Six-Month Isoniazid-Rifampin Treatment for Pulmonary Tuberculosis in Children1,2
FRANCISCO J. c. REIS, MARIA B. M. BEDRAN, JOSE A. R. MOURA, IRMGARD ASSIS, and MARY E. S. M. RODRIGUES
Introduction
Short-course chemotherapy represents an important advance in the treatment of tuberculosis (TBC) in the past few years. For 30 yr, the treatment of active TBC involved the use of two or three drugs daily for 18 to 24 months, usually isoniazid (INH) plus para-aminosalicylic acid (PAS), and streptomycin (SM) (1). This regimen was quite efficient, but often the drugs were irregularly used and patient compliance was poor. Since the introduction of rifampin (RIF) as an antituberculous drug (2-7), it has been possible to shorten the duration of the treatment. Rifampin and isoniazid have in vitro bactericidal effects on different populations of bacilli found in tuberculous lesions, including bacilli in an active or latent phase of multiplication and independent of extracellular or intracellular location (3, 4, 8-10). Short-course therapy in bacillary positive adults had shown a cure rate of up to 95070 (5, 10, 11). In children, tuberculous lesions are less extensive and contain less bacilli than adults. These shortcourse regimens, together with better tolerance to antituberculous drugs, seem to be sufficient to treat childhood tuberculosis (12, 13). Our hypothesis was that short-course daily chemotherapy for 6 months with two drugs (RIF and INH) would be an effective regimen for treatment of pulmonary tuberculosis in children. To our knowledge, this regimen has not been previously tested in children. Methods From July 1979to December 1988, 152 chilren (under 15 yr of age) were diagnosed as having pulmonary TBC in the Pediatric Respirology Section of the Hospital das Clinicas of Belo Horizonte, Universidade Federal de Minas Gerais, Brazil. The diagnosis was based on a chest roentgenogram that showed parenchymal or mediastinal lymph nodes suggestive of tuberculosis plus two or more of the followingcriteria: (1) suggestivesymptoms 996
SUMMARY One hundred and seventeen children with pulmonary tuberculosis underwent treatment with a 6-month dally regimen of rlfampln (15 mg/kgJday) and Isoniazid (10 mglkg/day). The criteria for the diagnosis of pUlmonary tuberculosis were (1)clinical symptoms and signs In 93 children (79%), (2) history of direct contact with an adult with tuberculosis In 106 children (91%), (3) tuberculin reaction of 5 mm or more, without previous bacillus Calmette-Gu'rln (BeG), In 45 children (38%), (4) suggestive radiologic alterations In all patients, and (5) positive bacteriology or histology In four patients (3%). The treatment was completed by W children (83%). The mean weight gain during therapy was 2,145g. There was an excellent cllnlcoradlologlc I'8Sponseto the tl'88tment, and Improvement In chest roentgenograms was observed In all patients at the end of therapy. No relapses occurred among the patients followed for an average of 21.4 months. This study Indicates that the treatment of primary pulmonary tuberculosis In children with a combination of rlfamph, and Isoniazid dally for 6 months Is efficacious and does not result In any relapse. AM REV RESPIR DIS 1990; 142:996-999
and signs (clinical), (2) direct contact with a tuberculous adult, bacillary positive or negative (epidemiologic), (3) tuberculin test (pPD RT-23 2UT) positive (5 mm induration or greater), without previous bacillus CalmetteGuerin history or a scar (immunologic), and (4) acid-fast bacilli in the sputum or gastric washings or the presence of specific granuloma in any histologic specimen (histobacteriologic), From the 152 children initially diagnosed, 117 werereceiving their first treatment for pulmonary tuberculosis and all had normal values of liver transaminases (SOOf, SGPn. The treatment used was RIF and INH daily with doses of 10to 15mg/kg/day and a maximum dose of 600 and 400 mg, respectively. The great majority of the children were treated on an outpatient basis, and the use of the medications was checked at each visit. Patients were evaluated before entering the study and routinely at the end of the first, third, and sixth months (final) of the treatment or at any time if necessary. Further follow-up was done at 6, 18, 30, 42, and 54 months after the end of the treatment. The nutritional status was recorded at the first ex-amination according to the following criteria: (1) well nourished - within 10070 of ideal weight (fiftieth percentile), (2) first-degree malnutrition - a deficit of 10to 25070 of ideal weight, (3) second-degree malnutrition - a deficit of 25 to 40070 of ideal weight, and (4) third-degree malnutrition - a deficit of more than 40070 of ideal weight. A complete blood count (CBC) was done before treatment. The serum levels of transaminases (SGOT and
SGPT) were recorded and had to be normal for admission to the study. Liver enzyme studies were repeated at the end of the first, third, and the sixth month of treatment. Chest roentgenograms were done at this same interval during treatment. We used the following criteria for exclusion from the study: (1) irregular use of the medications with an interruption for a period > 15 days, (2) changes in address so the patient could not be located, (3)a change from the initial regimen by other health service personnel, which was independent of a worsening in clinicoradiologic condition, and (4) patients who died from a nontuberculous cause.
Results
There were 63 males and 54 females (male:femaleratio of 1.17:1) ranging from 6 months to 15yr of age. The most commonly affected group was that between 2 and 7 yr of age (45.3070). Malnutrition,
(Receivedin originalform December22, 1989 and in revised form June 14, 1990) 1 From the Department of Pediatrics, University of Minas Gerais (UFMG), Section of Pediatric Respirology,Hospital das Clinicas, Belo Horizonte, Minas Gerais, Brazil. 2 Correspondence and requests for reprints should be addressed to Francisco J. C. Reis, M.D., Associate Professor of Pediatrics, Head, Section of Pediatric Respirology,Rua Ceara 161 gr. 103/106, 30.150 Belo Horizonte-MG-Brazil.
997
TREATMENT OF CHILDHOOD TUBERCULOSIS
patient died from retinoblastoma. Four families had moved and could not be located, and the treatment in four children was changed from the standard regimen after visiting the medical personnel at the official health service, who routinely add pyrazinamide for 2 months. Eleven children were excluded because of interruption of more than 15 days in their treatment. At 6 months there was radiologic improvement in all patients. X-raystudies of the chest were completely normal in 21 (21.6070) children and presented calcifications in 31 (32070). Mediastinalnodes persisted until the end of treatment in 42 children, and residual parenchymal lesions persisted in 12 patients (table 5). Three patients (2.6070) had side effects from the medications. All three had vomiting and two 'had abdominal pain. These side effects disappeared after reduction in the dose of medications to 5 to 10mg/kg/day, Serum levels of soar and SGPT were recorded during and at the end of the treatment in 77 children. An increase of less than twice from the initiallevels were observed in 23 patients. They were asymptomatic and the serum levels of SGar and SGPT returned to normal even with continuation of the treatment (table 7). At the end of the treatment, there was a mean weight gain of 2,145 g, ranging
TABLE 1 CHARACTERISTICS OF 117 CHILDREN WITH PULMONARY TUBERCULOSIS Number Male Female 6 months to 2 yr 2 yr, 1 month to 7 yr 7 yr, 1 month to 15 yr Normal Malnourished First degree Second degree Third degree
Sex Age
Nutrition
Total
mainly of the first degree, was present in 67 (57070) children (table 1). Twenty-four children (20.5070) had no respiratory symptoms at the time of the diagnosis. When symptoms werepresent, the most prominent werecough (59.0070), afternoon fever (28.2070), and weight loss in 32 out of 117 children (table 2). A history of contact with a tuberculous adult was present in 106children (90.6070), mainly inside their home (64 children). Of 106 children, 58 (54.7070) had contact with sputum-positive tuberculous adults (table 3). The tuberculin test was negative « 5 mm of induration) in three patients, two of whom had not been vaccinated with BeG. One hundred and ten (94070) patients were strong reactors (10 mm or greater), and in 91 (77.8070) the tuberculin test was 15 mm or greater (table 4). All patients had suggestive radiologic changes of tuberculosis. Enlarged hilar, parahilar, and paratracheallymph nodes were frequently involved (80.3070) followed by parenchymal consolidation in 43.6070 of the patients. In 15 children we observed partial hilar or parenchymal calcification, but they also had signs of current tuberculous infectious activity such as enlarged lymph nodes, lung densities, and clinical symptoms (table 5).
Percentage
63
53.8
54
46.2
28 53 36 50
23.9 45.3 30.8 42.7
34 24 9
29.0 20.5 7.8
117
100.0
No cases of miliary or cavitary tuberculosis were seen. Weconfrrmed tuberculosis in four children by performing a histologic or bacteriologic examination. Three children had tuberculous granulomas with acidfast bacilli on pulmonary biopsy, and the fourth had positive culture from gastric washings. The relative importance of the criteria utilized for the diagnosis oftuberculosis in our patients is shown in table 6. The tuberculin test criterion was used in 45 patients (38.5%) and histology and bacteriology in four children (3.4070). Ninety-seven children were compliant until the completion of treatment. One
TABLE 3 HISTORY OF CONTACT WITH TUBERCULOUS ADULTS
Negative Positive Location of contact Inside home Outside home (within family) Outside home (neighborhood) Bacteriologic condition of contacts Bacillary Nonbacillary No information
Number
Percentage
11 106
90.6
45
54.7 2.8 42.5
106
100.0
3
Total
TABLE 4 RESULTS OF TUBERCULIN TESTS
Cough Afternoon fever Weight loss Anorexia Increased mucus production Frequent colds Excessive sweating Other Asymptomatic
69
33 32 23 21 10 7 25 24
Percentage 59.0 28.2 27.4 19.7 17.9 8.5 6.0 21.4 20.5
Induration (mm) 0-4
5-9 10-14 15-19 20 or more Total
(n
No.
With BCG (n = 70)
Without BCG (n = 47)
Total = 117) Percentage
13.2
58
TABLE 2
Number
60.4 26.4
64 28 14
SIGNS AND SYMPTOMS Symptom
9.4
No.
Percentage
No.
Percentage
2 3
1 1 10 31 27
0.8 0.8 8.5 26.6 13.1
70
59.8
3 4 19 50 41
2.6 3.4 16.2 42.8 35.0
19 14
1.7 2.6 7.7 16.2 12.0
117
100.0
47
40.2
9
998
REIS, SEDRAN, MOURA, ASSIS, AND RODRIGUES
TABLE 5 RADIOLOGIC FINDINGS After &-Month Treatment (n == 97)
At Diagnosis (n ==
11n
Radiologic
No.
Percentage
No.
Percentage
Mediastinal nodes Parenchymal lesions Calcification Pleural effusion Collapse and cavitation Miliary lesions Fibrosis Normal
94 51 15 5 0 0 1 0
80.3 43.6 12.8 4.3 0.0 0.0 0.9 0.0
42 12 31 0 0 0 3 21
43.3 12.4 32.0 0.0 0.0 0.0 3.1 21.6
TABLE 6 CRITERIA FOR THE DIAGNOSIS OF CHILDHOOD PULMONARY TUBERCULOSIS Criteria Clinical (signs and symptoms) Epidemiologic Immunologic (tuberculin test) Radiologic Bacteriologic and histologic
from 800 to 4,500 g. The mean followup time after the end of treatment was 21.4 months (6 to 57 months). There was no recurrence of tuberculosis during this period of follow-up judged by clinical and radiologic criteria. Discussion
Short-course regimens for treatment of bacillary pulmonary tuberculosis in adults had been proved to be efficient in sterilization of the sputum and reduction of recurrence rate of the disease. Drugs have been administered daily or intermittently for 3 to 9 months (13-17). RIF and INH have better action with extracellular bacillary populations in comparison with its action on intracellular bacilli (4, 10, 14, 18). It has been shown that the association of INH and RIF for a period of 6 to 9 months of treatment has a cure rate of up to 95% of bacillary adults (5, 10, 11).This 9-month regimen has also been utilized for treatment of pulmonary tuberTABLE 7 SIDE EFFECTS OF DRUG THERAPY Side Effect Vomiting Abdominal pain Peripheral neuropathy Nausea Othert
Number·
Percentage
3 2 1 1 0
2.6 1.7 0.9 0.9 0.0
• Total number of patients with side effects, n = 3.
t Jaundice, diarrhea, somnolence, and melena.
Number
Percentage
93 106 45 117 4
79.5 90.6 38.5 100.0 3.4
culosis in children with good results and tolerance (4, 13, 18). It is common in children to have natural spontaneous improvement after primary infection. Nevertheless, we usually establish the diagnosis of primary pulmonary tuberculosis (disease) only in those children who persist with the abnormal radiologic findings (progressive primary pulmonary tuberculosis). The short-course regimens present many advantages: reduction of total cost of the treatment, including number of doctor's and nurse's visits and decreased cost of medications, and lesstoxicity than observed with the prolonged use of standard medications. The majority of reactions occur at the beginning of treatment. Compliance is better and parents normallyare more cooperative in their appointments. If there is an interruption of treatment, even of more than a couple of months after 3 or 4 months of treatment, the chances of maintaining negative cultures in adults are better (1). With recurrent disease, the same regimen can be used again with minimal risk ofresistance (4, 11). Usually, recurrences are caused by sensitive but persistent bacilli that are not accessible to the medications. Therefore, with these short-course regimens, it is necessary to use bactericidal drugs that will avoid the appearance of resistant mutants (1, 5, 14, 15). Weobserved a high percentage (79.3070) of symptomatic children. Abernathy and
colleagues (13) reported the presence of symptoms in only 32070 of their patients. They detected tuberculosis by mainly using tuberculin test surveys for children in contact with bacillary tuberculous adults, independent of respiratory symptoms. Our outpatient clinic is a referral center in our state, and children are referred for the assessment and diagnosis of symptomatic pulmonary diseases. We observed negative tuberculin tests in three and weak reactions in four children. Five of them had never had BCG vaccination before. All of them had malnutrition of the second and third degree. It has been shown that 5 to 10070 ofbacillary tuberculous patients never react to the tuberculin test (19). Approximately 800/0 of the children (91/117) had a tuberculin test of 15mm of induration or greater. Therefore, this parameter seems to be very important for the diagnosis of tuberculosis in infancy, even in those children already vaccinated with BCG. After BCG vaccination, only 2% of children have a tuberculin test greater than 15 mm of induration (20). However, the tuberculin test along with a positive chest roentgenogram were utilized as the criteria for the diagnosis of tuberculosis in our study in only 45 children (38.50/0), with no previous BCG history and/or a definitive BCG scar, because of the high rate of patients with previous BCG. Our limitation for better documentation on the bacteriologic criterion (table 6) was due to the fact that almost all patients were diagnosed at the outpatient care unit. It is very difficult to assessthe outcome and the efficacy of any regimen for treatment of tuberculosis in children because they rarely have positive sputum and gastric washings, and the best criteria would be clinical findings, such as weight gain, and radiologic follow-up studies (18,21). We observed in our patients that the weight gain was twice as much as that expected for a period of 6 months in a normal child. The use of a radiologic criterion alone may present some difficulties because of residual lymph node or parenchymal lesions that may be present at the end of the treatment (43% of the children). These residual markings have been observed by other authors, and they usually disappear at the end of 2 or 3 yr (13, 18).The initial fear about the association of two hepatotoxic drugs (RIF and INH) in children weredissipated with the progression of the study. These synergistic effects are usually dose related (22). The noncompliance rate of 17% is con-
999
TREATMENT OF CHILDHOOD TUBERCULOSIS
sistent with previous reports in the literature (12, 13). Conclusion
Ninety-seven children with primary pulmonary tuberculosis weretreated with rifampin and isoniazid daily for 6 months. Clinical and radiologic follow-up for a period of 21.4months demonstrated the absence of recurrence of disease. The combined use of two potentially hepatotoxic drugs was not associated with clinically significant hepatic disease. Weconclude that primary pulmonary tuberculosis in children would be successfully treated with daily doses of rifampin and isoniazid for 6 months. Acknowledgment The writers thank Dr. Victor Chernick for invaluable advice and Angela Brito and Shirley Alton for the typing and preparation of this manuscript. References 1. Fox W, Mitchison DA. Short-course chemotherapy for pulmonary tuberculosis. Am Rev Respir Dis 1975; 111:325-53. 2. Verbist L, Gyslen A. Antituberculous activity
of rifampin in vitro and in vivo and the concentrations attained in human blood. Am Rev Respir Dis 1968; 98:923-32. 3. Grumbach FM, Rist N. Aetivite antituberculeuse experimentale de la rifampicine derivee de la rifamicine SV. Revue Tuberc (Paris) 1967; 31: 749-62. 4. Stead WW, Dutt AK. Chemotherapy for tuberculosis today. Am Rev Respir Dis 1982; 125:94-101. 5. British Medical Research Council. Controlled clinical trials of four six month regimens of chemotherapy of pulmonary tuberculosis (Second East African Study). Am Rev Respir Dis 1976; 114:471-5. 6. British Thoracic and Tuberculosis Association. Short-course chemotherapy in pulmonary tuberculosis. Lancet 1976; 2:1102-4. 7. Dutt AK, Stead WW. Short-course treatment regimens for patients with tuberculosis. Arch Intern Med 1980; 140:827-9. 8. Batten J. Experimental chemotherapy for tuberculosis. Br Med J 1968; 3:75-82. 9. Kendig EL. Evolution of short-course antimicrobial treatment of tuberculosis in children 1951-1984. Pediatrics 1985; 74:684-6. 10. Grosset J. Bacteriologic basis for short-course chemotherapy for tuberculosis. Clin Chest Med 1980; 1:231-42. . 11. Dutt AK, Stead WW. Short-course chemotherapy: the Arkansas experience. Chest 1981;80:724-7. 12. Ibanez S, Ross BG. Quimioterapia abreviada de 6 meses en tuberculosis pulmonar infantil. Rev Chil Pediatr 1980; 51:249-52. 13. Abernathy RS, Dutt AK' Stead WW, Moers DJ. Short-course chemotherapy for tuberculosis in
children. Pediatrics 1983; 72:801-6. 14. Fox W. The current status of short course chemotherapy. Tubercle 1979; 60:177-90. 15. Hong Kong Tuberculosis Treatment Services and British Medical Research Council. Controlled trial of 6 and 9 month regimens of daily and intermittent streptomycin plus isoniazid plus pyrazinamide for pulmonary tuberculosis in Hong Kong. Thbercle 1975; 56:81-96. 16. 'Iuberculosis Research Institute Bucharest. Trial of two intermittent short-course regimens (78 doses) in the initial treatment of pulmonary tuberculosis. Tubercle 1977; 58:1-8. 17. Mehrotra ML, Gautam KD, Chaube CK. Shortest possible, acceptable, effective, ambulatory chemotherapy in pulmonary tuberculosis: preliminary report. Am Rev Respir Dis 1981; 124: 239-44. 18. Jacobs RF, Abernathy RS. The treatment of tuberculosis in children. Pediatr Infect Dis 1985; 4:513-7. 19. Kent D, Schwartz R. Active pulmonary tuberculosis with negative tuberculin skin reactions. Am Rev Respir Dis 1967; 95:411-8. 20. ReisFJC. Alergia tuberculinica e cicatriz vacinal em lactentes que tomaram a vacina BCG injetavelquando recemnascidos. J Pediatr (Rio) 1982; 52:23-8. 21. Longman Group. Smear negative pulmonary tuberculosis. Thbercle 1980; 61:113-6. 22. O'Brien RJ, Long MW, Cross FS, Lyle MA, Snider DE. Hepatotoxicity from isoniazid and rifampin among children treated for tuberculosis. Pediatrics 1983; 72:491-9.
FRANCISCO J. c. REIS, MARIA B. M. BEDRAN, JOSE A. R. MOURA, IRMGARD ASSIS, and MARY E. S. M. RODRIGUES
Introduction
Short-course chemotherapy represents an important advance in the treatment of tuberculosis (TBC) in the past few years. For 30 yr, the treatment of active TBC involved the use of two or three drugs daily for 18 to 24 months, usually isoniazid (INH) plus para-aminosalicylic acid (PAS), and streptomycin (SM) (1). This regimen was quite efficient, but often the drugs were irregularly used and patient compliance was poor. Since the introduction of rifampin (RIF) as an antituberculous drug (2-7), it has been possible to shorten the duration of the treatment. Rifampin and isoniazid have in vitro bactericidal effects on different populations of bacilli found in tuberculous lesions, including bacilli in an active or latent phase of multiplication and independent of extracellular or intracellular location (3, 4, 8-10). Short-course therapy in bacillary positive adults had shown a cure rate of up to 95070 (5, 10, 11). In children, tuberculous lesions are less extensive and contain less bacilli than adults. These shortcourse regimens, together with better tolerance to antituberculous drugs, seem to be sufficient to treat childhood tuberculosis (12, 13). Our hypothesis was that short-course daily chemotherapy for 6 months with two drugs (RIF and INH) would be an effective regimen for treatment of pulmonary tuberculosis in children. To our knowledge, this regimen has not been previously tested in children. Methods From July 1979to December 1988, 152 chilren (under 15 yr of age) were diagnosed as having pulmonary TBC in the Pediatric Respirology Section of the Hospital das Clinicas of Belo Horizonte, Universidade Federal de Minas Gerais, Brazil. The diagnosis was based on a chest roentgenogram that showed parenchymal or mediastinal lymph nodes suggestive of tuberculosis plus two or more of the followingcriteria: (1) suggestivesymptoms 996
SUMMARY One hundred and seventeen children with pulmonary tuberculosis underwent treatment with a 6-month dally regimen of rlfampln (15 mg/kgJday) and Isoniazid (10 mglkg/day). The criteria for the diagnosis of pUlmonary tuberculosis were (1)clinical symptoms and signs In 93 children (79%), (2) history of direct contact with an adult with tuberculosis In 106 children (91%), (3) tuberculin reaction of 5 mm or more, without previous bacillus Calmette-Gu'rln (BeG), In 45 children (38%), (4) suggestive radiologic alterations In all patients, and (5) positive bacteriology or histology In four patients (3%). The treatment was completed by W children (83%). The mean weight gain during therapy was 2,145g. There was an excellent cllnlcoradlologlc I'8Sponseto the tl'88tment, and Improvement In chest roentgenograms was observed In all patients at the end of therapy. No relapses occurred among the patients followed for an average of 21.4 months. This study Indicates that the treatment of primary pulmonary tuberculosis In children with a combination of rlfamph, and Isoniazid dally for 6 months Is efficacious and does not result In any relapse. AM REV RESPIR DIS 1990; 142:996-999
and signs (clinical), (2) direct contact with a tuberculous adult, bacillary positive or negative (epidemiologic), (3) tuberculin test (pPD RT-23 2UT) positive (5 mm induration or greater), without previous bacillus CalmetteGuerin history or a scar (immunologic), and (4) acid-fast bacilli in the sputum or gastric washings or the presence of specific granuloma in any histologic specimen (histobacteriologic), From the 152 children initially diagnosed, 117 werereceiving their first treatment for pulmonary tuberculosis and all had normal values of liver transaminases (SOOf, SGPn. The treatment used was RIF and INH daily with doses of 10to 15mg/kg/day and a maximum dose of 600 and 400 mg, respectively. The great majority of the children were treated on an outpatient basis, and the use of the medications was checked at each visit. Patients were evaluated before entering the study and routinely at the end of the first, third, and sixth months (final) of the treatment or at any time if necessary. Further follow-up was done at 6, 18, 30, 42, and 54 months after the end of the treatment. The nutritional status was recorded at the first ex-amination according to the following criteria: (1) well nourished - within 10070 of ideal weight (fiftieth percentile), (2) first-degree malnutrition - a deficit of 10to 25070 of ideal weight, (3) second-degree malnutrition - a deficit of 25 to 40070 of ideal weight, and (4) third-degree malnutrition - a deficit of more than 40070 of ideal weight. A complete blood count (CBC) was done before treatment. The serum levels of transaminases (SGOT and
SGPT) were recorded and had to be normal for admission to the study. Liver enzyme studies were repeated at the end of the first, third, and the sixth month of treatment. Chest roentgenograms were done at this same interval during treatment. We used the following criteria for exclusion from the study: (1) irregular use of the medications with an interruption for a period > 15 days, (2) changes in address so the patient could not be located, (3)a change from the initial regimen by other health service personnel, which was independent of a worsening in clinicoradiologic condition, and (4) patients who died from a nontuberculous cause.
Results
There were 63 males and 54 females (male:femaleratio of 1.17:1) ranging from 6 months to 15yr of age. The most commonly affected group was that between 2 and 7 yr of age (45.3070). Malnutrition,
(Receivedin originalform December22, 1989 and in revised form June 14, 1990) 1 From the Department of Pediatrics, University of Minas Gerais (UFMG), Section of Pediatric Respirology,Hospital das Clinicas, Belo Horizonte, Minas Gerais, Brazil. 2 Correspondence and requests for reprints should be addressed to Francisco J. C. Reis, M.D., Associate Professor of Pediatrics, Head, Section of Pediatric Respirology,Rua Ceara 161 gr. 103/106, 30.150 Belo Horizonte-MG-Brazil.
997
TREATMENT OF CHILDHOOD TUBERCULOSIS
patient died from retinoblastoma. Four families had moved and could not be located, and the treatment in four children was changed from the standard regimen after visiting the medical personnel at the official health service, who routinely add pyrazinamide for 2 months. Eleven children were excluded because of interruption of more than 15 days in their treatment. At 6 months there was radiologic improvement in all patients. X-raystudies of the chest were completely normal in 21 (21.6070) children and presented calcifications in 31 (32070). Mediastinalnodes persisted until the end of treatment in 42 children, and residual parenchymal lesions persisted in 12 patients (table 5). Three patients (2.6070) had side effects from the medications. All three had vomiting and two 'had abdominal pain. These side effects disappeared after reduction in the dose of medications to 5 to 10mg/kg/day, Serum levels of soar and SGPT were recorded during and at the end of the treatment in 77 children. An increase of less than twice from the initiallevels were observed in 23 patients. They were asymptomatic and the serum levels of SGar and SGPT returned to normal even with continuation of the treatment (table 7). At the end of the treatment, there was a mean weight gain of 2,145 g, ranging
TABLE 1 CHARACTERISTICS OF 117 CHILDREN WITH PULMONARY TUBERCULOSIS Number Male Female 6 months to 2 yr 2 yr, 1 month to 7 yr 7 yr, 1 month to 15 yr Normal Malnourished First degree Second degree Third degree
Sex Age
Nutrition
Total
mainly of the first degree, was present in 67 (57070) children (table 1). Twenty-four children (20.5070) had no respiratory symptoms at the time of the diagnosis. When symptoms werepresent, the most prominent werecough (59.0070), afternoon fever (28.2070), and weight loss in 32 out of 117 children (table 2). A history of contact with a tuberculous adult was present in 106children (90.6070), mainly inside their home (64 children). Of 106 children, 58 (54.7070) had contact with sputum-positive tuberculous adults (table 3). The tuberculin test was negative « 5 mm of induration) in three patients, two of whom had not been vaccinated with BeG. One hundred and ten (94070) patients were strong reactors (10 mm or greater), and in 91 (77.8070) the tuberculin test was 15 mm or greater (table 4). All patients had suggestive radiologic changes of tuberculosis. Enlarged hilar, parahilar, and paratracheallymph nodes were frequently involved (80.3070) followed by parenchymal consolidation in 43.6070 of the patients. In 15 children we observed partial hilar or parenchymal calcification, but they also had signs of current tuberculous infectious activity such as enlarged lymph nodes, lung densities, and clinical symptoms (table 5).
Percentage
63
53.8
54
46.2
28 53 36 50
23.9 45.3 30.8 42.7
34 24 9
29.0 20.5 7.8
117
100.0
No cases of miliary or cavitary tuberculosis were seen. Weconfrrmed tuberculosis in four children by performing a histologic or bacteriologic examination. Three children had tuberculous granulomas with acidfast bacilli on pulmonary biopsy, and the fourth had positive culture from gastric washings. The relative importance of the criteria utilized for the diagnosis oftuberculosis in our patients is shown in table 6. The tuberculin test criterion was used in 45 patients (38.5%) and histology and bacteriology in four children (3.4070). Ninety-seven children were compliant until the completion of treatment. One
TABLE 3 HISTORY OF CONTACT WITH TUBERCULOUS ADULTS
Negative Positive Location of contact Inside home Outside home (within family) Outside home (neighborhood) Bacteriologic condition of contacts Bacillary Nonbacillary No information
Number
Percentage
11 106
90.6
45
54.7 2.8 42.5
106
100.0
3
Total
TABLE 4 RESULTS OF TUBERCULIN TESTS
Cough Afternoon fever Weight loss Anorexia Increased mucus production Frequent colds Excessive sweating Other Asymptomatic
69
33 32 23 21 10 7 25 24
Percentage 59.0 28.2 27.4 19.7 17.9 8.5 6.0 21.4 20.5
Induration (mm) 0-4
5-9 10-14 15-19 20 or more Total
(n
No.
With BCG (n = 70)
Without BCG (n = 47)
Total = 117) Percentage
13.2
58
TABLE 2
Number
60.4 26.4
64 28 14
SIGNS AND SYMPTOMS Symptom
9.4
No.
Percentage
No.
Percentage
2 3
1 1 10 31 27
0.8 0.8 8.5 26.6 13.1
70
59.8
3 4 19 50 41
2.6 3.4 16.2 42.8 35.0
19 14
1.7 2.6 7.7 16.2 12.0
117
100.0
47
40.2
9
998
REIS, SEDRAN, MOURA, ASSIS, AND RODRIGUES
TABLE 5 RADIOLOGIC FINDINGS After &-Month Treatment (n == 97)
At Diagnosis (n ==
11n
Radiologic
No.
Percentage
No.
Percentage
Mediastinal nodes Parenchymal lesions Calcification Pleural effusion Collapse and cavitation Miliary lesions Fibrosis Normal
94 51 15 5 0 0 1 0
80.3 43.6 12.8 4.3 0.0 0.0 0.9 0.0
42 12 31 0 0 0 3 21
43.3 12.4 32.0 0.0 0.0 0.0 3.1 21.6
TABLE 6 CRITERIA FOR THE DIAGNOSIS OF CHILDHOOD PULMONARY TUBERCULOSIS Criteria Clinical (signs and symptoms) Epidemiologic Immunologic (tuberculin test) Radiologic Bacteriologic and histologic
from 800 to 4,500 g. The mean followup time after the end of treatment was 21.4 months (6 to 57 months). There was no recurrence of tuberculosis during this period of follow-up judged by clinical and radiologic criteria. Discussion
Short-course regimens for treatment of bacillary pulmonary tuberculosis in adults had been proved to be efficient in sterilization of the sputum and reduction of recurrence rate of the disease. Drugs have been administered daily or intermittently for 3 to 9 months (13-17). RIF and INH have better action with extracellular bacillary populations in comparison with its action on intracellular bacilli (4, 10, 14, 18). It has been shown that the association of INH and RIF for a period of 6 to 9 months of treatment has a cure rate of up to 95% of bacillary adults (5, 10, 11).This 9-month regimen has also been utilized for treatment of pulmonary tuberTABLE 7 SIDE EFFECTS OF DRUG THERAPY Side Effect Vomiting Abdominal pain Peripheral neuropathy Nausea Othert
Number·
Percentage
3 2 1 1 0
2.6 1.7 0.9 0.9 0.0
• Total number of patients with side effects, n = 3.
t Jaundice, diarrhea, somnolence, and melena.
Number
Percentage
93 106 45 117 4
79.5 90.6 38.5 100.0 3.4
culosis in children with good results and tolerance (4, 13, 18). It is common in children to have natural spontaneous improvement after primary infection. Nevertheless, we usually establish the diagnosis of primary pulmonary tuberculosis (disease) only in those children who persist with the abnormal radiologic findings (progressive primary pulmonary tuberculosis). The short-course regimens present many advantages: reduction of total cost of the treatment, including number of doctor's and nurse's visits and decreased cost of medications, and lesstoxicity than observed with the prolonged use of standard medications. The majority of reactions occur at the beginning of treatment. Compliance is better and parents normallyare more cooperative in their appointments. If there is an interruption of treatment, even of more than a couple of months after 3 or 4 months of treatment, the chances of maintaining negative cultures in adults are better (1). With recurrent disease, the same regimen can be used again with minimal risk ofresistance (4, 11). Usually, recurrences are caused by sensitive but persistent bacilli that are not accessible to the medications. Therefore, with these short-course regimens, it is necessary to use bactericidal drugs that will avoid the appearance of resistant mutants (1, 5, 14, 15). Weobserved a high percentage (79.3070) of symptomatic children. Abernathy and
colleagues (13) reported the presence of symptoms in only 32070 of their patients. They detected tuberculosis by mainly using tuberculin test surveys for children in contact with bacillary tuberculous adults, independent of respiratory symptoms. Our outpatient clinic is a referral center in our state, and children are referred for the assessment and diagnosis of symptomatic pulmonary diseases. We observed negative tuberculin tests in three and weak reactions in four children. Five of them had never had BCG vaccination before. All of them had malnutrition of the second and third degree. It has been shown that 5 to 10070 ofbacillary tuberculous patients never react to the tuberculin test (19). Approximately 800/0 of the children (91/117) had a tuberculin test of 15mm of induration or greater. Therefore, this parameter seems to be very important for the diagnosis of tuberculosis in infancy, even in those children already vaccinated with BCG. After BCG vaccination, only 2% of children have a tuberculin test greater than 15 mm of induration (20). However, the tuberculin test along with a positive chest roentgenogram were utilized as the criteria for the diagnosis of tuberculosis in our study in only 45 children (38.50/0), with no previous BCG history and/or a definitive BCG scar, because of the high rate of patients with previous BCG. Our limitation for better documentation on the bacteriologic criterion (table 6) was due to the fact that almost all patients were diagnosed at the outpatient care unit. It is very difficult to assessthe outcome and the efficacy of any regimen for treatment of tuberculosis in children because they rarely have positive sputum and gastric washings, and the best criteria would be clinical findings, such as weight gain, and radiologic follow-up studies (18,21). We observed in our patients that the weight gain was twice as much as that expected for a period of 6 months in a normal child. The use of a radiologic criterion alone may present some difficulties because of residual lymph node or parenchymal lesions that may be present at the end of the treatment (43% of the children). These residual markings have been observed by other authors, and they usually disappear at the end of 2 or 3 yr (13, 18).The initial fear about the association of two hepatotoxic drugs (RIF and INH) in children weredissipated with the progression of the study. These synergistic effects are usually dose related (22). The noncompliance rate of 17% is con-
999
TREATMENT OF CHILDHOOD TUBERCULOSIS
sistent with previous reports in the literature (12, 13). Conclusion
Ninety-seven children with primary pulmonary tuberculosis weretreated with rifampin and isoniazid daily for 6 months. Clinical and radiologic follow-up for a period of 21.4months demonstrated the absence of recurrence of disease. The combined use of two potentially hepatotoxic drugs was not associated with clinically significant hepatic disease. Weconclude that primary pulmonary tuberculosis in children would be successfully treated with daily doses of rifampin and isoniazid for 6 months. Acknowledgment The writers thank Dr. Victor Chernick for invaluable advice and Angela Brito and Shirley Alton for the typing and preparation of this manuscript. References 1. Fox W, Mitchison DA. Short-course chemotherapy for pulmonary tuberculosis. Am Rev Respir Dis 1975; 111:325-53. 2. Verbist L, Gyslen A. Antituberculous activity
of rifampin in vitro and in vivo and the concentrations attained in human blood. Am Rev Respir Dis 1968; 98:923-32. 3. Grumbach FM, Rist N. Aetivite antituberculeuse experimentale de la rifampicine derivee de la rifamicine SV. Revue Tuberc (Paris) 1967; 31: 749-62. 4. Stead WW, Dutt AK. Chemotherapy for tuberculosis today. Am Rev Respir Dis 1982; 125:94-101. 5. British Medical Research Council. Controlled clinical trials of four six month regimens of chemotherapy of pulmonary tuberculosis (Second East African Study). Am Rev Respir Dis 1976; 114:471-5. 6. British Thoracic and Tuberculosis Association. Short-course chemotherapy in pulmonary tuberculosis. Lancet 1976; 2:1102-4. 7. Dutt AK, Stead WW. Short-course treatment regimens for patients with tuberculosis. Arch Intern Med 1980; 140:827-9. 8. Batten J. Experimental chemotherapy for tuberculosis. Br Med J 1968; 3:75-82. 9. Kendig EL. Evolution of short-course antimicrobial treatment of tuberculosis in children 1951-1984. Pediatrics 1985; 74:684-6. 10. Grosset J. Bacteriologic basis for short-course chemotherapy for tuberculosis. Clin Chest Med 1980; 1:231-42. . 11. Dutt AK, Stead WW. Short-course chemotherapy: the Arkansas experience. Chest 1981;80:724-7. 12. Ibanez S, Ross BG. Quimioterapia abreviada de 6 meses en tuberculosis pulmonar infantil. Rev Chil Pediatr 1980; 51:249-52. 13. Abernathy RS, Dutt AK' Stead WW, Moers DJ. Short-course chemotherapy for tuberculosis in
children. Pediatrics 1983; 72:801-6. 14. Fox W. The current status of short course chemotherapy. Tubercle 1979; 60:177-90. 15. Hong Kong Tuberculosis Treatment Services and British Medical Research Council. Controlled trial of 6 and 9 month regimens of daily and intermittent streptomycin plus isoniazid plus pyrazinamide for pulmonary tuberculosis in Hong Kong. Thbercle 1975; 56:81-96. 16. 'Iuberculosis Research Institute Bucharest. Trial of two intermittent short-course regimens (78 doses) in the initial treatment of pulmonary tuberculosis. Tubercle 1977; 58:1-8. 17. Mehrotra ML, Gautam KD, Chaube CK. Shortest possible, acceptable, effective, ambulatory chemotherapy in pulmonary tuberculosis: preliminary report. Am Rev Respir Dis 1981; 124: 239-44. 18. Jacobs RF, Abernathy RS. The treatment of tuberculosis in children. Pediatr Infect Dis 1985; 4:513-7. 19. Kent D, Schwartz R. Active pulmonary tuberculosis with negative tuberculin skin reactions. Am Rev Respir Dis 1967; 95:411-8. 20. ReisFJC. Alergia tuberculinica e cicatriz vacinal em lactentes que tomaram a vacina BCG injetavelquando recemnascidos. J Pediatr (Rio) 1982; 52:23-8. 21. Longman Group. Smear negative pulmonary tuberculosis. Thbercle 1980; 61:113-6. 22. O'Brien RJ, Long MW, Cross FS, Lyle MA, Snider DE. Hepatotoxicity from isoniazid and rifampin among children treated for tuberculosis. Pediatrics 1983; 72:491-9.
