Tdwcle

and Lung Lhease (1992) 73. 219-224

Drug resistance of Mycobacterium tuberculosis in Korea S. J. Kim, Y. P. Hong Korean Institute of Tuberculosis,

Korean National Tuberculosis Association,

Seoul, Korea

S U M M A R Y. Drug resistance of Mycobucterium tuberculosis has been investigated with isolates from patients screened from a sample population of the nationwide tuberculosis prevalence surveys or from routine cultures. The results showed a close inverse relationship between the prevalence of drug resistance and the efficiency of the past or current National Tuberculosis Control Program (NTP) treatment regimens. Individual drug resistance also showed a close relationship with the extent of use of the relevant drugs. Drug resistance was found in 38.0% of M tuberculosis isolates from patients in the1965 survey and remained unchanged until it increased to 48.0% in 1980. The resistance prevalence, however, dropped to 25.3% in the 1990 survey. Such a decrease coincided fairly well with a continuous increase of the treatment efficiency (from 60% in 1984 to 77% in 1989) in the 1980s. Initial drug resistance (IR) also showed a similar trend, namely 26.2% in 1965, 23.9% in 1970, 20.1% in 1975,30.6% in 1980,17.4% in 1985 and 15.0% in 1990. A higher prevalence of IR was observed among urban patients than rural patients and among young as opposed to old patients.

& . La rksistance de Mycobacterium tuberculosis aux drogues a ete examinee avec des isolats provenant de malades s&&ion& a partir d’un Cchantillon de population concernee par les enquCtes nationales sur la prevalence tubercnleuse ou B partir de cultures de routine. Les rkmltats ont montre un lien etroit inverse entre la prevalence de la resistance aux drogues et l’efficacite des rGgimes de traitement anciens ou actuels du Programme National de Tuberculose (PNT). La r&stance individuelle aux drogues a aussi montre une relation etroite avec l’etendue de l’utilisation des drogues concern&es. Une r&stance aux drogues avait Cteobservee dam 38,0% des isolats de Mycobacterium tuberculosis h partir des malades de l’enquete de 1%5, taux qui Ctait restk inchange jusqu’a ce qu’il augmente a 48,0% en 1980. Cependant la prevalence de la resistance a baisse jusqu’l 25,3% dans l’enqu&e me& en 1990. Une telle baisse a sensiblement comcide avec une augmentation continue de l’efficacite du traitement (de 60% en 1984 a 77% en 1989) pendant les an&es 80. La resistance initiate (RI) aux medicaments a montre Cgalement une evolution similaire, c’est-a-dire de 26,2% en 1965, a 23,9% en 1970, 20,1% en 1975, 30,6% en 1980, 17,4% en 1985 et 15,0% en 1990. Une prevalence plus Bevke de la RI a ete observee chez les malades provenant des regions urbaines par comparaison avec celle des malades en provenance des regions rurales, et parmi les malades jeunes par comparaison avec les malades LgCs. R k S UM

R ES U M E N . En Corea, se investigo la resistencia a Lasdrogas, de Mycobacterium tuberculosis, con aislados

provenientes de enfermos seleccionados a partir de una muestra de poblaci6n, en el context0 de encuestas nacionales de prevalencia de la tuberculosis o a partir de cultivos de rutina. Los resultados mostraron una estrecha relachjn, inversamente proportional, entre la prevalencia de la resistencia a las drogas y la eficacia de 10s esquemas de tratamiento antiguos o actuales de1 Programa National de Tuberculosis (PNT). La resistencia individual a las drogas mostr6 igualmente una estrecha relaci6n con la extensi6n de1 uso de las drogas concernidas. Se habIa demostrado una resistencia a las drogas en un 38,0% de 10s aislados de M. tuberculosis provenientes de enfermos de la encuesta de 1965, tasa que se mantuvo hasta que aumento a 48,0% en 1980. Sin embargo, la prevalencia de la resistencia baj6 a 25,3% en la encuesta de 1990. Esta disminuci6n coincidio claramente con el aumento sostenido de la eficacia de1 tratamiento en la decada de1 80 (de 60% en 1984 a 77% en 1989). La resistencia initial (RI) a las drogas mostr6 una tendencia similar, es decir 26,2% en 1965,23,9%

Correspondence to: Dr Sang Jae Kim, Director, Dept of Bacteriology, Korean Institute of Tuberculosis, 14 Woomyundong, Sochogu, Seoul 137-140, Korea. 219

220

Tubercle and Lung Disease

en 1970,20,1 % en 1975,30,6 % en 1980,17,4 % en 1985 y 15,0 % en 1990. Se observ6 una resistencia initial mb elevada en 10s enfermos de las heas urbanas, en comparacih con 10s de las aireas rurales y en 10sj6venes en comparacidn a 10s de edad avanzada.

INTRODUCTION Drug resistance of Mycobucterium tuberculosis develops by the selective growth of resistant mutants evolving in a wild population when exposed to an amount of drug sufficient to inhibit growth of sensitive organisms.‘-3 Drug resistant mutants most likely evolve through a chromosomal mutation to escape or prevent drug-induced metabolic damage, probably by a decreased permeability of the cell wall or membrane against the drug, by decreased affinity of drug binding sites, by inactivation of the drug or by loss of an enzyme that activates the drug.1,2,4,5 The major concerns over drug resistance were a fear of the spread of drug-resistant organisms and the ineffectiveness of chemotherapy in patients infected with them. In view of the results so far observed, however, there is no clear evidence of an increase in primary drug resistance (PR) in most countries, although relatively high prevalences have often been reported from developing countries.2.4.6 It has also been shown in several reports that patients with initial drug resistance (IR) have responded as well as sensitive cases to the intensive short-course regimens.4s5 In the meantime it has been suggested elsewhere that a continuous survey of drug resistance would provide useful epidemiological information for planning treatment programmes and for monitoring the efficacy and/or efficiency of chemotherapy in national tuberculosis control programmes (NTP).2-5,7 In this regard we have investigated drug resistance of M. tubercdosis isolates from patients identified in the sample population of the nationwide tuberculosis prevalence surveys performed at 5year intervals since 1965 or isolates derived from routine laboratory cultures.

METHODS M. tuberculosis isolates From the surveys The nationwide tuberculosis prevalence surveys (NTPS) have been carried out at 5-year intervals since the first survey of 1965. All the isolates of M. tuberculosis from patients screened from the sample population were submitted to sensitivity tests against the various antituberculous drugs. The number of isolates tested were 71 in 1965,132 in 1970,270 in 1975,177 in 1980, 247 in 1985, and 189 in the 1990 survey. From routine laboratory cultures Clinical isolates from newly diagnosed patients with pulmonary tuberculosis in the chest clinics of the Korean

National Tuberculosis Association or in private or public hospitals were also submitted to drug sensitivity tests in our laboratory. The number of isolates tested were 233 in 1962-1963, 548 in 1964-1965, 329 in 1966-1967, 281 in 1968, 211 in 1970-1971, 189 in 1975, 1 194 in 1977-1978, 306 in 1980, 202 in 1982-1983, 563 in 1984-1985and149in19861987.

Drug sensitivity tests The procedure of drug sensitivity

tests was based on the absolute concentration method described by Canetti et a18,’with a small modification of inoculum preparation and size.” The tests were done in Lowenstein-Jensen medium. The pre-inspissation critical concentrations of the drugs by which likely resistant strains have been separated from sensitive strains were 0.2 pg/ml for isoniazid (INH), 32 pg/ml for rifampicin (RMP), 10 l.@-nl for streptomycin (SM), 2.8 p.g/ml for ethambutol (EMB), 100 pg/ml for pyrazinamide (PZA), 1 pg/ml for p-aminosalicylic acid (PAS), 50 pg/ml for kanamycin (KM), 50 l.tg/ml for tuberactinomycin (TUM), 56 pg/ml for prothionamide (PTH), and 40 @ml for cycloserine (CS). In the 1965 and 1970 surveys the strains were tested against only SM, PAS and INH, but in the subsequent surveys, all of the aforementioned drugs were used in the tests.

RESULTS Table 1 sets out the prevalence of drug resistance in new and old patients screened from the sample populations of the last 6 nationwide tuberculosis prevalence surveys. Resistance to one or more antituberculous drugs was found in 38.0% of M. tuberculosis isolates from patients found in the 1965 survey and the prevalence remained stable until the 1980 survey when it increased to 48.0%. Multiple drug resistance, namely resistance to three or more drugs, was also highest (14.7%) in the 1980 survey, suggesting that a considerable number of treatment failure cases might have accumulated, probably due to ineffective chemotherapy. However drug resistance decreased significantly in the 1985 survey to 30.8% and a further reduction was observed in the 1990 survey, to 25.3%. It may indicate a reduced influx of treatment failures into the pool of prevalent bacillary cases. In general it is well known that resistance to individual drugs parallels the extent of the use of the relevant drug and the efficacy of the regimens containing the corresponding drug. For example, INH resistance was found most frequently, amounting to 25.4% in

Drug resistance of Mycobacferium Table 1

Survey

Drug resistance of M. tuberculosis

in the nationwide

Total cases studied

Resistance 21 drugs

Old % Both %

42 100.0 29 100.0 71 100.0

11 26.2 16 55.2 27 38.0

New % Old % Both %

92 100.0 40 100.0 132 100.0

New % Old Both %

Patients

year

tuberculosis

prevalence

INH+

RMP+

SM+

1 3.4 1 1.4

7 16.7 11 37.9 18 25.4

4 9.5 7 24.1 11 15.5

22 23.9 28 70.0 50 37.9

6 6.5 11 27.5 17 12.9

18 19.6 28 70.0 46 34.8

13 14.1 12 30.0 25 18.9

189 100.0 81 270 100.0

38 20. I 60 98 36.3

10 5.3 20 30 11.1

34 18.0 59 93 34.4

New % Old R Both %

108 100.0 69 100.0 177 100.0

33 30.6 52 75.4 85 48.0

6 5.6 20 29.0 26 14.7

1985

New 8 Old % Both 8

161 100.0 86 100.0 247 100.0

28 17.4 48 55.8 76 30.8

1990

New % Old Ye Both %

127 100.0 62 1tXI.O 189 100.0

19 15.0 29 46.8 48 25.3

1965

1970

1975

1980

New %

EMB+

I

INH RMP+

1NH SM+

INH EMB+

I 1.4 2 6.9 3 3.2 10

10.9 12 30.0 22 16.7

2 1.1 2 4 1.5

16 8.5 28 44 16.3

2 1.1 2 4 1.5

27 25.0 50 72.5 77 43.5

3 4.3 3 1.7

5 4.6 20 29.0 25 14.1

6 3.7 14 16.3 20 8.1

22 13.7 41 47.7 63 25.5

4 2.5 22 25.6 26 10.5

6 4.7 8 12.9 14 7.4

16 12.6 26 41.9 42 22.2

9 14.5 Y 4.8

1965, 34.8% in 1970, 34.4% in 1975, 43.5% in 1980, 25.5% in 1985, and 22.2% in 1990, because this drug was used most widely. A continuous decrease after 1980 indicates a significant improvement in the efficacy of initial treatment regimens containing INH. The next most common drug to which resistance was found was streptomycin, as it increased to 18.9% in 1970 and then decreased before it increased again to some extent in 1990. Less use of streptomycin due to a rare incidence of anaphylactic shock might have resulted in a continuous decrease of SM resistance after 1970. However the increase in 1990 cannot be explained. On the contrary, RMP and EMB resistance was encountered most often in the most recent surveys because these drugs started to be widely used in the late 1970s and early 1980s. However resistances to these drugs is seen to be decreasing in the last survey. Drug resistance of smear-positive cases, who are the major infectious sources in the community,is set out in Figure 1. The estimated number of smear-positive case was approximately 166000 out of a total population of 28.6 million in 1965 (580/105) and dropped to 54000

22

surveys

to 23 drugs

in Korea

tuberculosis

2

2 1.1 1 3 1.1

13 6.9 27 40 14.8

2 4 I .5

6 5.6 17 24.6 23 13.0

3 4.3 3 1.7

3 2.8 18 26.1 21 11.9

4 3.7 15 21.7 19 10.7

6 3.7 8 9.3 14 5.7

6 3.7 14 16.3 20 8.1

4 2.5 I7 19.8 21 8.5

2 I.2 6 7.0 8 3.2

6 3.7 13 15.1 19 7.7

9 7.1 12 19.4 21 11.1

2 1.6 7 Il.3 9 4.8

9 14.5 9 4.8

7 Ii.5 IO 16.1 17 9.0

2 I .h 7 II.3 Y 4.x

1.1

out of a total population of 42.8 million in 1990 ( 126/105). The proportion of patients who had a history of past or current chemotherapy was 41.0% of smearpositive cases found in 1965, 38.7% in 1970, 48.0% in 1975, 61.7% in 1980, 52.8% in 1985 and 40.6% in 1990. From 1965 through 1985, over two-thirds (67.8% -77.8%) excreted organisms resistant to one or more antituberculous drugs. However it decreased to 46.4% in the 1990 survey. Including initial drug resistance (IR), 49.3% of smear-positive found in 1970 were resistant to one or more drugs, 47.3% in 1975, 57.6% in 1980, 46.5% in 1985 and 27.5% in 1990. Although nearly half the smear-positive cases excreted drug-resistant bacilli, IR did not increase continuously (Table 1). IR was 26.2% in 1965 and increased to 30.6% in 1980 but decreased in the subsequent surveys to 17.4% in 1985 and 15.0% in 1990. The decrease of IR between 1985 and 1990 was somewhat less remarkable than the decrease in acquired resistance (AR). A large decrease in AR (resistance of patients having past or current chemotherapy history) might have resulted from (1) considerable reduction of treatment failures and

222

Tubercle and Lung Disease 8URVEY 19so

1985

YEAR

1980

1975

l970

1965

cz!!s NEW CASES

21 DRUG RESISTANTCASES _22 DRUG RESISTANTCASES

128

Fig.l-Estimated

220

288 427 PER 100.000 OF

smear-positive

SURVEY YEAR 1985

1aso

TREATED FOR C 3 4-8

MONTHS MONTHS

7 - 11 MONTHS

212

qSENSITIVE Fig. 2-Drug chemotherapy

resistance history.

f$jRESISTANCE

of patients with the past or current

580 -

cases with or without drag resistance.

(2) increased proportion of sensitive relapse cases among patients with a previous chemotherapy history. As seen in Figure 2, drug resistance among patients with a history of more than 12 months treatment was 90.2% in 1980, 71.7% in 1985, and 56.5% in 1990. Drug resistance of patients treated for 3 months or less was also much lower in the last survey when compared with those of the previous surveys. It is matter of concern to have observed a significantly higher prevalence of IR among younger patients. For instance, the average prevalence in patients less than 30 years old was 3 1.8%, while it was 14.1% in the over 59 years age groups as seen in Table 2. Whether younger patients are more likely to be infected with drug-resistant organisms is not known. There was also a significant difference in IR prevalence between urban and rural patients, showing much higher level of resistance among former than the latter group. The main reason why IR is high among urban patients may be that urban patients have easier access to drugs than rural patients. However

1980

470 POPULATION

MONTHS

there was no significant difference between male and female patients. The IR of clinical isolates from routine cultures is set out in Table 3. Although there is some fluctuation, drug resistance seems to have increased up to the late 1970s or early 1980s especially multiple drug resistance, but it decreased rapidly thereafter. Resistance to one or more drugs was found in 17.6% of the strains tested in 1962-1963,23.5% in 19641965,27.7% in 1966-1967, 24.6% in 1968, 26.1% in 1970-1971, 20.1% in 1975, 31.7% in 1977-1978, 29.4% in 1980, 26.7% in 19821983, 18.8% in 1984-1985 and 20.8% in 1986-1987. Because INH has been used as a basic drug in all the initial treatment regimens, prevalence of resistance to this drug coincided fairly well with all drug resistances alone or in combination. Resistances to PAS and SM continuously decreased after 1970 as these drugs were used less and less. EMB resistance sharply increased in the early 1980s because wide use of this drug started in the late 197Os, replacing PAS in the then standard regimen SIWPAUINH, for initial treatment in the NTP. RMP resistance steadily increased after introduction of the short-course regimen into the NTP in the early 1980s. IR may provide useful information on the type of chemotherapy that should be used for the treatment of patients who come to health institutions for the first time, and may also serve as a useful parameter in the evaluation of current and past chemotherapy programmes. In the early 1960s the most widely used initial treatment regimen was either PAS/INH, SM/INH or SM/PAS/INH. Full activity of the NTP began in the late 1960s and three-drug regimens such as SM/PAS/INH or SM/INI-I/EMB were available for all new patients found in the health centres in 1970s. The efficacy (conversion rate of those who completed chemotherapy) of the SIWINIWEMB regimen under programme conditions was found to be 66%, and its efficiency (conversion rate of those who have started

Drug resistance of Mycobacterium tuberculosis in Korea Initial drug resistance by age, sex, and areas in the nationwide

Table 2

Survey year 1970 A@ 5 29 30-59 2 60 Area Urban Rural sex Male Female

Survey year 1962-1963 1964-1965 1966-1967 1968 1970-1971 1975 1977-1978 1980 1982-1983 1984-1985 1986-1987

233 548 329 281 211 189 1194 306 202 563 149

Total

15/45(33.3) 8/69( 11.6) 5/47(10.6)

4/13(30.8) 10/65(15.4) 5/48(10.2)

49/154(31.8) 64/324(19.8) 28/199(14.1)

18/49(36.7) 5/43(11.6)

21/68(30.9) 17/121(14.0)

22/49(44.9) 11/59(18.6)

22/82(26.8) 6/79(7.6)

13/70(18.6) 6/57(10.5)

96/3 18(30.2) 45/359( 12.5)

15/67(22.4) 8/25(32.0)

24/121(19.8) 14/68(20.6)

16/61(26.2) 17/47(36.2)

21/116(18.1) 7/45(15.6)

13/92(14.1) 6/35(17.1)

89/457( 19.5) 52/220(23.6)

cases/number

of cases studied. ( ) = %

of the patients diagnosed

Resistance (%) bl 22

23

drug

drugs

drugs

17.6 23.5 27.7 24.6 26.1 20.1 31.7 29.4 26.7 18.8 20.8

6.0 9.0 14.3 10.3 9.0 10.1 13.3 15.7 15.8 8.7 7.3

4.4 4.6 5.0 4.3 5.3 7.1 8.8 7.9 4.6 3.3

under the routine

>4 drugs

INH+

1.6 3.4 5.2 5.4 1.2 1.3

9.9 17.2 24.6 22.1 22.3 18.0 28.6 26.5 23.3 17.1 18.1

services

RMP+

EMB+

SM+

12.2

1.0 3.9 5.9 3.6 4.0

5.4 8.2 7.9 5.2 2.0

INH EMB+

0.2 3.6 5.4 3.4 3.4

0.7 1.0 0.5

7.8 7.1 6.9 3.1 4.9 5.9 3.6 3.4

4.5 6.2 7.9 4.8 2.0

INH/RMP/BMB, initially in 1982 and a subsequent nationwide extension in the following years. Efficacy of INH/RMP/EMB was over 95% and efficiency reached to 80% in 1988 as seen in Table 4. The improved efficacy might have led to a significant reduction of drug resistance in the surveys that followed. The extension of treatment coverage with INW RMP/EMB eventually led to a continuous increase in the overall efficiency of the NTP treatment from 60% in 1984 to 77% in 1989. The prevalence rate of drug resistance showed an inverse relationship with the overall efficiency of the NTP treatment programme as

in the Korean national tuberculosis

EFN (%)

Retreatments No EFC (%)

EFN (%)

174 380 4694 8922 14774 15045 15559 15882

571 923 292 5085 6763 3256 2745 2212 2113

41 41 52 52 56 57 58 58 55

493 1043 492 3619 6472 3541 3025 2744 2133

47 48 52 52 56 56 59 58 61

66 60 75 75 80 79 82 84 86

INH SM+

10.0 10.0 9.5 8.5 3.7 7.2 5.9 4.3 5.4

SIWINWEMB No. EFC (%)

73 75 75 75 77 79 80 81

INH RMP+

2.6 6.0

Initial treatments INH/RMP/EMB No. EFC EFN (%) (%)

87 92 92 94 94 95 95 96

PZA+

8.2 10.9

Table 4 Efficacy and efficiency of the initial treatment and retreatment programme

1981 1982 1983 1984 1985 1986 1987 1988 1989

1990

8/21(38.1) 15/49(30.6) 10/38(26.3)

chemotherapy) was 41% in the 1981 assessment as seen in Table 4. Before 1980, the efficiency of SM/INH/BMB or SM/PAS/INH was no better than that observed in the 1981 and 1982 assessment. Case-finding of tuberculosis patients in the health centres was extended considerably in the 1970s without any improvement in treatment efficiency. As a result, an accumulation of drug-resistant cases seemed to be unavoidable in the late 1970s and so this might have produced a high prevalence of drug resistance in the 1980 survey. However a remarkable improvement in both efficacy and efficiency was achieved by introduction of the short-course regimen,

Year assessed

surveys

12/53(22.6) 20/89(22.5) 6/47(12.8)

Initial drug resistance No. of cases studied

1985

1980

1975

prevalence

10/22(45.5) 11/51(21.6) 2/19(10.5)

Number of drug-resistant

Table 3

tuberculosis

223

No. = number of cases assessed, EFC = efficacy, EFN = efficiency

74 71 74 74 78 80 80 81 85

Overall EFC EFN (%) (%)