Utility of gastric aspirates for diagnosing tuberculosis in children in a low prevalence area: predictors of positive cultures and significance of non-tuberculous mycobacteria.

Abstract

In countries with low rates of tuberculosis (TB), yields of gastric aspirates (GAs) for Mycobacterium tuberculosis (MTB) culture are low. The significance of non-tuberculous mycobacteria (NTM) isolated from GA is uncertain. We reviewed clinical, microbiologic and radiologic data for children who underwent GA between 1999 and 2011 at Sick Kids, Toronto. Radiologic features of cases were compared with those of age matched controls. 785 GAs were obtained from 285 patients of whom 20 (7%) had positive MTB cultures: in 15 patients the GA was the only positive culture for MTB. Of 15 culture-positive patients who underwent exactly 3 GAs, MTB was isolated from the first lavage in 10 (67%), only from the second in 3 (20%) and only from the third in 2 (13%). On univariate analysis, miliary disease and intrathoracic lymphadenopathy were associated with a positive GA MTB culture. On multiple conditional logistic regression analysis, adenopathy remained significant (OR 10.2 [95% CI 2.0-51.4] p =0.005). Twelve patients had NTM isolated, most commonly M. avium complex: none had evidence of invasive NTM disease during a median duration of 12 months of follow-up. Causal pathogens different from the GA NTM culture were isolated from biopsies or bronchoalveolar lavage in 3. GAs continue to be important for TB diagnosis in children. Three GAs have a yield better than 1. Those with miliary or disseminated TB and intrathoracic lymphadenopathy have highest yields. NTM isolates from GA are likely unimportant and can be clinically misleading.