- Pan, Sheng-Wei;
- Syed, Rehan R;
- Catanzaro, Donald G;
- Ho, Mei-Lin;
- Shu, Chin-Chung;
- Tsai, Tsung-Yeh;
- Tseng, Yen-Han;
- Feng, Jia-Yih;
- Chen, Yuh-Min;
- Su, Wei-Juin;
- Catanzaro, Antonino;
- Rodwell, Timothy C
Objectives
Human mitochondrial cell-free DNA (Mt-cfDNA) may serve as a useful biomarker for infectious processes. We investigated Mt-cfDNA dynamics in patients with pulmonary mycobacterial infections to determine if this novel biomarker could be used to differentiate disease states and severity.Methods
Patients with pulmonary tuberculosis (PTB), latent tuberculosis infection (LTBI), and nontuberculous mycobacterial-lung disease (NTM-LD) were enrolled at a tertiary care hospital in Taiwan between June 2018 and August 2021. Human Mt-cfDNA and nuclear-cfDNA (Nu-cfDNA) copy numbers were estimated by quantitative polymerase chain reaction. Variables associated with PTB and 2-month sputum culture-positivity, indicating poor treatment response, were assessed using logistic regression.Results
Among 97 patients with PTB, 64 with LTBI, and 51 with NTM-LD, Mt-cfDNA levels were higher in patients with PTB than in LTBI (p=0.001) or NTM-LD (p=0.006). In the Mycobacterium tuberculosis-infected population, Mt-cfDNA levels were highest in smear-positive PTB patients, followed by smear-negative PTB (p<0.001), and were lowest in LTBI persons (p=0.009). A Mt-cfDNA, but not Nu-cfDNA, level higher than the median helped differentiate culture-positive PTB from culture-negative PTB and LTBI (adjusted OR 2.430 [95% CI 1.139-5.186], p=0.022) and differentiate PTB from NTM-LD (adjusted OR 4.007 [1.382-12.031], p=0.011). Mt-cfDNA levels decreased after 2 months of treatment in PTB patients (p=0.010). A cutoff Mt-cfDNA level greater than 62.62 x 106 copies/μL-plasma was associated with a 10-fold risk of 2-month culture-positivity (adjusted OR 9.691 [1.046-89.813], p=0.046).Conclusion
Elevated Mt-cfDNA levels were associated with PTB disease and failed sputum conversion at 2 months in PTB patients, and decreased after treatment.