Tubercle, Land., (1969), 50, 7
THE ETHIONAMIDE SENSJTIVITY OF EAST AFRICAN STRAINS OF MYCOBACTERIUM TUBERCULOSIS RESISTANT TO THIACETAZONE By M. J. LEFFORD from the Medical
Research Council’s Unit on Drug Sensitivity in Tuberculosis, Royal Postgraduate Medical School. London, W.12
SUMMARY The ethionamide sensitivity of cultures of M. tuberculosis isolated from East African patients has been measured in population-study sensitivity tests. The results have been expressed as MICs using large or small inocula, and by the proportion method. Pre-treatment cultures which were naturally sensitive or naturally resistant to thiacetazone did not differ appreciably in their sensitivity to ethionamide. Pairs of cultures obtained before and after six months treatment with isoniazid plus thiacetazone were found to differ significantly in their sensitivity to ethionamide. Although none of the pre-treatment cultures were resistant to ethionamide, six of 21 post-treatment cultures were definitely resistant. It is concluded that natural resistance to thiacetazone is not accompanied by resistance to ethionamide; whereas among cultures with acquired resistance to thiacetazone approximately 30 ‘A are also resistant to ethionamide.
La sensibilite a l’bthionamide de cultures de M. tuberculosis isolees en Afrique Orientale a Ct.5mesuree par l’etude de la sensibilitt des populations. Les resultats ont Ctt exprimes en M I C sur des inoculums faibles et forts, et par la mtthode des proportions. Les cultures isolees avant traitement et qui Ctaient naturellement sensibles ou rtsistantes a la thiacetazone ne differaient pas de facon appreciable quant a leur sensibilitt a l’tthionamide. Des cultures appariees obtenues avant et aprb six mois de traitement par isoniazide plus thiaccttazone ont montrt une difference significative de leur sensibilite a l‘ethionamide. Bien qu’aucune des cultures avant traitement n’etaientresistantes a l’ethionamide, six sur 21 cultures apres traitement se rtvelaient nettement rbistantes. On en conclut que la resistance naturelle a la thiacetazone ne s’accompagne pas dune resistance a l‘tthionamide; tandis qu’environ 30% des cultures avec resistance acquise a la thiacetazone sont aussi rtsistantes acquise a la thiacetazone sont aussi rtsistantes a l’ethionamide.
RESUMEN Se ha medido la sensibilidad frente a la ethionamida de M. tuberculosis aislados de enfermos de Africa de1 Este. Los resultados se ezpresaron por la concentration minima inhibitoria (MIC), con inoculos grandes y pequeiios, y por el metodo de las proporciones. Los cultivos previos al tratamiento que et-an naturalmente sensibles o resistentes a la thioacetazona no diferian en su sensibilidad frente a la ethionamida.
8
t-UHERCLI:
Pares de cultivos obtenidos antes y despues de 6 meses de tratamiento con isoniazida y thioacetazona se diferenciaban en forma significativa en cuanto a su sensibilidad frente a la ethionamida. Aunque ninguno des 10s cultivos previos al tratamiento eran resistentes a la ethionamida, 6 de 10s 21 cultivos post-tratamiento eran resistentes. Se concluye que la resistencia natural a la thioacetazona no se acompaiia de resistencia frente a la ethionamida; mientras que, de 10s cultivos con resistencia adquirida a la thioacetazona, alrededor de130 7,; son tambien resistentes a la ethionamida.
ZUSAMMENFASSUNG Die kulturelle Enpfindlichkeit von M. tuberculosis, isoliert von Patienten aus Ost-Afrika, ist in populationsbezogenen Sensibilittitspriifungen untersucht worden. Die Ergebnisse sind als “MICs” (= maximale Hemmungs-Konzentrationen) ausgedriickt, unter Verwendung kleiner oder groper Inocula und nach der Proportionsmethode. Kulturen, die vor der Behandlung prim& sensibel oder prim% resistent gegen Thiacetazon waren, unterschieden sich nicht nennenswert in ihrer Empfindlichkeit gegen Ethionamid. Signifikante Unterschiede in der Sensibilitgt gegen Ethionamid bestanden jedoch zwischen Kulturen, die vor und 6 Monate nach einer Behandlung mit Isoniazid undThiacetazon angelegt wurden. Obwohl keine der Kulturen vor Behandlung Resistenz gegen Ethionamid zeigte, waren sechs von 21 Kulturen nach Behandlung nachweislich resistent. Es wird daraus geschlossen, dap primlre Resistenz gegen Thiacetazon nicht mit Resistenz gegen Ethionamid kombiniert ist, wzhrend in Kulturen mit erworbener Resistenz gegen Thiacetazon ungefghr 30 “/, such resistent gegen Ethionamid sind.
Introduction Cross-resistance between ethionamide and thiacetazone was first demonstrated by Bartmann (1960). He found that cultures of Mycobacterium tuberculosis with acquired resistance to thiacetazone were frequently resistant to ethionamide in vitro. Similarly, cultures with acquired ethionamide resistance were often resistant to thiacetazone. These results have been confirmed by other workers (Rist, 1960; Tsukamura and Kasai, 1961; Stewart, 1962). The Japanese workers made an additional observation of great interest, namely, that strains which were naturally resistant to either ethionamide or thiacetazone did not exhibit cross-resistance to the other drug. The sensitivity to ethionamide of cultures of M. tuberculosis with natural or acquired resistance to thiacetazone has been studied among strains obtained from East African patients: the results are reported here. Methods Cultures of M. tuberculosis Cultures of M. tuberculosis were obtained from African patients admitted to two chemotherapy studies (East African/British Medical Research Council, 1966a,b). Four groups of cultures were examined, of which two groups comprised pre-treatment cultures, each from a separate patient. These two groups differed in thiacetazone sensitivity, as measured by sensitivity tests performed in Nairobi. In these tests, slopes of Liiwenstein-Jensen medium were inoculated with approximately 0.0002 mg. moist bacterial weight and incubated for four weeks. The results were expressed as minimal inhibitory concentrations (MICs), taking a IO-colony endpoint. Cultures yielding an MIC of 1 Erg/ml thiacetazone or less were regarded as sensitive (TS) and the remainder were regarded as resistant (TR). A third group of pre-treatment cultures (MO) was obtained from patients who produced positive cultures after six months’ treatment with thiacetazone and isoniazid. The
ETHIONAMIDE
SENSITIVITY
9
corresponding six-month cultures (M6) from the same patients comprised the fourth group. All cultures were shown to be M. tuberculosis by virtue of positive niacin tests and characteristic colonial morphology on 7H-10 agar plates. Sensitivity tests Sensitivity tests to ethionamide and thiacetazone were performed on Lowenstein-Jensen medium without potato starch (Cruickshank, 1965). Approximately 4 mg. moist weight of growth from a Lowenstein-Jensen slope was emulsified in 0.2 ml. water to which a further 0.8 ml. water was added, making the standard inoculum suspension. Four serial tenfold dilutions of this suspension were made in water, and Lbwenstein-Jensen slopes containing the following drug concentrations were inoculated with a 3 mm. loopful of the appropriate suspension: undiluted suspension (standard inoculum): 10, 20,40, 80, 160 pg/ml. ethionamide; 1 :I0 dilution: 0.12, 0.25,0.5 1, 2,4, 8, 16. 32 pg/ml. thiacetazone; I:100 dilution (small inoculum): 5, 10, 20, 40, 80 pg/ml. ethionamide: I :lOOOdilution: 5, 10, 20, 40 pg/ml. ethionamide. Three drug-free control slopes were inoculated from each of the 1 :lO,OOO,1 :lOOOand I : 100 dilutions: and one slope was inoculated from each of the 1: 10 and undiluted suspensions. Ethionamide was dissolved in ethylene glycol to give a final concentration of 2 % in the medium. and thiacetazone was dissolved in triethylene glycol to give a final concentration of 0.5 % in the medium. All cultures were incubated at 37°C. Ethionamide sensitivity tests (Lefford and Mitchison, 1966) were read at four and six weeks and expressed as follows : I. Standard inoculum MIC, using a 20-colony endpoint ; 2. Small inoculum MIC, (1 :lOOdilution of the standard inoculum) using a 5-colony endpoint; and 3. The proportion of growth on medium containing 10 and 20 pg/ml. ethionamide. This was estimated by expressing the viable count on drug-containing medium as a percentage of the viable count on drug-free medium. Thiacetazone sensitivity tests were read at four weeks and expressed as MICs, taking a IO-colony endpoint. Each batch of sensitivity tests included an approximately equal number of TS and TR cultures, and the MO culture from a given patient was always tested with the M6 culture from the same patient. The standard strain H37Rv was titrated in the above manner in each batch of tests. Results
Comparison oj‘TS with TR cultures Thiacetazone and ethionamide sensitivity tests were performed on 30 TS and 34 TR cultures. The thiacetazone sensitivity test results are reported fully in an accompanying paper (Lefford, Dickinson and Mitchison, 1969). Briefly, the difference in thiacetazone sensitivity between the TS and TR strains is indicated by the geometric mean MICs, which were 0.48 and 2.87 [rg/ml. thiacetazone respectively. The standard and small inoculum MICs of the ethionamide sensitivity tests read at four weeks are shown in Table I. The geometric mean of the standard inoculum MICs was 25.9 pg/ml. for the TS cultures and 31.4 pg/ml. ethionamide for the TR cultures, a non-significant difference. The corresponding values for the small inoculum results were 14.5 and 17.7 pg/ml. ethionamide, a statistically significant difference (P
IO
I-UBERCLE TABLE I.-ETHIONAMIDE
SENSITIVITY TESTS: MINIMAL INHIBITORY CONCENTRATIONS ?r 4 WEI:KS
Standard inoculum
Smull inoculum
Standard inoculum
Small inoculum
MIC
(pgjml.)
TS*
10 20 40 80 160
TR*
I 18 10 1 0
~ ~
Mean
Q-x/W
2 8 24 0 0
25.9
Total cultures
TS
31.4 _
30
TR
14 16 0 0 0
_~
34
MOi
7 26 I 0 0
.MO
M6
0 s 9 5 2
6 15 0 0 0
3 II 5 2 0
16.4
24.5
71
21
0 II 10 0 0
14.5 17.7 ~~__________ 30
!Mht
27.9
34
45.6
21
21
* TS -
thiacetazone sensitive. TR - thiacetazone resistant. t see text
fourth and sixth weeks of incubation, so that the MlCs and proportions were higher at the sixweek reading. However, this increase was relatively greater for the TS cultures and there was no significant difference in the mean MIC values between the two groups. Despite the general increase of growth on ethionamide medium, there was still a significant difference in the distribution of the proportion results between the two groups. Only 12 of 30 TS cultures yielded 5 % or more growth on 10 pg/ml. ethionamide, compared with 27 of 34 TR cultures (P ~0.01). The corresponding numbers yielding 1% or more growth on 20 pg/ml. ethionamide were 11 of 30 cultures, and 25 of 34 cultures, respectively (P cO.01). Thus it appeared that the TR cultures were slightly less sensitive to ethionamide than the TS cultures as judged by the MIC results; and appreciably less sensitive in terms of the proportion TABLE II.-ETHIONAMIDE
SENSITIVITY TESTS: PROPORTIONSAT 4 WEEKS
Ethionamide Proportion %
/
/ :
O.OlO.l-
10
concentration ,p
20
TS
TR
TS
~
I 2 9
0 1 4
18 4 7
i
2
5 7 1
i
30
34
(pg/ml.) IO
20
TR 6’ 6
‘,p-I Mo 0 ‘, 0
M6
_““““--
0 0
8 6
2 4
21
21
I2slo205&100 p/
Total cultures
~ /
0 0 0
30
34
~
21
21
:
ETHIONAMIDE
11
SENSITIVITY
results. However, the interpretation of proportion results is complicated by the finding that low viable counts on drug-free medium are associated with high proportions (Lefford and Mitchison. 1966). The viable counts on drug-free medium were therefore examined. The geometric mean viable counts at four weeks were 4.77 and 4.35 log units for the TS and TR cultures, respectively, a highly significant difference (P = O=OOl),which may account for part, if not all, of the difference in the proportion results of the two groups of cultures. When the available measures of sensitivity (MICs and proportions) and the influence of the viable counts on drug-free medium were evaluated for individual cultures, not one culture could be regarded as unequivocally resistant to ethionamide in the light of previous experience (Lefford. 1966). Comparison of A40 and M6 Cultures
Pairs of cultures obtained pre-treatment and at six months from 21 patients were examined. The sensitivity test results were expressed as MICs and proportions as in previous sections (Tables 1 and II). In addition, the difference in sensitivity between the cultures comprising each pair was estimated by dividing the sensitivity test result of each M6 culture by the corresponding result of the MO culture from the same patient. In this way, differences in sensitivity were expressed as a series of ‘sensitivity ratios’. The estimated change in sensitivity of a given strain could have been caused either by a real change in sensitivity or by experimental error. It was considered improbable that experimental variation could account for a four-fold increase in an MIC estimate or a ten-fold increase in a proportion estimate. An MIC ratio of 4 or more, or a proportion ratio of 10 or more were therefore taken to represent real changes in sensitivity. TABLE III.-THIACETAZONE AND ETHIONAMIDESENSJTJVITYOF CULTURES OBTAINED DURJNG THJACETAZONETREATMENT
Ethionamide sensitivity ratios Patient HO.
Thiacetazone A4IC
Standard inoculum MIC
Wml)
Small inoculum MIC
~ 1O~gglml.
20 pgglml. proportion
proportion
~_ K 307 K16 K 13 K 26 K 36 J 13 J 55 J 106 J 126 J 129 J 205 J 233 J 260 K153 J 254 K409 J 446 J 157 K 33 K 208 K 314
I ’
0.5
I 4 4 4 8 8 8 8 8 8 8 8 8 16 16 16 ~32 ~32 >32 >32
2 0.5
2 2 1 4 1 4 1 4 2
1 2 1 2 1 4
I I I
0.5 2 2 4
I
4 1 2
I i F5 ,
0.5 4 2
1
4 1 2 1 2 2 2
0.59 0.14 8.7 15 5.9 41 56 12 2.3 1.9 2.3 OGO4 589 4.7 14 1.5 4.3 5.0 2.3 30 4.7
~
1.7 13 162 11
1.1 ~
:
32 3.6 141 20 170 35 lOOO 331
>looo 1.2 5.4
I.9 I
4.2 0.21 11
I2
I UHERCLE
When the thiacetazone sensitivity test results were expressed as sensitivity ratios, it was found that I8 of the 21 M6 cultures had MIC ratios of four or more, of which 17 gave ratios of at least eight. In only three patients was there no significant change in thiacetazone sensitivity during treatment. Cultures from patients K307 and K16 remained sensitive throughout, and the MO culture from patient J 254 was as resistant (MIC of 16 pg/ml.) as the M6 culture (Table III). The distributions of the ethionamide sensitivity test results are shown in Tables 1 and II. It is evident both from the MIC and the proportion results that the M6 cultures were less sensitive than the MO cultures. Thus the geometric mean standard inoculum MIC was 27.9 pg/ml. for the MO cultures and 45.6 yg/ml. ethionamide for the M6 cultures (P
TABLE IV.-E.
AFRICANCULTURESWITH ACQUIREDRESWANC~. TO THIACETAZONE CLASSIFIED AS RESISTANT TO ETHIONAMIDE
Sensitivity Patient number
Four week readings
Six week readings
10pg/ml. proportion
20 pg/ml. proportion
IO pg/ml. proportion
20 pg/ml. proportion
(%)
(%)
(%I
(%I
Standard inoculum MIC
Small inoculum MIC
80 160 160 40 80 80
40 80 80 40 40 40
60.7 33.3 115 82.9 48.6 88.3
I .25 21.2 76.9 18.9 I .90 2.21
76.4 34.4 78.0 47.4 54.0 75.8
I .60 21.9 52.0 9.48 2.38 6.70
1
160
I
40
80 40
70.0 39.9
23.4 4.10
67.2 42.0
23.4 15.6
J 13 J 106 J 254 J 260 K26 K314 878* 720*
to ethionamide
* Initial sensitivity test on British pretreatment strains shown to be resistant to ethionamide (Lefford, 1966).
ETHIONAMIDE
13
SENSITIVITY
ficant association was found between the degree of thiacetazone resistance, measured either as an absolute MIC or as an MIC sensitivity ratio, and the emergence of ethionamide resistance. measured either as an MIC sensitivity ratio or as a proportion sensitivity ratio. Discussion
In this study it has been found that a proportion of cultures of M. tuberculosis obtained from patients treated with thiacetazone was resistant to ethionamide. However, since these patients were also treated with isoniazid, the possibility arises that cross-resistance between isoniazid and ethionamide may account for the results. Tan Thiam Hok (1964) reported cross-resistance between isoniazid and ethionamide whereas Rist (1960) and Stewart and others (1962) failed to find such an association. More recently Canetti and others (1967) have resolved this apparent contradiction by showing that cultures with low degrees of resistance to isoniazid (defined as less than 1 % growth on 0.5 pg/ml. isoniazid) are frequently resistant to ethionamide. However, cultures which are highly resistant to isoniazid are much less likely to be resistant to ethionamide. Of the 21 M6 cultures in the present study only one was sensitive to isoniazid (patient K409) and the remainder, including the six ethionamideresistant cultures, were highly resistant to isoniazid, having MICs of 5-50 pg/ml. It may reasonably be concluded that the ethionamide resistance among the M6 cultures was secondary to thiacetazone resistance, in agreement with other studies (Bartmann, 1960; Rist, 1960). The occurrence of ethionamide resistance following treatment of patients with thiacetazone is of importance since it affects the usefulness of ethionamide in secondary treatment regimens. Bartmann (1960) found 15 of 22 thiacetazone-resistant cultures to be resistant to ethionamide compared with six of 21 in the present study, in which stringent criteria of resistance have been applied. The validity of these criteria derives support from a previous study (Lefford, 1966). It would appear that two-thirds of patients with strains having acquired thiacetazone resistance might be expected to respond to treatment which includes ethionamide. This estimate may prove to be optimistic since reference to Table Ill indicates that almost all the M6 cultures were less sensitive to ethionamide, to a greater or lesser degree, than the corresponding MO cultures. It is notable that the degree of ethionamide resistance cannot be predicted from the degree of thiacetazone resistance. REFERENCES BARTMANN,K..
(1960). Kreuzresistenz zwischen a-.%thylthioisonicotanamid (1314 Th) und Thiosemicarbazon (Conteben). Tuberkuloseartz, 14,525. CANETTI, G., KREIS, B., THJBIER, R., GAY, P., LE LIRZIN, P. (1967). Don&es actuelles SW la resistance primaire dans la tuberculose pulmonaire de l’adulte en France. Rev. Tuberc., Paris, 31,433. CRUJCKSHANK, R., (1965). MedicalMicrobiology. 11th edition, p. 753. E. & S. Livingstone, Edinburgh. EAST AFRICAN/BRITISHMEDICAL RESEARCHCOUNCILTHIRD THJACETAZONEINVESTIGATION (1966a). Isoniazid with thiacetazone (thioacetazone) in the treatment of pulmonary tuberculosis in E. Africa - Third Investigation: The effect of an initial streptomycin supplement. Tuber&, Load., 47,1. EAST AFRICAN/BRITISHMEDICALRESEARCHCOUNCILFOURTHTHIACETAZONE INVESTIGATION (1966b). Isoniazid with
thiacetazone (thioacetazone) in the treatment of pulmonary tuberculosis in E. Africa The effect of increasing the dose of isoniazid. Tubercle, Lond.. 47,315.
Fourth Investigation
:
LEFFORD. M. J. 0966). The ethionamide sensitivitv of British oretreatment strains of Mvcobacterium tuberculosis. Tubercle, Loid., 4?, 198. LEFFORD, M. J., DICKINSON,J. M. & MIXHISON, D. A. (1969). The sensitivity to thiacetazone and para-aminosalicylic acid and the virulence in the guinea pig of East African strains of Mycobacterium tuberculosis. Tubercle, Lond. 50,l. LEFFORD, M. J. & MITCHISON,D. A. (19661. Comnarison of methods for testing the sensitivitv of Mvcobac?erium tuberculosis to ethionamide. Tubercie, L&d., 47.; 250. RISE, N. (1960). The anti-tuberculosis activity of ethlonamide. Adv. Tuberc. Res., 10,69. STEWART,S. M., HALL, E., RIDDELL, R. W. & SUMNER, A. R. (1962). Bacteriological aspects of the use of ethionamide. pyrazinamide and cycloserine in the treatment of chronic pulmonary tuberculosis. Tubercle, Lond., 43,417. TAN THIAM HOK (1964). A comparative study of the susceptibility to ethionamide, thiosemicarbazone and isoniazid of tubercle bacilli from patients never treated with ethionamide or thiosemicarbazone. Am. Rev. resp. Dis.. 90,468. TSUKAMURA,M. & KASAI, E. (1961). Clinical observations on the nature and acquired resistance of tubercle bacilli to ethionamide. Jap. J. Chest Dis., 20,671.