Abstract
Diagnosing non-tuberculous mycobacterial pulmonary disease (NTM-PD) in patients unable to produce sputum spontaneously requires invasive procedures to obtain valid respiratory specimens. In this retrospective study, we evaluated the results of microbiological tests performed on respiratory samples of 132 patients affected by NTM-PD. In the diagnostic workout, 98 patients performed both induced sputum (IS) and bronchoalveolar lavage (BAL) and were enrolled in our study. A total of 93 out of 98 BAL samples (95%) were culture-positive for mycobacteria, whereas only 67/153 (44%) induced sputum cultures were positive for NTM (p < 0.001). Molecular identification of NTM with real-time polymerase chain reaction (PCR) was positive in 48/64 BAL (75%) and in 47/139 (34%) IS samples (p < 0.001). Patients affected by nodular-bronchiectatic form were 65/98 (66%): BAL culture was positive in 95% of cases (62/65 BAL), while only 30/99 IS cultures were positive (30%; p < 0.001). PCR was positive in 76% of BAL samples examined (26/34) and in 26% of the IS samples (24 out of 91) (p < 0.001). Among 33 patients with a fibro-cavitary radiological pattern, 65% of IS (35/54) were culture-positive for NTM, whereas 94% of cases (31/33) had a positive culture for NTM from BAL (p = 0.002). PCR was positive in 73% of BAL samples tested (22/30) and 48% of IS samples tested (23/48) (p = 0.031). Our results confirm BAL mycobacterial culture as the gold standard for the diagnosis of pulmonary mycobacteriosis. FBS with BAL should be performed in every patient with a strong suspicion of NTM-PD, if other respiratory samples are repeatedly negative. Sputum induction is a useful technique to obtain valid respiratory samples when patients are unable to produce spontaneous sputum, especially in the outpatient setting. However, during the diagnostic workup of NTM-PD, we should not forget that PCR and mycobacterial culture of induced sputum have a lower yield than when performed on BAL, especially in the nodular-bronchiectatic form of the disease.