Abstract:

Objective: To establish a method to improve the positive rate of culture for Mycobacterium tuberculosis (MTB) in gastric fluid specimens from pediatric patients with pulmonary tuberculosis, and to investigate the effect of culture medium pH on the correct interpretation of a tuberculosis antigen-colloidal gold assay and the effect of phosphate buffer solution (PBS) buffer on the pH of the NaOH-sample mix-solution. Methods: Seventy gastric juice samples received from suspected tuberculosis children were detected by self-established two-step culture method, traditional culture method and Xpert Ultra at the same time. The two-step culture method was to inoculate the gastric fluid specimens directly on the Lowenstein-Jensen (L-J) media first, and then, collect the suspected MTB colonies after about 20 to 30 days and treat them with sodium hydroxide-neutralization-centrifugation (traditional method) for MTB isolation. The difference of positive detection rate between traditional culture method and two-step method was compared. Take the Xpert Ultra test results of the sample as the bacterial load standard, the difference of positive detection rate of the two-step method in different bacterial load samples was analyzed. Meanwhile, a series of solutions with gradient pH were prepared to determine the optimal pH value for MTB antigen-colloidal gold diagnosis kit; the mixture of 3% or 4% NaOH and equal volume sample were neutralized with PBS buffer (pH=6.8) of a series of volumes, and the volume dilution ratio required when the pH value reaches 7-9 were determined. Results: The MTB positive rates were 26.5% (18/68) by the traditional method and 23.1% (15/65) by the two-step culture method, there was no statistically significant difference (χ²=0.125, P>0.05). The two-step method raised the positive detection rate from 26.5% (18/68) to 32.4% (22/68) based on the traditional method. The positive isolation rate of the two-step method in samples with bacterial load ≥low grade (44.0%, 11/25) was significantly higher than that in samples with load ≤very low grade (11.8%, 4/34) (χ²=7.896, P=0.005). The optimal pH range for the chosen tuberculosis antigen-colloidal gold diagnosis kit was 6-9. After being treated with 3% or 4% NaOH, adjustment of pH to 7-9 required at least 10- or 14-fold dilution with PBS buffer (pH=6.8) for gastric fluid simulative samples, and at least 12- or 16-fold dilution for sputum simulative samples. Conclusion: The positive detection rate of the two-step isolation culture method was not found better than the traditional method, however, it could be used as a supplement to the conventional method to improve the overall positive isolation rate of gastric fluid samples, and was especially recommended for the MTB detection in children with highly suspected but difficultly diagnosed pulmonary tuberculosis. In addition, this study investigated the optimal pH range for the use of MTB antigen-colloidal gold diagnosis kit and the optimal dilution ratio of neutralizing the NaOH-sample solution with PBS (pH=6.8), which could provide scientific basis for standardizing the isolation of MTB in China.

Key words: Child, Tuberculosis, pulmonary, Gastric juice, Diagnostic techniques and procedures