| |||||||||||||||||||||||||||||||
We describe two patients with congenital tuberculosis. We also review the 29 cases reported since 1980 and propose revised criteria for distinguishing congenital from postnatally acquired tuberculosis.
Case Reports
Patient 1
A 24-day-old Hispanic male infant presented with a one-day history of respiratory distress, vomiting, and poor feeding. A physical examination revealed hepatosplenomegaly. A chest radiograph showed bilateral infiltrates. The white-cell count was 13,900 per cubic millimeter, with 23 percent polymorphonuclear cells, 39 percent band forms, 6 percent myelocytes, 24 percent lymphocytes, and 8 percent monocytes. Serologic testing for human immunodeficiency virus (HIV) was negative. The serum concentration of total bilirubin was normal, alanine aminotransferase was 133 U per liter, and aspartate aminotransferase was 98 U per liter. The cerebrospinal fluid was normal. Bacterial cultures of blood, cerebrospinal fluid, and urine were negative. The patient required mechanical ventilation and received intravenous antibiotic therapy and treatment for shock and disseminated intravascular coagulation. Smears of endotracheal and gastric aspirates contained acid-fast bacilli, and cultures were positive for Mycobacterium tuberculosis. Mycobacterial cultures of urine and blood were negative. The infant initially responded to isoniazid, rifampin, and pyrazinamide and was discharged after 15 days. Subsequent problems with failure to thrive and nonadherence to therapy resolved after foster-care placement.
The infant's mother, a 26-year-old HIV-negative Hispanic woman, had fever and back pain for nine days after delivery. Her chest radiograph showed blunting of the left costophrenic angle. She denied exposure to tuberculosis. After the infant was given a diagnosis of tuberculosis, her Mantoux skin test was found to be positive, and chest radiography showed a small left pleural effusion. Examination of a sputum specimen revealed acid-fast bacilli, and M. tuberculosis was cultured from sputum and urine. An endometrial biopsy revealed caseating granulomas containing acid-fast bacilli (Figure 1), and cultures grew M. tuberculosis. Examination of the placenta revealed acid-fast bacilli in a small thrombus (Figure 2). The mother responded to antituberculous therapy. The M. tuberculosis isolates from the infant and the mother were sensitive to all antituberculous medications and had the same pattern of restriction-fragment-length polymorphism.
|
|
Patient 2
A 15-day-old Native American female infant presented with a two-day history of fever, poor feeding, and respiratory distress. A physical examination revealed hepatosplenomegaly. A chest radiograph showed a right-lung infiltrate. The white-cell count was 26,000 per cubic millimeter, with 51 percent polymorphonuclear cells, 7 percent band forms, 26 percent lymphocytes, and 16 percent monocytes. The serum concentration of total bilirubin was 1.4 mg per deciliter (24 µmol per liter), and serum concentrations of alanine and aspartate aminotransferases were normal. The cerebrospinal fluid was normal. Bacterial cultures of blood, cerebrospinal fluid, and urine were negative. The patient required mechanical ventilation and received intravenous antibiotic therapy. Smears of gastric and endotracheal aspirates contained acid-fast bacilli, and cultures were positive for M. tuberculosis. Cultures of cerebrospinal fluid and urine were negative. A Mantoux test was negative. The infant responded to treatment with isoniazid, rifampin, and pyrazinamide for two months, with streptomycin added for the first two weeks, followed by four months of isoniazid and rifampin.
The maternal grandmother was a 41-year-old Native American who had tuberculous meningitis eight months before the infant's birth. M. tuberculosis was isolated from her cerebrospinal fluid, sputum, and urine. She responded to antituberculous therapy.
The infant's mother was a 20-year-old Native American who had a positive Mantoux test at the time the grandmother's contacts were investigated, eight months before delivery. A chest radiograph showed a minimal reticulonodular infiltrate. Sputum was not obtained. The mother was treated with isoniazid as preventive therapy, which was stopped after two weeks when the pregnancy was discovered. She was asymptomatic during pregnancy and thereafter. After the infant was given a diagnosis of tuberculosis, an endometrial biopsy was performed that revealed acid-fast bacilli; the specimen was culture-positive for M. tuberculosis. The mother responded to antituberculous therapy. The M. tuberculosis isolates from the infant, the grandmother, and the mother were sensitive to all antituberculous medications and had the same pattern of restriction-fragment-length polymorphism.
All the household contacts had negative Mantoux tests except two adults with tests that had previously been positive. Both had normal chest radiographs and were not given preventive therapy.
Discussion
Tuberculous bacillemia during pregnancy may result in infection of the placenta or the maternal genital tract1,2. Such infection may then be transmitted to the fetus by hematogenous spread from the placenta to the umbilical vein or by the aspiration or ingestion of amniotic fluid contaminated by placental or genital infection. Hematogenous spread leads to the formation of one or more primary complexes in the liver or lungs, whereas the aspiration or ingestion of infected amniotic fluid results in primary complex formation in the lungs or gastrointestinal tract, respectively1,2,3,4. Hematogenous spread and aspiration of infected amniotic fluid account for approximately half the cases of congenital tuberculosis each5. Since the localization of the primary complex requires an open surgical procedure or autopsy, the mode of congenital transmission often cannot be determined, as is illustrated by the two patients described here. Coinfection with HIV, an important contributing factor to the recent rise in tuberculosis, increases the likelihood of extrapulmonary tuberculosis6. Women of childbearing age who have HIV infection may therefore be at increased risk for placental or genital tuberculosis, resulting in an increased risk of congenital transmission.
Congenital tuberculosis is rare. Blackall found 3 affected patients among 100 infants of mothers with active tuberculosis,3 but in two other series no affected patients were found among 260 and 1369 such infants7,8. Overall, fewer than 300 cases have been reported,4,5,9,10,11,12 with the most recent case in the United States reported in 198213.
When the patients described here are included, 29 cases of congenital tuberculosis have been reported in the English-language literature since the 1980 review by Hageman et al.11,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33. The median age at presentation was 24 days (range, 1 to 84). The presenting signs and symptoms were often nonspecific (Table 1). Chest radiographs were abnormal in 23 infants, although 18 had nonspecific infiltrates. Tuberculin skin tests at the time of diagnosis in nine infants were all negative; tests performed later in two of the nine were positive. Both infants tested for HIV infection were negative. Both biopsy and noninvasive procedures were often diagnostic (Table 2). Only 12 mothers had had active tuberculosis. Fifteen mothers were asymptomatic and were given a diagnosis of tuberculosis only after the disease was diagnosed in their infants. Mortality was 38 percent overall and 22 percent (5 of 23) among the infants who received chemotherapy.
|
|
Patients with congenital tuberculosis and those with postnatally acquired disease should be treated similarly1,2. Neonates should receive isoniazid (10 to 15 mg per kilogram of body weight per day), rifampin (10 to 20 mg per kilogram per day), pyrazinamide (15 to 30 mg per kilogram per day), and either streptomycin (20 to 30 mg per kilogram per day) or ethambutol (15 to 25 mg per kilogram per day) for the first 2 months,1,35 followed by isoniazid and rifampin for 4 to 10 months, depending on the severity of the disease1,2. Because of the potential for optic neuritis, streptomycin is generally preferred over ethambutol in infants35. In communities where rates of resistance to isoniazid are 4 percent or less, initial regimens of three drugs may be acceptable35.
The only standardized criteria for distinguishing congenital from postnatally acquired tuberculosis were established by Beitzke in 19359. These criteria require that the infant have proved tuberculous lesions and one of the following: lesions in the first few days of life, a primary hepatic complex, or the exclusion of postnatal transmission by the separation of the infant at birth from the mother and other sources of infection. Patients meeting these criteria almost certainly have true congenital tuberculosis. These criteria have become increasingly difficult to apply in current practice, however, and have been met in only three cases since 1980. In current practice, the exclusion of postnatal transmission relies on a thorough investigation of contacts, including the infant's hospital attendants, and adherence to recommendations for the treatment of infants exposed to tuberculosis, rather than separation of the infant from the mother at birth.
The demonstration of a primary hepatic complex requires an open surgical procedure or autopsy to confirm liver and regional lymph-node involvement9. Primary hepatic complexes are seldom demonstrated in current practice (and have been found in only two cases since 1980, both at autopsy), because of reduced mortality and increased use of percutaneous liver biopsy. The presence of caseating hepatic granulomas permits congenital and postnatally acquired tuberculosis to be distinguished on the basis of liver-biopsy findings alone. In one series, caseation was seen in only 1 of 79 patients with postnatally acquired tuberculosis with hepatic granulomas,11 as compared with 7 of 14 patients with congenital tuberculosis with hepatic granulomas studied by Hageman et al.11 and 6 of 9 such patients reported since 1980.
Many authors have inferred that infection of the placenta or the maternal genital tract is necessary for congenital transmission1,2,3,4,5,9. Female genital tuberculosis is uncommon, with an average annual incidence of 0.036 case per 100,000 women of childbearing age in the United States from 1985 through 1991 (CDC: unpublished data). Appropriate clinical testing of the maternal genital tract was performed in 14 mothers of infants with congenital tuberculosis described since 1980. All 14 mothers were found to have genital tuberculosis. The placentas of two other infants also showed evidence of tuberculous infection.
On the basis of these considerations, we propose revised diagnostic criteria for congenital tuberculosis. The infant must have proved tuberculous lesions and at least one of the following: (1) lesions in the first week of life; (2) a primary hepatic complex or caseating hepatic granulomas; (3) tuberculous infection of the placenta or the maternal genital tract; or (4) exclusion of the possibility of postnatal transmission by a thorough investigation of contacts, including the infant's hospital attendants, and by adherence to existing recommendations for treating infants exposed to tuberculosis. Different patterns of restriction-fragment-length polymorphism in the M. tuberculosis isolates from the mother and the infant would exclude congenital transmission, although identical patterns could result from either congenital or postnatal transmission.
Some cases of postnatally acquired tuberculosis with caseating hepatic granulomas or tuberculous infection of the placenta or the maternal genital tract may be misclassified by these criteria. The advantages of the revised criteria include increased diagnostic sensitivity (e.g., 23 cases reported since 1980 met the revised criteria) and applicability in current practice, eliminating the potential for misclassification that has resulted from the lack of usefulness of the 1935 criteria. A further advantage of the revised criteria is their emphasis on evaluating the mothers of infants with suspected congenital tuberculosis, potentially improving the detection of subclinical tuberculosis in these women.
We are indebted to Chris Fox, B.S.N., Pima County Health Department, Tucson, Ariz., Carol D. Harriman, M.P.H., M.S.N., New Mexico Department of Health, Santa Fe, and Michelle P. Webster, M.D., University of New Mexico School of Medicine, Albuquerque, for clinical information regarding these patients; and to the following CDC staff members: Charles L. Woodley, Ph.D., Mycobacteriology Laboratory, for the analyses of restriction-fragment-length polymorphism in the M. tuberculosis isolates; Ms. Billie L. Swisher, Scientific Resources Program, for histotechnology support; and Mr. James Gathany, Scientific Resources Program, for photomicrography.
Source Information
From the Division of Tuberculosis Elimination, National Center for Prevention Services (M.F.C., S.E.V., I.M.O.), and the Scientific Resources Program, National Center for Infectious Diseases (E.P.E.), Centers for Disease Control and Prevention, Atlanta; the Department of Pediatrics and the Steele Memorial Children's Research Center, University of Arizona, Tucson (Z.M.S.); University Medical Center, Tucson, Ariz. (A.M.C.); Arizona Department of Health Services, Phoenix (L.S.); and the Public Health Service Indian Hospital, Albuquerque, N.M. (W.F.G.).
Address reprint requests to Dr. Cantwell at the Division of Tuberculosis Elimination, Centers for Disease Control and Prevention, 1600 Clifton Rd., E-10, Atlanta, GA 30333.
References
| |||||||||||||||||||||||||||||||
Related Letters:
Congenital Tuberculosis
Lackmann G. M., Kabra S. K., Jain Y., Madhulika , Bagga A., Schiffman R., Cantwell M. F., Valway S. E., Onorato I. M.
Extract |
Full Text
N Engl J Med 1994;
331:548-549, Aug 25, 1994.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
