To the Editors:
We read with interest the case report by Shah et al,1 which summarizes the current difficulties with regard to diagnosis of congenital tuberculosis (TB).
We have reviewed patient’s records of pediatric TB cases diagnosed at Hospital Universitario La Paz in the period 1978 to 2014. Of the 555 children younger than 14 years with TB, 3 patients (0.5%) fulfilled Cantwell diagnostic criteria for congenital TB.2 There was no maternal history of TB in any of them, and initial diagnoses were sepsis (2 patients) and whooping cough (1 patient). Clinical symptoms had started between 8 and 36 days after birth. Two patients presented with fever and progressive respiratory distress and 1 with persistent cough. All infants had abnormal chest radiograph (1 hilar adenopathy and 2 miliary pattern).
One patient died within 24 hours of hospital admission, and diagnosis was confirmed at necropsy, which showed caseating granulomas in the lungs, liver, spleen, kidneys, adrenal glands and testicles. Direct smear revealed acid-fast bacilli (AFB) and culture grew Mycobacterium tuberculosis. His mother developed miliary disease with meningitis shortly afterward.
The other 2 patients had positive gastric aspirate AFB smears and cultures and were successfully treated with antitubercular drugs and corticosteroids. Tuberculin skin test (TST) was initially negative (0 mm) in one of them, but after 6 weeks, the induration was 8 mm. The other patient’s TST was 7 mm at diagnosis. Both mothers were immigrants from high TB-burden countries and had a positive TST, a normal chest radiograph and negative sputum AFB smears and cultures. Endometrial biopsies showed tubercular granulomas, and M. tuberculosis was identified by polymerase chain reaction. One of the mothers had been diagnosed with primary sterility because of tubal obstruction, and pregnancy had been achieved by in vitro fertilization. The other had had menstrual irregularities but no other symptoms suggestive of TB and no history of infertility.
These cases highlight the importance of considering congenital TB in infants with sepsis or respiratory tract infection not responding to antibiotics, especially in those born to immigrants from TB-endemic countries or conceived by in vitro fertilization. In these cases, AFB smear, culture and polymerase chain reaction of gastric aspirates (tracheal if intubated) should be performed. Maternal endometrial biopsy should be considered when the mother has a positive TST and a normal chest radiograph, particularly in immigrants from TB-endemic countries or in women with history of tubal infertility. In our series, none of the mothers had significant symptoms during pregnancy, and the diagnosis of TB was made only after their infants became ill.
Teresa del Rosal, MD, PhD
Fernando Baquero-Artigao, MD
Ana Méndez-Echevarría, MD, PhD
María José Mellado, MD, PhD
General Paediatrics and Infectious and Tropical Diseases Department Hospital Universitario La Paz Madrid, Spain
1. Shah G, Tse-Chang A, Cooper R, et al. Nodular lung lesions in a 10-week-old infant. Pediatr Infect Dis J. 2015;34:912
2. Cantwell MF, Shehab ZM, Costello AM, et al. Brief report: congenital tuberculosis. N Engl J Med. 1994;330:1051–1054