Multidrug-resistant and rifampicin-resistant tuberculosis, collectively known as MDR/RR-TB, remains one of the most formidable challenges in global infectious disease control. When the bacteria that cause tuberculosis become resistant to isoniazid and rifampicin, the two most powerful first-line antibiotics, treatment options narrow dramatically and the regimens that remain are longer, more toxic, and less effective. For patients whose disease progresses to pre-extensively or extensively drug-resistant tuberculosis, the therapeutic arsenal shrinks even further, making every usable drug precious.

Delamanid, a nitroimidazo-oxazole derivative, was developed specifically to address this gap. The drug works through a mechanism distinct from most classical anti-tuberculosis agents: it blocks the synthesis of mycolic acids, essential components of the mycobacterial cell wall, thereby killing the bacillus while interfering with its energy metabolism. Controlled clinical trials demonstrated its therapeutic benefit, and it has since become a cornerstone of optimized background regimens, the individually tailored combinations of drugs assembled around core agents for each resistant case. Yet despite its growing importance, large-scale real-world safety data have remained limited, particularly for Chinese patients, who represent a substantial share of the global burden of drug-resistant tuberculosis.

To close this evidence gap, researchers led by teams at Beijing Chest Hospital, Capital Medical University, conducted a prospective observational study at 26 hospitals across China between July 2020 and August 2023. They enrolled 608 patients with microbiologically confirmed MDR/RR-TB, pre-XDR-TB, or XDR-TB, all of whom received delamanid alongside optimized background regimens. The study was registered at clinicaltrials.gov under identifier NCT04421495 on June 9, 2020, and was conducted in accordance with the Declaration of Helsinki, with ethics approval from the leading center and written informed consent from all participants.

The safety evaluation was comprehensive, encompassing adverse events, serious adverse events, and adverse drug reactions, together with targeted cardiac monitoring. Because delamanid is known to prolong the QT interval, the portion of the electrocardiographic cycle representing ventricular repolarization, the investigators systematically measured the QT interval corrected using Fridericia’s formula, abbreviated QTcF. Prolongation of this interval matters because extreme QT stretching can predispose patients to torsades de pointes, a potentially life-threatening arrhythmia. All data were captured through a standardized electronic data management system and analyzed with SAS 9.4, ensuring consistency across the many participating centers.

The headline numbers paint a picture of tolerability broadly consistent with prior knowledge. Among the 608 participants, 352 patients, or 57.9 percent, reported at least one adverse event, while 331 patients, or 54.4 percent, experienced adverse drug reactions. Serious adverse events were documented in 39 patients, representing 6.4 percent of the cohort. When adverse events were grouped by system organ class, the most frequently affected categories were blood and lymphatic system disorders at 24.5 percent, investigations at 22.7 percent, driven mainly by electrocardiographic QT prolongation, and hepatobiliary disorders at 15.6 percent. These patterns reflect the cumulative toxicity of multidrug regimens rather than delamanid alone, since patients typically take five or more medications for many months.

Drilling down to reactions specifically attributed to delamanid, the investigators found that 167 patients, or 27.5 percent, experienced at least one. The most common was QT prolongation, affecting 17.1 percent of the cohort, followed by gastrointestinal disturbances at 3.3 percent and hepatotoxicity at 3.1 percent. QTcF prolongation, defined either as an increase of at least 60 milliseconds from baseline or as an absolute value of at least 500 milliseconds, was observed in 138 individuals, or 22.7 percent of participants. This rate falls within the range documented in earlier trials, reinforcing the message that cardiac monitoring, rather than avoidance of the drug, is the appropriate clinical response.

Perhaps the most intriguing analytical finding concerned who develops QT prolongation. Using multivariable logistic regression, the researchers identified age as the only significant factor associated with QTcF prolongation, and the direction of the effect ran counter to many clinicians’ intuitions. Compared with patients younger than 25 years, those aged 25 to 44 years had significantly lower odds of developing QTcF prolongation, with an adjusted odds ratio of 0.232 and a 95 percent confidence interval of 0.150 to 0.361. Patients aged 45 years or older also had lower odds, with an adjusted odds ratio of 0.386 and a 95 percent confidence interval of 0.243 to 0.614. In other words, the youngest patients in the cohort were the most likely to experience electrocardiographic changes, a pattern that suggests heightened vigilance and more frequent ECG surveillance may be warranted for adolescents and young adults receiving delamanid.

The study’s scale and design give its conclusions considerable weight. By enrolling more than 600 patients across 26 hospitals in a routine care setting, the investigators captured a far broader spectrum of patients than typical randomized trials do, including individuals with comorbidities, varying disease severity, and diverse background regimens that would often exclude trial participants. The prospective structure and standardized electronic data capture minimized the reporting inconsistencies that plague retrospective pharmacovigilance studies. The result is one of the clearest real-world portraits to date of how delamanid behaves outside the controlled environment of a clinical trial.

For global tuberculosis programs, the implications are practical and timely. The World Health Organization has encouraged the adoption of shorter, all-oral regimens for drug-resistant tuberculosis, and combinations that include both delamanid and bedaquiline, another QT-prolonging agent, are increasingly used despite theoretical concerns about additive cardiac effects. The Chinese cohort’s finding that no new safety signals emerged during available follow-up supports the continued expansion of delamanid-containing therapy, provided that the established monitoring framework, periodic electrocardiography and serum potassium assessment, is maintained. The authors conclude that the observed tolerability provides supportive real-world safety evidence for delamanid use in patients with MDR/RR-TB, while emphasizing that careful QT interval monitoring remains essential. As drug-resistant tuberculosis continues to pose a substantial public health threat, evidence of this kind, generated in the clinics where patients are actually treated, is what ultimately transforms a promising experimental drug into a dependable public health tool.

Subject of Research: Real-world safety of delamanid-containing regimens in patients with multidrug-resistant tuberculosis in China

Article Title: Delamanid-containing regimens in patients with pulmonary multidrug-resistant tuberculosis in China: focus on the safety

Article References: Zhu, Y.-Q., Gao, M.-Q., Du, J., Liang, R.-X., Wu, G.-H., Pei, Y., Yan, X.-F., Cai, Q.-S., Huang, Y.-P., Huang, Z.-F., Wu, Y.-Q., Li, X.-J., Liu, Q.-Q., Jin, L., Wu, Q.-H., Xiong, Y., Li, M.-W., Wang, J.-Y., Kuang, H.-B., … Gao, J.-T. (2026). Delamanid-containing regimens in patients with pulmonary multidrug-resistant tuberculosis in China: focus on the safety. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14540-8

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14540-8

Keywords: delamanid, multidrug-resistant tuberculosis, MDR/RR-TB, QT prolongation, drug safety, tuberculosis treatment, China, adverse drug reactions, pharmacovigilance, XDR-TB, optimized background regimens, cardiac monitoring

The DprE1–DprE2 complex performs a deceptively simple but absolutely essential chemical transformation. It converts decaprenylphosphoribose into decaprenylphosphoarabinose, the activated sugar donor required to build arabinan, a critical polysaccharide component of the mycobacterial cell wall. The reaction is a two-step epimerization: DprE1, a flavin-dependent oxidase, oxidizes the ribose moiety of the lipid-linked substrate, and DprE2, an NADH-dependent reductase belonging to the short-chain dehydrogenase/reductase superfamily, then reduces the intermediate back to yield the arabinose epimer. Because arabinan is indispensable for the bacterium’s survival and integrity, blocking either half of this two-enzyme relay starves the cell wall of its building blocks and kills the pathogen. This vulnerability has made the epimerase one of the most intensively pursued targets in modern anti-tubercular drug discovery, yet the physical architecture of the complex and the mechanism of its inhibition by clinical compounds had remained poorly defined.

The structural work now published reveals that the functional unit of the epimerase is far more elaborate than a simple pair of enzymes. Cryo-EM analysis shows that DprE1 and DprE2 assemble into a membrane-associated tetramer in which two DprE2 molecules form a central homodimer, flanked on each side by a single DprE1 subunit. Each DprE2 monomer cradles a molecule of NADH in its catalytic core, while the neighboring DprE1 subunits carry their own FAD cofactors and bind the lipid-like decaprenylphosphoribose substrate. The interfaces holding this four-subunit assembly together were mapped in detail: the DprE2 dimer is stabilized through contacts involving its alpha-5 and alpha-7 helices and its C-terminal region, while the heterodimeric contacts between DprE1 and DprE2 involve the N-terminal loop of DprE2, residues 2 through 10, reaching into the FAD-binding and substrate-binding domains of DprE1. This arrangement positions the two active sites in close proximity, providing a structural rationale for how the oxidized intermediate can be handed efficiently from DprE1 to DprE2 without diffusing away into the membrane.

Perhaps the most consequential finding concerns pretomanid and delamanid, two nitroimidazole prodrugs that have become cornerstones of therapy for multidrug-resistant tuberculosis. Both compounds require activation by the bacterial deazaflavin F420-dependent nitroreductase Ddn, a process that generates reactive des-nitro metabolites. Earlier work had established that the activated derivatives of these drugs ultimately target DprE2, but the structural basis of that interaction was unknown. The new structures show that both activated pretomanid and activated delamanid bind to DprE2 in the form of covalent NADH adducts. Mass spectrometric analysis confirmed the formation of these adducts, with the drugs attached to the nicotinamide ribose of NADH, producing an inhibitory species that occupies an extended, conserved binding pocket in DprE2. Strikingly, this pocket stretches all the way from the NADH-binding site at the catalytic center to the substrate-binding site, meaning the drug adduct simultaneously blocks cofactor chemistry and substrate access in one motion.

This mechanism is a striking echo of the way isoniazid, the oldest frontline tuberculosis drug, kills mycobacteria. Isoniazid is also a prodrug, activated by the catalase-peroxidase KatG, and its activated form forms a covalent NAD adduct that potently inhibits the enoyl reductase InhA, halting mycolic acid synthesis. Pretomanid and delamanid, it now appears, deploy essentially the same chemical strategy against a different cell wall pathway: they hijack the cell’s own NADH to forge a tight-binding inhibitor inside the DprE2 active site. The structural overlays show that the drug portion of each adduct sits in a hydrophobic channel adapted to accept the long lipophilic tail of the decaprenyl substrate, with hydrogen-bonding interactions anchoring the drug in place. The catalytic triad of DprE2, comprising serine 147, tyrosine 160 and lysine 164, is highly conserved across the SDR superfamily, and the structures explain how the adduct wedges into the space adjacent to this catalytic machinery.

The third drug interrogated in the study, quabodepistat, acts on the other half of the complex. Unlike the nitroimidazoles, quabodepistat is a noncovalent inhibitor of DprE1, the oxidase component, and it is currently being evaluated in phase 2b/c clinical trials in combination with delamanid and bedaquiline. The cryo-EM structure of the quabodepistat-bound complex reveals a distinctive binding mode. The inhibitor occupies the substrate-binding site of DprE1, with its head, trunk and tail regions engaging separate polar and nonpolar pockets, and it induces a conformational change in two surface loops of DprE1 that adopt an open, ordered conformation not seen in structures with other noncovalent DprE1 inhibitors. Differential scanning fluorimetry experiments showed that wild-type DprE1 binds quabodepistat completely, whereas a catalytically relevant C387S mutant shows only partial binding, underscoring the specificity of the interaction and distinguishing quabodepistat’s binding mode from that of covalent inhibitors such as the benzothiazinones, which attack the same cysteine residue.

The technical achievement underlying these structures should not be understated. DprE1 and DprE2 are membrane-associated proteins, and capturing the intact tetramer in a homogeneous, grid-ready state required careful biochemical optimization, including crosslinking of the complex. The researchers collected cryo-EM datasets for four distinct states: the substrate-bound complex, the complex bound to the NADH-pretomanid adduct, the complex bound to the NADH-delamanid adduct, and the quabodepistat-bound complex. Atomic coordinates were deposited in the Protein Data Bank under accession codes 21GR, 21GQ, 43FX and 21GT, with corresponding maps in the Electron Microscopy Data Bank. The resulting models reveal not only ligand-binding details but also the electrostatic character of DprE2’s surface, including a loop region that likely helps position the enzyme at the membrane interface where its long lipid substrate resides.

The therapeutic implications are significant. By showing exactly which residues line the DprE2 drug pocket and how the NADH-adduct form of pretomanid and delamanid nestles within it, the structures give medicinal chemists a template for designing next-generation nitroimidazoles with improved potency and a reduced likelihood of resistance. Resistance to nitroimidazoles in the clinic is already known to arise through mutations in the F420 biosynthetic pathway or in the nitroreductase Ddn, which prevent drug activation; the new structures now make it possible to anticipate and monitor resistance mutations directly in DprE2 as well. Similarly, the unique binding mode of quabodepistat in DprE1, with its ordered surface loops and its three-part engagement of the substrate pocket, offers a scaffold for optimizing noncovalent DprE1 inhibitors, a class that includes the advanced clinical candidate OPC-167832 and has been the focus of intense structure-based drug design for over a decade.

More broadly, the study closes a long-standing gap in mycobacterial cell wall biology. Since the discovery in 2005 that decaprenylphosphoarabinose is formed by a two-step epimerization, the field has studied DprE1 and DprE2 largely as separate enzymes, crystallizing DprE1 with numerous inhibitors and probing DprE2 biochemically. The tetrameric complex captured here, a DprE2 dimer flanked by two DprE1 subunits, provides the first complete architectural picture of the functional epimerase and suggests how its two active sites are coordinated in a single membrane-associated assembly. As tuberculosis continues to claim more than a million lives each year and drug resistance spreads, structures of this quality do more than satisfy curiosity: they convert a validated but opaque drug target into an engineerable one, and they explain at atomic resolution why some of the newest weapons against the world’s deadliest bacterial infection work the way they do. The path from these electron density maps to better tuberculosis regimens now runs directly through rational design, guided by the pocket architecture the researchers have laid bare.

Subject of Research: Structural and mechanistic analysis of the Mycobacterium tuberculosis DprE1–DprE2 epimerase complex and its inhibition by anti-tubercular drugs

Article Title: Mechanism of the M. tuberculosis DprE1–DprE2 epimerase complex and inhibition by anti-tubercular drugs

Article References: Gao, S., Wu, F., Zhang, Y., Batt, S. M., Wei, C., Wang, Z., Yang, X., Abrahams, K. A., Fütterer, K., Liu, X., Wang, P., Besra, G. S., Rao, Z., & Zhang, L. (2026). Mechanism of the M. tuberculosis DprE1–DprE2 epimerase complex and inhibition by anti-tubercular drugs. Nature Chemical Biology. https://doi.org/10.1038/s41589-026-02317-8

Image Credits: AI Generated

DOI: 10.1038/s41589-026-02317-8

Keywords: tuberculosis, Mycobacterium tuberculosis, DprE1, DprE2, cryo-EM, pretomanid, delamanid, quabodepistat, cell wall synthesis, drug resistance, structural biology, enzyme mechanism