Determination of ethambutol MICs for Mycobacterium tuberculosis
and Mycobacterium avium isolates by resazurin microtitre assay
G. P. S. Jadaun, Chhaya Agarwal, Hirdesh Sharma, Zafar Ahmed, Prashant Upadhyay,
Jaya Faujdar, Anuj Kumar Gupta, Ram Das, Pushpa Gupta,
D. S. Chauhan, V. D. Sharma and V. M. Katoch*
Department of Microbiology and Molecular Biology, National JALMA Institute for Leprosy
and Other Mycobacterial Diseases (Indian Council of Medical Research), PO Box 1101,
Dr M. Miyazaki Marg, Tajganj, Agra 282001, India
Received 21 January 2007; returned 23 February 2007; revised 27 March 2007; accepted 28 March 2007
Objectives: To test susceptibilities of Mycobacterium tuberculosis (MTB) isolates to ethambutol by the
Lo ¨wenstein–Jensen (LJ) proportion method and resazurin microtitre assay (REMA) and to evaluate
REMA for the determination of ethambutol MICs for MTB and Mycobacterium avium isolates.
Methods: A total of 50 MTB and 20 M. avium isolates were tested to determine the MICs of ethambutol
by REMA and agar dilution method. MTB isolates were also tested by the LJ proportion method.
Results: REMA provided ethambutol susceptibility results for all the isolates within 8–9 days. For MTB
isolates, REMA showed 96.7% sensitivity, 100.0% specificity and 98.0% accuracy when LJ proportion
results were taken as ‘gold standard’. For both MTB and M. avium isolates, the MICs determined by
REMA were lower than those determined in agar medium, indicating that MIC values determined by
REMA are closer to the actual MICs for the isolates.
Conclusions: REMA can be used as a rapid and inexpensive method for mycobacterial drug suscepti-
bility testing against ethambutol. In comparison with the agar method, the MICs determined by REMA
can more accurately be correlated with achievable plasma concentrations of antimycobacterial agents.
Keywords: mycobacteria, drug susceptibility, M. tuberculosis, M. avium
Ethambutol is a narrow-spectrum antimycobacterial agent that is
used for the treatment of tuberculosis as well as infections caused
by Mycobacterium avium. Ethambutol has been shown to have
bactericidal action against Mycobacterium tuberculosis (MTB)
and M. avium.1Ethambutol is an important antimycobacterial
drug as it enhances the effect of other companion drugs including
aminoglycosides, rifamycins and quinolones. Moreover, this drug
has also been shown to significantly decrease the levels of bacter-
aemia in patients with AIDS,2showing its potential role in treat-
ment of infections caused by M. avium strains.
Availability of rapid results of mycobacterial culture and
drug susceptibility testing is a prerequisite for the design and
success of treatment regimens. Conventional methods available
for mycobacterial drug susceptibility testing are based on solid
media that take longer turnaround times to give final results.
Various rapid methods have also been developed for this
purpose, but these methods require expensive instruments and/or
tedious procedures. Palomino et al.3had proposed an assay
based on the oxidation–reduction dye resazurin, called the resa-
zurin microtitre assay (REMA), to test drug susceptibilities of
MTB isolates. This assay is simple, inexpensive and gives rapid
results in comparison with conventional methods. REMA has
been used in various studies to test susceptibilities of MTB iso-
lates to rifampicin and isoniazid. However, the experience about
its performance with other first-line antitubercular drugs, such as
ethambutol, is limited.4We have used REMA for the determi-
nation of MICs of ethambutol for MTB and M. avium isolates.
Objectives of the present study were: (i) to compare ethambutol
susceptibility of MTB isolates by the Lo ¨wenstein–Jensen (LJ)
proportion method and REMA; and (ii) to evaluate the utility of
*Corresponding author. Tel: þ91-562-2331756; Fax: þ91-562-2331755;
E-mail: email@example.com or firstname.lastname@example.org
Journal of Antimicrobial Chemotherapy
Page 1 of 4
# The Author 2007. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved.
For Permissions, please e-mail: email@example.com
Journal of Antimicrobial Chemotherapy Advance Access published May 4, 2007
by guest on May 31, 2013
MIC determination of ethambutol for MTB and M. avium iso-
lates by REMA and the agar dilution method.
Materials and methods
A total of 50 MTB and 20 M. avium isolates were analysed for their
susceptibilities to ethambutol. All the mycobacterial isolates
included in this study were from the collection of the Mycobacterial
Repository Centre of our institute. MTB isolates were obtained from
sputum samples of pulmonary tuberculosis patients collected during
the period September 2004–August 2005 from different parts of
India. M. avium isolates were from clinical (sputum) as well as
environmental (soil and water) sources collected from the Agra
region. Isolates were maintained on LJ medium and freshly subcul-
tured before being used for further microbiological investigations.
Susceptibility testing by the proportion method
MTB isolates were tested for their susceptibilities to ethambutol
(Sigma Chem. Co., USA) by the standard proportion method on LJ
medium.5Briefly, a bacterial suspension was prepared and several
10-fold dilutions (1021, 1022and 1023) of this suspension were
inoculated on drug-containing (critical ethambutol concentration
was 2 mg/L) and drug-free LJ slopes. Media bottles were incubated
at 378C. Final readings were noted after 6 weeks and an isolate
showing 1% or more growth on ethambutol-containing slopes
in comparison with drug-free slopes was considered resistant,
MIC determination by the REMA plate method
The REMA plate method was performed as described previously3
to determine the MICs of ethambutol for all the mycobacterial
isolates. Briefly, a 100 mL volume of Middlebrook 7H9 broth
(Difco, USA) was dispensed in each well of a 96-well cell
culture plate (Nunc, Denmark). Ethambutol concentrations pre-
pared directly in the medium were 1.25, 2.5, 3.75, 5.0, 6.25, 7.5,
8.75 and 10.0 mg/L. Perimeter wells of the plate were filled with
sterile water to avoid dehydration of the medium during incu-
bation. A standard bacterial suspension equivalent in turbidity to
that of a no. 1 McFarland standard was prepared and diluted 1:20
in 7H9 broth; a 100 mL inoculum was used to inoculate each well
of the plate. A growth control containing no ethambutol and a
sterile control without inoculum were also included for each
isolate. Plates were sealed and incubated at 378C for 1 week.
Twenty-five microlitres of 0.02% resazurin (Sigma Chem. Co.)
solution was added to each well; plates were re-incubated for an
additional 2 days. A change in colour from blue to pink indicated
the growth of bacteria, and the MIC was read as the minimum
ethambutol concentration that prevented the colour change in resa-
zurin solution. It is important to note that we have used the final
ethambutol concentration of each well (by taking half values of
the ethambutol concentrations prepared initially in the wells of
the microtitre plates) during interpretation of the results.
MIC determination by the agar dilution method
MTB and M. avium isolates were tested on Middlebrook 7H10 and
7H11 agar media (Difco), respectively, to determine their MICs of
ethambutol as per the described procedure.1Ethambutol concen-
trations prepared in agar media were 0, 1.0, 5.0, 7.5, 10.0, 15.0, 20.0
and 30.0 mg/L. A standard bacterial suspension equivalent in turbid-
ity to that of a no. 1 McFarland standard was prepared and diluted
1:100. One hundred microlitres of this diluted suspension was used
to inoculate agar media. Final readings were taken after 3 weeks of
incubation at 378C. MIC was defined as the minimum ethambutol
concentration resulting in no visible growth on agar medium.
Costs (shipping, custom taxes and labour not included) involved in
the tests were calculated by reviewing purchase records or catalogue
prices from commercial suppliers.
Results and discussion
Among the 50 MTB isolates, 19 were ethambutol-susceptible
and 31 were ethambutol-resistant by the LJ proportion method.
The MICs for MTB and M. avium isolates determined by
REMA are presented in Table 1. Ethambutol-resistant and
ethambutol-susceptible MTB isolates showed a clear distinction
between their MICs in REMA. Ethambutol MICs for MTB iso-
lates determined to be resistant by the proportion method were
at least 3.125 mg/L (except for one isolate with an MIC value of
2.5 mg/L). Ethambutol MICs of 2.5 mg/L or lower were found
in REMA for isolates determined to be susceptible by the pro-
portion method. On the basis of these results, the tentative
breakpoint concentration for ethambutol resistance was defined
as3.125 mg/L inREMA.Attheproposedbreakpoint
Table 1. Results of ethambutol susceptibility testing for MTB and M. avium isolates by REMA
Cumulative percentage of isolates inhibited at MIC (mg/L) of
aResults are based on the LJ proportion method.
Jadaun et al.
Page 2 of 4
by guest on May 31, 2013
concentration, out of 31 ethambutol-resistant isolates, only one
MTB isolate (3.2%) was misclassified as susceptible in REMA.
The proportion of resistant population on LJ medium showed
that the isolate showing the discordant result was a case of bor-
derline resistance as only 1.04% population of this isolate was
resistant to ethambutol. Taking LJ results as the ‘gold standard’,
accuracy, sensitivity and specificity of REMA were calculated to
be 98.0%, 96.7% and 100.0%, respectively. Similarly, predictive
values of susceptibility and resistance for REMA were 95.0%
and 100.0%, respectively.
Previously, Montoro et al.4had suggested a 4 mg/L etham-
butol concentration (initially prepared in the wells of microti-
tre plates) as the cut-off in REMA; however, at this cut-off,
the specificity of the assay was very low. In our study, at a
3.125 mg/L cut-off for ethambutol (equivalent to 6.25 mg/L
ethambutol prepared initially in the microtitre well), we
obtained 100% specificity without compromising sensitivity.
Early determination of ethambutol MICs for MTB isolates is
valuable as a large proportion of high-level ethambutol-
resistant MTB isolates have also been reported to be isoniazid
resistant.7This observation needs to be studied in different
settings. Our results also implicate that, by testing narrow
differences in drug concentrations, exact cut-off levels for anti-
tubercular drugs can be determined. This approach can be
applied in future studies for testing susceptibilities to other
antitubercular drugs, such as streptomycin, whose results have
been shown to display low reliability.4
In this study, REMA results were available for all the isolates
after 2 days of addition of resazurin solution. The current esti-
mated cost for ethambutol susceptibility testing per sample with
REMA is $0.576, which is about 10 times higher than the LJ
proportion method (Table 2). However, REMA can be per-
formed with minimum labour inputs when compared with the
other two methods used in this study. The cost of REMA can be
reduced further by decreasing the number of tested drug concen-
trations, such as one optimum concentration (i.e. 5 mg/L etham-
butol) used previously by Palomino and Portaels8in a simplified
microplate Alamar Blue assay (MABA). One disadvantage of
REMA is associated with biosafety since the plates use liquid
medium and could generate aerosols. However, these concerns
can be taken care off by adapting this assay to screw-cap tube
The MICs determined for M. avium isolates by REMA
were 1.25 mg/L for 15.0%, 1.875 mg/L for 5.0%, 3.75 mg/L
for 5.0%, 4.375 mg/L for 10.0% and .5 mg/L for 65.0%.
When these MICs were compared with the 3.125 mg/L break-
point concentration adopted for MTB, most of the M. avium
isolates fell into the resistant group (Table 1). However, it has
been previously reported that in contrast to MTB isolates,
M. avium isolates show broad ranges of MICs of various
drugs.9Hence, M. avium susceptibility testing based on the
breakpoint concentration adopted for MTB will have little
clinical significance. However, correlation of the ethambutol
MICs with peak plasma levels of ethambutol (2–5 mg/L)6
1.875 mg/L) could be grouped as susceptible, 15% of isolates
(MICs within the range of 3.75–4.375 mg/L) as moderately
susceptible and 65% of isolates (MICs more than 5 mg/L) as
resistant. Hence, interpretation of ethambutol MICs in compa-
rison with the achievable plasma concentration and therapeutic
response would be more useful for assessing the susceptibi-
lities of M. avium isolates.
Previously, Vanitha and Paramasivan10reported usefulness of
MABA for drug susceptibility testing of M. avium isolates
against clarithromycin. Our study showed that REMA can also
be used for drug susceptibility testing of M. avium isolates. In
contrast to Alamar Blue, resazurin is a non-proprietary com-
pound and hence REMA would be an inexpensive alternative to
MABA for susceptibility testing of M. avium isolates.
In our study, for those mycobacterial isolates having their
ethambutol MICs within the tested concentration range, the
broth-determined MICs were usually lower than the agar-
determined MICs (data not shown). The only exception was one
MTB isolate with an MIC of ethambutol of 5 mg/L by both
methods. These data are in agreement with a previous investi-
gation by Heifets et al.1on ethambutol. The differences in the
MICs are attributed to higher levels of absorption, binding and
degradation of the drug in solid media, particularly because of
the longer incubation period required for obtaining sufficient
growth in solid media.6Moreover, drug activity is more pro-
nounced in liquid medium because drug remains in direct
contact with bacterial growth. It is because of these reasons that
broth-determined MICs of antimycobacterial agents are more
accurate and closer to the true MICs and have been suggested to
be more accurately correlated with their achievable plasma
Tentative criteria for mycobacterial susceptibilities have been
proposed on the basis of the comparison of the MICs determined
radiometrically in 7H12 broth and achievable drug levels in
serum.6,9Our results showed that the REMA plate method can
also be used to establish such tentative criteria. More impor-
tantly, we suggest careful correlation between the MICs deter-
mined by REMA and the achievable plasma concentrations as
the actual MIC values must be read for such comparisons.
Quantitative measurement of degree of susceptibility can help in
management of infection as an increase in MIC during the
course of chemotherapy may be an indirect marker that a drug is
indeed affecting the bacterial population.9However, the exact
clinical relevance of MIC determination for monitoring the treat-
ment of mycobacterial diseases can only be established on the
basis of clinical trials.
In summary, REMA was able to overcome the problems of
longer turnaround times and higher MICs associated with drug
susceptibility testing on solid media by giving rapid results and
utilizing Middlebrook 7H9 broth as culture medium. REMA
could be particularly useful for proper management of mycobac-
terial infections in low resource countries as it provides more
accurate MICs at a reasonable cost.
Table 2. Comparison of cost involved in ethambutol susceptibility
testing by different methods
Ethambutol susceptibility testing by REMA
Page 3 of 4
by guest on May 31, 2013
Acknowledgements Download full-text
We are thankful to the medical and technical staff of DOTS
centres of Agra and Kanpur districts for their support during the
collection of sputum samples. Contribution of mycobacterial
strains to the Mycobacterial Repository Centre by collaborating
groups is also duly acknowledged. Financial support for the
present study was provided by Central TB Division, Government
of India and Indian Council of Medical Research. G. P. S. J. and
A. K. G. were supported by Senior Research Fellowship
grants from the Council of Scientific and Industrial Research,
New Delhi, India.
None to declare.
1. Heifets LB, Iseman MD, Lindholm-Levy PJ. Ethambutol MICs
and MBCs for Mycobacterium avium complex and Mycobacterium
tuberculosis. Antimicrob Agents Chemother 1986; 30: 927–32.
2. Kemper CA, Havlir D, Haghighat D et al. The individual microbio-
logic effect of three antimycobacterial agents, clofazimine, ethambutol,
and rifampin on Mycobacterium avium complex bacteremia in patients
with AIDS. J Infect Dis 1994; 170: 157–64.
3. Palomino JC, Martin A, Camacho M et al. Resazurin microtiter
assay plate: simple and inexpensive method for detection of drug
Chemother 2002; 46: 2720–2.
4. Montoro E, Lemus D, Echemendia M et al. Comparative evalu-
ation of the nitrate reduction assay, the MTT test, and the resazurin
microtitre assay for the drug susceptibility testing of clinical isolates of
Mycobacterium tuberculosis. J Antimicrob Chemother 2005; 55:
5. Canetti G, Fox W, Khomenko A et al. Advances in techniques of
testing mycobacterial drug sensitivity, and use of sensitivity tests in
tuberculosis control programmes. Bull WHO 1969; 41: 21–43.
6. Heifets LB. Qualitative and quantitative drug susceptibility tests
in mycobacteriology. Am Rev Respir Dis 1988; 137: 1217–22.
7. Gupta P, Jadaun GPS, Das R et al. Simultaneous ethambutol
and isoniazid resistance in clinical isolates of Mycobacterium tuberculo-
sis. Indian J Med Res 2006; 123: 125–30.
8. Palomino JC, Portaels F. Simple procedure for drug suscepti-
bility testing of Mycobacterium tuberculosis using a commercial colori-
metric assay. Eur J Clin Microbiol Infect Dis 1999; 18: 380–3.
9. Heifets L. Susceptibilitytesting
complex isolates. Antimicrob Agents Chemother 1996; 40: 1759–67.
10. Vanitha JD, Paramasivan CN. Evaluation of microplate Alamar
blue assay for drug susceptibility testing of Mycobacterium avium
complex isolates. Diagn Microbiol Infect Dis 2004; 49: 179–82.
Jadaun et al.
Page 4 of 4
by guest on May 31, 2013