Why oral antiseptic mouth rinsing before sputum collection cannot reduce contamination rate of mycobacterial culture in Burkina-Faso

The threat of persistent bacteria and fungi contamination in tuberculosis sputum cultures

Background

Tuberculosis (TB) sputum culture contaminants make it difficult to obtain pure TB isolates.We aimed to study and identify persistent TB sputum culture contaminants post the standard laboratory pre-culture sample decontamination techniques.

Methods

This was a longitudinal study of TB sputum culture contamination for a cohort of TB patients on standard treatment at: baseline, during TB treatment and post TB treatment. Sputum samples were decontaminated with 1.5%NaOH and neutralized using 6.8 Phosphate buffer solution.Sputum was then inoculated into MGIT (mycobactrial growth indicator tube) supplemented with 0.8ml PANTA. A drop of each positive MGIT culture was sub cultured onto blood agar and incubated for 48 hours at 35 -37OC.Any growth was identified using growth characteristics and colony morphology.

Results

From October 2017 through May 2019;we collected 8645 sputum samples of which 8624(99.8%) were eligible and inoculated into MGIT where 2444(28.3%)samples were TB culture positive and 255(10.4%)were positive for contaminants: 237 none-tuberculosis bacteria, 12 fungi and 6 mixed(none-tuberculous bacteria+fungi). There was no statistically significant difference between none tuberculosis bacteria and fungi in the treatment (OR=1.4,95%CI:0.26–7.47,p=0.690) and the post treatment TB phases(OR=2.02,95%CI:0.38–10.79,p=0.411)Vs baseline.

Conclusion

None-tuberculous bacteria and fungi dominate the plethora of TB sputum culture contamination and persist beyond the standard laboratory pre-culture decontamination algorithm.

[Vancomycin-based Lowenstein-Jensen selective medium for reducing contamination of mycobacterial cultures by spore-forming bacteria]

INTRODUCTION

despite the development of new methods, culture on solid medium is the gold standard for the diagnosis of tuberculosis. However, this method is associated with increased rates of contamination of cultures by spore-forming bacteria. These bacteria are generally sensitive to vancomycin and to a combinsation of vancomycin, colistin, nystatin, and trimethoprim (VCNT). The purpose of this study was to assess the effectiveness of VCNT-based selective Lowenstein-Jensen (LJ) medium in reducing contamination of cultures by spore-forming bacteria.

METHODS

sputum samples, collected from the 120 TB and non-TB patients included in the study between October 2016 and May 2017, were decontaminated with the modified Petroff method. Decontamination pellets were inoculated onto conventional LJ media and selective VCNT-based LJ medium containing 10µg/ml vancomycin. Fifteen strains of spore-forming bacteria were inoculated onto the same media in order to assess their sensitivity to VCNT.

RESULTS

the contamination of cultures on VCNT-based LJ medium containing 10µg/ml of vancomycin and LJ medium were 11.66% (14/120) and 39.16% (47/120) with p <0.0001, respectively. Sensitivity of spore-forming bacteria to VCNT decreased with the increasing of culture incubation time.

CONCLUSION

VCNT-based selective LJ medium containing 10µg/ml vancomycin led to a significant reduction in the rate of culture contamination. This environment could contribute to improve the quality of mycobacterial cultures and thus bacteriological diagnosis of tuberculosis.

Identification of spore-forming bacteria isolated from contaminated Lowenstein Jensen media and effectiveness of Vancomycin to reduce Mycobacterial culture contamination in Burkina-Faso

The type of commensal microorganisms can influence the efficiency of sputum decontamination for TB diagnosis. A basic characterization of contaminants from LJ contaminated media showed that Gram positive Spore Forming Bacteria (SFB) were the major contaminants. This study aims to identify the species of this contaminants and to evaluate the effectiveness of VCNT at 10 µg of vancomycin to reduce mycobacterial culture contamination mainly linked to SFB. Fifty-three SFB isolated between February 2016 and May 2017 were used. The effectiveness of LJ with VCNT at 10 µg of Vancomycin were evaluated with sputum collected in the same period. SFB had been stored at -20 °C and identified after subculture onto 5% sheep blood Columbia agar and incubated at 37 °C during 24 h. Bacteria cells and isolated colonies were described. API 50CH/B was performed and MALDI-TOF MS was used for external quality control. Thirty- five (66%) isolates representing 4 genera (Bacillus, Paenibacillus, Brevisbacillus and Lysinibacillus) including 10 species were identified. The most important species were Bacillus cereus (30%) and Bacillus licheniformis (21%). Eighteen (34%) isolates were non-reactive Bacillus. The overall contamination rate on LJ with VCNT at 10 µg of vancomycin was statistically lower than which without VCNT (18.7% versus 43.8%) (p = 0.01). The most important SFB identified were B. cereus and B. licheniformis. Almost all identified strains were similar to those currently isolated in fermented traditional food suggesting in part food related contaminants. VCNT containing 10 µg of vancomycin is a good alternative method to reduce mycobacterial culture contamination.