Shandong study maps NTM susceptibility and links cavities to lower cure

A single-center retrospective study in Shandong Province, China, found that the Mycobacterium avium complex accounted for nearly three-quarters of 606 consecutive nontuberculous mycobacterial pulmonary disease isolates, while most isolates tested susceptible to four antimicrobial agents with CLSI clinical breakpoints. In a separate cohort of 127 patients who completed at least 12 months of susceptibility-informed therapy, cavitary disease was associated with substantially lower odds of cure among patients with M. avium complex pulmonary disease. The outcome estimates describe only this selected completer cohort and should not be interpreted as population-wide treatment effectiveness.

The researchers analyzed one index isolate from each of 606 patients whose samples were collected at Shandong Public Health Clinical Center between February 2021 and October 2024. Species identification combined DNA microarray hybridization with MALDI-TOF mass spectrometry, with sequencing used to resolve discrepant results. The isolate collection was dominated by the M. avium complex at 73.1%, followed by the M. abscessus complex at 12.4% and M. kansasii at 8.1%; M. fortuitum, M. chelonae and rarer species made up the remainder. The overall group was 54.8% male and had a mean age of 58.6 years, but demographic patterns differed by species. M. abscessus complex cases were predominantly female and younger on average, whereas M. kansasii disease was strongly male-predominant.

Susceptibility was measured by broth microdilution against 15 antimicrobial agents. Although the procedure followed the principle of CLSI M24, the concentration series came from a single commercial panel rather than the doubling dilutions specified by CLSI. The laboratory revised its interpretive criteria twice as successive CLSI standards were adopted, creating three periods whose categorical results could not always be compared directly. To address this problem, the investigators reclassified the four agents with CLSI clinical breakpoints for the M. avium complex—clarithromycin, amikacin, linezolid and moxifloxacin—from the individual MIC values using one consistent set of criteria.

After this reclassification, 98.9% of M. avium complex isolates were susceptible to clarithromycin, 99.8% to amikacin, 99.3% to linezolid and 96.2% to moxifloxacin. These results indicate high in vitro activity among the isolates tested; they do not establish that any particular regimen was superior or that susceptibility testing improved clinical outcomes. The most striking apparent discordance involved the two macrolides. Azithromycin received a higher category than clarithromycin in 97 isolates, but 93 of those isolates remained in clarithromycin’s lowest category. The difference was therefore attributable to the thresholds applied to the two drugs rather than evidence of separate resistance mechanisms.

Ethambutol illustrated the limits of categorical interpretation. Among M. avium complex isolates, 16.5% occupied the lowest ethambutol MIC category, 38.9% the intermediate category and 44.6% the highest, with an MIC50 of 2.5 mg/L and MIC90 of 10 mg/L. No CLSI clinical breakpoint exists for ethambutol against this species complex, so these bins describe the measured MIC distribution and do not demonstrate clinical resistance. The distribution varied significantly between years, including an unusual concentration of isolates in the highest category in 2023, but there was no monotonic trend across years. The available records did not identify a methodological or case-mix explanation for that temporary excursion.

The changing criteria had a measurable effect even though the assay concentrations did not change. Among seven of 15 agents tested against the M. avium complex, category thresholds were revised at least once. The same MIC value was assigned to different categories at different points in 18 of 41 species–agent pairs, and between-period differences were concentrated among agents without a CLSI clinical breakpoint. In contrast, MIC values themselves remained stable and could serve as the primary surveillance data. The findings therefore show why an apparent change in NTM “resistance” cannot be interpreted without the criteria used to assign each category.

For the clinical analysis, the investigators used a per-protocol cohort of adults with M. avium complex, M. abscessus complex or M. kansasii pulmonary disease who received at least 12 months of continuous susceptibility-informed therapy and had complete follow-up. Only 127 of 606 patients qualified, with 39 excluded because they had rare species and 440 excluded for other eligibility reasons. Cure required two consecutive negative sputum cultures at least four weeks apart, sustained for at least 12 months during therapy, together with stable or improved imaging. Because patients who died, stopped treatment or were lost to follow-up were absent from the analysis, the resulting cure proportions are upper estimates rather than intention-to-treat estimates.

Among the completers, cure was recorded in 50 of 85 patients with M. avium complex disease, 12 of 22 with M. abscessus complex disease and 19 of 20 with M. kansasii disease, corresponding to 58.8%, 54.5% and 95.0%, respectively. The two smaller species groups are descriptive because of their limited numbers. Microbiological recurrence after conversion occurred in nine of the 81 patients classified as cured, including four of the 12 cured patients with M. abscessus complex disease. Median culture conversion occurred after 3.7 months for M. avium complex and 3.6 months for M. abscessus complex, while median treatment lasted 16 and 18 months, respectively. The M. kansasii subgroup had the shortest median treatment duration, 14 months, and no recurrence was observed after conversion in that group.

The exploratory regression model for M. avium complex disease identified radiographic pattern as the main correlate of outcome. Cure declined from 79.4% in patients with a bronchiectasis-nodular pattern to 50.0% with mixed disease and 40.0% with cavitary disease. In a model containing radiographic pattern, regimen category and total drug count, cavitation was associated with an adjusted odds ratio of 0.146 for cure, with a 95% confidence interval of 0.040–0.538 and P=0.004. Similar estimates were obtained in a parsimonious model and a Firth penalized-likelihood model. Regimen composition was not associated with cure, but the analysis cannot establish that regimens can safely be narrowed: the model was exploratory, had only 5.8 events per variable, included terms that were collinear by construction, and lacked a concurrent comparison group receiving empirical rather than susceptibility-informed therapy.

Several additional limitations constrain the findings. M. abscessus complex subspecies and erm(41) sequence variants were not determined, although these factors can affect macrolide activity. The rrl genotype was also not measured. M. abscessus complex patients did not receive a strict intravenous induction phase, and their outcome subgroup contained only 22 selected patients. Exposure data were retrospective, adverse events were abstracted from clinical notes, and culture-conversion times were approximate because sputum collection followed routine care rather than a fixed schedule. Below-range MIC results were assigned half the lowest tested concentration for analysis, which may compress the lower tail of distributions. The study used one referral center and one commercial assay panel, limiting generalizability to laboratories with different methods and to species less prominent in this collection. Taken together, the results support numeric MIC-based NTM surveillance with explicit documentation of interpretive criteria, while leaving the clinical effectiveness of susceptibility-guided treatment selection to prospective controlled evaluation.

Subject of Research: Nontuberculous mycobacterial species distribution, antimicrobial susceptibility and clinical outcomes in Shandong, China

Article Title: Species distribution and antimicrobial susceptibility profiles of nontuberculous mycobacteria, with an exploratory 12-month completer outcome analysis: a single-center retrospective study from Shandong Province, China

Article References: Zhang, J., Ru, X., Shao, Y., Xu, L., & Fan, H. (2026). Species distribution and antimicrobial susceptibility profiles of nontuberculous mycobacteria, with an exploratory 12-month completer outcome analysis: a single-center retrospective study from Shandong Province, China. Frontiers in Microbiology, 17, Article 1917292. https://doi.org/10.3389/fmicb.2026.1917292

Image Credits: AI Generated

DOI: 10.3389/fmicb.2026.1917292

Keywords: nontuberculous mycobacteria, Mycobacterium avium complex, antimicrobial susceptibility, minimum inhibitory concentration, CLSI breakpoints, pulmonary disease, cavitary disease, treatment outcomes, China, surveillance, Species, distribution