Detection of rifampicin susceptibility in mycobacterium tuberculosis by flow cytometry

Abstract

Objective To explore the feasibility of flow cytometry (FCM) in detecting rifampicin (RFP) drug sensitivity to mycobacterium tuberculosis (MTB). Methods Mycobacterium tuberculosis clinical isolates, including drug sensitive strains and RFP monoresistant strains, were treated by RFP under concentrations of 12.5 μg/ml, 25.0 μg/ml, 50.0 μg/ml, and 100.0 μg/ml for 3, 7, or 10 days, then Fluorescein diacetae (FDA) staining was performed. Propidium iodide (PI) staining was also performed in the above strains, which were treated with RFP under the concentrations of 1.56 μg/ml, 6.25 μg/ml, 25.00 μg/ml, and 100.00 μg/ml for 1, 3, or 7 days. Mean fluorescence intensity (MFI) was evaluated, and the the sensitivity index (experimental tube MFI / control tube MFI) was calculated. Base on the Receiver Operating Characteristic Curve (ROC), the diagnostic performance of FCM in detecting RFP drug sensitivity to MTB was evaluated at different drug concentrations and treatment process. The optimal conditions for identifying MTB RFP sensitivity based on FDA and PI were selected and the threshold of sensitivity index for distinguishing MTB RFP sensitivity from drug resistance was determined. The susceptibility to RFP of 95 MTB clinical isolates and 78 MTB clinical isolates were explored by FCM based on FDA and PI stains, and the results were determined and compared with those by the routine proportional method. Results The optimum condition for detecting RFP sensitivity to MTB by FDA-based FCM was under treatment of RFP at the concentration of 12.5 μg/ml for 7 days. The threshold value of sensitivity index was 0.75. The accuracy, sensitivity, and specificity of the results were 88.4%, 90.2%, and 87.0%, respectively. No statistical differences were found between the two methods (χ2=0.364, P=0.549). While for PI-based FCM, the best condition of RFP treatment was under the concentration of 1.56 μg/ml for one day. The threshold value of sensitivity index was 1.12. And the accuracy, sensitivity, and specificity were 84.6%, 85.7%, and 83.7%, respectively. Still no statistical difference was found between the two methods (χ2=0.083, P=0.774 5). Conclusion FCM based on FDA or PI can be applied in the detection of RFP sensitivity to MTB clinical isolates. Key words: Flow cytometry; Mycobacterium tuberculosis; Rifampicin sensitivity test; Fluorescein diacetate; Propidium iodide