Primary Drug-resistant Tuberculosis in Children

INT J TUBERC LUNG DIS 4(12):1149–1155© 2000 IUATLD

Primary drug-resistant tuberculosis in children

H. S. Schaaf,* R. P. Gie,* N. Beyers,* F. A .Sirgel,† P. J. de Klerk,‡ P. R. Donald*

* Department of Paediatrics and Child Health, Tygerberg Hospital and the University of Stellenbosch, † TuberculosisResearch Programme of the MRC, ‡ Department of Microbiology, Tygerberg Hospital and the University of Stellenbosch,

S U M M A R Y

Tygerberg, South Africa

SETTING: The Western Cape Province of South Africa,an area with a high tuberculosis (TB) incidence whereinitial isoniazid (INH) resistance and multidrug resis-tance (MDR) among adults was 3.9% and 1.1%, respec-tively, during 1992–1993.OBJECTIVE: To determine the drug resistance incidenceamong children as compared to adults, to compare theclinical features of drug-resistant and drug-susceptible TB,and the degree of INH resistance in isoniazid-resistantisolates.METHODS: All Mycobacterium tuberculosis culturesobtained from children (0–13 years) at a regional hospi-tal were prospectively collected from August 1994 toApril 1998 and susceptibility testing done on eachchild’s specimens. Degree of INH resistance was deter-

mined in available resistant isolates. The children’s clin-ical records were reviewed.RESULTS: Susceptibility results were available in 306/338 children with cultures of M. tuberculosis; 21 iso-lates (6.9%) were INH-resistant, and seven were MDR.Taking into account study limitations, the incidence ofINH resistance was 5.6% and MDR 1% in childrenaged �5 years. Clinical features were similar in chil-dren with drug-susceptible and drug-resistant TB.CONCLUSION: The incidence of drug resistance in child-hood tuberculosis in Western Cape is low, and probablyreflects the level of primary drug resistance amongstorganisms currently circulating in the community.KEY WORDS: children; tuberculosis; drug resistance;incidence; isoniazid

IN THE LAST DECADE there has been increasedawareness that drug-resistant tuberculosis (TB)poses a major threat to patients as well as to tuber-culosis control programmes. Surveillance of drugresistance is therefore essential, because trends inprimary drug resistance (i.e., resistance in culturesfrom patients with no previous tuberculosis treat-ment) or initial drug resistance (i.e., primary resis-tance plus undisclosed acquired resistance) providean indication of the effectiveness of treatment regi-mens, while drug resistance rates in patients with ahistory of previous treatment (acquired resistance)can indicate failures in the management of the dis-ease.1 Accuracy in determining whether patientshave initial/primary or acquired drug resistance isoften unsatisfactory, as obtaining a history of previ-ous treatment is difficult.2 Van Rie et al.3 demon-strated that some patients who previously haddrug-susceptible TB and thereafter presented withdrug-resistant tuberculosis were reinfected with drug-resistant strains and thus had primary drug-resistanttuberculosis, despite having had previous tuberculo-sis treatment.

Children have mainly primary drug-resistant TB,

and previous treatment is easier to exclude. Theyshould therefore be a good source for surveillance oftrue primary drug-resistant TB, and may accuratelyreflect the transmission of these organisms in thecommunity.4 There are few studies of drug-resistantTB in children or surveillance data that include chil-dren in significant numbers,5,6 and no data fromdeveloping countries could be traced.

The use of isoniazid (INH) in isoniazid-resistantTB patients is controversial.7 There is reason to believethat adding INH to the treatment of newly diagnosedchildren with INH-resistant TB would be beneficial,as the degree of INH resistance differs in patientswith primary and acquired resistance, and in earlystudies some benefit was derived from adding INHto treatment regimens in adults with primary INH-resistant TB.8,9

The aim of this study was to determine the preva-lence of drug resistance among children with tubercu-losis and to compare the clinical and radiological fea-tures of drug-resistant and drug-susceptible TB inthese children. A secondary aim was to determine theminimal inhibitory concentration (MIC) of INH-resistant isolates.

Correspondence to: H Simon Schaaf, Department of Paediatrics and Child Health, University of Stellenbosch, PO Box19063, Tygerberg 7505, South Africa. Tel: (�27) 21 938-9506. Fax: (�27) 21 938-9138. e-mail: [email protected] submitted 28 February 2000. Final version accepted 22 August 2000.

1150 The International Journal of Tuberculosis and Lung Disease

PATIENTS AND METHODS

SettingThis prospective study was conducted betweenAugust 1994 and April 1998 at Tygerberg Hospital, asecondary and tertiary referral centre in the WesternCape Province of South Africa. This area had areported TB notification rate of 589 new cases per100 000 population per year in 1998 (Department ofHealth: Directorate Health Systems Research andEpidemiology). The rate of initial resistance to INHdetermined in adult TB cases in the Western CapeProvince during 1992–1993 was 3.9% (95% confi-dence intervals [CI] 3.3–4.6%) and initial multidrugresistance (MDR—defined as cultures resistant toINH and rifampicin [RMP], with or without resis-tance to other anti-tuberculosis drugs) was found in1.1% of isolates (95%CI 0.7–1.4%).1 Prevalence ofhuman immunodeficiency virus (HIV) infection inwomen attending antenatal clinics in the WesternCape Province rose from 1.16% (95%CI 0.76–1.56)in 1994 to 5.21% (95%CI 4.2–7.2) in 1998 (Depart-ment of Health: Directorate Health Systems Researchand Epidemiology).

Drug susceptibility testingAll cultures of Mycobacterium tuberculosis obtainedat Tygerberg Hospital from children 0 to 13 years ofage were collected prospectively. A single specimenfrom each patient was sent to the South African Insti-tute for Medical Research in Cape Town for suscepti-bility testing. Initial screening was done for INH resis-tance only, because RMP resistance was seldom foundwithout INH resistance in primary (initial) drug resis-tance in the Western Cape.1 When resistance to INHwas found, susceptibility testing for RMP, streptomy-cin and ethambutol was performed.

Laboratory procedures for determining drug resis-tance were as follows: Middlebrook 7H12 (Bactec)culture medium was used for selective primary isola-tion of mycobacterial strains. The niacin productiontest was used to identify M. tuberculosis. Drug sus-ceptibility testing was performed by the economicvariant of the indirect proportion method currently inuse by the majority of laboratories in South Africa.This method entails the incorporation of the requireddrug concentration into Löwenstein-Jensen (L-J) egg-based medium before coagulation, while the slantswere subsequently inoculated with a standardizedinoculum. The following drugs were tested at theindicated concentrations: INH 0.2 �g/ml L-J; RMP30.0 �g/ml L-J; streptomycin 5 �g/ml L-J and etham-butol 2 �g/ml L-J. The susceptibility of a strain wasjudged by determining the proportion of bacilli resis-tant to a specific drug in comparison with growth ona specific control, using international criteria. Resis-tance was defined as 1% or more bacterial growth.Quality assurance for drug susceptibility results is

done locally with every batch and quarterly by thenational tuberculosis reference laboratory.10

In addition, nine of the clinical isolates that wereresistant to the critical concentrations of INH (0.2�g/ml), and which were still available at the time,were processed to determine the degree of their resis-tance in MIC values (�g/ml). Briefly, precultures wereprepared in 7H9 Middlebrook medium and incu-bated at 37�C for approximately 14 days before theywere adjusted to an optical density of a No. 1 McFar-land standard, which is equivalent to 2 � 107 organ-isms/ml. The latter suspensions were further diluted(1/20) to a concentration of about 1 � 106 colonyforming units (cfu)/ml, and 20 �l aliquots were inoc-ulated on to 7H10 Middlebrook medium, with orwithout appropriate dilutions of INH, to yield finalconcentrations of approximately 2 � 104 cfu. INHwas incorporated into 7H10 plates at final concentra-tions of 2, 4, 5, 10 and 20 �g/ml, and M. tuberculosisH37Rv (ATCC 27294) was used as a drug-sensitivecontrol. An organism was considered to be resistantto a specific concentration of the drug if it failed toinhibit more than 99% of the bacterial population,indicating whether the isolate has a high or low levelof resistance to INH.

Clinical dataThe clinical records of all children with positive cul-tures for M. tuberculosis were reviewed, and allchildren with drug-resistant TB were recalled. A his-tory was obtained from all children regarding previ-ous TB prophylaxis or treatment, and whether theyhad close contact with adults with pulmonary TB.Weight at diagnosis and site of tuberculosis wererecorded. Results of tuberculin skin test (Mantouxtest by intradermal injection, 5 tuberculin units,Japanese purified protein derivative), HIV serology,and sites of positive M. tuberculosis specimensobtained were noted where available. An area ofinduration of �15 mm after Mantoux skin testingwas regarded as significantly positive in accordancewith World Health Organization criteria, as morethan 90% of children in this area receive BCG.11

Chest radiographs were read according to a standard-ized method.12 All children with drug-susceptible TBwere treated according to the National TuberculosisProgramme (NTP) policy with INH, RMP and pyr-azinamide. Children with drug-resistant TB weretreated according to the drug susceptibility patternof their M. tuberculosis strain and followed up for2 years.

Categorical data was analyzed using the �2 test tocompare groups, and Fisher’s exact test was appliedwhere appropriate. Statistical analysis was doneusing Epi-Info version 6.04.

The study was approved by the Ethics Commit-tee of the Faculty of Medicine of the University ofStellenbosch.

Drug-resistant tuberculosis in children 1151

RESULTS

During the 3 year 9 month period, 538 M. tuberculo-sis isolates were obtained from 338 children. Thechildren, 193 (57%) boys and 145 (43%) girls, had amedian age of 2.6 years (range 0.06–13 years). Theywere divided into two groups; �5 years of age (n 241) and 5–13 years of age (n 97). In respectively23 (9.5%) and nine (9.3%) children, drug susceptibil-ity testing was not done due to lost specimens (n 13), contamination (n 15) or loss of viability (n 4).

Susceptibility resultsSusceptibility results were available in 90.5% of cases(Table 1). Of the 21 isolates (6.9%) that were resis-tant to INH, seven (2.3%) were multidrug-resistant.Of 56 (19.6%) M. tuberculosis isolates that were sus-ceptible to INH and also tested for RMP susceptibil-ity, none was resistant to RMP.

Study limitationsThirty children whose M. tuberculosis strains hadbeen tested for drug susceptibility received previousanti-tuberculosis treatment. Of these, 16 defaultedfrom treatment. Drug resistance was found in two:one child had received 5 months of INH, RMP andpyrazinamide and did not respond to treatment. Theinitial culture was done before the study began, butsusceptibility testing was not done. Follow-up culturewas multidrug-resistant. No close contact could beidentified. The second child had previous TB and wastreated for more than a year due to extensive pulmo-nary TB. Her father had INH-resistant TB but diedafter her initial TB diagnosis. She again presented 4years later with culture-confirmed TB and bron-chiectasis with no known TB source case. The M.tuberculosis strain was resistant to INH only. Both ofthese cases could therefore still have had primarydrug-resistant TB.

The second limitation of the survey is that it wasconducted concurrently with a study of childhoodcontacts �5 years of age of adults with MDR pulmo-nary TB at the same hospital. Four of seven childrenidentified with MDR-TB were part of the latter study.

If these children were excluded from the study, only17 (5.6%) children had INH-resistant and three (1,0%)MDR-TB.

Finally, laboratory inaccuracies could have led tofalse positive drug-resistant results. However, qualityassurance procedures were strictly adhered to, andfurthermore, M. tuberculosis isolates obtained fromfour known MDR-TB adult source cases were identi-cal in drug susceptibility patterns and DNA finger-printing results compared to their childhood con-tacts’ isolates.

Clinical dataClinical data of children with drug-susceptible TB,drug-resistant TB and those in whom susceptibilitytesting was not done are compared and summarizedin Table 2.

In 138 (46%) of 297 children an adult pulmonaryTB source case could be identified. Specific attentionwas paid to source cases of the 21 children with drug-resistant TB. Six (28%) children (four from the MDRcontact study) had a known drug-resistant sourcecase, five of whom had the same drug susceptibilitypattern. A further five (24%) children had knownadult source cases, but the drug susceptibility patternsof the adults’ M. tuberculosis strains were not known.Three of the 11 children had additional contact withadults who had drug-susceptible TB. No source caseswere found in nine (43%) children, and one (5%)child could not be traced.

HIV test results were available in 166 (49%) chil-dren, of whom 13 (7.8%) were positive. PulmonaryTB was present in 89% of the children. A diagnosisof pulmonary TB was based on chest radiographchanges and/or culture of M. tuberculosis from gas-tric aspirate, sputum or bronchial aspirate. Extra-pulmonary TB (including miliary TB and pleural effu-sion) was found in 57% of all of the children, and inseven of the 13 (54%) HIV-infected children.

Degree of INH resistanceThe MICs of INH against five of the nine M. tubercu-losis isolates were between 0.2 and 2.0 �g/ml; these

Table 1 Susceptibility test results

�5 years oldn 218

5–13 years oldn 88

All n 306

Susceptible to H*† 202 (92.7%) 83 (94.3%) 285 (93.1%)Resistant to H only 5 4 9Resistant to H � S 3 1 4Resistant to H � E 1 — 1Resistant H � R 2 — 2Resistant to H � R � S 3 — 3Resistant to H � R � S � E � ETH 2 — 2

Total resistant to H 16 (7.3%) 5 (5.7%) 21 (6.9%)Total MDR 7 (3.2%) — 7 (2.3%)

* 56 (20%) also tested for R susceptibility; all were susceptible.† 24 (8%) also tested for S and E susceptibility; all were susceptible.H isoniazid; R rifampicin; S streptomycin; E ethambutol; ETH ethionamide.


results imply that the specific strains were moderatelysusceptible to the drug. However, as MIC values 20�g/ml were observed in the remaining four isolates,this group was considered to be very resistant to INH.

DISCUSSION

Surveillance of anti-tuberculosis drug resistance in acommunity provides a measure of the success of theNTP and gives an indication of suitable drug regi-mens for future use.13 Primary or initial resistancepatterns reflect transmission of drug-resistant strains,but as tuberculosis in children is a mark of recent TBtransmission in a community, the frequency of drugresistance in children, particularly those �5 years ofage, reflects a precise evaluation of the current situa-tion.4,13 Few surveillance studies are available, how-ever, on drug resistance in childhood TB.

This was a hospital-based study, and therefore rep-resents mainly children referred for secondary or ter-tiary care problems. Specimens for culture of M.tuberculosis are difficult to obtain in children, how-ever, and will often be done only when children arehospitalised. The most comprehensive study to date,by Steiner et al., was also a hospital-based study con-ducted over a 24-year period (1961–1984) in KingCounty Hospital Center, New York.5,14 Of 374 chil-dren screened for primary drug resistance during this

period, 16.3% had resistance to one or more anti-tuberculosis drugs. INH resistance was relatively sta-ble at 10%, but RMP resistance, which was intro-duced during this period, was on the increase.5 As faras we could determine, no studies on the prevalenceof drug resistance in children from developing coun-tries are available.

Isolated RMP resistance is reported to be rare.1,15,16

Weyer et al. showed initial RMP resistance and multi-drug resistance both to be 1.1% in adult patients intheir survey,1 and Bohmer et al. found two of 808(0.2%) cases with monoresistance to RMP in thegreater Cape Town area in 1990–1991.17 Due to costconstraints we therefore tested initially for INH sus-ceptibility only; none of the 20% M. tuberculosisstrains which were susceptible to INH, and weretested for RMP susceptibility, was resistant to RMP.

Initial INH resistance is a sensitive indicator of theoverall success of a treatment programme in a coun-try.1 Failure to control its availability adequately willsoon lead to the appearance of INH resistance. Theprevalence of primary INH resistance in this study isbelow the median reported for the 35 countries in theglobal surveillance for anti-tuberculosis drug resis-tance from 1994 to 1997 (7.3%).18 This prevalencenevertheless still requires the use of a four-drug regi-men in newly diagnosed cases of adult type pulmo-nary TB, as the level of primary INH resistance is

Table 2 Clinical data and special investigations

Drug susceptibility results

Susceptible to H*†

n 285

Resistant to H � other drugs*

n 21Not donen 32

Age (median [range]) in years 2.64 (0.15–13) 2.67 (0.22–10.3) 2.43 (0.06–12.3)Sex: male 157 (55%) 14 (67%) 22 (69%)Previous anti-tuberculosis drugs

(including defaulters) 28 (10%) 2 (10%) 2 (6%)Household TB contact 116/248 (47%) 11/20 (55%) 11/29 (38%)Weight for age �3rd percentile

(NCHS) 116 (41%) 8 (38%) 11 (34%)Pulmonary TB 248 (87%) 21 (100%) 32 (100%)Extra pulmonary TB (including pleural

effusion and miliary TB) 171 (60%) 8 (38%) 15 (47%)Mantoux test �10 mm 185/222 (83%) 13/18 (72%) 17/22 (77%)Mantoux test �15 mm 168/222 (76%) 10/18 (56%) 17/22 (77%)HIV-seropositive 12/137 (8.8%) 1/20 (5%) 0/9CXR

Done and read 275 (96%) 20 (95%) 32 (100%)Hilar/mediastinal lymphadenopathy 152 (55%) 9 (45%) 9 (28%)Opacification lobar/segmental 112 (41%) 8 (40%) 12 (38%)Bronchopneumonic opacification 43 (16%) 3 (15%) 4 (13%)Perihilar opacification 27 (10%) 4 (20%) 6 (19%)Cavities 32 (12%) 3 (15%) 0Miliary 24 (9%) 1 (5%) 5 (16%)Pleural effusion 30 (11%) 2 (10%) 6 (19%)Calcification 12 (4%) 0 0Normal 40 (15%) 4 (20%) 2 (6%)

Died in hospital 11 (4%) 0 3 (9%)

* The differences between drug-susceptible and drug-resistant groups were not statistically significant (P � 0.05).† Of 56 (20%) children tested for susceptibility to RMP, all were susceptible. As primary RMP monoresistance in West-ern Cape is 0.0–0.2%, RMP resistance is unlikely.H isoniazid; NCHS US National Center for Health Statistics; HIV human immunodeficiency virus; CXR chestX-ray; RMP rifampicin.


more than the cut-off of 4% recommended by the USCenters for Disease Control and Prevention.19 TheWestern Cape Province had an initial INH resistancerate of 3.9% among adult pulmonary tuberculosispatients during 1992–1993.1 The increase found withthe current survey may be due to several factors.Firstly, age is said to be inversely related to drug resis-tance. A survey of drug resistance in the USA duringthe first quarter of 1991 showed a decreasing inci-dence of drug resistance with increasing age, with21.6% of children �14 years of age resistant to oneor more drugs compared to 9.4% adults �65 years ofage.6 Age was shown not to be significantly associ-ated with drug resistance in developing countries,however.13,20 Secondly, hospital-based studies arebiased towards patients referred with problems suchas possible drug resistance, and will therefore proba-bly predict the worst case scenario.21 When theknown study limitations were removed, the preva-lence of INH and MDR is not different from theresults obtained in the adult study. Thirdly, primaryresistance may be higher than determined by Weyer etal.,1 as some of the patients in the acquired resistancegroup may well have had primary drug-resistant TB.3Lastly, the possibility exists that the increase is a trueincrease in prevalence of drug resistance in our com-munity. Our results suggest that the observed inci-dence of drug resistance in childhood tuberculosis,even though determined in a hospital-based popula-tion, is a good reflection of the incidence of primarydrug resistance in a community.

The clinical presentation and chest radiographresults among those infected with a resistant strainwere not significantly different from those infectedwith a drug-susceptible strain (Table 2). This is inagreement with findings from previous studies.14,22

An unexpected finding was the high rate of extra-pulmonary tuberculosis (57% overall), which ishigher than the expected 25%–35% previouslydescribed in children.23,24 This can be explained bythat fact that more complicated cases of suspectedtuberculosis are referred to our hospital, and thatthere is a higher yield from tissue cultures than frompulmonary sources.15 Eighty-nine per cent of all ourpatients did, however, have pulmonary tuberculosison chest radiography and/or culture of M. tuberculo-sis from gastric or bronchial aspirate or sputum.

The use of INH for the treatment of INH-resistantM. tuberculosis infection is controversial.7 A distinc-tion could probably be made between the use of INH inprimary or acquired (retreatment) cases of INH resis-tance. Canetti25 and Tripathy et al.8 have both shownlarger numbers of low level INH-resistant strains(MIC �5 �g/ml) in primary resistance compared toacquired resistance. Several studies have shown somebenefit by adding INH to the treatment regimen.Devadatta et al.9 and Tripathy et al.8 showed an ini-tial improvement in patients with primary INH resis-

tance treated with INH alone or with PAS, and Pettyand Mitchell26 demonstrated significant benefit by theaddition of high dose INH to a regimen of ethiona-mide and pyrazinamide in the retreatment of INH-resistant cases. An International Union AgainstTuberculosis (IUAT) study, however, indicated noadvantage in adding INH in conventional doses to aretreatment regimen for INH-resistant cases.27 Morerecently, a mouse model study suggests that INH maybe useful in combination therapy of M. tuberculosisinfection caused by low-level INH-resistant organ-isms (INH MIC 0.2–5 �g/ml).7 Our data showed thatfive of nine (56%) strains from children with primarydrug-resistant tuberculosis had low level INH resis-tance. This is similar to findings in previous studies inprimary INH-resistant strains.8,27 These concentra-tions are attainable in children at normal to high dos-ages (10–20 mg/kg/day) of INH. Further studies toevaluate a possible role of INH in primary INH-resis-tant patients are needed.

In conclusion, our data suggest that the incidenceof drug resistance in childhood tuberculosis in theWestern Cape is low, and probably reflects the level ofprimary drug resistance amongst organisms currentlycirculating in the community.

AcknowledgementsThe authors would like to thank the Medical Research Council ofSouth Africa for financial assistance, Ms Sophia Carlini for sub-culturing of M. tuberculosis organisms, and Dr Etienne Nel for sta-tistical advice.

This article was written in partial fulfilment of a registeredMD-thesis.

References1 Weyer K, Groenewald P, Zwarenstein M, Lombard C J. Tuber-

culosis drug resistance in the Western Cape. S Afr Med J 1995;85: 499–504.

2 Horne N W. Drug-resistant tuberculosis: a review of the worldsituation. Tubercle 1969; 50 (Suppl): 2–12.

3 Van Rie A, Warren R, Richardson M, et al. Classification ofdrug-resistant tuberculosis in an epidemic area. Lancet 2000;356: 22–25.

4 Rieder H L. Drug-resistant tuberculosis: issues in epidemiologyand challenges for public health. Tubercle Lung Dis 1993; 75:321–323.

5 Steiner P, Rao M, Mitchell M, Steiner M. Primary drug-resistanttuberculosis in children. Emergence of primary drug-resistantstrains of M. tuberculosis to rifampin. Am Rev Respir Dis1986; 134: 446–448.

6 Bloch A B, Cauthen G M, Onorato I M, et al. Nationwide sur-vey of drug-resistant tuberculosis in the United States. JAMA1994; 271: 665–671.

7 Cynamon M H, Zhang Y, Harpster T, Cheng S, DeStefano M S.High-dose isoniazid therapy for isoniazid-resistant murineMycobacterium tuberculosis infection. Antimicrob AgentsChemother 1999; 43: 2922–2924.

8 Tripathy S P, Menon N K, Mitchison D A, et al. Response totreatment with isoniazid plus PAS of tuberculosis patientswith primary isoniazid resistance. Tubercle 1969; 50: 257–268.

9 Devadatta S, Bhatia A L, Andrews R H, et al. Response of pa-


tients infected with isoniazid-resistant tubercle bacilli to treat-ment with isoniazid plus PAS or isoniazid alone. Bull WorldHealth Organ 1961; 25: 807–829.

10 Schaaf H S, Vermeulen H A S, Gie R P, Beyers N, Donald P R.Evaluation of young children in household contact with adultmultidrug-resistant pulmonary tuberculosis cases. Pediatr In-fect Dis J 1999; 18: 494–500.

11 Harries A D, Maher D. Diagnosis of tuberculosis in children.In: TB/HIV: a clinical manual. WHO/TB/96.200. Geneva;WHO, 1996: 61–68.

12 Smuts N A, Beyers N, Gie R P, et al. Value of the lateral chestradiograph in tuberculosis in children. Pediatr Radiol 1994;24: 478–480.

13 Glynn J R, Jenkins P A, Fine P E M, et al. Patterns of initial andacquired antituberculosis drug resistance in Karongo District,Malawi. Lancet 1995; 345: 907–910.

14 Steiner P, Rao M, Victoria M S, Hunt J, Steiner M. A continu-ing study of primary drug-resistant tuberculosis among chil-dren observed at the Kings County Hospital Medical Centerbetween the years 1961 and 1980. Am Rev Respir Dis 1983;128: 425–428.

15 Hurley J C, Andrew J H. Bacteriology and drug susceptibilityof tuberculosis at St Vincent’s Hospital, Melbourne 1962–1991. Tubercle Lung Dis 1993; 74: 163–166.

16 Swanson D S, Starke J R. Drug-resistant tuberculosis in pedi-atrics. Pediatr Clin North Am 1995; 42: 553–581.

17 Bohmer P D, MacNab M F, Coetzee G J. Primary drug resis-tance in Mycobacterium tuberculosis in the Greater CapeTown area. S Afr Med J 1992; 81: 382–383.

18 Pablos-Méndez A, Raviglione M C, Laszlo A, et al. Global sur-

veillance for antituberculosis-drug resistance, 1994–1997. NEng J Med 1998; 338: 1641–1649.

19 American Thoracic Society/Centers for Disease Control andPrevention. Treatment of tuberculosis and tuberculosis infec-tion in adults and children. Am J Respir Crit Care Med 1994;149: 1359–1374.

20 Grandes G, Lopez-de-Munain J, Diaz T, Rullan T V. Drug-resistant tuberculosis in Puerto Rico, 1987–1990. Am RevRespir Dis 1993; 148: 6–9.

21 Nunn P, Felten M. Surveillance of resistance to antituberculosisdrugs in developing countries. Tubercle Lung Dis 1994; 75:163–167.

22 Debré R, Noufflard H, Brissaud H E, Gerbeaux J. Infection ofchildren by strains of tubercle bacilli initially resistant to strep-tomycin or to isoniazid. Am Rev Respir Dis 1959; 80: 326–339.

23 Rieder H L, Snider D E Jr, Cauthen G. Extrapulmonary tuber-culosis in the United States. Am Rev Respir Dis 1990; 141:347–351.

24 Biddulph J. Short course chemotherapy for childhood tubercu-losis. Pediatr Infect Dis J 1990; 9: 794–801.

25 Canetti G. Present aspects of bacterial resistance in tuberculo-sis. Am Rev Respir Dis 1965; 92: 687–703.

26 Petty T L, Mitchell R S. Successful treatment of advanced iso-niazid and streptomycin-resistant pulmonary tuberculosis withethionamide, pyrazinamide and isoniazid. Am Rev Respir Dis1962; 86: 503–512.

27 International Union Against Tuberculosis. A comparison ofregimens of ethionamide, pyrazinamide and cycloserine in re-treatment of patients with pulmonary tuberculosis. Bull IntUnion Tuberc 1969; 42: 7–57.

R É S U M É

CADRE : La province du Cap Ouest d’Afrique du Sud,une région à haute incidence de tuberculose (TB) où larésistance initiale à l’isoniazide (INH) et la multirésis-tance (MR) chez les adultes atteignent respectivement3,9% et 1,1% pendant la période 1992–1993.OBJECTIF : Déterminer l’incidence de la résistance auxmédicaments parmi les enfants par comparaison avecles adultes, comparer les caractéristiques cliniques destuberculoses à germes résistants ou à germes sensibles etétablir le degré de résistance à l’INH dans les isolatsisoniazido-résistants.MÉTHODES : Toutes les cultures de Mycobacteriumtuberculosis provenant d’enfants (0–13 ans) dans unhôpital régional ont été colligées prospectivement entreaoût 1994 et avril 1998 et les tests de sensibilité réaliséssur les échantillons de chaque enfant. Le degré de résis-tance à l’INH a été déterminé dans les isolats résistants

disponibles. On a passé en revue les dossiers cliniquesdes enfants.RÉSULTATS : Les tests de sensibilité ont pu être obtenuschez 306 des 338 enfants dont les cultures de M. tuber-culosis étaient positives. Une résistance à l’INH a étéobservée dans 21 isolats (6,9%) ; il y avait sept cas deMR. Si l’on prend en compte les limitations de cetteétude, l’incidence de la résistance à l’INH est de 5,6%et celle des MR de 1% chez les enfants âgés de moins de5 ans. Le tableau clinique était similaire que les enfantssoient atteints de tuberculose à germes sensibles ourésistants.CONCLUSION : L’incidence de la tuberculose à germesrésistants chez l’enfant dans la province du Cap Ouestest faible, et pourrait indiquer l’incidence actuelle de larésistance médicamenteuse primaire dans la collectivité.

R E S U M E N

MARCO DE REFERENCIA : La provincia Western Capede Sud Africa, un área con una alta incidencia de tuber-culosis (TB), donde la resistencia inicial a la isoniacida(INH) y la multirresistencia (MDR) en los adultos fuede 3,9% y de 1,1%, respectivamente, entre 1992 y1993.OBJETIVO : Determinar la incidencia de la drogorresis-tencia en los niños comparada con la de los adultos,comparar los datos clínicos de las TB resistentes y de las

TB sensibles y el grado de resistencia a la INH en lascepas isoniacido-resistentes.MÉTODOS : Desde agosto de 1994 hasta abril de 1998 seestudiaron prospectivamente todos los cultivos deMycobacterium tuberculosis obtenidos de niños (0–13años) en un hospital regional, así como los tests de sen-sibilidad efectuados en cada muestra. En las cepas resis-tentes a la INH se determinó el grado de resistencia a lamisma. Se revisaron las historias clínicas de los niños.


RESULTADOS : Se analizaron los tests de sensibilidad en306 de 338 niños con cultivos de M. tuberculosis. Veinti-una cepas (6,9%) eran isoniacido-resistentes ; siete eranMDR. Teniendo en cuenta las limitaciones de este estu-dio, la incidencia de isoniacido-resistencia fue del 5,6%y la de MDR del 1%, en los niños menores de 5 años. Las

características clínicas fueron similares en los niños con TBsensibles a las drogas y con drogorresistencia.CONCLUSIÓN : La incidencia de la drogorresistenciainfantil en la provincia Western Cape es baja, y puedereflejar la incidencia de la drogorresistencia primaria enesta comunidad.

Documents

Primary Drug-resistant Tuberculosis in Children