Molecular Mechanisms of Resistance and Treatment Efficacy of Delamanid Against Mycobacterium tuberculosis: A Systematic Review Authors: Md Mahmudul Islam, Md Zahid Hasan, Md Touki Tahamid Tusar, Md Yeamin Hossain, Md Motaher Hossain, Md Abdulla Al Jubayed, Md Jubaer‐Al‐Abedin, Sheikh Soikot, Shanzida Akther, Jahid Bhuyian, Hafizur Rahman Gazi, B M Mahmudul Hasan, Md Shofiul Azam, Md Enamul Haque, Abdullah‐Al‐Jubayer, Md Faruk Hasan, F M Ali Haydar, Md Khalekuzzaman, Md Torequl Islam, Sohel Hasan Published: 2026 ABSTRACT ABSTRACTBackground and AimsTuberculosis (TB) remains a major global health problem, and treatment progress is increasingly threatened by rising multidrug‐resistant tuberculosis (MDR‐TB). Delamanid (DLM), a nitroimidazole drug, has shown good efficacy and safety against both drug‐susceptible and drug‐resistant Mycobacterium tuberculosis (Mtb) strains. However, data on its resistance mechanisms, drug susceptibility testing (DST), clinical effectiveness, safety, and pharmacokinetics remain limited. This review aims to summarize the most recent molecular, structural, and clinical evidence related to DLM.MethodsA comprehensive literature search was performed using WHO publications and major scientific databases, including PubMed, Web of Science, Embase, Scopus, and the Cochrane Library. Studies published through 2024 and early 2025 on DLM resistance, mechanisms of action, DST, pharmacokinetics, safety, and treatment outcomes were included. Structural analyses of key proteins involved in DLM activation were carried out using crystal structures and AlphaFold models.ResultsRecent research identified multiple mutations in the F420‐dependent activation pathway, particularly in ddn, fgd1, fbiA, fbiB, fbiC, and fbiD that contribute to DLM resistance. Structural modeling demonstrated how these mutations affect protein stability and cofactor binding. Clinical studies showed that DLM‐containing regimens improve culture conversion and treatment success, especially when combined with oral agents such as bedaquiline and linezolid. Safety data indicate that DLM is generally well tolerated, with QT prolongation being the main but manageable adverse effect.ConclusionDLM is an important and effective component of MDR‐/XDR‐TB treatment. A clearer understanding of its resistance mechanisms, pharmacological properties, and clinical outcomes can support better regimen design and help prevent the development of further resistance.Keywords: delamanid, drug‐resistance, molecular mechanism, treatment efficacy, tuberculosis Abbreviations bedaquiline broth microdilution method clofazimine deazaflavin‐dependent nitroreductase deletion delamanid drug‐resistant drug‐resistant tuberculosis drug‐susceptible drug susceptibility testing European Medicines Agency enoylpyruvoyl‐2‐diphospho‐5′‐guanosine fuoroquinolone resistance frameshift insertion linezolid microplate alamar blue assay multidrug‐resistant tuberculosis mycobacteria growth indicator tube Mycobacterium tuberculosis not mentioned non‐tuberculous mycobacteria optimized background regimen pretomanid pre‐extensively drug‐resistant Research Collaboratory for Structural Bioinformatics Protein Data Bank resazurin microtiter assay rapidly growing mycobacteria rifampicin‐resistant slowly growing mycobacteria second‐line injectable resistance tuberculosis World Health Organization extensively drug‐resistant tuberculosis Introduction Tuberculosis (TB) is caused by the harmful organism Mycobacterium tuberculosis (Mtb) and is a serious global public health issue that claims over a million lives annually. Based on estimates, approximately 25% of people worldwide may have contracted TB [1]. Significantly, the global increase in TB incidence that began during the COVID‐19 epidemic has decreased and begun to stabilize. According to the World Health Organization (WHO) Global Tuberculosis Report 2024, an estimated 10.8 million people developed TB in 2023, with approximately 33% in women, 12% in children and young adolescents, and 55% in men. Although the 10.8 million cases in 2023 were greatly higher than the 10,400 thousand in 2021 and the 10,100 thousand in 2020, they were marginally higher than the 10,700 thousand in 2022. TB affects people of all age groups and is present in every country in some form. It is important to note that among the 1.25 million TB‐related deaths in 2023, approximately 0.161 million occurred in HIV‐positive individuals. This figure remains below the prepandemic estimate of 1.34 million deaths in 2019, as well as the projected 1.32, 1.42, and 1.40 million deaths for 2022, 2021, and 2020, respectively (https://www.who.int/teams/global-programme-on-tuberculosis-and-lung health/tb‐reports/global‐tuberculosis‐report‐2024). Mtb strains that are multidrug‐resistant (MDR), defined as Mtb strains resistant to isoniazid and rifampicin, pose a significant threat to the worldwide efforts to reduce TB. The WHO reported in 2023 that there were an expected 0.4 million new cases of rifampicin‐resistant (RR) and MDR TB worldwide [2]. The developmental stage of adult TB makes diagnosis and treatment especially challenging; thus, it is critical to comprehend the potential and problems associated with adult TB and with MDR/RR‐TB in particular. It is necessary to create precise and effective therapies as well as early diagnosis approaches to mitigate the impact of this disease. Treatment of MDR‐TB is a difficult task because there are few effective treatments available, and those that are can be costly, prolonged, or have side effects [3]. Globally, national TB programs reported to the WHO that, based on the cohort that commenced treatment in 2017, just 57% of those diagnosed with RR‐/MDR‐TB and 47% of those who were resistant to fluoroquinolones experienced positive treatment outcomes (https://www.who.int/publications/i/item/9789240013131). The rate of therapeutic success remains unacceptable, notwithstanding a minor increase in recent years. Significantly, research revealed that drugs used to treat MDR‐TB, like aminoglycosides, can have neurological adverse effects, hepatotoxicity, and irreversible ototoxicity. Thus, the discovery of novel anti‐TB drugs is a crucial approach for controlling the spread of TB, particularly in cases of TB that are drug‐resistant (DR). After 50 years of inaction in the search for novel TB targets and treatments, the new drug delamanid (DLM) has demonstrated encouraging efficacy in adult patients [4]. However, despite increasing global use of DLM since 2020, important gaps remain in understanding its molecular mechanisms of action and resistance, drug‐susceptibility testing parameters, pharmacokinetics, and updated clinical performance as of 2025. This gap limits optimal use of DLM within MDR‐TB regimens and underscores the need for an updated review focused specifically on its molecular basis of resistance and treatment efficacy. The drug DLM, a nitroimidazole, has been suggested as part of new treatment plans for MDR–TB. DLM has emerged as a key component of MDR‐TB therapy as well as part of the shorter regimens. DLM was authorized by the European Medicines Agency (EMA) in 2014 as a new treatment for DR‐TB. It is a strong option for combination with other recently marketed anti‐TB drugs, like bedaquiline (BDQ), due to its tolerability and mode of action, which involves inhibition of mycolic acid formation [5]. DLM works against DR and drug‐susceptible (DS) strains of Mtb, notably MDR‐Mtb strains, by mainly preventing the formation of mycolic acid [6, 7]. According to the WHO report from 2020, the clinical studies of chemotherapy regimens containing DLM have progressed to phase III, and reports have been made regarding the clinical efficacy of both DLM and DLM‐containing regimens (https://www.who.int/publications/i/item/9789240013131). A brief update of information for the development of the new drug DLM is summarized in Figure 1 [8, 9, 10, 11, 12, 13, 14, 15, 16]. A brief update on the information for the development of the new drug delamanid. An increasing global concern is drug resistance [17, 18, 19]. Bacteria evolve drug resistance as a response to the challenges presented by therapeutic drugs. All specific pathogens are still sensitive to an antibiotic when it is first introduced from a therapeutic perspective; nevertheless, bacteria eventually develop resistance to the drug. Bacteria and antibiotics share an ecological niche, and microbes evolve defense mechanisms that fight off the damaging effects of drug compounds. Drugs have four primary targets in bacterial cells: protein synthesis, nucleic acid synthesis, the cell wall, and the membrane. Principal mechanisms behind drug resistance include inactivating a drug, altering a drug target, limiting drug uptake, and increasing active drug efflux [17]. The first step of developing drug use guidelines is to comprehend the mechanisms underlying the development of resistance. Resistance to DLM has been associated with gene mutations, including ddn, fbiA, fbiB, fbiC, fbiD, and fgd1. DLM resistance could result from any alteration in one of the above genes. It is already known that resistance to DLM has been associated with mutations in ddn and fgd1 [20]. Unfortunately, little is understood about the mechanisms underlying resistance, treatment efficacies, DST, and pharmacokinetics of DLM against the Mtb strain. Considering these concerns, we summarize and discuss in detail recent updates on DLM, including drug identification, mechanism of action, resistance, clinical efficacy, drug susceptibility, and pharmacokinetic profiles that have contributed to this drug being used in clinical or therapeutic settings. Unlike earlier reviews that summarized only the basic mechanisms and clinical data available up to 2020–2022, this review incorporates the most recent findings from 2023 to early 2025, including newly identified mutations in ddn, fbiA, fbiB, fbiC, fbiD, and fgd1, along with a detailed structural analysis of how these mutations influence protein stability and F420/guanosine diphosphate (GDP) binding. Additionally, this review integrates updated evidence on DST, pharmacokinetics, safety, and real‐world treatment outcomes from newly published cohort studies in multiple countries. By combining molecular, structural, and clinical perspectives, this review provides the most comprehensive and up‐to‐date synthesis of DLM resistance and efficacy to date. Methodology The relevant information and data for our review manuscript were collected through an extensive search of WHO publications and various databases, including Web of Science, PubMed/MEDLINE, Embase, Scopus, and the Cochrane Library. We searched these sources to identify studies related to DLM against Mtb strains published throughout 2024 and early 2025. A combination of keywords, such as “TB,” “MDR‐TB,” “XDR‐TB,” “mechanism of action of DLM”, “resistance mechanisms of DLM,” “treatment efficacy of DLM,” “DST of DLM against Mtb,” “pharmacokinetics of DLM,” and “drug interactions,” was used to gather the information and data for this review manuscript. The review process adhered to the PRISMA (preferred reporting items for systematic reviews and meta‐analyses) 2020 guidelines, and the selection procedure was documented accordingly. A total of 2200 records were identified across all databases, of which 62 studies were included. The structure files of Ddn, FbiA, FbiB, FbiC, FbiD, and Fgd1 were retrieved from the RCSB Protein Data Bank and AlphaFold for structural analysis [21, 22]. The crystal structure of the Mtb Ddn core bound to F420 (PDB ID: 3R5R) was used to model additional Ddn mutations, while the AlphaFold structure AF‐P9WP15‐F1 was used for the R23P variant. The structures of Mtb Fgd1 (PDB ID: 3B4Y) and the C‐terminal region of FbiB bound to F420 (PDB ID: 4XOQ) served as templates for their respective mutants. SWISS‐MODEL was used to generate the wild‐type M. smegmatis FbiA structure using the FbiA–FO–GDP complex (PDB ID: 6UW5) as a template [23]. AlphaFold‐predicted structures of Mtb FbiC (AF‐P9WP77‐F1) and FbiD (AF‐A5WRN6‐F1) were used for modeling their variants. Free energy changes associated with point mutations were estimated using PremPS to assess effects on protein stability [24]. Protein–ligand interactions were visualized and analyzed using Discovery Studio Visualizer v4.5 and UCSF Chimera [25, 26]. Inclusion Criteria 1.Original research articles, reviews, clinical trials, case reports, or surveillance studies focused on DLM activity, mechanisms of action, resistance mechanisms, pharmacokinetics, drug interactions, or drug susceptibility testing (DST) against Mtb. Original research articles, reviews, clinical trials, case reports, or surveillance studies focused on DLM activity, mechanisms of action, resistance mechanisms, pharmacokinetics, drug interactions, or drug susceptibility testing (DST) against Mtb. 2.Studies published in English. Studies published in English. 3.Publications appearing between January 2024 and February 2025. Publications appearing between January 2024 and February 2025. Exclusion Criteria 1.Studies not specifically addressing DLM or its molecular targets. Studies not specifically addressing DLM or its molecular targets. 2.Non‐peer‐reviewed materials (e.g., conference abstracts without full text, commentaries, editorials). Non‐peer‐reviewed materials (e.g., conference abstracts without full text, commentaries, editorials). 3.Duplicate records. Duplicate records. 4.Studies lacking sufficient methodological detail or relevance to the scope of this review. Studies lacking sufficient methodological detail or relevance to the scope of this review. Results and Discussion Molecular Mechanisms of DLM DLM A new anti‐TB drug is DLM (OPC‐67683), a bicyclic nitroimidazole that exhibits strong TB action both in vivo and in vitro. Since 2014, DLM has been an important and effective anti‐TB drug that is used to treat DR and DS Mtb strains, including MDR‐TB [27]. DLM inhibits cell wall formation by blocking the biosynthesis of mycolic acid, which kills aerobically replicating Mtb. In anaerobic conditions, it causes respiratory toxicity through the emission of nitric oxide. The products of the ddn, fbiA, fbiB, fbiC, fbiD, and fgd1 genes mediate the nitro‐reduction of the drug within the mycobacterial cell, which is necessary for these actions. DLM resistance has been linked to mutations in any of these six genes, at least in vitro and in mice. However, as of right now, little is understood regarding the prevalence and significance of the various genetic variations discovered in clinical settings. Notably, DLM exhibits partial cross‐resistance with pretomanid (PMD) and shares a similar activation pathway with PMD. Structure of DLM DLM, also known as OPC‐67683, is formulated by Otsuka Pharmaceutical Co. to treat MDR‐TB and belongs to the bicyclic nitroimidazoles group with strong anti‐TB efficacy in vitro and in vivo against Mtb strains that are both DS and DR [28, 29, 30]. Over the past three decades, only two novel drugs have been invented as anti‐TB drugs, and DLM is one of them [31]. The chemical name of DLM is (2R)‐2‐Methyl‐6‐nitro‐2‐[(4‐{4‐[4(trifluoromethoxy)phenoxy]‐1‐piperidinyl} phenoxy) methyl]‐2,3‐dihydroimidazo[2,1‐b] [1, 3] oxazole. A class of molecules known as nitroimidazoles contains a five‐membered heterocycle in their structure, with a nitro group that can be found at positions 2, 4, or 5, and a pair of nitrogen atoms at positions 1 and 3 (Figure 2A). The nitroimidazole molecule, also referred to as 2‐nitroimidazole, is an analog of azomycin and a naturally occurring antibiotic derived from Streptomyces eurocidicus in the 1950s. The racemic mixture of 6‐nitro‐2,3‐dihydroimidazo[2,1‐b] oxazole acts on the Mtb complex through right‐handed agonists but not left‐handed ones. The chemical structure of delamanid (A) and how the drug is activated inside Mycobacterium tuberculosis (B). [Delamanid is a pro‐drug that requires the F420 cofactor system for activation. Fgd1 produces F420H₂, which donates electrons to Ddn. Activated delamanid then inhibits mycolic‐acid synthesis and generates reactive nitrogen species that damage the bacterial respiratory chain.]. Mechanism of Action of DLM In 2014, DLM was authorized by the EMA as a new treatment for DR‐TB [5]. Because of its mode of action (blocking the synthesis of mycolic acid) and its tolerability, it is an attractive option for combination with other recently approved anti‐TB drugs, such as BDQ [5]. By targeting various pathways in Mtb, this combination may improve results for DS‐ and DR‐TB and allow for the phase‐out of other second‐line anti‐TB drugs that are less effective and more toxic [32]. Two oral drugs that could be crucial for new, shorter regimens against DR‐TB are DLM and BDQ [33, 34]. The prodrug DLM needs to be bioactivated through the mycobacterial F420 coenzyme system, which contains the deazaflavin‐dependent nitroreductase Ddn (Rv3547), in order to have an antimycobacterial effect on Mtb. DLM has a dual bactericidal mechanism of action, that is, (1) disturbance with the synthesis of mycolic acid and (2) respiratory poisoning [35, 36]. Disturbance With the Synthesis of Mycolic Acid Mycolic acids are important and unique lipid components of the cell envelope of Mtb and related species, which are essential for growth and survival [37]. Enzymes that participate in the synthesis of mycolic acids are crucial for viability and virulence, making them new targets for drug discovery [36]. Three different mycolate species are produced in Mtb: α mycolates, which have two cis‐cyclopropane rings, and methoxy and keto series, which have a cis‐ or trans‐cyclopropane in the adjacent position along with the proximal oxygen function and an adjoining methyl branch [38]. Significantly, the synthetic process of ketomycolates and methoxymycolates is inhibited by DLM, but not that of α‐mycolates [28]. Methoxy‐ and ketomycolates are typically less abundant than α‐mycolates in Mtb and related organisms, though their precise ratios vary depending on growth conditions. Respiratory Poisoning DLM, a prodrug, needs to be metabolically activated by the mycobacterial deazaflavin (F420)‐dependent nitroreductase (Ddn) (Figure 2B). Of note, in vitro and clinical Mtb resistant isolates show mutations in the ddn gene or one of five other genes (fgd1, fbiA, fbiB, fbiC, fbiD) that are essential for prodrug activation. Like PMD, DLM is activated by the deazaflavin (F420) via the transfer of hydrides, generating unstable intermediates that ultimately result in the production of reactive nitrogen species such as nitrous acid and nitric oxide [39, 40]. Finally, the poisoning or interruption in the respiratory system is caused by reactive nitrogen species intermediates in the bicyclic nitroimidazole metabolic pathway [35, 39, 40, 41]. Of note, DLM exposure has been shown through transcriptional profiling of Mtb to possibly contribute to respiratory poisoning [35]. Recently, Hayashi et al. [42] found that DLM resistance can result from mutations in type II NADH dehydrogenase (ndh). According to this study, instead of nitric oxide, a NAD‐DLM adduct may be responsible for its anti‐TB activity. Mechanism of Resistance to DLM DLM, a relatively new anti‐TB drug that is part of Group C, is recommended for use in lengthier MDR‐TB regimens [43]. Known methods of resistance to DLM against Mtb strains are caused by mutations in the ddn, fbiA, fbiB, fbiC, fbiD, and fgd1 genes, which are connected to the F420‐dependent bioactivation pathway, according to earlier studies [43, 44, 45]. The epidemiology of DLM resistance is summarized in Table 1 [43, 45, 46, 47, 48, 49]. In a recent study by Aghababa et al. [45], all 38 MDR‐TB isolates were sensitive to BDQ, while 7 (18.5%) of them showed resistance to DLM. According to an investigation of the sequencing data, the DLM‐resistant isolates had the most changes in the ddn gene, which included 18 nonsynonymous variants and 1 indel that resulted in frameshift mutations. Another study showed that 65 (65.00%) of the spontaneous DLM‐resistant strains had mutations in the known DR genes ddn, fbiA, fbiB, fbiC, fbiD, and fgd1, either through the F420 biosynthetic route (16.44% in fbiA, 5.48% in fbiB, and 21.92% in fbiC) or DLM prodrug activation (39.73% in ddn, 16.44% in fgd1) [43]. Significantly, 35% (35/100) of DLM‐resistant Mtb progeny strains had no known mutations in the ddn, fbiA, fbiB, fbiC, fbiD, and fgd1 genes. A total of 45 mutations were identified, 38 of which had never been reported before. Among these 38 novel mutations, 3 were stop codon mutations, and 17 were frameshift mutations, both of which had a significant impact on protein structure. Nearly all of these mutations exhibited high levels of resistance to the DLM drug. The remaining 18 mutations were identified as point mutations, displaying varying levels of resistance to DLM. Similarly, Pang et al. [50] reported that 50% (2/4) of XDR‐Mtb clinical isolates had no known mutations in the six genes linked with DLM resistance. Another study from Iran demonstrated that 9/35 (25.7%) RR Mtb strains were resistant to DLM, including three RR, four MDR, and two pre‐XDR strains [51]. In contrast to the fact that this new compound (DLM) has not yet been adopted by Iranian TB patients as a treatment option, the high DLM resistance rate was astonishingly significant in the initial investigation. The R321S in the fbiA gene was the prevalent mutation found in 22.2% (2/9) DLM‐resistant strains. Notably, the Trp678Gly mutation in the fbiC gene was linked to a DLM‐susceptible clinical isolate [51]. A refractive index–based circular ring biosensor was developed to rapidly detect four types of Mtb cells, achieving high sensitivity (1500 nm/RIU) across the 2000–2050 nm wavelength range. It demonstrated excellent performance metrics, including a quality factor of 40,280, a figure of merit of 30,000, and a detection limit as low as 0.000005 for the TBC2 strain [52]. DLM and PMD share overlapping resistance mechanisms because both require activation through the F420‐dependent pathway, involving Ddn, Fgd1, and the F420‐biosynthetic genes (fbiA, fbiB, fbiC, fbiD). Mutations in any of these genes can disrupt prodrug activation and lead to cross‐resistance to both agents. However, this overlap is not complete; some ddn and fbiC mutations reduce susceptibility to DLM more than PMD, and vice versa. These differences reflect subtle variations in how each drug interacts with Ddn and F420. As shown in Table 2, we have summarized mutations in the genes fgd1, ddn, fbiA, fbiB, fbiC, and fbiD associated with resistance to DLM that have been identified by several recent studies [43, 45, 47, 49, 53, 54, 55, 56]. Subsequently, we analyzed point mutations, summarizing findings from various recent studies to investigate the impact of these mutations on protein structure in relation to DLM resis tance. Table 1Epidemiology of delamanid resistance.ReferencesCountryPublished timeNo. of Mtb isolatesNo. of DLM‐resistant isolatesMIC (µg/mL)Method for MICInvestigating genesMechanism of resistance for mutation in genesResistance rateMutations rateStudy designAghababa et al. [45]Iran202438 MDR clinical strains70.016 μg/mL7H11 Middlebrook Agar ddn, fgd1 ddn, fgd1 18.5%57.1%Observational (cross‐sectional)The CRyPTIC Consortium [46]23 different countries202315,211 Mtb clinical strains850.12 μg/mLBroth microdilution plate assayddn, fgd1, fbiA, fbiB, fbiC, fbiD, Rv3249c11 ddn, 7 fbiA, 1 fbiC 0.56%22.35%International cohortMok et al. [47]Ireland2023900 Mtb clinical isolates1140.06 mg/LBACTEC MGIT 960 system fgd1, ddn, fbiA, fbiB, fbiC, Rv2983 fgd1, ddn, fbiA, fbiB, fbiC, Rv2983114/900 (12.67%)69/114 (60.53%)Retrospective cohortLiu et al. [43]China202260 (20 DS, 20 MDR, and 20 XDR clinical strains)1000.125Resazurin microtiter assay (REMA) ddn, fgd1, fbiA, fbiB, fbiC, and fbiD ddn, fgd1, fbiA, fbiB, fbiC —65%Experimental + clinicalHe et al. [48]11 provinces in China20211603 Mtb clinical strains110.12 μg/mLUKMYC6 plates assay ddn, fbiA, fbiB, fbiC and fgd1 No mutation any gene0.7%0Retrospective surveillanceWen et al. [49]China2019220 (110 MDR‐ and 110 XDR‐TB strains)70.2 mg/LThe microplate alamar blue assay (MABA) ddn, fbiA, fbiB, fbiC and fgd1 fbiA 6.36%14.29%ObservationalAbbreviations: DLM, delamanid; DS, drug‐susceptible; MDR, Multi‐drug resistant; Mtb; Mycobacterium tuberculosis; NM, not mentioned; NTM, non‐tuberculous mycobacteria; RGM, rapidly growing mycobacteria; SGM, slowly growing mycobacteria; XDR, extensively drug. Epidemiology of delamanid resistance. Abbreviations: DLM, delamanid; DS, drug‐susceptible; MDR, Multi‐drug resistant; Mtb; Mycobacterium tuberculosis; NM, not mentioned; NTM, non‐tuberculous mycobacteria; RGM, rapidly growing mycobacteria; SGM, slowly growing mycobacteria; XDR, extensively drug. Table 2Summary of recently identified mutations in ddn, fgd1, fbiA, fbiB, and fbiC associated with resistance to DLM in Mtb clinical, in vitro, and in vivo isolates.Resistance genesNucleotide substitutionAmino acid changeDLM breakpointMIC methodS/RLevel of resistanceNo. of isolatesStrains typesReferencesddn—W20Stop0.06 mg/LBactec MGIT 960 systemR—11ClinicalMansjö et al. [53] [51]—Del G in aa 39—Broth Macrodilution assayR> 16 mg/L1In vivo mutantsRifat et al. [54]68 G > CR23P0.125 μg/mLREMA0.25 μg/mL1In vitro mutantsLiu et al. [43]Del TAC at 86‐88Y29del0.06 mg/LBactec MGIT 960 systemR> 0.24 mg/L1ClinicalMok et al. [47]Del G at 92R31fs0.06 mg/LBactec MGIT 960 systemR> 0.24 mg/L1ClinicalMok et al. [47]113delG38fs0.125 μg/mLREMAR16 μg/mL1In vitro mutantsLiu et al. [43]116_117insGbG39fs0.125 μg/mLREMAR16 μg/mL1In vitro mutantsLiu et al. [43]117_118insCT40fs0.125 μg/mLREMAR0.5‐1 μg/mL2In vitro mutantsLiu et al. [43]134delP45fs0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]ctg/cCgL49P0.12 μg/mLUKMYC6 plateR—1ClinicalZhao et al. [55]A151CT51P0.016 μg/mL7H11 Middlebrook AgarR—2ClinicalAghababa et al. [45]157 G > TG53C0.125 μg/mLREMAR0.5 μg/mL1In vitro mutantsLiu et al. [43]161delR54fs0.125 μg/mLREMAR16 μg/mL1In vitro mutantsLiu et al. [43]188 C > AP63Q0.125 μg/mLREMAR> 16 μg/mL2In vitro mutantsLiu et al. [43]ctc/cCcL64P0.12 μg/mLUKMYC6 plateR—ClinicalZhao et al. [55]195 C > GY65*0.125 μg/mLREMAR16 μg/mL2In vitro mutantsLiu et al. [43]G241AG81S0.016 μg/mL7H11 Middlebrook AgarR—4ClinicalAghababa et al. [45]gcc/AccA77T0.12 μg/mLUKMYC6 plateR—ClinicalZhao et al. [55]G229CA77P0.016 μg/mL7H11 Middlebrook AgarR—1ClinicalAghababa et al. [45]233 C > AS78Y0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]G241AG81S0.2 mg/LBMDR0.4 mg/L21ClinicalYang et al. [56]G241AG81S0.2 mg/LBMDR0.8 mg/L9ClinicalYang et al. [56]G241AG81S0.2 mg/LBMDR> 1.6 mg/L1ClinicalYang et al. [56]G242AG81A0.2 mg/LBMDR> 1.6 mg/L2ClinicalYang et al. [56]A266CY89S0.016 μg/mL7H11 Middlebrook AgarR—1ClinicalAghababa et al. [45]A272CD108A0.016 μg/mL7H11 Middlebrook AgarR—1ClinicalAghababa et al. [45]A328CT110P0.016 μg/mL7H11 Middlebrook AgarR—1ClinicalAghababa et al. [45]R112W—Broth Macrodilution assayR> 16 mg/L1In vivo mutantsRifat et al. [54]A343CT115P0.016 μg/mL7H11 Middlebrook AgarR—1ClinicalAghababa et al. [45]C348GD116E0.016 μg/mL7H11 Middlebrook AgarR—1ClinicalAghababa et al. [45]352 G > AE118KREMAR> 16 μg/mLIn vitro mutantsLiu et al. [43]A418CT140P0.016 μg/mL7H11 Middlebrook AgarR—2ClinicalAghababa et al. [45]A422CD141A0.016 μg/mL7H11 Middlebrook AgarR—1ClinicalAghababa et al. [45]434_435insCP145fs0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]447_448insCE150fs0.125 μg/mLREMAR0.5‐16 μg/mL3In vitro mutantsLiu et al. [43]C447GC149W0.016 μg/mL7H11 Middlebrook AgarR—1ClinicalAghababa et al. [45]Del‐G at 92Fs31a 0.016 μg/mL7H11 Middlebrook AgarR—1ClinicalAghababa et al. [45]fgd1—K9N—Broth Macrodilution assayR0.5 μg/mL1In vivo mutantsRifat et al. [54]209 T > GL70R0.125 μg/mLREMAR16 μg/mL4In vitro mutantsLiu et al. [43]278 T > GM93R0.125 μg/mLREMAR16 μg/mL1In vitro mutantsLiu et al. [43]603_604insGL202fs0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]851 G > CW284S0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]870delP290fs0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]911 G > TG304V0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]962 T > CL321P0.125 μg/mLREMAR0.25 μg/mLIn vitro mutantsLiu et al. [43]fbiAG8A0.06 mg/LBactec MGIT 960 systemR0.24 mg/L1ClinicalMok et al. [47]22 G > AG8S0.125 μg/mLREMAR1 μg/mL1In vitro mutantsLiu et al. [43]—Del G in aa 47—Broth Macrodilution assayR> 16 μg/mL1In vivo mutantsRifat et al. [54]—D49G—Broth Macrodilution assayR> 16 μg/mL1In vivo mutantsRifat et al. [54]193 T > CC65R0.125 μg/mLREMAR0.25 μg/mL1In vitro mutantsLiu et al. [43]309delE103fs0.125 μg/mLREMAR16 μg/mL1In vitro mutantsLiu et al. [43]—Q120P—Broth Macrodilution assayR> 16 μg/mL1In vivo mutantsRifat et al. [54]428 T > CL143P0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]596 C > AA199E0.125 μg/mLREMAR1 μg/mL1In vitro mutantsLiu et al. [43]—S219G—Broth Macrodilution assayR0.03 μg/mL1In vivo mutantsRifat et al. [54]745_809delE249fs0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]—E249K0.2 mg/LMABAR> 161ClinicalWen et al. [49]747 A > G, 748 A > TK250*0.125 μg/mLREMAR16 μg/mL1In vitro mutantsLiu et al. [43]—D286A—Broth Macrodilution assayR> 16 μg/mL1In vivo mutantsRifat et al. [54]864 G > AW288*0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]—L308P—Broth Macrodilution assayR> 16 μg/mL1In vivo mutantsRifat et al. [54]fbiB—L15P—Broth Macrodilution assayR0.125 μg/mL1In vivo mutantsRifat et al. [54]201delP67fs0.125 μg/mLREMAR4 μg/mL1In vitro mutantsLiu et al. [43]—L173P—Broth Macrodilution assayR0.125 μg/mL1In vivo mutantsRifat et al. [54]735_736insCA246fs0.125 μg/mLREMAR8 μg/mL1In vitro mutantsLiu et al. [43]991 A > TI331L0.125 μg/mLREMAR16 μg/mL1In vitro mutantsLiu et al. [43]—Del T in aa 684—Broth Macrodilution assayR0.125 μg/mL1In vivo mutantsRifat et al. [54]fbiC—‐C in aa 20—Broth Macrodilution assayR2 μg/mL1In vivo mutantsRifat et al. [54]177 C > AC59*0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]343_344insCR115fs0.125 μg/mLREMAR2‐16 μg/mL6In vitro mutantsLiu et al. [43]—G194D—Broth Macrodilution assayR1 μg/mL1In vivo mutantsRifat et al. [54]765delF255fs0.125 μg/mLREMAR8 μg/mL1In vitro mutantsLiu et al. [43]773 G > AG258D0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]1130 T > GL377R0.125 μg/mLREMAR8‐16 μg/mL2In vitro mutantsLiu et al. [43]—L377P—Broth Macrodilution assayR> 16 μg/mL1In vivo mutantsRifat et al. [54]—C562W—Broth Macrodilution assayR> 16 μg/mL1In vivo mutantsRifat et al. [54]1792 G > CE598Q0.125 μg/mLREMAR16 μg/mL1In vitro mutantsLiu et al. [43]1810G > TG604C0.125 μg/mLREMAR16 μg/mL1In vitro mutantsLiu et al. [43]2003_2007delA668fs0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]—K684T—Broth Macrodilution assayR> 16 μg/mL1In vivo mutantsRifat et al. [54]2092 C > GH698D0.125 μg/mLREMAR> 16 μg/mL1In vitro mutantsLiu et al. [43]—A827G—Broth Macrodilution assayR> 16 μg/mL1In vivo mutantsRifat et al. [54]ggt/ggCG839A0.12 μg/mLUKMYC6 plateR—1ClinicalZhao et al. [55]Abbreviations: aa, amino acid; BMD, broth microdilution method; Del, deletion; Fs, frameshift; Ins, insertion; MABA, the microplate alamar blue assay; NC, no change; R, Resistance; REMA, Resazurin microtiter assay; S, Susceptible; *, Stop codon. Summary of recently identified mutations in ddn, fgd1, fbiA, fbiB, and fbiC associated with resistance to DLM in Mtb clinical, in vitro, and in vivo isolates. Abbreviations: aa, amino acid; BMD, broth microdilution method; Del, deletion; Fs, frameshift; Ins, insertion; MABA, the microplate alamar blue assay; NC, no change; R, Resistance; REMA, Resazurin microtiter assay; S, Susceptible; *, Stop codon. Impact of Point Mutations on DLM Resistance (Structure Stability and Ligand Binding) During the DLM activation process, the Ddn protein utilizes the cofactor F420‐H2 to catalyze the reduction of nitroimidazole prodrugs, resulting in the intracellular release of reactive nitrogen species. Consequently, we conducted a structural analysis of the Ddn protein using the Ddn‐F420 complex (PDB ID: 3R5R) as a template (Figure 3A–C). As shown, residues I18, S22, R23, N25, T26, V46, P59, R60, V61, N62, P63, L64, Y65, A76, S78, K79, P86, M87, W88, and N91 are located near the binding site of F420. Notably, the mutations R23P, T51P, G53C, P63Q, L64P, and S78Y are in close proximity to the F420 binding site (Figure 3A–C). The 3D structure of the Ddn enzyme bound to its cofactor F420. Important amino acid residues surrounding the F420 binding pocket are highlighted, along with mutations known to cause delamanid resistance. These mutations often reduce F420 binding or disrupt enzyme stability (A–C). According to a thorough analysis of the interaction between Ddn and F420, the following mutations prevent receptor‐ligand binding: R23P, T51P, G53C, P63Q, L64P, and S78Y. Interestingly, compared to the wild type, these mutations (R23P, T51P, G53C, P63Q, L64P, and S78Y) have a greater impact on catalytic activity and show a less pronounced hydrogen bonding pattern (Figure 4A–G). The 3D and 2D diagrams displaying the interactions between Ddn protein and cofactor F420: (A) wild type Ddn, (B) mutant R23P, (C) mutant T51P, (D) mutant G53C, (E) mutant P63Q, (F) mutant L64P, and (G) mutant S78Y. The F420 molecule (blue color) is shown in the middle of the 3D and 2D diagrams using a stick representation. These figures compare how F420 interacts with normal Ddn and clinically important mutants. Mutations weaken hydrogen bonding and alter the structure of the binding site, reducing DLM activation. Furthermore, the mutations G53C, P63Q, and S78Y had a more significant effect on Ddn protein stability than R23P, as demonstrated by the lower DLM MIC value. The Eris method indicated that the T51P, G53C, P63Q, and L64P mutations in Ddn have a more substantial impact on protein stability than R23P and S78Y, as shown by the low ∆∆G values (Table S1). In contrast, the mutations L49P, A77T, A77P, G81S, G81D, Y89S, D108A, T110P, R112W, T115P, D116E, E118K, T140P, and D141A are located more than 5 Å from the F420 binding site (Figure 3B). Significantly, the calculated free energy (∆∆G) for L49P, A77T, A77P, G81S, G81D, Y89S, T110P, R112W, T115P, E118K, and T140P was higher, suggesting that these mutations alter drug resistance by modifying the stability of the protein structure and its interactions rather than by affecting cofactor F420 binding. The ∆∆G values for D108A, D116E, and D141A were low, and these mutations are far from the F420 binding site, indicating that they do not induce unfavorable interactions. Therefore, the impact of these mutations (D108A, D116E, and D141A) on drug resistance remains unclear, warranting further investigation to elucidate their role in DLM drug resistance. Mycobacteria contain Fgd1, which is involved in energy metabolism. F420‐dependent Fgd1 (glucose‐6‐phosphate [G6P] dehydrogenase) in Mtb uses the oxidized F420 cofactor, which is reduced to F420H2 for the activation of DLM, to catalyze the process that converts G6P to 6‐phosphogluconolactone. In this review, we have summarized mutations (K9N, L70R, M93R, G191D, W284S, G304V, and L321P) in the fgd1 gene associated with DLM resistance that have been identified by several recent studies [43, 54]. Consequently, we conducted a structural analysis of the Fgd1 protein using the Fgd1‐F420 complex (PDB ID: 3B4Y) as a template (Figure 5). As shown, residues S38, D39, H40, S73, V74, T76, T107, G108, E109, N112, G177, G178, P179, A180, V181, E230, and H260 are located near the F420 binding site. According to a study by Liu et al. [43], the mutations L70R, M93R, G191D, W284S, G304V, and L321P in Fgd1 are positioned more than 8 Å from the F420 binding site. As a result, these mutations are likely to influence drug resistance by affecting structural stability rather than direct cofactor binding (Figure 5A–C). The K9N mutation in Fgd1, located approximately 5 Å from the F420 binding site, was found to have a high calculated free energy change (∆∆G). Additionally, this mutation introduced unfavorable interactions (Table S1). Considering these findings, the K9N mutation may contribute to drug resistance by destabilizing the protein structure rather than interfering with cofactor binding. This figure shows the overall structure of Fgd1 bound to F420. Resistance‐related mutations are displayed, most of which lie outside the binding pocket and likely contribute to resistance by destabilizing the protein (A–C). It is now known that a transferase encoded by FbiA catalyzes the formation of dehydro‐F420 by converting a phosphoenolpyruvyl moiety to 7,8‐didemethyl‐8‐hydroxy‐5‐deazariboflavin (FO). The FbiA protein utilizes enoylpyruvoyl‐2‐diphospho‐5’‐guanosine and FO as substrates to facilitate the biosynthesis of the redox cofactor F420. To model the crystal structure of M. smegmatis FbiA in complex with GDP and FO, we used the available structure (PDB ID: 6UW5) as a template to predict the FbiA‐FO‐GDP complex (Figure 6A–C). As summarized in Table 2, we identified several nonsynonymous mutations—G8A, G8S, D49G, C65R, Q120P, L143P, A199E, S219G, E249K, D286A, and L308P—that have been associated with DLM resistance in Mtb, based on findings from different studies [43, 47, 49, 54]. Therefore, only mutations G8A, G8S, S219G, and L308P were situated close to the GDP substrate binding site among our detected mutations in FbiA from several studies (Figure 6A–C). FbiA interacts with GDP and FO during F420 biosynthesis. Mutations near the substrate‐binding region reduce cofactor formation and contribute to delamanid resistance (A–C). GDP, guanosine diphosphate. Mutations in G8A, G8S, S219G, and L308P caused barriers to receptor‐ligand interaction, according to our comprehensive in silico analysis of the interaction between FbiA and GDP. Remarkably, the Gly8Ala, G8S, S219G, and L308P mutations affect catalytic activity more than the wild type, and their hydrogen bonding pattern is less noticeable (Figure 7A–E). Other mutations, D49G, C65R, Q120P, L143P, A199E, Glu249Lys, and D286A, are located far away, 5 Å from the GDP binding site, and their computed ΔΔG values were extremely positive, indicating that this mutation may have an allosteric effect on reducing the stability of the FbiA‐GDP complex (Table S1). The 3D and 2D diagrams illustrate the interactions between the FbiA protein and GDP: (A) wild‐type FbiA, (B) mutant G8A, (C) mutant G8S, (D) mutant S219G, and (E) mutant L308P. This figure compares GDP binding between normal FbiA and resistant mutants. Altered bonding and structural distortions weaken GDP association and impair F420 synthesis. GDP, guanosine diphosphate. The coenzyme F420 plays a crucial role in various biological processes, including methanogenesis in archaea, prodrug activation in Mtb, and antibiotic production in streptomycetes. Its biosynthesis is catalyzed by FbiB, which utilizes glutamate and dehydro‐F420 as substrates. The crystal structure of the F420‐bound Mtb‐FbiB C‐terminal domain (PDB ID: 4XOQ) is shown in Figure S1A. As depicted in Figure S1B, residues H291, P354, D355, G356, A357, S359, Y360, T366, E369, and H370 are positioned near the F420 binding site. Several FbiB mutations—L15P, L173P, I331L, and W397R—have been reported in different studies, with I331L being the only mutation located close to the F420 binding site (Figure S1C). The AlphaFold‐predicted structure (AF‐P9WP79‐F1) suggests that L15 is part of a β‐strand at the N‐terminus of FbiB, and the L15P mutation (associated with a DLM MIC of 0.125 μg/mL) likely disrupts this β‐strand, leading to structural instability. The DLM MIC of 0.125 μg/mL is linked to the L173P mutation, which is situated in a coil area and is far from the F420 binding site. Its calculated ΔΔG values were extremely positive, indicating that it might have an allosteric impact and lessen the FbiB‐F420‐H2 complex's stability (Table S1). Notably, the I331L mutation caused undesirable bumps with the ligand F420 in addition to the residues R334 and R337 of the Mtb FbiB protein (Figure 8A,B). The 3D and 2D diagrams depict the interactions between the FbiB protein and the cofactor F420 for (A) wild‐type FbiB and (B) mutant I331L. The FbiC protein catalyzes the radical‐mediated synthesis of FO. Since no crystal structure for FbiC was available, structural analysis was performed using the AlphaFold‐predicted model (AF‐P9WP77‐F1) (Figure 9). The structure of FbiC in Mtb, along with the identified mutated residues (black), is depicted using a blue‐ribbon representation. Free energy calculations (Table S1) suggest that mutations G194D, G258D, L377R, L377P, C562W, E598Q, G604C, K684T, H698D, A827G, and G839A may result in structural instability. Therefore, we focused on analyzing the potential interactions that impact the stability of the structure. Liu et al. [49] reported that mutations G258D, L377R, and G604C introduced unfavorable bumps. The calculated ∆∆G energy for G194D, L377P, C562W, E598Q, K684T, H698D, A827G, and G839A in FbiC was high, and these mutations introduced unfavorable interactions. Taking into account that these alterations may change resistance to drugs by modifying structural stability. DST and DLM Standardized and repeatable procedures for in vitro DST should be established concurrently with the release of novel antimicrobials to halt the early development of drug resistance and the transmission of resistant strains. Schena et al. [57] reported that spontaneous resistance to DLM underscores the need for DST before its clinical use in treatment regimens. This precaution helps prevent the selection and spread of DLM‐resistant strains, a pattern observed with other anti‐TB drugs. Various DST techniques are available for DLM, with liquid media‐based methods providing results in under 10 days, around 3 weeks quicker than the proportional approach based on agar [57]. This accelerated detection enables earlier treatment adjustments and helps curb the spread of DR‐TB. Both the REMA and Mycobacteria growth indicator tube (MGIT) methods effectively and accurately differentiate between Mtb strains susceptible or resistant to DLM, using a breakpoint concentration of 0.125 mg/L. Treatment Efficacies of DLM Treating DR‐TB poses significant challenges for physicians and national TB programs due to the limited availability of effective drugs and inexperience with more recent regimens. Managing MDR‐ and XDR‐TB is particularly difficult, as treatment options are scarce, costly, and require prolonged therapy, often accompanied by severe side effects. Research has shown that drugs used for MDR‐TB, such as aminoglycosides, can cause neurological complications, hepatotoxicity, and irreversible ototoxicity [29]. The novel drug DLM (OPC‐67683), a member of the nitro‐dihydro‐imidazooxazole class, prevents mycolic acid production. It has shown potent early bactericidal activity in clinical trials and strong preclinical efficacy, both in vivo and in vitro, against DR and DS Mtb strains [58]. After decades of stagnant research on new TB treatments, the recently developed drugs BDQ and DLM have shown promising success in treating TB in adults [4]. Research indicates that adding BDQ or DLM to an optimized background regimen (OBR) improves treatment outcomes and reduces mortality risk. Additionally, antibacterial drugs like clofazimine (CFZ) and linezolid (LZD), known for their antimycobacterial properties, are being repurposed for DR‐TB treatment [59, 60]. A few additional antibacterial drugs, such as CFZ and LZD, have anti‐mycobacterial features and are being repurposed for the treatment of DR‐TB. DLM, BDQ, and PMD are the key new oral agents for MDR‐/XDR‐TB therapy. All three improve culture conversion and treatment success, with BDQ showing strong bactericidal activity and PMD contributing to high cure rates in BPaL/BPaLM regimens. DLM provides significant early bactericidal activity and enhances outcomes in longer MDR‐TB regimens. Resistance differs among the drugs: BDQ resistance is rising in some settings, while DLM and PMD share F420‐dependent activation, leading to overlapping resistance mutations. Safety profiles vary BDQ and DLM may prolong QTc, whereas PMD is generally well tolerated but depends on companion‐drug toxicity. These agents are complementary, and their selection should be guided by susceptibility testing and safety considerations. Rather than taking each drug alone with OBR, country programs should think about efficiently controlling DR‐TB by mixing new drugs with repurposed drugs [61]. After 6 months of treatment, treatment regimens combining BDQ and DLM showed adequate culture conversion with no additive or synergistic QTc‐prolongation [62, 63]. The WHO has also regrouped anti‐TB drugs and advised combining drugs from groups A and B such that a regimen contains at least four effective drugs [64]. As shown in Table 3, studies on DLM‐containing regimens conducted in various countries, including Nigeria, China, India, and Congo, have reported high treatment success rates ranging from 70% to 91% [64, 65, 66, 67]. Table 3Summary of observational research studies describing for DR‐TB/RR‐TB/Pre‐XDR‐TB patients treated with DLM‐containing individual or standardized shorter/longer treatment regimen others drug.AuthorStudy designCountryYearMean ageHIVClinical strainsNo. of TB patients receiving DLMDrugs used in the regimenTreatment durationOutcomesTreatment successFailureDeathFadeyi et al. [65]Prospective single‐armNigeria2024355% (1/20)DR‐TB/RR‐TB/Pre‐XDR‐TB20DLM, BDQ, CFZ, LZD9 months70% (14/20)025% (5/20)Chen et al. [66]Retrospective cohortChina2024370MDR/RR/pre‐XDR‐TB25WHO‐recommended and China22 weeks84% (21/25)00Padmapriyadarsini et al. [64]Prospective cohortIndia202327NRMDR‐TBFQ + /MDR‐TBSLI+165DLM, BDQ, CFZ, LZD24–36 weeks91% (139/153)1.3% (2/153)2.6% (4/153)Kashongwe et al. [67]Retrospective studyCongo202035.838.3% (1/12)MDR‐TB/pre‐XDRTB12DLM, Bdq, Km, Pas, Lzd, H, Cfz, Z, Lfx, Mpm‐Clv, Am, Ipm‐clv, Cs20 months83.3% (10/12)08.3% (1/12)Abbreviations: Am, amikacin; BDQ, bedaquiline; CFZ, clofazimine; Cs: Cycloserine; DLM, delamanid; FLQ, fluoroquinolone; H, Isoniazid; H: Isoniazid; Ipm‐clv, Imipenem‐clavulanate acid; Km, Kanamycin; Lfx, Levofloxacin; LZD, linezolid; MDR, multidrug‐resistant; MDR‐TBFQ + , MDR‐TB with additional resistance to fluoroquinolones; MDR‐TBSLI + , second‐line injectable; Mpm‐Clv, Meropenem‐clavulanate; NR, not reported; PAS, para‐amino salicylic acid; Pre‐XDR, Pre‐extensively drug‐resistant; RR, rifampin‐resistant; XDR, extensively drug‐resistant; Z, Pyrazinamide. Summary of observational research studies describing for DR‐TB/RR‐TB/Pre‐XDR‐TB patients treated with DLM‐containing individual or standardized shorter/longer treatment regimen others drug. Abbreviations: Am, amikacin; BDQ, bedaquiline; CFZ, clofazimine; Cs: Cycloserine; DLM, delamanid; FLQ, fluoroquinolone; H, Isoniazid; H: Isoniazid; Ipm‐clv, Imipenem‐clavulanate acid; Km, Kanamycin; Lfx, Levofloxacin; LZD, linezolid; MDR, multidrug‐resistant; MDR‐TBFQ + , MDR‐TB with additional resistance to fluoroquinolones; MDR‐TBSLI + , second‐line injectable; Mpm‐Clv, Meropenem‐clavulanate; NR, not reported; PAS, para‐amino salicylic acid; Pre‐XDR, Pre‐extensively drug‐resistant; RR, rifampin‐resistant; XDR, extensively drug‐resistant; Z, Pyrazinamide. According to a recent systematic review and meta‐analysis study, patients with MDR‐TB receiving DLM had high rates of culture conversion and treatment success (72.5%; 95% CI: 44.2–89.8 and p < 0.001) despite widespread resistance and few adverse effects [68]. Another significant study from India found that patients with fluoroquinolone resistance (MDR‐TBFQ+) or second‐line injectable resistance (MDR‐TBSLI+) had a 91% favorable outcome when using DLM and BDQ in combination with other drugs in a short‐term, completely oral regimen, and 69% when treated with resistance to both FQ and SLI [64]. When combined with moxifloxacin, CFZ, or pyrazinamide, the BDQ‐DLM regimen demonstrated more favorable outcomes than those observed in a South African cohort study [61], although it was comparable to the EndTB observational study [69]. Even though the majority of the patients in the cohort trial had significant bacterial loads in their sputum and substantial bilateral illness, they were still able to obtain higher culture conversion at the end of the treatment period [64]. This finding is probably due to the favorable treatment response observed when BDQ and DLM were combined with repurposed drugs. The outcomes of this study are likewise similar to the NixTB experiment, which included PMD, LZD, and BDQ in combination and had a 90% success rate [70]. A noteworthy study by Skripconoka et al. [11] shows that when DLM is used for at least 6 months combined with an OBR that is advised by the WHO, a greater percentage of patients obtain favorable outcomes and a lower death rate. Compared to 126 (55%) of 229 patients who had DLM for at least 2 months, 143 (74.5%) of the 192 patients who received it for at least 6 months experienced it. Similarly, among the subset of patients with XDR‐TB, a higher percentage observed favorable results and a decreased death rate after receiving DLM for at least 6 months. Interestingly, after using DLM for at least 6 months, all 44 XDR‐TB patients survived [11]. According to this investigation, individuals with both MDR and XDR TB may benefit from treatment with DLM for 6 months when combined with an OBR, and this approach can also lower mortality rates. Treatment regimens containing DLM have demonstrated increased efficacy in the management of XDR‐TB and MDR‐TB. However, despite relatively low rates, treatment failure and mortality remain concerning and should not be disregarded. One significant factor contributing to these adverse outcomes is inadequate adherence to treatment. For optimal therapeutic efficacy, patients must complete the full course of DLM. A higher risk of treatment failure and death is linked to nonadherence to the recommended regimen. Suboptimal drug consumption may also contribute to the development of resistance in the population. Evidence consistently supports that DLM is a potent anti‐TB agent that improves culture conversion rates, which correlates with favorable treatment outcomes. A short course, fully oral regimen containing DLM has shown efficacy in curing MDR‐TB and fluoroquinolone‐resistant MDR‐TB (MDR‐TBFQ+). Additionally, DLM has been shown to be safe for prolonged use. For MDR‐TB patients who exhibit inadequate response to standard therapies or have limited treatment options, DLM should be considered as a valuable therapeutic option. Drug Interactions The anti‐TB drug DLM, which belongs to the nitro‐dihydro‐imidazole class of compounds, prevents the Mtb cell wall from synthesizing mycolic acid. Significantly, DLM showed strong in vitro and in vivo activity against DR and DS strains of Mtb during preclinical development [28]. DLM demonstrated significant efficacy in early bactericidal studies during clinical development in patients with DS TB [58]. Compared to treatment with a control group with an OBR, treatment with DLM for 2 months dramatically increased 2‐month sputum culture conversion by roughly 50% in MDR‐TB patients [10]. Furthermore, a longer‐term observational trial found that treatment with DLM plus an improved background regimen for at least 6 months was linked to considerably decreased mortality and higher favorable treatment outcomes (74.5% vs. 55%) when compared to treatment for at least 2 months [11]. DLM received authorization in 2014 for the treatment of pulmonary multidrug‐resistant tuberculosis (MDR‐TB) in adult patients in the European Union, Japan, and the Republic of Korea based on these findings. Adults are currently advised to take 100 mg twice daily of DLM with food [71]. DLM drug is mostly metabolized by albumin to DM‐6705; cytochrome P450 (CYP)‐mediated pathways are thought to be involved in the metabolism of DM‐6705 to additional metabolites [72]. Albumin is a unique part of the major drug metabolism route, despite the fact that some drugs are partially metabolized by it. DLM did not appear to have any clinically significant drug‐drug interactions with CYP isoenzyme inducers or inhibitors, but it was not possible to rule out interactions resulting from some degree of CYP involvement. In line with the in vitro data, the drug‐drug interaction investigations involving DLM demonstrate little effect on DLM absorption because of CYP enzyme activation or inhibition [73]. The metabolism of DLM is not greatly affected by CYP3A4, according to the outcomes of investigations using CYP3A4 inducers (efavirenz and rifampin) or inhibitors (ritonavir). This conclusion is consistent with the in vitro data. Furthermore, coadministering DLM with other drugs whose metabolism is mediated by CYP450 pathways is safe, as there is little chance of clinically important CYP‐related drug‐drug interactions [73]. The studies concluded DLM is a promising way to reduce drug–drug interactions since it is primarily digested by albumin rather than CYP enzymes [74]. With antiretroviral drugs, there are no known drug–drug interactions [73]. In addition, the concomitant use of DLM does not appear to pose any new safety issues, according to a recent data assessment conducted for the WHO recommendations [75]. Safety and Tolerability of DLM The MDR/RR‐TB treatment has a terrible track record due to its long duration, high costs, high toxicity, large pill burden, little efficacy, serious adverse events (AEs), and high mortality rates [76, 77]. Even though clinicians are aware of the AEs that patients experiencing various combinations of drugs regularly suffer, they are typically poorly reported and mistakenly believed to be unavoidable. A number of drugs have been in use for many years, including injectables (amikacin, kanamycin, and capreomycin), prothionamide/ethionamide, terizidone/cycloserine, and para‐aminosalicylic acid (PAS); however, patients have rarely been given options, even though it is known that they can suffer AEs. However, the approach to treating MDR/RR‐TB has changed for patients as well as physicians due to the rising usage of BDQ, DLM, and so‐called repurposed drugs like CFZ and LZD. The incidence and frequency of clinically important AEs of special interest in MDR/RR‐TB patients on longer BDQ and/or DLM‐containing regimens were investigated in a very recent study by [78]. They discovered that AEs generally associated with injectable drugs and LZD were more prevalent than those commonly associated with BDQ and DLM. Similar findings by Huerga et al. [79] reported that using Bdq and Dlm concurrently with LZD and CFZ is safe and efficacious for patients with MDR/RR‐TB severe disease. It is a viable therapeutic option for people who are resistant to many anti‐TB drugs to use these drugs concurrently. A significant study reported that DLM was tolerated very well even in people who had used the drugs for more than 6 months. Another trial showed that DLM was well‐tolerated and had a very well‐defined safety profile [31]. A very recent study reported by Schönfeld et al. [80] found that even in patients receiving DLM for longer than 24 weeks, no additional safety issues were identified. Prolonged QT intervals were effectively treated and did not result in any cardiac complications that were clinically relevant. The results of the treatment met the WHO goal for Europe. It can be concluded that DLM has been proven to be safe, efficacious, and well‐tolerated in clinical trials involving adults with MDR/XDR‐TB who received 100 or 200 mg of the drug twice a day in addition to a standardized background therapy. Pharmacokinetics of DLM Mtb possesses the ability to tolerate anti‐TB drug resistance even when strong drugs are used in different combination therapy regimens [81]. A well‐designed pharmacological regimen can support clinical and bacteriological treatments while preventing the development and spread of resistance. The care of MDR‐TB and other chronic TB cases is clinically problematic due to limited treatment choices, which also raises concerns for public health. Similarly, inadequate administration of prescribed drugs during the intensive and continuation stages of therapy not only affects patient survival but also develops increased resistance. New drugs, such as DLM, have brought about a significant change in the treatment of TB. DLM was conditionally approved by the EMA as a first‐in‐class bicyclic nitroimidazole, based on the need to treat MDR‐TB and encouraging phase IIb study findings [30]. A prodrug called DLM is used to treat MDR/RR‐TB. It functions by activating the drug by decreasing the nitro group of the F420‐dependent nitroreductase. The nitroimidazole molecule DLM is digested by albumin to produce DM‐6705. Although this metabolic route has been identified in vitro [15, 29], there has never been a clear connection established between the levels of plasma albumin in patients and the metabolism of DLM in humans [5]. The pharmacokinetic diversity observed in clinical practice is a result of multiple factors influencing the bioavailability of various drugs, such as drug–herb interactions, drug–food interactions, and CYP3A4 interactions. A recent study with 744 patients receiving DLM found that hypoalbuminemia appeared to improve DLM clearance [82]. This could be explained by the strong protein binding (> 99.5%) of DM‐6705 and DLM [15]. Thus, hypoalbuminemia may result in a reduction in protein binding and an increase in the overall clearance of DLM [82]. Other significant aspects of DLM pharmacokinetics include the fact that food has a significant impact on absorption (a meal has an approximate two‐fold increase in bioavailability when taken with food) [29] and that CYP enzymes are only necessary for the digestion of metabolites [15, 73]. Because DLM is not very soluble in water, the body may find it difficult to absorb. However, oral bioavailability varies from 35% to 60% and rises with meal intake, especially with high‐fat foods, according to animal studies [14]. According to Liu et al. [29], DLM has little chance of interfering with antiretroviral drugs and no interactions with CYP enzymes. Moreover, DM‐6705 has a prolonged terminal half‐life of 121–322 h, whereas DLM has an elimination half‐life of 30–38 h [29]. Age, race, sex, or renal impairment do not appear to impact DLM pharmacokinetics [5]. It can be concluded that when DLM is administered with meals as compared to when it is taken without, its relative bioavailability increases. In addition, humans have an oral bioavailability of 25%–47% for DLM, which rises dramatically with a high‐fat diet. Along with this, DLM has a lengthy half‐life of 30–38 h due to its strong protein binding (99.5%). Novelty, Limitations, and Future Perspectives The current evidence on DLM offers important advancements in understanding its utility for MDR‐TB management. Notably, recent studies have elucidated the drug's molecular activation pathway via the F420‐dependent system, highlighting how mutations in ddn, fgd1, and fbiA/B/C/D directly impair prodrug activation. These mechanistic insights represent a significant step in predicting and diagnosing resistance, strengthening the rationale for integrating both phenotypic and targeted genotypic susceptibility testing into routine practice. Additionally, updated clinical data demonstrate meaningful improvements in treatment outcomes when DLM is administered for extended durations (≥ 6 months) as part of optimized combination regimens, with safety profiles remaining acceptable under appropriate QTc monitoring. This positions DLM as a valuable add‐on agent, especially for patients with limited effective drug options. Despite these strengths, several limitations persist. Much of the clinical evidence supporting DLM effectiveness stems from observational cohorts rather than randomized controlled trials, which restricts the certainty of comparative efficacy. DST testing remains incomplete because a proportion of resistant isolates do not exhibit mutations in known F420‐pathway genes, indicating additional, undiscovered mechanisms of resistance. Pharmacokinetic variability, particularly in populations with hypoalbuminemia, pediatric patients, pregnant women, and individuals with hepatic impairment, remains insufficiently characterized. Moreover, the long‐term safety of DLM in combination with other QT‐prolonging agents requires continued vigilance, given the possibility of rare but clinically significant cardiac events. Future research should focus on expanding rapid DST capacity using standardized phenotypic assays and targeted sequencing panels to improve early detection of resistance. Comparative clinical trials are needed to define the optimal role of DLM alongside newer regimen backbones, such as BDQ‐ and PMD‐based therapies. Further PK/PD studies in special populations are essential to refine dose optimization and monitoring strategies. Mechanistic investigations exploring non–F420‐dependent resistance pathways will fill current knowledge gaps and strengthen surveillance systems. Continued postmarketing safety monitoring and implementation research will also be critical to ensuring safe, effective, and programmatically feasible integration of DLM into global MDR‐TB treatment strategies. Conclusions DLM has been incorporated into treatment options for MDR‐ and XDR‐TB, offering an additional mechanism of action for managing DR Mtb. This article provides data on the effectiveness of DLM both alone and in combination with other drugs for the treatment of DR and MDR‐TB patients. For MDR‐TB patients, the addition of DLM to therapeutic regimens significantly improves outcomes, with high culture conversion rates and low mortality observed when DLM‐containing regimens are used. Additionally, this review supports and details findings on drug resistance and DST for DLM in both in vitro and clinical Mtb strains. Drug resistance substantially reduces the effectiveness of treatment, highlighting the need for phenotypic and genotypic testing before treating MDR‐TB patients with DLM alone or in combination. While DLM offers important therapeutic benefits, it should not be interpreted as a core MDR‐TB drug. WHO classifies DLM as a Group C drug, meaning it should be added only when the core Group A and B agents, such as BDQ, LZD, levofloxacin/moxifloxacin, and CFZ cannot construct a fully effective regimen. Therefore, the therapeutic impact of DLM in practice is dependent on its use within optimized combination regimens, rather than as a standalone contributor. This review summarizes the challenges surrounding the use of DLM against Mtb strains, which could assist in the development and discovery of new anti‐TB drugs and drug combinations. A thorough understanding of DST, pharmacokinetics, resistance‐associated mutations, and the therapeutic efficacy of DLM against Mtb can provide valuable insights into improving treatment outcomes, reducing mortality, preventing drug resistance, and halting the spread of TB. It will also facilitate the development of novel TB drugs and rapid molecular diagnostic methods. Author Contributions Md. Mahmudul Islam: conceptualization, investigation, methodology, software, data curation, writing – original draft, resources. Md. Zahid Hasan: investigation, writing – original draft. Md. Touki Tahamid Tusar: writing – original draft, data curation. Md. Yeamin Hossain: investigation, validation, writing – original draft. Md. Motaher Hossain: writing – original draft, visualization, validation. Md. Abdulla Al Jubayed: Writing – original draft, investigation. Md. Jubaer‐Al‐Abedin: methodology, visualization, validation, writing – original draft, investigation. Sheikh Soikot: writing – original draft, software, visualization, validation. Shanzida Akther: writing – original draft, validation, software. Jahid Bhuyian: investigation, validation, formal analysis. Hafizur Rahman Gazi: writing – original draft, visualization, methodology. B. M. Mahmudul Hasan: writing – original draft, software. Md Shofiul Azam: writing – original draft, formal analysis, validation. Md. Enamul Haque: writing – original draft, formal analysis, visualization. Abdullah‐Al‐Jubayer: writing – original draft, methodology, validation. Md. Faruk Hasan: writing – original draft, software. F. M. Ali Haydar: writing – original draft, visualization. Md. Khalekuzzaman: writing – original draft, conceptualization. Md. Torequl Islam: writing – review and editing, visualization, validation. Sohel Hasan: supervision, project administration, writing – review and editing. Funding The authors have nothing to report. Conflicts of Interest The authors declare no conflicts of interest. Transparency Statement The lead author Md. Torequl Islam, Sohel Hasan affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained. Supporting information Figure S1: Structure of FbiB of the Mtb complexed with F420 cofactors. (A) Ribbon representation of Mtb FbiB bound cofactor F420 (PDB ID: 4XOQ). The F420 (orange) is demonstrated with displaying style of stick. (B) Residues in FbiB are close to the cofactor F420 binding site. (C) A ball and stick display were used to represent the recently identified mutant residues (purple). The BIOVIA Discovery Studio Visualizer v.4.5 program was used to acquire these images. Table S1: The effect of point mutations on the proteins (Ddn, Fgd1, FbiA, FbiB, and FbiC) stability and MIC range of mutants associated with DLM resistance against Mtb. Acknowledgments We would like to thank Md. Shah Alam (Guangzhou Institutes of Biomedicine and Health, Chinese Academy of Sciences, Guangzhou, China) for his thoughtful suggestions and assistance in enhancing our manuscript. We are grateful to Md. Sakib Al Hasan, an undergraduate student of the Department of Pharmacy, Gopalganj Science and Technology University, Gopalganj, for his help during the revision phase of this manuscript. Islam M. M., Hasan M. Z., Tusar M. T. T., et al., “Molecular Mechanisms of Resistance and Treatment Efficacy of Delamanid Against Mycobacterium tuberculosis: A Systematic Review,” Health Science Reports 9 (2026): e72481, 10.1002/hsr2.72481. Md. Mahmudul Islam and Md. Zahid Hasan contributed equally to this work. Contributor Information Md. Torequl Islam, Email: dmt.islam@gstu.edu.bd. Sohel Hasan, Email: sohel_bio@ru.ac.bd. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. Data will be made available on request to the authors. References Associated Data Supplementary Materials Figure S1: Structure of FbiB of the Mtb complexed with F420 cofactors. (A) Ribbon representation of Mtb FbiB bound cofactor F420 (PDB ID: 4XOQ). The F420 (orange) is demonstrated with displaying style of stick. (B) Residues in FbiB are close to the cofactor F420 binding site. (C) A ball and stick display were used to represent the recently identified mutant residues (purple). The BIOVIA Discovery Studio Visualizer v.4.5 program was used to acquire these images. Table S1: The effect of point mutations on the proteins (Ddn, Fgd1, FbiA, FbiB, and FbiC) stability and MIC range of mutants associated with DLM resistance against Mtb. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. Data will be made available on request to the authors.