populations via ents an excellent opportunity, as well as a unique challenge, for
vaccine or immunotherapy remains a work in progress.6–­8 That “nipping TB in the bud.” Chemotherapy during LTBI can reduce
immensely important topic is beyond the scope of this tutorial. the lifetime risk of active disease in the host by over 90%.3,9 Most
Instead, this tutorial focus on the pharmacology of TB drugs, treated patients with LTBI will never become sick with TB, and
their uses in TB patients, and currently available avenues for op- thus will never spread TB, breaking the cycle of new infections.
timizing pharmacotherapy. A note about abbreviations: Among The challenge comes from the need to convince people who feel
US clinicians, RIPE stands for Rifampin, Isoniazid, Pyrazinamide, perfectly well to take drugs with important toxicities for extended
and Ethambutol. Researchers and many non-­US clinicians use periods of time. Nevertheless, without a concerted global effort
HRZE for isoniazid (isonicotinic acid Hydrazide), Rifampin, pyr- to reduce this reservoir of millions of cases of LTBI, the cycle of
aZinamide, and Ethambutol. Depending on context, P can stand disease cannot be broken.

1
College of Pharmacy and Emerging Pathogens Institute, University of Florida, Gainesville, Florida, USA; 2Institute of Infection and Global
Health, University of Liverpool, Liverpool, UK; 3Institute of Translational Medicine, University of Liverpool, Liverpool, UK. *Correspondence: Charles A.
Peloquin (peloquin@cop.ufl.edu)
Received November 20, 2020; accepted March 24, 2021. doi:10.1002/cpt.2261

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 110 NUMBER 6 | December 2021 1455
 15326535, 2021, 6, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.2261 by Ircss Burlo Garofolo, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TUTORIAL

Historically, the main option for the treatment of LTBI had 18 months, toxicity was common, and success was good but not
been isoniazid (INH).9 George Comstock demonstrated the effi- great. Pyrazinamide (PZA), ethionamide (ETA), cycloserine (CS)
cacy of INH with pioneering studies performed in Alaska, USA, and ethambutol (EMB) were developed over the ensuing decade,
during the 1960s.10,11 INH 300 mg given daily for six months had making modest contributions at the time. The game-­changer was
considerable benefit, 9 months even more benefit, and 12 months RIF.17,18
near-­maximal benefit.12 When used, 9 months of INH is the pre- Adding RIF to INH-­containing regimens made it possible
ferred duration.9 The difficulties come with the fact that INH is to cure TB in 9 months, half the prior duration. Adding PZA
not the best drug for the job –­it is not considered a potent “ster- to INH and RIF made it possible to cure drug-­susceptible TB
ilizing” drug, described later, and it carries a risk of drug induced (DS-­T B) in 6 months, ushering in the so-­called “short course”
liver injury (DILI). So, while effective, better alternatives to INH era.17,18 Short, that is, compared to 18 months, but not short
are being sought. compared to 10 days of penicillin for strep throat. The current
Rifampin (RIF) is the best “sterilizing” drug (a drug that pre- regimen, known as RIPE, is 2 months of RIF, INH, PZA, and
vents post-­treatment relapse in patients with active TB disease by EMB. EMB is added to protect RIF while clinicians await sus-
eliminating so-­called “persisters.”) To the extent that the physiolog- ceptibility data on the slow-­growing Mtb. Currently, roughly 1
ical state of a “latent” TB bacillus resembles the physiological state in 10 Mtb isolates in the US is resistant to INH, with that fre-
of a “persister” TB bacillus, RIF should be the preferred drug for quency varying by nation. Since PZA is not effective at suppress-
LTBI. Studies have shown that RIF 600 mg is quite effective, even ing the development of resistance (i.e., it does not protect RIF if
when given for only 4 months.9,13 the isolate is INH resistant), the 4th drug, EMB, serves that pur-
Combination therapy also has been effective for LTBI. Some pose. EMB can be stopped once the isolate is known to be INH
countries use 3 months of daily INH plus RIF, which now has susceptible, or after 2 months of effective treatment, whichever
entered the US guidance.9 Alternatively, INH plus rifapentine comes first. PZA also can be stopped after 2 months of effective
(RPT, or cyclopentyl-­rifampin) can be used. 3HP is the abbre- treatment, because historical data suggest most of the benefit
viation for 3 months of once weekly INH (shortened to H) has been accrued by then.3,17–­19 RIPE then reduces to RI for
900 mg and RPT (shortened to P) 900 mg.9 A newer modifi- an additional 4 months, assuming that the patient responds to
cation is daily INH and RPT for one month has been studied treatment, and assuming the patient is not at an elevated risk for
in persons living with HIV.14 A one month treatment clearly treatment failure. Patients who are substantially underweight,
has the advantage of shortest duration (from 9 to 4 to 3 down patients with large cavitary lung lesions, patients with bone or
to 1 month). Most patients could be engaged for that period of central nervous system disease typically are treated longer than
time, and observed for possible adverse effects. With further 6 months.19,20 Some clinicians extend treatment in patients who
study, this can be scaled up to more patient populations, and are immunocompromised, depending on clinical circumstances.
could start to reduce the pool of future TB cases. Other agents Treatment gets much more complicated when the isolate is
are being considered for LTBI, including fluoroquinolones or resistant to INH and RIF (so called multidrug-­resistant TB, or
the new drug bedaquiline (BDQ) which is described in detail MDR-­TB).21 Additional resistance to fluoroquinolones and ami-
below.15 However, clinical data do not exist yet to comment on noglycosides is called extensively drug resistant TB (XDR-­TB). For
the role of BDQ for LTBI. Should that prove to be effective, it many years, the “second-­line” TB drugs PAS, ETA, and CS were
would be preferred in cases where the patient likely had been used in varying combinations with whatever other drugs showed
exposed to forms of TB that were resistant to the standard “first-­ in vitro susceptibility and / or had not been used in a given pa-
line” TB drugs. Additional research is focused on long acting tient previously. This approach is called an individualized, or opti-
injectable dosage forms that would allow as little as one dose mized, MDR-­TB regimen which will typically contain 5–­7 drugs.
to block future transmission of TB.16 Considerable research is Typically, these regimens lasted 18–­24 months.17–­22 The WHO
needed, however, before that becomes a clinical reality. has categorized drugs into groups based on preferred order of use
(Figure 1).22 More recently, shortened standardized MDR-­TB reg-
Active TB disease imens have been tried, under selected conditions.21–­23 Such regi-
LTBI presents with few live organisms, so the probability of se- mens can be as short as 9 months. Clofazimine (CLOF) is a leprosy
lecting for drug resistance using monotherapy is considered very drug that has found new life as a TB drug within such regimens.
small. The situation changes dramatically with active TB disease. New drugs have been tested, initially as add-­ons to an opti-
For active TB, monotherapy is unacceptable. Within cavitary mized background regimen (OBR). There is a desire to replace
pulmonary lesions, there can be 109 organisms or more, a number the most toxic older TB drugs, including the injectable drugs
that exceeds typical wild-­t ype mutation frequencies for many TB (aminoglycosides and capreomycin), since they can cause dis-
drugs. That is, there could be 103 or more pre-­existing mutants at abilities that last a lifetime. The TB Alliance has moved forward
the onset of therapy, primed to survive any chosen monotherapy.3 with trials using two new drugs, BDQ and pretomanid (PMD),
This was proven conclusively in the 1940s with the introduction along with the repurposed drug linezolid (LZD).24 These regi-
of streptomycin (SM).17,18 The lesson was learned quickly, but mens are only 6 months long, bringing the treatment duration of
options were in short supply. Para-­aminosalicylic acid (PAS) and MDR-­T B in line with the treatment duration of drug-­susceptible
eventually INH were added to SM by the early 1950s, producing TB. Work continues on optimizing the dose of linezolid, which
the first effective 3 drug regimens. The duration was typically is the most toxic of the 3 drug combination.25,26 Delamanid

1456 VOLUME 110 NUMBER 6 | December 2021 | www.cpt-journal.com

 15326535, 2021, 6, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.2261 by Ircss Burlo Garofolo, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TUTORIAL

Figure 1 WHO grouping of medications for multidrug-­resistant TB. WHO consolidated guidelines on drug-­resistant tuberculosis treatment.
Geneva: World Health Organization; 2019. License: CC BY-­NC-­SA 3.0 IGO. Cataloguing-­in-­Publication (CIP) data. CIP data are available at
http://apps.who.int/iris. The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed
or recommended by WHO in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of
proprietary products are distinguished by initial capital letters.

(DLM) is another new TB drug that has some similarities to TB drug pharmacology
PMD.27,28 No head-­to-­head trials comparing PMD to DLM Extensive publications exist about the older TB drugs, with more
exist, and clinical safety data for PMD are limited. Beta-­lactam –­ limited data available for the newer drugs. Excellent resources
beta-­lactamase inhibitor combinations also are being evaluated include, but are not limited to, the 1988 book Anti-­tuberculosis
as potential TB drugs, with clavulanic acid and avibactam show- Drugs, edited by K. Bartmann,30 the 1991 book Drug
ing good suppression of mycobacterial enzymes.21,29 Additional Susceptibility in the Chemotherapy of Mycobacterial Infections,
drugs are at different stages of development (Figure 2). edited by Leonid Heifets,31 the 2000 book A Clinician’s Guide to

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 110 NUMBER 6 | December 2021 1457
 15326535, 2021, 6, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.2261 by Ircss Burlo Garofolo, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TUTORIAL

Figure 2 Working Group on New TB Drugs 2021 pipeline of TB drugs. Credit: Stop TB Partnership’s Working Group on New TB Drugs (Pipeline
as of March 2021).

Tuberculosis, written by Michael D. Iseman,19. There are many N-­acetyl-­transferase 2 (NAT2).33,34 Fast (half-­ l ife, or t1/2 ,
TB drug chapters in the 2005 book Antimicrobial Therapy and < 2 hours), intermediate (i.e. heterozygous fast), and slow acetyl-
Vaccines, Volume II, 2nd edition: Antimicrobial Agents, edited ators (t1/2 greater than 2 hours) have been identified. Some Asian
by Victor Yu and colleagues (with subsequent on-­line updates populations approach 80% fast genotype, while other popula-
at antimicrobe.org),32 and a complete description of TB in the tions trend towards 50% of each genotype. Some data suggest
2020 book Clinical Tuberculosis, 6th edition, edited by Lloyd higher doses should be used in fast acetylators. Renal excretion
N. Friedman, Martin Dedicoat, and Peter D. O. Davies.33 The accounts for up to 40% of INH, and it has good cerebrospinal
authors of this Tutorial are the authors of the chapter Clinical fluid penetration.31–­33 INH accumulates in epithelial lining
Pharmacology of the Anti-­Tuberculosis Drugs from the latter fluid and in alveolar macrophages while concentrations in le-
book, and we present a condensed summary here.33 We give sions are similar to plasma and may not extend into caseum.33,35
credit to these original publications, where full details and many INH can inhibit CYP1A2, 2A6, 3A4 and 2C19 to varying de-
more references may be found.31–­33 Figure 3 shows a depiction grees.33 Most clinically reported interactions involve CYP3A4
of the mechanisms of action for the TB drugs, from the NIAID. or 2C19, leading to higher plasma concentrations of the com-
panion drugs, including anticonvulsants and benzodiazepines.
Traditional first-­line drugs INH pharmacodynamics (PD) is driven by the (free drug)
The majority of TB patients continue to be treated with the RIPE area under the curve to MIC ratio (AUC/MIC).33,34 Even a few
regimen, so it is important to understand the basic features of days of INH is very effective at rapidly reducing the recoverable
these four drugs. number of live Mtb in sputum, a phenomenon known as early
bactericidal activity (EBA).17–­20 The main toxicity of INH is
Isoniazid DILI, affecting from <1% to about 2% of patients. Peripheral
INH is a pro-­d rug which is activated within Mtb by the enzyme neuropathy and central nervous toxicity also occur, and these can
KatG.33 A reactive intermediate of INH inhibits the synthesis be mitigated by the use of small doses of pyridoxine (vitamin B6).
of mycolic acids, which are needed for cell wall construction.18,33 These adverse events may be more common in slow acetylators.
Resistance is conferred by mutations in katG (high-­level) and/
or inhA (low-­level).33 INH 300 mg daily has high oral bioavail- Rifampin
ability, which is modestly reduced by food.33,34 Some RIPE reg- RIF currently is the most important TB drug. It binds to the
imens are given 3 times weekly, using INH 900 mg, although beta subunit of mycobacterial DNA-­dependent RNA poly-
intermittent regimens have fallen out of favor. N-­acetylation to merase enzyme, inhibiting transcription.33 Resistance is con-
inactive metabolites occurs through the polymorphic enzyme ferred by mutations clustered in an 81-­bp region of the rpoB

1458 VOLUME 110 NUMBER 6 | December 2021 | www.cpt-journal.com

 15326535, 2021, 6, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.2261 by Ircss Burlo Garofolo, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TUTORIAL

Figure 3 NIH depiction of TB drug mechanisms of action. The photo of Mycobacterium tuberculosis is from the Centers for Disease Control and
Prevention (CDC; http://wayba​ck.archi​veit.org/7761/20160​90919​2435/http://phil.cdc.gov/phil/home.asp), CDC/Dr. Ray Butler, Janice Carr.
This illustration is in the public domain. Credit: the National Institute of Allergy and Infectious Diseases (NIAID).

gene, and this knowledge can be used to great effect within longer, and generally less effective.17–­19,33,34 This ability to act
rapid genotypic tests for RIF susceptibility.33 RIF is approxi- against what generally are termed non-­replicating persisters
mately 70% bioavailable; absorption is modestly affected by (NRP) is the key feature of RIF and other rifamycins.17–­20,33,34
food.33 Absorption increases non-­linearly with dose through Importantly, the maximum effective RIF dose has never been
40 mg/kg, with AUC higher than otherwise expected.33,34 RIF defined in any PD model, with increasing AUC/MIC causing
has complex absorption involving the hepatic organic anion further reductions in the colony forming units (CFU). This has
transporter SLCO1B1.33,34 RIF is metabolized by hepatic ester- led to a renewed interest in high-­dose RIF therapy for all forms
ases, possibly arylacetamide deacetylase (AADAC), primarily to of TB.33,34,36,37
the partially active 25-­O -­desacetyl-­rifampin.33 Auto-­i nduction Transient rises in bilirubin and/or transaminases are com-
of non-­c ytochrome P450 pathways reduces RIF’s AUC by 30–­ mon, while true DILI is much less common.19,20,33 RIF may
40% over the first 2 weeks of treatment. RIF and its metabolites cause cutaneous hypersensitivity or serious hypersensitivity
are excreted in the bile, and may undergo enterohepatic recircu- reactions, especially with intermittent dosing.19,20,33 A flu-­like
lation. Concentrations in cerebrospinal fluid at doses of 10 mg/ syndrome, or rarely, respiratory distress, hemolytic anemia, leu-
kg are very low.31–­33 RIF does accumulate in aveolar macro- copenia and severe thrombocytopenia may occur. RIF should be
phages.33 Penetration into caseum may be slow, but is good.33,35 permanently discontinued in the face of serious hypersensitivity
RIF displays a modest EBA, but it is the most important ster- reactions. RIF is a strong inducer of multiple CYP isoforms in-
ilizing drug available for TB, meaning that the bacillary popula- cluding CYP3A4, CYP2A6, CYP2B6, CYP2C9 and CYP2C19
tion is sufficiently reduced to prevent relapse. Until very recently, but not CYP2D6 in vitro.19,20,32,33 RIF also induced Phase 2
regimens lacking RIF generally were 12–­18 months, occasionally enzymes and various transporters, including Pgp. RIF therefore

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 110 NUMBER 6 | December 2021 1459
 15326535, 2021, 6, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.2261 by Ircss Burlo Garofolo, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TUTORIAL

produces numerous drug-­drug interactions (DDI) since the ma- ensure 24-­hour trough concentrations are less than 1 mcg/mL.32–­34
jority of prescription drugs are metabolized by one or more of Anecdotal experience shows EMB should be avoided in patients
these isoforms. receiving peritoneal dialysis.

Pyrazinamide Alternative drugs

Pyrazinamide (PZA) is a pro-­drug, activated by human and myco- Under certain circumstances, one cannot or should not use the
bacterial amidases to form pyrazinoic acid (POA). It appears that standard RIPE regimen. These include known allergy or intoler-
protonated POA accumulates within Mtb and damages many im- ance, extensive DDI, or known drug resistance. Under these con-
portant cellular processes.32,33 Resistance primarily is associated ditions, and in consultation with experienced TB clinicians, one
with mutations in the pncA gene. Phenotypic and genotypic resis- can derive an effective alternative regimen to meet most clinical
tance testing for PZA can be challenging. The typical PZA dose situations. Every effort should be made to include a rifamycin in
is 25 mg/kg once daily, which is lower than the 35 mg/kg dose the regimen. The reader is referred to recently published guide-
studied by the British Medical Research Council (BMRC).17,18 A lines and their supplements for further details.3,19–­22
dose of 50–­70 mg/kg three times per week occasionally is used
when intermittent therapy is required.20,33 PZA is well-­absorbed Rifabutin
and not affected by food or antacids. PZA is metabolized to pyr- Rifabutin (RBN) is a lipophilic, synthetic spiro-­piperidyl deriva-
azinoic acid, and subsequently to 5-­hydroxy-­pyrazinoic acid.31–­33 tive of RIF, with a similar mechanism of action. Most mutations
Concentrations in the cerebrospinal fluid are similar to those in in the rpoB gene associated with RIF resistance also produce RBN
plasma.31–­33 PZA accumulates strongly in epithelial lining fluid resistance.20,31–­34 Some controversy exists as to whether “RIF-­
but not in alveolar macrophages.33 PZA does penetrate pulmo- resistant, RBN-­susceptible” isolates can be effectively treated with
nary lesions reasonably well.33,35 RBN, since clinical experience is severely limited. RBN absorp-
PZA has a weak EBA but is considered a key sterilizing drug, tion is not significantly affected by food.32,33 The primary metab-
shortening the treatment for DS-­TB from 9 to 6 months when olite is the partially active 25-­O-­desacetyl-­R BN.32,33 Like RIF,
combined with rifampin.17,18,20 PZA appears to have a maximum RBN shows concentration-­dependent killing. Cmax values with
impact within the first two months of treatment, targeting a sub- the typical 300 mg daily dose above 0.45 mcg/ml are associated
population of Mtb within acidified portions of the lesions.33 PZA’s with better clinical outcomes.32–­34 RBN does accumulate in aveo-
efficacy is predicted by AUC/MIC.34 DILI is the main concern lar macrophages.33 Penetration data into caseum are not available.
with PZA. Current guidelines do not recommend reintroducing Like RIF, RBN may occasionally cause DILI. RBN also may
PZA after a severe episode of DILI. Increased plasma uric acid, but cause reversible anterior uveitis, typically with combined Cmax val-
very rarely gout, and gastrointestinal intolerance may occur. PZA ues of the parent plus des-­acetyl metabolite above 1 mcg/ml, usu-
is free of significant drug interactions.20,32,33 ally as the result of DDI.31–­33 Arthralgia, skin discoloration and
leucopenia also have been reported. RBN is a less potent inducer of
Ethambutol CYP isoforms than RIF, and this is why it is used.20,31–­34 However,
The D-­isomer of ethambutol (EMB) is an inhibitor of myco- unpredictable bi-­directional interactions may occur, because RBN
bacterial cell wall arabinosyltransferases, leading to depletion of is a substrate for CYP3A4, as well as an inducer.
arabinogalactan and lipoarabinomannan. Resistance primarily
is associated with mutations in the embB gene, which codes for Rifapentine
the major arabinosyltransferase enzyme.32,33 Oral bioavailability Like RIF, RPT is an inhibitor of DNA-­ dependent RNA-­
is approximately 80%, with minimal effects of food or antacids.33 polymerase.31–­33 Complete cross-­resistance with RIF is expected.
70% is excreted unchanged in the urine, and elimination is related High-­fat meals increase the AUC by 30–­80%.20,32,33 Plasma
to renal function.31–­33 Concentrations in CSF are typically 50% protein binding is around 99%, and this may reduce penetration
or less than those in plasma, generally less that 1 mcg/ml.31–­33 into cavitary TB lesions.33 RPT is primarily metabolized to the
ETH does not accumulate in epithelial lining fluid but concen- partially active 25-­desacetyl-­R PT.32,33 Like RIF, RPT shows
trates highly in alveolar macrophages and in pulmonary lesions.33 concentration-­dependent killing.33,34. Initially, RPT was dosed
EMB has a modest EBA.33 EMB’s current first-­line use is in- at 600 mg, but more recent studies have used 1,200 mg daily,
tended to prevent emergence of RIF resistance. EMB is dosed at with food, for active TB disease. Dosing for LTBI is noted above.
15–­25 mg/kg daily or 50 mg/kg thrice weekly. EMB’s efficacy is RPT does accumulate in aveolar macrophages.33 Penetration into
predicted by Cmax/MIC and by AUC/MIC.34 The most serious caseum is very limited.33,38
toxicity of EMB is optic neuritis and is dose-­related, and which is Like RIF, RPT may be associated with DILI, and RPT may
consistent with EMB delivery to the CNS.20,31–­33 Screening of acu- cause serious rifamycin hypersensitivity syndromes. RPT strongly
ity and color vision prior to and during treatment is essential, and induces CYP3A4, especially with daily doses. Unlike RBN, RPT
EMB should be discontinued as soon as this complication is sus- has no advantage over RIF regarding DDI.20,32,33
pected. EMB is free of significant drug interactions. The half-­life of
EMB is prolonged in renal failure and EMB is minimally removed Fluoroquinolones
by dialysis (~1%).32–­34 Three-­times weekly dosing post-­dialysis Some of the fluoroquinolones have established roles as TB drugs,
is recommended with therapeutic drug monitoring (TDM) to even though they have not been approved by regulatory agencies

1460 VOLUME 110 NUMBER 6 | December 2021 | www.cpt-journal.com

 15326535, 2021, 6, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.2261 by Ircss Burlo Garofolo, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TUTORIAL

for that indication. 20,33,34 Moxifloxacin (MFX) and levofloxa- duration at 9 months, and possibly less. Various combinations of new
cin (LFX) are most commonly used. Limited head to head com- and investigational drugs will continue to be studied.28,29,33,34,39
parisons in humans with TB have been conducted, and debate
continues regarding the correct dose for each. Both MFX and Bedaquiline
LFX concentrate in alveolar macrophages and in pulmonary le- Bedaquiline (BDQ) is a synthetic diarylquinoline that inhibits my-
sions, as measured by microdialysis and MALDI-­ToF imaging.33 cobacterial ATP synthase.28,29,33,34,39 Mutations in the atpE gene
Early on, it was hoped that fluoroquinolones could approach the selected in vitro are clearly associated with high-­level resistance.33
sterilizing activity of the rifamycins, but they do not appear to Bioavailability is increased 2-­fold by food, and plasma protein
be as potent. Nevertheless, meta-­a nalyses show fluoroquinolo- binding is >99%.33,34 BDQ is metabolized by CYP3A4, CYP2C8
nes are key drugs in the treatment of MDR-­T B. 28,29,34,39 and 2C19 to N-­monodesmethyl-­BDQ and further demethlyated
metabolites, which are excreted in the feces.33 A loading dose of
Moxifloxacin 400 mg daily is administered for 2 weeks, followed by a mainte-
MFX and other fluoroquinolones inhibit the enzymes DNA gy- nance dose of 200 mg three times a week with food. BDQ’s Cmin
rase and topoisomerase IV, although the latter is lacking in Mtb. and AUC have a significant effect on response.34 In mouse models,
This blocks the supercoiling of DNA. Mutations in both the gyrA BDQ preferentially accumulated within the highly cellular regions
and less commonly gyrB genes confer resistance.21,22,32,33 in the lungs, with lower concentrations within the central caseum.41
Oral bioavailability of MFX is greater than 90%, and it has good BDQ exhibits a modest and delayed EBA. Culture conversion
CNS penetration.32,33 Elimination involves N-­sulfo conjugation at 8 weeks was improved by the addition of BDQ in a Phase II
and glucuronidation. The t½ of MFX is 11–­15 hours in healthy trial in MDR-­T B.28,29,33,34,39 Also, there was evidence of a mor-
volunteers, but is shorter, around 7 hours, in TB patients.33,34 tality benefit from observational data in South Africa. BDQ at
MFX’s efficacy is predicted by AUC/MIC.34 MFX concentrates steady state prolongs the QTc interval, but has not been clearly
in epithelial lining fluid, alveolar macrophages, and pulmonary associated with serious dysrhythmias.28,29,33,34,39 It may cause
lesions33,34A. MFX 400 mg daily has a moderate EBA and mod- gastrointestinal disturbance, arthralgia, headache and dizziness.
estly accelerated culture conversion in several Phase IIB studies in Phospholipidosis seen in preclinical models has not been ob-
DS-­TB.21,22,33,34 Preliminary results show that a 4 month regimen served in humans to date.33 Since it is principally metabolized
including RPT 1,200 mg and MFX 400 mg was non-­inferior to the by CYP3A4, BDQ plasma concentrations may be significantly
standard 6 month regimen.40 decreased by inducers of the enzyme (rifamycins, 75% or more
MFX prolongs the QTc interval and should be used with caution for RIF and RPT) and increased by inhibitors (ritonavir, keto-
in conjunction with other QTc prolonging agents, and in patients conazole and clarithromycin).33,34
with pro-­arrythmic conditions.32,33,34 Like other fluoroquinolones,
MFX also is associated with psychiatric disturbance, tendinopathy, Pretomanid
and increases in transaminases. RIF reduces the MFX AUC by 27–­ Pretomanid (PMD) is a pro-­drug that gained FDA approval in
32%, with RPT expected to do the same.33,34 Co-­administration of late 2019, in combination with BDQ and linezolid (BPaL), for the
di-­or trivalent cations should be avoided. treatment of highly-­resistant tuberculosis. 28,29,34,39,42 It is a nitro-
imidazole that inhibits mycolic acid synthesis in actively multiply-
Levofloxacin ing bacilli, with additional affects against NRP likely based on the
Levofloxacin shares many of the characteristics described above release of intracellular nitric oxide.28,42 Like delamanid below, it
for MFX. Oral bioavailability of LFX is 99–­100% with minimal appears that mutations in F420 coenzymes Rv3547, FGD, FbiA,
food effect.32–­34 Nearly 90% of the parent drug is excreted un- FbiB and FbiC are associated with in vitro resistance.43 Food in-
changed in the urine. The t½ of LFX is about 7 hours, and CSF creases pretomanid absorption. 28,34 Pretomanid is metabolized
penetration is about 74%.32–­34 LFX’s efficacy is predicted by by multiple reductive and oxidative pathways, with CYP3A4 re-
AUC/MIC.34 LFX concentrates in epithelial lining fluid, alveolar sponsible for approximately 20%.44 Efavirenz reduces the AUC
macrophages, and pulmonary lesions.33 of pretomanid by 35%, and rifampin by 66%.28,34 PMD’s Cmin
At a dose of 1,000 mg, LFX had a slightly higher EBA than and AUC have a significant effect on response in murine mod-
MFX.33 LFX typically is dosed at 750–­1,000 mg once daily for els.45 Detailed tissue penetration studies have yet to be published.
TB, although higher doses are being investigated. LFX can pro- The most common adverse events observed during pretomanid
long the QTc interval.32–­34 It shares other potential adverse drug therapy included peripheral neuropathy, anemia, GI upset, and el-
effects with MFX. LFX does not depend upon, or interfere with, evated liver enzymes. 28,34 Monitoring the latter is very important,
metabolizing enzymes. and studies continue regarding effects on male fertility.

Newer replacement regimens Linezolid

When RIPE or single drug modifications cannot be used, novel Linezolid (LZD) has been used for drug-­resistant Gram-­positive
regimens must be considered. Newer regimens are displacing the infections since 2000.32 It is a synthetic inhibitor of ribosomal
older second-­line drugs, with the “BPaL” regimen, BDQ, PMD, translation, binding to the 23S subunit and formation of the
and LZD currently of interest. Other regimens can be built around initiation complex.32,33 Resistance is associated in vitro with
these drugs, especially BDQ, to eliminate injectables, and keep the point mutations in the peptidyl transferase domain of the 23S

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 110 NUMBER 6 | December 2021 1461
 15326535, 2021, 6, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.2261 by Ircss Burlo Garofolo, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TUTORIAL

Rrna (rrs gene). Oral bioavailability is approximately 100% and pretomanid. It is a prodrug activated by the F420 dependent nitro-­
is not affected by food.32,33 Biotransformation occurs by non-­ reductases, and resistance is associated with mutations in the
enzymatic oxidation of the morpholine ring, and renal elimi- genes coding for these coenzymes.33 DLM inhibits mycolic acid
nation of parent drug plus metabolites.32–­3 4 Concentrations synthesis and releases reactive oxygen species in mycobacteria.27,33
in lesions determined by ex vivo dialysis were 49% of serum, Oral bioavailability is estimated to be 25–­47% and is increased
with good distribution by lesion type.33,35 CSF exposure is 57% 3–­4 fold by food.27,33,34 Plasma protein binding >99.5% to al-
of plasma.33 While twice daily dosing is common for Gram-­ bumin and lipoproteins.33 The metabolism of DLM is complex,
positive infections, 600 mg once daily is the most frequently initiated by albumin, with subsequent oxidation by CYP3A4 and
used dose for TB. Higher doses are being explored, but appear possibly other enzymes.33 DLM is dosed at 100 mg twice daily for
to be more toxic. 21,22,25 AUC appears to be the main driver of two months and then 200 mg daily with food.27–­29,33,34 DLM’s
efficacy, especially with combination therapy.34 Trough appears PK/PD driver is not well described. It may be reasonable to expect
closely associated with mitochondrial toxicity.34 Cmin and AUC have a significant effect on response, based on
LZD monotherapy achieved 87% culture conversion at six PMD.43
months in a small trial of patients with XDR-­TB.33 LZD inhib- DLM prolongs the QTc interval. Caution should be exercised
its protein synthesis in human mitochondria resulting in clinically in patients with hypoalbuminaemia. DLM also may cause nausea,
significant toxicities including lactic acidosis, myelosupression, and vomiting, tremor, anxiety and paraesthesia. RIF reduces the DLM
peripheral and optic neuropathy with chronic dosing, as is required AUC 47%, while ritonavir increases AUC by 25%.33 Use cau-
for TB.26,28,32 Trough concentrations of LZD correlate with mi- tiously with other inducers and inhibitors of CYP3A4.
tochondrial dysfunction and risk of toxicity, and therapeutic drug
monitoring may be useful.26,34 LZD does not depend heavily upon, Beta-­lactams
or interfere with, metabolizing enzymes. However, RIF can reduce Imipenem, Meropenem, Ceftazidime
LZD plasma exposure by about a third.33 LZD is a weak inhibitor of Imipenem (IMP) and Meropenem (MRP) are synthetic
monoamine oxidase and should usually not be used with MAOIs, carbapenems, and ceftazidime (CTZ) is a 3rd generation
SSRIs or triptans due to reports of serotonergic syndrome.33 cephalosporine.32,33 They target multiple transpeptidases
(penicillin-­binding proteins) involved in bacterial cell-­ wall
Clofazimine synthesis.32,33 Resistance occurs by enzymatic degradation.
Clofazimine (CFZ) is a semi-­s ynthetic rimino- Specifically, Mtb possesses an extended spectrum class A β-­
phenazine. 21,22,33,34 The mechanism of action has been pos- lactamase (BlaC), which must be inactivated.33 Clavulanic acid
tulated to be disruption of electron transport and generating and avibactam are candidates for this role. 29,33,34,46 Clavulanic
reactive oxygen species.33 Additional mechanisms have also acid can be given orally (only in combination with amoxicillin);
been suggested, including disruption of the membrane potential the other drugs are given intravenously. Dosing has been em-
through inhibition of potassium uptake channels.33 Resistance piric, and is not a settled matter. For rapidly multiplying bac-
mutations and the PK/PD driver for CFZ have not been clearly teria, time above MIC is the PK/PD driver. For slow-­g rowing
identified.33,34 Since CFZ has a very long half-­life, like BDQ, TB, clinical proof is pending.34 To date, most experience has
one can posit that Cmin and AUC will correlate with activity. been with IMP dosed at 1000 mg 12 hours and MRP dosed at
Oral bioavailability is 45–­62% and is increased 45% by a high 1,000 mg 8 hours, intravenously –­a route of administration
fat meal.33,34 Though the initial t½ is 7.8–­15.9 hours this reflects that limits widespread application in outpatient clinics.33 Most
distribution and the terminal elimination t½ is 70 days. 29,33,34 of the drugs are excreted unchanged in the urine.32–­3 4 No data
CFZ concentrations in the rim of lesions are 10 times higher are available on penetration of lesions. Concentrations in CSF
than in plasma, though lower in caseum.33,35 are 10–­30% of plasma.32,33 No comparative or controlled clini-
CFZ is usually dosed at 100 mg daily with food, though 200 mg cal studies have been reported, though a meta-­a nalysis of obser-
may be used for short periods in severe disease.20,21,33,34 CFZ crys- vational studies in MDR-­T B observed more favorable response
tals are deposited throughout the body during treatment, and rates with MRP than with IMP. 29,33 Adverse effects include hy-
may be associated with epithelial and body fluid discoloration, persensitivity reations, abnormal liver enzymes, and DILI. CNS
dry skin, photosensitivity and corneal deposits. The discoloration side-­effects including seizures may occur, especially with renal
is reversible, but only after six months or more off treatment. dysfunction.32,33 Beta-­lactams do not depend upon, or interfere
Gastrointestinal discomfort also is common, and gastrointestinal with, metabolizing enzymes.
obstruction and bleeding have been reported. CFZ has not been
shown to provoke drug-­drug interactions. However, CFZ may sig- Reserve drugs
nificantly prolong the QTc interval, and could potentiate this effect
A number of older TB drugs can be used in certain situations.
by other drugs.20,21,33,34 These drugs tend to have weaker activity against TB, higher rates
of adverse drug reactions, or both. Therefore, these drugs are used
Delamanid out of necessity, to round out regimens that otherwise are lacking.
Delamanid (DLM) is the R-­enantiomer of a synthetic nitro-­ With the advent of newer agents, these drugs likely will be used
dihydro-­
imidazo-­
oxazole derivative, a chemical relative of less. A brief summary is provided.

1462 VOLUME 110 NUMBER 6 | December 2021 | www.cpt-journal.com

 15326535, 2021, 6, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.2261 by Ircss Burlo Garofolo, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TUTORIAL

Aminoglycosides Hypothyroidism may occur in up to 30% of patients on ETA.32,33

Streptomycin (SM), kanamycin (KM), and amikacin (AK) all Gynaecomastia also may occur.
have been used for TB, usually at a dose of 15 mg/kg once daily.
Currently, amikacin appears to be the preferred drug, although Cycloserine and Terizidone
treatment is moving away from the use of injectable agents. D-­Cycloserine (CS) is a small, water-­ soluble weak acid.30–­33
Clearance is highly dependent on renal function. Aminoglycoside Terizidone (TZ) is a condensation product of two cycloserine
Cmax and AUC have a significant effect on response.34 SM and molecules with terephtalaldehyde, which acts as a pro-­drug. CS is
AK have weak EBA but the clinical efficacy of aminoglycosides a natural product analogue of D-­A lanine. It is a competitive inhib-
in both DS and MDR-­TB is supported by clinical trial and ob- itor of alanine racemase and D-­a lanine:D-­a lanine ligase, disrupt-
servational data.17–­19,30,32,33 Penetration into human lung lesions ing peptidoglycan synthesis.33 CSF concentrations are 80–­100%
was poor.35 of plasma in meningitis.14 Analogous to beta-­lactams, Cmin/
Aminoglycosides are associated with ototoxicity (hearing loss MIC or time above MIC have been suggested as the driver of effi-
or vestibular effects) which are related to the duration of dosing, cacy.34 Penetration into human lung lesions is unknown.
but not to dose size or interval.30–­34 Nephrotoxicity is less com- No EBA studies of CS or TZ are available. Weak evidence for
mon and usually reversible. Hypokalemia and hypomagnesemia their efficacy derives from combination studies in retreatment
also occur.30–­33 from the 1960s.30–­34. The usual starting dose of CS and TZ is
250 mg once daily, increased by 250 mg every few days to a maxi-
Capreomycin mum of 500 mg twice daily, though many patients tolerate less than
Capreomycin (CM) is an injectable macrocyclic polypeptide. that. The therapeutic index of CS is low. Neuropsychiatric side-­
Cross-­resistance with KM / AK is common but not with SM. It is effects are common (~6%) including drowsiness, anxiety, mood
excreted unchanged in the urine.30–­33 The typical dose is 15 mg/ disturbance, psychosis and seizures.30–­34 Pyridoxine 50–­100 mg is
kg daily IM or IV. Overall, CM is similar to aminoglycosides re- suggested to counteract neurotoxicity, but proof of efficacy is lack-
garding dose, pharmacokinetics, and adverse effects. It is assumed ing.33 Dosing should be guided by therapeutic drug monitoring.
that the PK/PD drivers also are similar. Based on recent meta-­
analyses, its use has fallen out of favor.20–­22 Problems and Solutions
Heteroresistance and multiple isolates
Para-­aminosalicylic acid Knowing what you are treating is one key to success. It is well
Para-­aminosalicylic acid (PAS) is a synthetic analogue of para-­ known that patients can have more than one Mtb isolate infecting
aminobenzoic acid.17–­19,30–­34 Absorption of the sustained release them at the same time. This might occur due to more than one in-
dosage forms is improved by food.32,33 The usual dose is 4000 mg fectious event, or generated within the host due to variable growth
twice daily. The major metabolite is N-­acetyl PAS, produced by conditions or gradients of drug exposure. Molecular approaches
NAT1. PAS has a modest EBA that may be dose dependent.30–­34 can detect heteroresistance, i.e. the presence of a resistant subpop-
The PK/PD drivers for PAS are not certain. Based on limited in ulation in otherwise phenotypically drug susceptible mycobacte-
vitro activity, time above MIC can be suggested.47 In contrast, ria.49,50 This is an active area of research focused on refining tests
limited clinical data suggest high peaks may produce better ef- to be both sensitive and highly specific.
fects.48 Penetration into human lung lesions is unknown. GI upset
and diarrhea are the main adverse effects. PAS has been associ- Drug penetration into cavitary lesions
ated with hypothyroidism and goiter, especially when used with It is widely recognized that Mtb provokes an immune response
thioamides.30–­34 that can seriously damage the lung parenchyma.3,6,19,33 Cavitary
lesions can wax and wane over the course of the disease, and can
Thioamides provide sanctuary to organisms, especially those in the acidified
Ethionamide, Prothionamide areas and in the caseum. Drug delivery into these areas is needed,
Ethionamide (ETA) and Prothionamide (PTA) are synthetic but must rely on passive diffusion and the chemical properties of
structural analogues of nicotinamide. Both are prodrugs activated the drug.33 Several models have been evaluated, including in vivo
by the mycobacterial mono-­oxygenase EthA and target the same and ex vivo animal models, and ex vivo human models. It appears
enoyl-­acyl reductase enzyme InhA as INH, disrupting mycolic that the drug with the most favorable penetration profile is rifam-
acid synthesis.30–­34 Oral bioavailability is nearly complete and pin.35,38,51,52 Information of this type can assist in the design of
is not significantly affected by food.32,33 Clearance involves sul- next-­generation TB drugs.
phoxidation, desulphuration and deamination followed by meth-
ylation. The sulphoxide metabolites are active. AUC has been Drug Penetration into the central nervous system
suggest as the driver of efficacy, similar to that seen with INH.34 The deadliest form of TB in central nervous system TB, most
Penetration into human lung lesions is unknown. often meningitis but sometimes tuberculoma.3,19,33 As with
No EBA data are available for ETA or PTA. ETA and PTA are other central nervous system infections, CNS penetration is
dosed at 15–­20 mg/kg per day up to a maximum of 1g. To im- limited for some important drugs, including rifampin and
prove tolerability, the drugs are usually dose-­escalated over several ethambutol.31–­33,53–­56 Treatment typically is extended up to
days. Gastrointestinal tolerability of ETA and PTA is poor.30–­34 a year.3,19,20 Higher doses, and alternative agents, including

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 110 NUMBER 6 | December 2021 1463
 15326535, 2021, 6, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.2261 by Ircss Burlo Garofolo, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TUTORIAL

fluoroquinolones and linezolid, are being studied for this condi- The interested reader also can evaluate TB research programs at
tion.33,53–­56 Advances for this condition can contribute signifi- other institutions, including Johns Hopkins University, Colorado
cantly to lowering TB morbidity and mortality. State University, Hackensack Meridian Health, the University of
California, San Francisco, and several others outside of the United
Drug Penetration into bone States, including the University of Liverpool, Radboud University,
TB osteomyelitis is another condition frequently requiring ex- the University of Cape Town, and Stellenbosch University.
tended treatment.3,19,33 Because of the limited vascular supply to Preclinical animal models of infection are employed to develop
parts of various bones, drug penetration can be limited as well as new agents, but also to screen among molecules to rank them.78,79
difficult to study.19,31–­33,57–­59 This is an area where alternative dos- Limitations include variable natural susceptibility to TB, different
age forms might be able to expedite the elimination of Mtb within immune responses, different lesion pathology, and so forth. Also,
bone lesions. animal models of infection may have very different pharmacoki-
netics compared to humans, leading to substantial differences in ef-
Inhaled drug therapy ficacy.78,79 Otherwise useful drugs could be discarded if a particular
Given that TB typically presents as a chronic pneumonia, it is model does not look favorable. The HFIM can be used to evaluate
tempting to consider direct chemotherapy in the lungs, thus spar- the impact of different pharmacokinetic profiles on bacterial cell
ing systemic toxicity.19,32 Several drugs have been prepared as in- kill, as well as resistance suppression.78,79 Then, the in vitro results
haled dosage forms.60–­6 4 The challenges to this approach include are examined with mathematical modeling, allowing one to estab-
delivery into the damaged areas of the lung, extra-­pulmonary or- lish links between HFIM reproductions of plasma and bronchoa-
ganisms, patient acceptance, adherence, and proper technique, and veolar lavage (BAL) concentrations from animals or humans, and
cost. All of these problems can be overcome, some more easily than the pharmacodynamics of test drugs.78,79 This information allows
others. It will remain true that TB is not confined to the lungs.3,19 one to select the dose and frequency needed to optimize therapy.
Therefore, inhaled therapy may be an add-­on component of ther- As more repurposed drugs and new agents are considered, how
apy, potentially rendering patients non-­infectious more quickly, does one prioritize among the hundreds of potential regimens
and potentially shortening the duration of treatment. to explore? Even 6–­8 potential agents translate into 64-­256 pos-
sible 2-­drug combinations.80 There is neither time nor resources
Are there biological barriers to shortening regimens? to give an extensive evaluation for all combinations.80 A screen-
It is only with immense difficulty that randomized clinical trials ing procedure is needed to identify combinations with a high
have been able to show reasonable success (non-­inferiority) with a likelihood of achieving good bacterial burden decline.80 The
standardized 4 month regimen.38 Many other potentially good reg- University of Florida research team examined pretomanid, moxi-
imens were not able meet this goal.65–­67 Part of the puzzle lies with floxacin, linezolid and bedaquiline in Log-­phase growth, Acid-­
the heterogeneity among patients, some having new disease, others phase growth and non-­replicating persister (NRP)-­phase, using
well-­entrenched or recurrent, some having minimal lesions, others the Greco interaction model.80 The primary metric was bacterial
having large cavities. Another part of the puzzle is that there is no kill.80 The combination of pretomanid plus moxifloxacin emerged
good way, currently, to assess the degree of immune function specif- as the clear frontrunner, as the largest bacterial declines were seen
ically against TB, and no good way to enhance it. Yet another piece in Log-­phase and Acid-­phase with this regimen and it was second
lies in standardized dosing –­one size fits all –­even though wide best in NRP-­phase, while bedaquiline also produced good kill.80
interpatient pharmacokinetic variability is known to exist.34,68 This screening process may identify optimal combinations that can
Finally, it is not known if the pathogen within the host, especially be further evaluated in both the HFIM and in animal models of
acid-­phase and NRP Mtb, might have defense mechanisms that Mycobacterium tuberculosis infection.80
simply require months of chemotherapy to effectively overcome. In The next phase is to develop multidrug regimens, combining the
other words, are there biological barriers to shortening regimens, best candidates from the initial steps.81 First, 2-­drug regimens with
such as phenotypic antibiotic tolerance, that prevent further short- a synergistic or additive interaction are determined, using a full fac-
ening of regimens until they are understood and overcome? torial study design.81 Next, optimal 3-­drug regimens are explored.
Total bacterial burdens and the less-­susceptible or resistant subpop-
Understanding drugs and doses ulations can be examined.81 Once again, mathematical models are
Preclinical models allow for explorations that would not be possi- fit to the data, followed by Monte Carlo simulations. The time and
ble in humans. Several animal species have been used for TB drug conditions needed to drive the total bacterial burden to extinction
development, each with advantages and disadvantages. These in- with multidrug regimens can be determined.81 This model-­based
clude various types of mice, Guinea pigs, rabbits, and non-­human system approach to evaluating combinations of 3 agents shows
primates.69–­74 In vitro systems, including Hollow Fiber Infection promise to rapidly identify the most promising combinations to
Model (HFIM), allow one to control drug exposure, and one can take forward in human trials.81
mimic either animal or human pharmacokinetics.75–­77 Through We may look at the combination of these citations 78–­81 and rec-
this process, important insights have been learned. A few of these ognize that, taken together, they lay out an algorithm to identify
are shared here, reflecting our experience with an NIH-­funded potential high value multi-­drug combinations. As more and more
program project grant for TB regimen development, based at the agents are discovered or repurposed, this algorithm will be increas-
University of Florida. This is one example of such an approach. ingly useful.

1464 VOLUME 110 NUMBER 6 | December 2021 | www.cpt-journal.com

 15326535, 2021, 6, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.2261 by Ircss Burlo Garofolo, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TUTORIAL

Conclusions 16. Swindells, S. et al. Long-­acting formulations for the treatment of

This tutorial should provide the reader with a concise summary latent tuberculosis infection: opportunities and challenges. Int. J.
Tuberc. Lung Dis. 22, 125–­132 (2018).
of major trends in the area of TB treatment. We attempted to be 17. Fox, W., Ellard, G.A. & Mitchison, D.A. Studies on the treatment
judicious in our selection of topics, and this tutorial should not be of tuberculosis undertaken by the British Medical Research
considered encyclopedic. We do hope that the references provided Council tuberculosis units, 1946–­1986, with relevant subsequent
publications. Int. J. Tuberc. Lung Dis. 3(10 Suppl 2), S231–­S279
offer the reader an opportunity to examine the past and to look (1999).
into the future of TB treatment. 18. Mitchison, D.A. Role of individual drugs in the chemotherapy of
tuberculosis. Int. J. Tuberc. Lung Dis. 4, 796–­806 (2000).
FUNDING 19. Iseman, M.D. A clinician’s guide to tuberculosis (Lippincott
No funding was received for this work. Williams & Wilkins, Philadelphia, PA, 2000); 448 pages.
20. Nahid, P. et al. Official American Thoracic Society/Centers for
Disease Control and Prevention/Infectious Diseases Society
CONFLICT OF INTEREST of America Clinical Practice Guidelines: treatment of drug-­
The authors declared no competing interests for this work. susceptible tuberculosis (and supplementary materials). Clin.
Infect. Dis. 63, e147–­e195 (2016).
© 2021 The Authors. Clinical Pharmacology & Therapeutics 21. Nahid, P. et al. Treatment of drug-­resistant tuberculosis. an
© 2021 American Society for Clinical Pharmacology and Therapeutics official ATS/CDC/ERS/IDSA clinical practice guideline. Am. J.
Respir. Crit. Care Med. 200, e93–­e142 (2019).
22. WHO consolidated guidelines on drug-­resistant tuberculosis
1. Global Tuberculosis Report 2020. Geneva: World Health treatment. Geneva: World Health Organization <https://apps.who.
Organization; 2020 <https://apps.who.int/iris/bitst​ream/handl​ int/iris/bitst​ream/handl​e/10665/​31138​9/97892​41550​529- ­eng.
e/10665/​33606​9/97892​40013​131-­eng.pdf?ua=1>. Accessed pdf?ua=1> (2019). Accessed November 3, 2020
November 3, 2020. 23. Nunn, A.J. et al. A trial of a shorter regimen for rifampin-­resistant
2. Hargreaves, J.R., Boccia, D., Evans, C.A., Adato, M., Petticrew, tuberculosis. N. Engl. J. Med. 380, 1201–­1213 (2019).
M. & Porter, J.D. The social determinants of tuberculosis: from 24. Conradie, F. et al. Treatment of highly drug-­resistant pulmonary
evidence to action. Am J Public Health. 101, 654–­662 (2011). tuberculosis. N. Engl. J. Med. 382, 893–­902 (2020).
3. Namdar, R. & Peloquin, C.A. Tuberculosis. In: DiPiro, J.T., 25. Safety and efficacy of various doses and treatment durations of
Yee, G.C., Posey, L., Haines, S.T., Nolin, T.D. & Ellingrod, V. linezolid plus bedaquiline and pretomanid in participants with
eds. Pharmacotherapy: A Pathophysiologic Approach, 11e. pulmonary TB, XDR-­TB, Pre-­XDR-­TB or non-­responsive/intolerant
McGraw-­Hill <https://acces​sphar​m acy.mhmed​ical.com/conte​ MDR-­TB (ZeNix) <https://clini​caltr​ials.gov/ct2/show/NCT03​
nt.aspx?booki​d =2577&secti​o nid=22339​8 313>. Accessed June 086486>. Accessed November 3, 2020
16, 2020. 26. Brown, A.N. et al. Preclinical evaluations to identify optimal
4. Treatment Action Group. Public Investment in Bedaquiline Up linezolid regimens for tuberculosis therapy. MBio 6,
to Five-­F old That of Johnson and Johnson <https://www.treat​ e01741–­e1815 (2015).
menta​c tion​g roup.org/state​m ent/publi​c -­inves​t ment​-­in-­b edaq​ 27. Liu, Y. et al. Delamanid: from discovery to its use for pulmonary
uilin​e -­up-­to-­f ive-­fold-­t hat- ­of-­johns​o n-­johns​o n/>. Accessed multidrug-­resistant tuberculosis (MDR-­TB). Tuberculosis 111, 20–­
November 3, 2020. 30 (2018).
5. Working Group on New TB Drugs. Clinical Pipeline <https://www. 28. Ignatius, E.H. & Dooley, K.E. New drugs for the treatment of
newtb​drugs.org/pipel​ine/clinical>. Accessed November 3, 2020. tuberculosis. Clin. Chest Med. 40, 811–­827 (2019).
6. Huang, H. & Russell, D.G. Protective immunity against 29. Tiberi, S., Muñoz-­Torrico, M., Duarte, R., Dalcolmo, M.,
tuberculosis: what does it look like and how do we find it? Curr. D’Ambrosio, L. & Migliori, G.B. New drugs and perspectives for
Opin. Immunol. 48, 44–­50 (2017). new anti-­tuberculosis regimens. Pulmonology. 24, 86–­98 (2018).
7. Kaufmann, S.H.E. & Weiner, J. Fordham von Reyn C. Novel 30. Bartmann, K. ed. Anti-­tuberculosis Drugs (Springer-­Verlag, Berlin,
approaches to tuberculosis vaccine development. Int. J. Infect. 1988).
Dis. 56, 263–­267 (2017). 31. Heifets, L. ed. Drug Susceptibility in the Chemotherapy of
8. Tang, J., Yam, W.C. & Chen, Z. Mycobacterium tuberculosis Mycobacterial Infections (CRC Press, Boca Raton, 1991).
infection and vaccine development. Tuberculosis 98, 30–­41 32. Antimicrobial Chemotherapy and Vaccines Vol II: Antimicrobial
(2016). Agents 2005 (and subsequent on-­line only updates). Pittsburgh;
9. Sterling, T.R., Njie, G., Zenner, D. et al. Guidelines for the Esun Technologies, LLC <http://www.antim​icrobe.org>.
treatment of latent tuberculosis infection: recommendations from 33. Friedman, L.N., Dedicoat, M. & Davies, P.D.O. eds. Clinical
the National Tuberculosis Controllers Association and CDC, 2020. Tuberculosis, 6th edn (CRC Press, Boca Raton, 2020) <https://
MMWR Recommend. Rep. 69, 1–­16 (2020). www.taylo​r fran​cis.com/books/​e/97813​51249980>.
10. Comstock, G.W., Ferebee, S.H. & Hammes, L.M. A controlled 34. Märtson, A.G., Burch, G., Ghimire, S., Alffenaar, J.C. & Peloquin,
trial of community-­wide isoniazid prophylaxis in Alaska. Am. Rev. C.A. Therapeutic drug monitoring in patients with tuberculosis
Respir. Dis. 95, 935–­943 (1967). and concurrent medical problems. Expert Opin. Drug Metab.
11. Comstock, G.W., Baum, C. & Snider, D.E. Jr Isoniazid prophylaxis Toxicol. 17, 23–­39 (2021).
among Alaskan Eskimos: a final report of the bethel isoniazid 35. Strydom, N. et al. Tuberculosis drugs’ distribution and emergence
studies. Am. Rev. Respir. Dis. 119, 827–­8 30 (1979). of resistance in patient’s lung lesions: a mechanistic model
12. Comstock, G.W. How much isoniazid is needed for prevention of and tool for regimen and dose optimization. PLoS Medicine 16,
tuberculosis among immunocompetent adults? Int. J. Tuberc. Lung e1002773 (2019).
Dis. 3, 847–­850 (1999). 36. Boeree, M.J. et al. High-­dose rifampicin, moxifloxacin, and SQ109
13. Menzies, D. et al. Four months of rifampin or nine months of for treating tuberculosis: a multi-­arm, multi-­stage randomized
isoniazid for latent tuberculosis in adults. N. Engl. J. Med. 379, controlled trial. Lancet Infect. Dis. 17, 39–­49 (2017).
440–­453 (2018). 37. Svensson, E.M. et al. The potential for treatment shortening with
14. Swindells, S. et al. One month of rifapentine plus isoniazid to higher rifampicin doses: relating drug exposure to treatment
prevent HIV-­related tuberculosis. N. Engl. J. Med. 380, 1001–­ response in patients with pulmonary tuberculosis. Clin. Infect. Dis.
1011 (2019). 67, 34–­41 (2018).
15. Worley, M.V. & Estrada, S.J. Bedaquiline: a novel antitubercular 38. Rifat, D. et al. Pharmacokinetics of rifapentine and rifampin in
agent for the treatment of multidrug-­resistant tuberculosis. a rabbit model of tuberculosis and correlation with clinical trial
Pharmacotherapy 34, 1187–­1197 (2014). data. Sci. Transl. Med. 10, eaai7786 (2018).

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 110 NUMBER 6 | December 2021 1465
 15326535, 2021, 6, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.2261 by Ircss Burlo Garofolo, Wiley Online Library on [13/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TUTORIAL

39. Chang, K.C., Nuermberger, E., Sotgiu, G. & Leung, C.C. New drugs 60. Tsapis, N. et al. Direct lung delivery of para-­aminosalicylic acid by
and regimens for tuberculosis. Respirology 23, 978–­990 (2018). aerosol particles. Tuberculosis 83, 379–­385 (2003).
40. CDC. Media statement: Landmark TB trial identifies shorter-­ 61. Sung, J.C. et al. Dry powder nitroimidazopyran antibiotic PA-­824
course treatment regimen. Oct. 22, 2020 <https://www.cdc.gov/ aerosol for inhalation. Antimicrob. Agents Chemother. 53, 1338–­
nchhs​tp/newsr​oom/2020/landm​ark-­tb-­trial​-­media​-­state​ment. 1343 (2009).
html>. 62. Sung, J.C. et al. Formulation and pharmacokinetics of self-­
41. Irwin, S.M. et al. Bedaquiline and pyrazinamide treatment assembled rifampicin nanoparticle systems for pulmonary
responses are affected by pulmonary lesion heterogeneity in delivery. Pharm. Res. 26, 1847–­1855 (2009).
Mycobacterium tuberculosisInfected C3HeB/FeJ mice. ACS Infect. 63. Garcia-­Contreras, L. et al. Pharmacokinetics of ethionamide
Dis. 2, 251–­267 (2016). delivered in spray-­dried microparticles to the lungs of guinea pigs.
42. Lee, S.F.K., Laughon, B.E., McHugh, T.D. & Lipman, M. New J. Pharm. Sci. 106, 331–­337 (2017).
drugs to treat difficult tuberculous and nontuberculous 64. Braunstein, M., Hickey, A.J. & Ekins, S. Why wait? The case for
mycobacterial pulmonary disease. Curr. Opin. Pulm. Med. 25, treating tuberculosis with inhaled drugs. Pharm. Res. 36, 166
271–­280 (2019). (2019).
43. Dookie, N., Rambaran, S., Padayatchi, N., Mahomed, S. & Naidoo, 65. Gillespie, S.H. et al. Four-­month moxifloxacin-­based regimens
K. Evolution of drug resistance in Mycobacterium tuberculosis: for drug-­sensitive tuberculosis. N. Engl. J. Med. 371, 1577–­1587
a review on the molecular determinants of resistance and (2014).
implications for personalized care. J. Antimicrob. Chemother. 73, 66. Merle, C.S. et al. A four-­month gatifloxacin-­containing regimen for
1138–­1151 (2018). treating tuberculosis. N. Engl. J. Med. 371, 1588–­1598 (2014).
44. Pretomanid Full Prescribing Information <https://www.acces​ 67. Jindani, A. et al. High-­dose rifapentine with moxifloxacin for
sdata.fda.gov/drugs​atfda_docs/nda/2019/21286​2Orig​1s000​ pulmonary tuberculosis. N. Engl. J. Med. 371, 1599–­1608 (2014).
Lbl.pdf>. Accessed November 5, 2020. 68. Peloquin, C. The role of therapeutic drug monitoring in
45. Ahmad, Z. et al. PA-­824 exhibits time-­dependent activity in a mycobacterial infections. Microbiol. Spectrum. 5, TNMI7-­0029
murine model of tuberculosis. Antimicrob. Agents Chemother. 55, (2016).
239–­245 (2011). 69. Gumbo, T., Lenaerts, A.J., Hanna, D., Romero, K. & Nuermberger,
46. Deshpande, D. et al. Ceftazidime-­avibactam has potent sterilizing E. Nonclinical models for antituberculosis drug development: a
activity against highly drug-­resistant tuberculosis. Sci. Adv. 3, landscape analysis. J. Infect. Dis. 211(Suppl 3), S83–­S95 (2015).
e1701102 (2017). 70. Nuermberger, E.L. Preclinical efficacy testing of new drug
47. Peloquin, C.A., Berning, S.E., Huitt, G.A., Childs, J.M., Singleton, candidates. Microbiol. Spectr. 5 (2017). https://doi.org/10.1128/
M.D. & James, G.T. Once-­daily and twice-­daily dosing of p-­ micro​biols​pec.TBTB2​- ­0034-­2017
aminosalicylic acid granules. Am. J. Respir. Crit. Care Med. 159, 71. Blanc, L., Lenaerts, A., Dartois, V. & Prideaux, B. Visualization
932–­934 (1999). of mycobacterial biomarkers and tuberculosis drugs in infected
48. Abulfathi, A.A., Donald, P.R., Adams, K., Svensson, E.M., Diacon, tissue by MALDI-­MS imaging. Anal. Chem. 90, 6275–­6282
A.H. & Reuter, H. The pharmacokinetics of para-­aminosalicylic (2018).
acid and its relationship to efficacy and intolerance. Br. J. Clin. 72. Zhang, N. et al. Mechanistic modeling of Mycobacterium
Pharmacol. 86, 2123–­2132 (2020). tuberculosis infection in murine models for drug and vaccine
49. Hofmann-­Thiel, S. et al. Mechanisms of heteroresistance to efficacy studies. Antimicrob. Agents Chemother. 21, e01727-­19
isoniazid and rifampin of Mycobacterium tuberculosis in Tashkent, (2020).
Uzbekistan. Eur. Respir. J. 33, 368–­374 (2009). 73. Ernest, J.P. et al. Development of new tuberculosis drugs:
50. Rigouts, L. et al. Fluoroquinolone heteroresistance in translation to regimen composition for drug-­sensitive and
Mycobacterium tuberculosis: detection by genotypic and multidrug-­resistant tuberculosis. Annu. Rev. Pharmacol. Toxicol.
phenotypic assays in experimentally mixed populations. Sci. Rep. 61, 495–­516 (2020).
9, 11760 (2019). 74. Kim, S. et al. Pharmacokinetics of tedizolid, sutezolid, and
51. Kempker, R.R. et al. Cavitary penetration of levofloxacin among sutezolid-­M1 in non-­human primates. Eur. J. Pharm. Sci. 151,
patients with multidrug-­resistant tuberculosis. Antimicrob. Agents 105421 (2020).
Chemother. 59, 3149–­3155 (2015). 75. Drusano, G.L. From lead optimization to NDA approval for
52. Dheda, K. et al. Drug-­penetration gradients associated with a new antimicrobial: Use of pre-­clinical effect models and
acquired drug resistance in patients with tuberculosis. Am. J. pharmacokinetic/pharmacodynamic mathematical modeling.
Respir. Crit. Care Med. 198, 1208–­1219 (2018). Bioorg. Med. Chem. 24, 6401–­6408 (2016).
53. Wilkinson, R.J. et al. Tuberculous meningitis. Nat. Rev. Neurol. 13, 76. de Miranda, S.C. et al. Effect of linezolid plus bedaquiline against
581–­598 (2017). Mycobacterium tuberculosis in log phase, acid phase, and
54. Abulfathi, A.A., Decloedt, E.H., Svensson, E.M., Diacon, nonreplicating-­persister phase in an in vitro assay. Antimicrob.
A.H., Donald, P. & Reuter, H. Clinical pharmacokinetics and Agents Chemother. 62, e00856-­18 (2018).
pharmacodynamics of rifampicin in human tuberculosis. Clin. 77. Louie, A. et al. Activity of moxifloxacin against Mycobacterium
Pharmacokinet. 58, 1103–­1129 (2019). tuberculosis in acid phase and nonreplicative-­persister phenotype
55. Ruslami, R. et al. Intensified regimen containing rifampicin and phase in a hollow-­f iber infection model. Antimicrob. Agents
moxifloxacin for tuberculous meningitis: an open-­label, randomised Chemother. 62, e01470-­18 (2018).
controlled phase 2 trial. Lancet Infect. Dis. 13, 27–­35 (2013). 78. Drusano, G.L. et al. Developing new drugs for the therapy of
56. Donald, P.R. Cerebrospinal fluid concentrations of Mycobacterium tuberculosis: What information do we get from
antituberculosis agents in adults and children. Tuberculosis 90, pre-­clinical animal models? Antimicrob. Agents Chemother. 64,
279–­292 (2010). e01376-­20 (2020).
57. Donald, P.R. The chemotherapy of osteo-­articular tuberculosis 79. Drusano, G.L. et al. Dose-­fractionation of moxifloxacin for
with recommendations for treatment of children. J. Infect. 62, the treatment of tuberculosis: impact of dosing interval and
411– ­4 39 (2011). elimination half-­life on microbial kill and resistance suppression.
58. Kizilbash, Q.F. & Seaworth, B.J. Multi-­drug resistant tuberculous Antimicrob. Agents Chemother. 65, e02533-­20 (2021).
spondylitis: a review of the literature. Ann. Thorac. Med. 11, 233–­ 80. Drusano, G.L. et al. The funnel: a screening technique for
236 (2016). identifying optimal two-­drug combination chemotherapy regimens.
59. Pandita, A., Madhuripan, N., Pandita, S. & Hurtado, R.M. Antimicrob. Agents Chemother. 65, e02172-­20 (2021).
Challenges and controversies in the treatment of spinal 81. Drusano, G.L. et al. Building optimal three-­drug combination
tuberculosis. J. Clin. Tuberc. Other Mycobact. Dis. 19, 100151 chemotherapy regimens. Antimicrob. Agents Chemother. 64,
(2020). e01610-­20 (2020).

1466 VOLUME 110 NUMBER 6 | December 2021 | www.cpt-journal.com