Tuberculosis (TB) is a common infectious disease of the human respiratory system caused by Mycobacterium tuberculosis. While it has a high mortality rate, it can be treated with timely detection. Hence the rapid detection of mycobacteria is an essential component of efforts to control the spread of TB. The current gold standard for detection of TB (especially in low-resource environments) involves culture-based diagnostic techniques. These techniques are able to infer the presence of live microorganisms in a sample by detecting changes to the properties of the suspension (such as O2/CO2 levels, pH, electrical conductivity, temperature etc) brought about by bacterial metabolism. Although these methods are relatively inexpensive and reliable, they are time consuming. For slow growing mycobacteriumspecies, which have a doubling time of ~24 hours, this method of detection can take anywhere from 7-56 days. The objective of our research is to reduce the time to detect mycobacterium in a sample and thereby prevent the delay in treatment.
We achieve this using a multi frequency electrical impedance spectroscopy (EIS) to detect the presence of mycobacterium in a sample that has been treated to eliminate all organisms except mycobacteria. Under the influence of an AC field, the membrane of a living cell, which encloses a cell at a different potential than that of its surroundings (membrane potential), acts as a capacitor. Thus the “bulk capacitance” of a suspension is proportional to the number of living cells present. Consequently, any increase in the bacterial number is reflected as an increase in the bulk capacitance of the system – which we monitor over time. A “significant” change in the value of the bulk capacitance from its baseline (time t=0) value indicates the presence of growing microorganisms (presumably mycobacteria) in the sample of interest.
This basic approach (multi-frequency EIS) has been previously used by our group to perform blood cultures (wherein blood suspected of being infected with bacteria is added to growth media and the presence of live organisms is inferred) significantly faster than currently methods that rely on bacterial metabolism. In this study, we demonstrate the applicability of this technique towards detecting mycobacteria, and obtain estimates of the amount of time that could be saved in detectingmycobacteria upon using our method.
For our study, we used Mycobacterium bovis (a zootonic analog of M. tuberculosis). M. bovis was cultured in modified Middlebrook 7H9 media, and samples were prepared with three different initial loads of M. bovis: “low” (2-5 CFU/ml), “medium” (100-200 CFU/ml) and “high” (~3000 CFU/ml). Each sample (and a control) was assayed using both the established technique (BACTEC MGIT TM culture) and our method, which involves periodic (once every 12-24 hrs) sampling of the culture fluid, measurement of electrical impedance at frequencies ranging from 10KHz to 100 MHz, an analysis of EIS data to calculate the bulk capacitance of the sample, and tracking the calculated bulk capacitances over time.
We found that for control samples, the bulk capacitance of the suspensions did not change in a statistically significant manner from their baseline values. On the other hand, the bulk capacitance values different samples containing Mycobacteria increased over time and become statistically significantly different from baseline when concentrations rose to ~ 10,000 CFU/ml. Thus our times to detection (TTDs) were about 60-70 hrs (< 3 days) for “high” concentration samples, 84-108 hrs (< 5 days) for “medium” concentration samples, and about 132-156 hrs (< 7 days) for “low” concentration samples. In contrast the TTDs for the laboratory standard was ~ 6 days for “high” concentration samples, and > 40 days for “medium” and “low” concentration samples (we are still awaiting results from the University Hospital microbiology lab).
Hence our method promises to deliver results significantly faster than current automated systems used in commercial clinical microbiology laboratories.