Mycobacteriology: Acid-Fast Staining, Culture, and Identification

Spread two to three drops of concentrated sediment thinly, air-dry, and heat-fix (15 minutes at 80 °C, or 1 to 2 hours at 65 to 70 °C on a slide warmer) before staining. Heat fixation does not reliably kill M. tuberculosis: tubercle bacilli have remained viable in flame-fixed and hot-plate-fixed smears in controlled studies, so handle unstained slides as potentially infectious and keep every smear preparation step inside a certified biological safety cabinet.^{2,32 Clinical and Laboratory Standards Institute. M48, Laboratory Detection and Identification of Mycobacteria. 2025.3 Allen BW. Survival of tubercle bacilli in heat-fixed sputum smears. 1981.} The staining reagents, including phenol-containing carbol fuchsin and auramine mixtures, inactivate bacilli. For added safety, fix unstained smears with 5% phenol-ethanol inside the cabinet, which both fixes and inactivates them.^{33 Allen BW. Survival of tubercle bacilli in heat-fixed sputum smears. 1981.}

Fluorochrome screening covers far more of the smear per unit time than carbol fuchsin and is roughly 10% more sensitive than Ziehl-Neelsen against the same specimens. The Centers for Disease Control and Prevention (CDC) accepts either fluorochrome dye. Auramine-O and auramine-rhodamine perform equivalently, and the advantage comes from the fluorochrome method whichever dye is used. The World Health Organization (WHO) recommends light-emitting-diode fluorescence microscopy to replace mercury-vapor fluorescence microscopy and as a phased-in alternative to conventional Ziehl-Neelsen microscopy. That recommendation concerns the light source.^{1,41 Centers for Disease Control and Prevention. Core Curriculum on Tuberculosis. 2021.4 World Health Organization. Fluorescent light-emitting diode (LED) microscopy for diagnosis… 2011.} Stained fluorochrome slides fade in light, so keep them covered and read them the same day.

Acid-fast bacilli (AFB) are red rods against the blue methylene-blue or green counterstain background with carbol fuchsin, and yellow-orange bacilli on a dark field with fluorochrome dyes. They are slightly curved rods, about 1 to 10 µm by 0.2 to 0.6 µm, often beaded or banded. Morphology does not identify species, but a few patterns are suggestive: serpentine cording in broth smears for M. tuberculosis complex (MTBC), driven by cord factor, trehalose 6,6'-dimycolate; long, broad, beaded, cross-barred bacilli, sometimes described as ladder-like, for M. kansasii; and short, pleomorphic, sometimes coccobacillary forms for M. avium complex (MAC). In tissue sections, macrophages packed with MAC are periodic acid-Schiff-positive and can mimic Whipple disease. An acid-fast stain distinguishes the two.^{1,51 Centers for Disease Control and Prevention. Core Curriculum on Tuberculosis. 2021.5 Martin I, et al. Mycobacterium. 2019.}

MTBC includes M. tuberculosis (the dominant human pathogen), M. bovis (cattle and other warm-blooded animals, including humans; the bacille Calmette-Guérin (BCG) vaccine strain is derived from it and can cause disseminated disease), M. africanum (West Africa), M. caprae (mainly goats and cattle), M. microti (voles and other small mammals), M. canettii, M. pinnipedii (pinnipeds), M. mungi, M. suricattae, and M. orygis, which is increasingly recognized in human disease.^{7,87 Centers for Disease Control and Prevention. Clinical and laboratory diagnosis for tuberculosis. 2025.8 Centers for Disease Control and Prevention. Reported tuberculosis in the United States, 2022. 2022.} The proposal to split the genus into names such as Mycobacteroides and Mycolicibacterium has not been adopted by CDC, the Clinical and Laboratory Standards Institute (CLSI), or clinical laboratory practice, and United States laboratories report Mycobacterium names. Transmission is mainly inhalation of 1 to 5 µm droplet nuclei generated by coughing, shouting, or singing. Surface contact does not transmit MTBC. Direct inoculation of abraded skin is an occupational risk when infected tissue is handled. M. bovis also spreads through unpasteurized milk and from infected humans back to cattle via urine when genitourinary infection is present.^{11 Centers for Disease Control and Prevention. Core Curriculum on Tuberculosis. 2021.}

Close contact with known or suspected active tuberculosis, birth in or travel to a high-prevalence country, congregate settings, health-care work with at-risk patients, and drug or alcohol use disorder increase the chance of acquiring infection. HIV, age under 5 years, immunosuppressive therapy including tumor necrosis factor-alpha antagonists, infection acquired in the past 2 years, prior inadequate treatment, silicosis, diabetes, chronic renal failure, certain cancers, gastrectomy or jejunoileal bypass, weight under 90% of ideal, and cigarette smoking increase the chance of progression to disease. About 5% to 10% of infected people progress to active tuberculosis over a lifetime, roughly half of them within the first 2 years after infection, and untreated HIV coinfection raises that risk sharply.^{11 Centers for Disease Control and Prevention. Core Curriculum on Tuberculosis. 2021.}

Inhaled bacilli reach alveoli and are phagocytosed by macrophages that cannot kill them, because the bacilli block phagosome-lysosome fusion and survive intracellularly. Infected macrophages migrate to tracheobronchial nodes, present antigen, and seed the lung apices and distant sites (kidney, long-bone epiphyses, vertebral bodies, meninges). Cytokine-activated macrophages then drive caseation. A residual apical nidus can reactivate years later. The primary Ghon lesion is a 1 to 2 cm subpleural focus in the lower part of an upper lobe or the upper part of a lower lobe. Together with the draining-node lesion it is the Ghon complex.^{11 Centers for Disease Control and Prevention. Core Curriculum on Tuberculosis. 2021.}

Pulmonary disease usually begins gradually, with cough lasting more than 2 weeks, fever, night sweats, weight loss, fatigue, hemoptysis, or pleuritic chest pain. Extrapulmonary disease concentrates in young children, older adults, and immunocompromised persons, especially with HIV coinfection.^{11 Centers for Disease Control and Prevention. Core Curriculum on Tuberculosis. 2021.}

Runyon grouped NTM by pigment and growth rate on solid media: group I photochromogens pigment only after light exposure; group II scotochromogens pigment in dark or light; group III nonchromogens remain colorless; group IV rapid growers form visible colonies within 7 days on subculture. The scheme is obsolete for identification. Unrelated species share a Runyon phenotype, pigmentation varies with growth conditions, and liquid culture excludes light, so Runyon grouping cannot be done from a broth system. Current guidelines require species-level identification of clinically significant NTM by molecular methods or matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry. Runyon groups remain in use as descriptive terms. M. kansasii and M. marinum are group I photochromogens. M. gordonae, M. scrofulaceum, and M. szulgai are group II scotochromogens. MAC and M. xenopi are group III nonchromogens. M. abscessus, M. chelonae, and M. fortuitum are group IV rapid growers. More than 150 NTM species are now recognized.^{99 Daley CL, et al. Treatment of nontuberculous mycobacterial pulmonary disease. 2020.}

Swabs are unacceptable for mycobacterial culture when fluid or tissue can be obtained, because they trap organisms in fibers, dry out, and provide too little material for digestion and decontamination. Specimens must never be submitted in formalin or any fixative, which kills organisms and is a rejection criterion for culture. Submit tissue fresh in a sterile container with a small amount of sterile saline. A separate formalin-fixed sample may accompany it for histology only. Blood for mycobacterial culture requires a dedicated method, a mycobacterium-specific bottle such as Myco/F Lytic or a lysis-centrifugation tube, held for up to 42 days. Sodium polyanetholesulfonate and heparin are acceptable anticoagulants. EDTA is inhibitory and unacceptable. Collect urine as a first-morning voided specimen. A 24-hour pooled collection is unacceptable because it dilutes the specimen and allows overgrowth. Stool culture for AFB is an ancillary test. When disseminated MAC is suspected, mycobacterial blood culture is the recommended method. A positive stool culture may reflect passage of organisms through the gut without disseminated disease.^{11,511 Association of Public Health Laboratories. Mycobacterium tuberculosis. 2019.5 Martin I, et al. Mycobacterium. 2019.}

CDC asks for at least three consecutive sputum specimens, each collected 8 to 24 hours apart, with at least one early-morning specimen. The target volume is 5 to 10 mL, and pooled 24-hour sputum collections are unacceptable. Collect them in an airborne-infection isolation room or other well-ventilated space. Refrigerate specimens that cannot be processed promptly and process them within 24 hours, because mycobacteria lose viability and contaminants overgrow in delayed specimens. Sterile-site specimens may be inoculated directly. Specimens from sites with normal flora must be liquefied, decontaminated, and concentrated, or faster-growing flora will overgrow the mycobacteria. Perform all processing inside a certified Class II biological safety cabinet. MTBC specimen processing follows biosafety level 2 (BSL-2) practice. Manipulations of cultures and concentrated suspensions follow biosafety level 3 (BSL-3) practices and containment per Biosafety in Microbiological and Biomedical Laboratories (BMBL). Ventilation, respirator, and room requirements are covered in the biosafety section of the mycobacterial susceptibility lesson.^{1,21 Centers for Disease Control and Prevention. Core Curriculum on Tuberculosis. 2021.2 Clinical and Laboratory Standards Institute. M48, Laboratory Detection and Identification of Mycobacteria. 2025.}

The standard NALC-NaOH method uses an equal volume of freshly prepared digestant: equal parts 4% NaOH and 2.9% sodium citrate, with 0.5 g of NALC powder per 100 mL added immediately before use, used within 24 hours. NaOH is the decontaminant that kills faster-growing flora. NALC liquefies mucus so the digestant reaches organisms trapped in it. A shorter, gentler NaOH exposure is then sufficient. Sodium citrate binds heavy-metal ions in the specimen that would otherwise inactivate NALC. Vortex 15 to 30 seconds and let stand 15 minutes at room temperature. Do not exceed 15 minutes, because NaOH also kills some mycobacteria and longer contact kills more.^{1212 Kent PT, et al. Public Health Mycobacteriology. 1985.} Add sterile phosphate buffer pH 6.8 to about 45 to 50 mL to stop decontamination, then centrifuge at 3,000 to 3,900 × g for 15 to 20 minutes, refrigerated, in aerosol-tight sealed safety cups. Centrifuging a digest that may contain TB in unsealed cups is a recognized aerosol hazard. Decant and resuspend the sediment in 1.5 to 2 mL of sterile buffer. Never use non-sterile or deionized water, which introduces environmental mycobacteria such as M. gordonae.^{2,112 Clinical and Laboratory Standards Institute. M48, Laboratory Detection and Identification of Mycobacteria. 2025.11 Association of Public Health Laboratories. Mycobacterium tuberculosis. 2019.}

Specimens from cystic fibrosis or bronchiectasis patients overgrown with Pseudomonas receive a second decontamination with 5% oxalic acid after the NALC-NaOH step, because Pseudomonas survives standard NaOH exposure. Gastric aspirates must be processed within 4 hours or neutralized with sodium carbonate (about 100 mg per 5 to 10 mL), since gastric acid kills mycobacteria during any delay. Stool is diluted and filtered through gauze before decontamination. Urine sediments are concentrated by centrifugation at 3,000 to 3,600 × g for 30 minutes before processing.^{11,511 Association of Public Health Laboratories. Mycobacterium tuberculosis. 2019.5 Martin I, et al. Mycobacterium. 2019.}

Contamination rate is the routine quality indicator for this bench. Some contamination is expected because the decontaminant must be mild enough for mycobacteria to survive it. The usual expected rate is about 2 to 5% of inoculated solid media and 7 to 8% of liquid media containing antimicrobials. A rate below the expected range suggests over-decontamination that kills mycobacteria along with contaminants, usually with a rise in smear-positive, culture-negative specimens. A rate above the range suggests under-decontamination, delayed transport, or contaminated reagents. Investigate and act on observed rates outside the expected range.^{11,1611 Association of Public Health Laboratories. Mycobacterium tuberculosis. 2019.16 Forbes BA, et al. Practical guidance for clinical microbiology laboratories. 2018.}

Inoculate both broth and solid media. Broth detects growth sooner and in slightly higher overall yield. Solid media show colony morphology and pigmentation, provide isolated colonies for identification and susceptibility testing, and detect mixed cultures. In a large multicenter comparison, solid media detected roughly one in five positive cultures that liquid media missed. A broth-only approach lowers yield and is not endorsed for routine practice.^{11,511 Association of Public Health Laboratories. Mycobacterium tuberculosis. 2019.5 Martin I, et al. Mycobacterium. 2019.} Culture remains the reference method for laboratory confirmation of tuberculosis and is required for full susceptibility testing and genotyping even when a nucleic acid amplification test (NAAT) is positive.^{77 Centers for Disease Control and Prevention. Clinical and laboratory diagnosis for tuberculosis. 2025.}

Solid media are egg-based (Löwenstein-Jensen) and agar-based (Middlebrook 7H10, 7H11, selective 7H11). Malachite green in Löwenstein-Jensen suppresses bacteria that survive decontamination, and mycobacteria tolerate it. Middlebrook 7H11 is 7H10 with casein hydrolysate, which supports fastidious and drug-damaged strains, including isoniazid-resistant isolates, that grow poorly on plain 7H10. Selective media add antibiotics for heavily contaminated specimens. Cutaneous specimens need hemin or ferric ammonium citrate for M. haemophilum and mycobactin J for M. genavense. Incubate at 35 to 37 °C in 5 to 10% CO₂, keeping tubed media slanted with caps loosened for about the first week for gas exchange, then tightened to limit dehydration. Protect cultures from light throughout, because light degrades malachite green and induces pigment in photochromogens, which invalidates pigment-based observations. Skin, soft-tissue, joint, and bone specimens are also cultured in a second set at 30 °C (28 to 32 °C) to recover M. marinum, M. haemophilum, M. ulcerans, and some M. chelonae. These species grow preferentially or exclusively at temperatures below the routine 35 to 37 °C set, and 30 °C falls within every growth optimum listed in the tables above.^{11,511 Association of Public Health Laboratories. Mycobacterium tuberculosis. 2019.5 Martin I, et al. Mycobacterium. 2019.}

Three types of liquid system are in current United States use. The BACTEC MGIT 960 uses tubes of modified Middlebrook 7H9 broth with a fluorogenic, oxygen-quenching sensor at the bottom. Growth consumes dissolved oxygen, quenching is relieved, and fluorescence rises. VersaTREK bottles are sealed with a headspace pressure sensor that registers gas consumption and production. BacT/ALERT MP bottles carry a colorimetric CO₂ sensor at the bottle base that changes color as growth produces CO₂. The radiometric BACTEC 460TB is discontinued. All liquid systems add an enrichment supplement (OADC, supplying oleic acid, albumin, dextrose, and catalase) and an antimicrobial mixture (PANTA) to suppress contaminant regrowth. Omitting the enrichment delays or prevents MTBC growth, and omitting the antimicrobials raises contamination rates sharply.^{11,511 Association of Public Health Laboratories. Mycobacterium tuberculosis. 2019.5 Martin I, et al. Mycobacterium. 2019.}

Time to detection and the hold time before a negative report are separate values. MTBC in liquid culture is detected in about 7 to 21 days, faster from smear-positive than smear-negative specimens. On solid media it averages 3 to 6 weeks. A culture may be reported negative only after a full hold: a minimum of 42 days (6 weeks) for liquid systems and 8 weeks for solid media, with extended holds when M. ulcerans or M. genavense is suspected. Do not report a positive broth directly. Make an AFB smear of the broth, Gram stain turbid bottles as well, and subculture the broth to solid mycobacterial medium and to a blood agar purity plate. Growth on the purity plate usually represents a nonmycobacterial contaminant or a mixed culture. Rapidly growing mycobacteria also grow on blood agar, so identify the growth under the platform-specific procedure before you call it contamination. Treat a broth that signals positive with no AFB on smear as a probable instrument false signal or contaminant, and re-incubate it per the manufacturer's protocol. Mycobacterial growth is reported only after AFB are seen.^{11,511 Association of Public Health Laboratories. Mycobacterium tuberculosis. 2019.5 Martin I, et al. Mycobacterium. 2019.}

Laboratory cross-contamination of cultures is uncommon but documented, accounting for a few percent of MTBC-positive cultures in genotyping studies. Clues include a single positive culture from one of several specimens, scant growth, an unusually long time to detection, and a smear-negative specimen positive by culture after being processed in the same batch as a strongly positive specimen. Investigation follows the CDC false-positive toolkit and uses genotyping, which is also the method CDC applies nationally to every culture-positive United States TB case to detect outbreaks, laboratory cross-contamination, and unsuspected transmission.^{66 Centers for Disease Control and Prevention. False-positive tuberculosis laboratory results. 2024.}

The classical biochemical battery (catalase at 68 °C, Tween 80 hydrolysis, arylsulfatase, iron uptake, tellurite, urease) has been superseded as a primary identification method. Niacin and nitrate are still used to presumptively differentiate M. tuberculosis from other MTBC members, notably M. bovis, and from NTM while molecular confirmation is pending. Current methods are faster and more specific.^{55 Martin I, et al. Mycobacterium. 2019.}

Direct-specimen NAAT gives the fastest laboratory result when pulmonary TB is suspected. CDC recommends NAAT on at least one respiratory specimen from each patient with symptoms of pulmonary TB when the diagnosis is being considered and the result would change case management or control activities.^{77 Centers for Disease Control and Prevention. Clinical and laboratory diagnosis for tuberculosis. 2025.} Real-time PCR assays such as Xpert MTB/RIF, the cartridge cleared by the Food and Drug Administration (FDA) and used in United States laboratories, detect MTBC and, in the same run, rpoB mutations associated with rifampin resistance, usually in under 2 hours. Transcription-mediated amplification of rRNA is the other cleared format, reporting positive, equivocal, or negative. The successor Xpert MTB/RIF Ultra is WHO-recommended and adds multicopy targets (IS6110 and IS1081), a lower limit of detection, better sensitivity in smear-negative and paucibacillary disease, and a trace result category, but it is not FDA-cleared for the United States market. NAAT sensitivity is higher on smear-positive than smear-negative sputum. No format differentiates individual MTBC species. NAAT detects DNA from nonviable organisms, so it cannot be used to monitor treatment response. NAAT does not replace smear, culture, growth-based susceptibility testing, or genotyping.^{13,1413 Centers for Disease Control and Prevention. Xpert MTB/RIF assay. 2024.14 World Health Organization. The use of Xpert MTB/RIF Ultra and Xpert MTB/RIF assays for… 2017.}

Interpretation follows CDC categories. MTBC detected with rifampin resistance detected has a high probability of rifampin resistance, requires immediate notification of the clinician and the public health authority, and needs additional rapid confirmation, because silent or low-level rpoB mutations lower the positive predictive value in low-prevalence settings. MTBC detected with resistance not detected still needs growth-based first-line testing. An indeterminate rifampin result also needs growth-based testing. When a smear-positive specimen gives a negative NAAT, retest the specimen, including for inhibitors. When NAAT is negative but clinical suspicion persists, repeat testing on an additional specimen.^{7,137 Centers for Disease Control and Prevention. Clinical and laboratory diagnosis for tuberculosis. 2025.13 Centers for Disease Control and Prevention. Xpert MTB/RIF assay. 2024.}

Chemiluminescent DNA probes identify a species or complex within 1 to 2 hours once adequate growth is available. Known probe limitations moved laboratories toward MALDI-TOF and sequencing: occasional failure to identify MAC grown in liquid medium, false-positive MAC results from broth aliquots, and rare false-positive MTBC results. Hologic discontinued its AccuProbe culture-identification probes in the United States in December 2022, so isolates are now usually identified by MALDI-TOF or sequencing.^{5,185 Martin I, et al. Mycobacterium. 2019.18 Armstrong DT, et al. Diagnostic accuracy of the Abbott SD Bioline MPT64 antigen test… 2024.}

MALDI-TOF requires extraction to disrupt the mycolic-acid-rich wall. Perform the extraction on live growth inside the containment laboratory. The extraction inactivates MTBC, but each laboratory must validate inactivation under its own conditions before material is handled outside containment. A dedicated mycobacterial reference library is required because routine bacterial databases lack validated mycobacterial spectra. VITEK MS includes mycobacteria in its FDA-cleared database. The Bruker mycobacteria library has not been FDA-cleared in the United States. Most protocols use solid-medium growth. MALDI-TOF cannot resolve individual MTBC species, cannot resolve M. abscessus subspecies, and cannot reliably separate M. chimaera from M. intracellulare. For this reason, sequencing was needed to recognize the heater-cooler-associated M. chimaera outbreak isolates.^{55 Martin I, et al. Mycobacterium. 2019.}

Subspecies identification of M. abscessus is clinically important. M. abscessus subsp. abscessus and subsp. bolletii typically carry a functional erm(41) gene that confers inducible macrolide resistance, although the gene in a T28C sequevar within subsp. abscessus is nonfunctional. Subsp. massiliense usually carries a nonfunctional gene and remains macrolide-susceptible. The subspecies predicts the usual pattern. The susceptibility lesson explains why the extended-incubation read is required to detect inducible resistance. Subspecies identification requires sequencing of targets such as rpoB, hsp65, or erm(41) itself.^{99 Daley CL, et al. Treatment of nontuberculous mycobacterial pulmonary disease. 2020.}

Line-probe reverse-hybridization assays hybridize labeled PCR product to immobilized probes. WHO-endorsed MTBDRplus and MTBDRsl detect first- and second-line drug resistance from colonies, positive broth, or AFB smear-positive specimens. Genus and species line-probe panels identify NTM. These assays are used mainly outside the United States, where they are part of national testing algorithms.

Sequencing has the highest identification accuracy among these methods. 16S rRNA is the usual first target, but it cannot resolve individual MTBC members, M. abscessus subspecies, or M. chimaera from M. intracellulare, so many species need additional loci (rpoB, secA1, hsp65, gyrB, or the internal transcribed spacer). CLSI M48 is the standard for sequencing-based identification of mycobacteria. Whole-genome sequencing adds resistance markers and is the basis of CDC's national genotyping of United States TB isolates and of the WHO catalog of MTBC mutations. Phenotypic susceptibility testing is still performed alongside it because uncataloged mutations are missed.^{9,159 Daley CL, et al. Treatment of nontuberculous mycobacterial pulmonary disease. 2020.15 World Health Organization. Catalogue of mutations in Mycobacterium tuberculosis complex and… 2021.}

Overall turnaround is longer than the assay run time. A cartridge or probe run takes about an hour or two, but identification from culture requires growth first. The expected sequence is a positive smear the day of receipt, a NAAT result within about a day of specimen receipt, a positive MTBC culture signal in 1 to 3 weeks, and identification and susceptibility testing following after that, with final negative cultures incubated the full 42 days (liquid) or 8 weeks (solid).^{1,111 Centers for Disease Control and Prevention. Core Curriculum on Tuberculosis. 2021.11 Association of Public Health Laboratories. Mycobacterium tuberculosis. 2019.} Interpret every AFB smear result, positive or negative, with the specimen type and the stain and magnification used, and correlate it with the culture result when that is available.