Epidemiological cut-off values for a 96-well broth microdilution plate for high-throughput research antibiotic susceptibility testing of M. tuberculosis

Abstract

Drug susceptibility testing of M. tuberculosis is rooted in a binary susceptible/resistant paradigm. While there are considerable advantages in measuring the minimum inhibitory concentrations (MICs) of a panel of drugs for an isolate, it is necessary to measure the epidemiological cut-off values (ECOFF/ECVs) to permit comparison with qualitative data. Here we present ECOFF/ECVs for 13 anti-tuberculosis compounds, including bedaquiline and delamanid, derived from 20 637 clinical isolates collected by 14 laboratories based in 11 countries on five continents. Each isolate was incubated for 14 days on a dry 96-well broth microdilution plate and then read. Resistance to most of the drugs due to prior exposure is expected and the MIC distributions for many of the compounds are complex, and therefore a phenotypically wild-type population could not be defined. Since a majority of samples also underwent genetic sequencing, we defined a genotypically wild-type population and measured the MIC of the 99th percentile by direct measurement and via fitting a Gaussian using interval regression. The proposed ECOFF/ECVs were then validated by comparing with the MIC distributions of high-confidence genetic variants that confer resistance and with qualitative drug susceptibility tests obtained via the Mycobacterial Growth Indicator Tube (MGIT) system or Microscopic-Observation Drug Susceptibility (MODS) assay. These ECOFF/ECVs will inform and encourage the more widespread adoption of broth microdilution: this is a cheap culture-based method that tests the susceptibility of 12-14 antibiotics on a single 96-well plate and so could help personalise the treatment of tuberculosis.

FIGURE 1

The CRyPTIC Consortium has collected 20 637 clinical tuberculosis (TB) samples worldwide. a, b) Layout and concentrations of the anti-TB drugs on the a) UKMYC5 and b) UKMYC6 microdilution 96-well plates. All concentrations are in mg·L⁻¹; for clarity only the first and last concentrations in each doubling series are given. The two unlabelled wells in the bottom right-hand corner contain no antibiotic and are therefore positive controls. Note that all doubling dilution series are based around 1 mg·L⁻¹ with the exception of INH which is based around 0.1 mg·L⁻¹. c) Geographical distribution of CRyPTIC laboratories and samples collected. 14 laboratories from 11 countries collected data from 27 countries. Each country is coloured depending on the number of originating samples using a logarithmic scale. PAS: para-aminosalicylic acid.

FIGURE 2

Minimum inhibitory concentration (MIC) histograms for the 13 antibiotics on the UKMYC6 plate. Only MICs which have passed the quality assurance process described in the Methods are shown. ECOFFinder was used to fit a log-normal distribution to each histogram; this is drawn in blue and the resulting 99th percentile is labelled. ECOFFinder was unable to fit a log-normal to both RIF and RFB. See figure 1 for drug abbreviations. See supplementary figure S6 for the UKMYC5 histograms and the TSV file in the supplementary material for the numerical data. The histograms can be reproduced online [21].

FIGURE 3

Directly measuring the epidemiological cut-off values (ECOFF/ECVs) from the genotypically wild-type population on the UKMYC6 plate. To illustrate the sensitivity to the precise percentile used in the definition, the 95th, 97.5th and 99th percentiles are all shown. See figure 1 for drug abbreviations. The analysis and figure can be reproduced online [21].

FIGURE 4

Interval regression is able to fit a log-normal distribution to the minimum inhibitory concentration (MIC) histograms of the genotypically wild-type (gWT) isolates for all 13 drugs on the UKMYC6 plate. Data from both plate designs were considered simultaneously, hence the resulting distributions are those the algorithm considers to best describe both the UKMYC5 (supplementary figure S10) and UKMYC6 datasets. See figure 1 for drug abbreviations. See the TSV file in the supplementary material for the numerical data. The data can be reproduced online [21].

FIGURE 5

99th percentiles of the wild-type populations for the 13 drugs on the a) UKMYC6 and b) UKMYC5 plate designs as calculated by ECOFFinder, direct measurement and interval regression. The epidemiological cut-off values (ECOFF/ECVs) are drawn on each graph as a horizontal line. See figure 1 for drug abbreviations.

FIGURE 6

Minimum inhibitory concentrations (MICs) of isolates containing genetic variants known to confer resistance to different drugs tend to lie above the epidemiological cut-off value (ECOFF/ECV) on the UKMYC6 plate: a) isoniazid, b) rifampicin, c) ethambutol, d) moxifloxacin, e) levofloxacin, f) kanamycin, g) amikacin and h) ethionamide. The number of isolates lying above and below the ECOFF/ECV is annotated. The dashed line indicates the margin of a proposed “borderline” category for isoniazid, ethambutol and ethionamide. The same analysis has been repeated on the UKMYC5 dataset (supplementary figure S11) and can be reproduced online [21].

FIGURE 7

Binary (or ternary) classification derived from the minimum inhibitory concentration (MIC) using the epidemiological cut-off values (ECOFF/ECVs) and MIC-based categorisation in table 3 agrees well with Mycobacterial Growth Indicator Tube (MGIT) results for the samples for a) isoniazid (INH), b) rifampicin (RIF), c) ethambutol (EMB), d) kanamycin (KAN), e) amikacin (AMI) and f) ethionamide (ETH). S: susceptible; R: resistant. The data and figure panels can be reproduced online [21].