Time to Positive Culture Detection Predicts Mycobacterium avium Pulmonary Disease Severity and Treatment Initiation.

Abstract

Rationale: Additional biomarkers are needed to guide initiationof treatment for Mycobacterium avium pulmonary disease (Mav-PD). Time to positive sputum culture detection(TTP) may offer potential prognostic and monitoring value. Objectives: To determine whether TTP is associated with infection severity and early treatment response in Mav-PD. Methods: We undertook a retrospective cohort study of patientswith two or more sputum cultures positive for M. avium, an "index" sputum M. avium isolate during 2015-2019, a computed tomographic scan within 6 months, and no treatmentfor at least 6 months before index sputum. TTP wasestimated from the date of laboratory receipt of the specimento the date of culture positivity confirmation. TTP was tested for association with markers of infection severity(Mav-PD, bronchiectasis, cavitary disease, treatment initiation by 3 and 6 months, and acid fast bacilli [AFB] smear) and treatment response using Mann-Whitney U, Spearman's correlation coefficient, and Wilcoxon signed-rank tests. We explored a threshold TTP that could identify significant M. avium disease. Results: We included 125 patients with mean (standard deviation) age 68.5 (12.5) years and 65% fulfilled disease criteria. Median TTP was 12 (interquartile range 10-15; range 6-44) days. TTP and AFB smear grade were negatively correlated (ρ -0.58; P < 0.001). TTP was associated with nontuberculous mycobacteria (NTM) disease (P = 0.03), AFB smear positivity (P < 0.001), and treatment initiation by 3 (P = 0.01) and 6 (P = 0.03) months. A threshold TTP of 10 days or less was associated with Mav-PD (80.6% vs. 58.4%; ð [95% confidence interval (CI)] 22.1% [5.6-38.6%]; P = 0.02), AFB smear positivity (83.3% vs. 20.2%, ð [95% CI] 63.1% [48.3-77.9%]; P < 0.001), treatment by 3 (38.9% vs. 13.5%; ð [95% CI] 25.4% [8.0-42.8%]; P = 0.003) and 6 (47.2% vs. 19.1%; ð [95% CI] 28.1% [9.9-46.4%]; P = 0.003) months. After 3 and 6 months of treatment, the median (interquartile range) change in TTP was 8 (1 undefined; P < 0.001) and 7 (0undefined; P = 0.001) days, respectively. Conclusions: TTP is associated with bacterial burden and infection severity and increases in response to treatment. A threshold of 10 days or less may be useful in predicting significant Mav-PD. As a readily available biomarker, further exploration of TTP is imperative.