Tuberculosis (TB) has been described in HIV-infected children. However, the burden, disease spectrum, diagnosis and barriers to treatment of TB disease have been less well-elucidated in children with other major medical comorbidities, such as oncology patients,1 rheumatology patients,2 stem cell and organ transplant recipients.3 As a consequence, TB screening guidelines have not uniformly emphasized the importance of testing for and treating latent tuberculosis infection (LTBI) in these high risk children. We describe a single tertiary care pediatric hospital’s experience in a low TB incidence country with TB exposure, infection and disease among children with underlying medical conditions over 3 decades.
This was a retrospective descriptive case series of children (≤18 years of age) with medical comorbidities seen at a children’s tuberculosis clinic in Houston, Texas from 1983 to 2013. This clinic is the site of referral for all children followed at Texas Children’s Hospital. Authors have cared for over 9000 children in this clinic. Comorbidity was defined as any child who was immunocompromised or receiving immunosuppressing medication, or with underlying cardiac, pulmonary, hepatic, endocrine or renal disease. Children with mild, well-controlled disease (eg, asthma, hemodynamically insignificant murmurs) were excluded.
TB exposure occurred when a child <5 years of age or an immunocompromised child of any age who was in close contact (as identified as part of contact investigations) with a person with infectious TB disease (termed the “source case”) had examination findings, chest radiography and tuberculin skin test (TST) or interferon gamma release assay that were negative. American Thoracic Society criteria were used to define a positive TST.4 LTBI was defined as a normal physical examination and chest radiograph in a child with a positive TST or interferon gamma release assay. Confirmed disease was defined as growth of Mycobacterium tuberculosis with compatible clinical, radiographic and/or histopathologic findings.5 Probable disease was defined as compatible clinical, radiographic and/or histopathologic findings in children who had a positive test result for infection, epidemiologic risk factors for TB and improvement of radiographic findings or symptoms and physical examination findings during treatment in whom cultures were negative or not obtained.
Patients were identified from the microbiology laboratory and TB clinic databases. Institutional review board approval was obtained. Frequencies were calculated for demographic variables and culture results.
Sixty-nine children were identified (Table 1): 7 exposed, 40 infected and 22 with disease (13 confirmed and 9 probable). Seven children were treated for TB exposure: 5 were <5 years of age; 2 older children (8 and 18 years of age with HIV and renal transplantation, respectively) also were treated for exposure. Six of the 7 exposed children had household contacts who had been diagnosed with pulmonary TB. The last child was exposed in a refugee camp immediately before arrival in the United States. All completed 2–3 months of isoniazid monotherapy without complication; none have progressed to disease.
Forty children were treated for LTBI: 9 had identified source cases (8/9 in the home); 31 were identified during routine LTBI screening for high risk patients (30 had positive TST results and 1 had a positive interferon gamma release assay result). Risk factors for the 31 LTBI children without source cases included foreign birth (18) or foreign travel (7). Six (15%) children had no identifiable TB risk factors (their TST indurations ranged from 10 to 23 mm). Of the 40 children treated for LTBI, 37 (93%) were started on isoniazid, and 34/37 (92%) tolerated a 9-month regimen well. Two children with hepatic risk factors (1 taking anti-epileptic medication and 1 with hepatitis B) developed hepatitis while taking isoniazid which resolved when the drug was stopped; both successfully completed 6 months of rifampin. In the third case, the caregiver was confused about the dosing and was giving isoniazid twice daily, and the child had asymptomatic transaminitis. Two children received rifampin initially because of hepatic dysfunction; both tolerated rifampin well and completed therapy. A 9-year-old girl with LTBI whose father had multidrug-resistant TB was started on pyrazinamide and levofloxacin and developed transaminitis, jaundice and coagulopathy which persisted for 3 months after medication cessation. She was later found to have hepatitis C infection. She did not receive further TB medications and was followed clinically for progression to disease. None of these children have progressed to disease.
Twenty-two children were treated for TB disease (13 confirmed and 9 probable); cultures were obtained for all 22. Only 3 children had known source cases, and 10 others were born in high-prevalence countries; thus, almost 60% of cases represented potentially missed screening opportunities. Conversely, over 40% lacked any apparent risk factors as identified on the American Academy of Pediatrics guidelines; of these, 5 of 9 were microbiologically confirmed (their TST indurations ranged from 0 to 18 mm). Eighteen (82%) had TSTs ≥ 5 mm. Twenty had pulmonary disease, and each one had scrofula and cutaneous TB.
Overall, 4 (18%) children with TB disease died. For 2, TB directly contributed to their deaths. One was an 18-year-old sickle cell patient who had a bone marrow transplant and developed respiratory distress syndrome after engraftment. TB was a postmortem diagnosis; M. tuberculosis grew on multiple acid-fast blood cultures. The other was a 14-month-old child with repaired congenital heart disease who had miliary TB and multiple pulmonary blebs; she died of a tension pneumothorax when 1 of the blebs ruptured. Two children died while on treatment for disease: a 3-year-old with Krabbe’s disease who died 2 months into treatment and a 16-year-old with sickle cell disease had a cardiac arrest 3 months into treatment, but death was thought to be unrelated to TB. The remaining 18 children completed therapy and tolerated multidrug therapy uneventfully, and none are known to have relapsed.
There are several reasons to think that LTBI screening strategies need to be revisited for medically complex children. First, use of epidemiologic risk factors alone to screen for LTBI risk, as is recommended for targeted skin testing of otherwise healthy children,4 would have missed 40% of high risk children who developed disease. Although targeted skin testing guidelines are very appropriate for otherwise healthy children, they may not be appropriate for children with underlying comorbidities or treatments that predispose them to TB. Second, many of these patients had comorbidities that would not have deemed them at highest risk for LTBI according to current guidelines. Third, the mortality rate from TB was very high in this cohort, and all-cause mortality was almost 20% while children were receiving TB treatment.
A recent review6 of US pediatric TB cases demonstrated that 25% of children would not have been identified as having TB risk factors when an enhanced version of the American Academy of Pediatrics-endorsed screening guidelines were used.4 In our cohort, over 40% of children with disease would have been missed by this screening tool. Given the high all-cause mortality rate we observed, currently recommended targeted testing as a stand-alone LTBI screening modality had an unacceptably low sensitivity.
Current recommendations identify certain groups of children with comorbidities as being high risk of progression from LTBI to disease. These include children with HIV infection, malignancy and children who are immunosuppressed or immunocompromised.7 However, 2 of the most common comorbidities seen in children with TB in this cohort were children with congenital heart disease and those with sickle cell disease. Neither condition would have prompted either more frequent or more thorough evaluation for LTBI. There are no epidemiologic data to suggest that children with heart disease or sickle cell are predisposed to TB disease; however, they seemed to have a worse outcome when they did progress to TB disease. However, the child who died with sickle cell disease had the additional risk factor of post-bone marrow transplant immunosuppression; as such, it is unclear on which was the larger risk factor for disease progression and severity. Although the small sample size and the geographically limited nature of the study preclude generalization, these findings do make us wonder if certain children with comorbidities should be screened for LTBI differently.
One potential concern for treatment with TB medication in medically complex children is the risk of adverse events, as these children may not have normal end-organ function and receive multiple, often hepatically metabolized, medications, to treat their underlying medical conditions. Only 1.5% of our cohort had intolerance to standard TB regimens. Although slightly higher than the incidence of hepatotoxicity in healthy children reported in 1 meta-analysis,8 this rate is substantially <10% incidence of transaminitis reported in adult solid-organ transplant recipients.9 However, both children had other hepatic risk factors, and thus required careful monitoring.
The mortality rate for children with TB disease was very high in this series. In 2011, 0.2% of persons treated for TB disease in the United States died of TB.10 In our series, TB- and all-cause mortality rates were 9% and 18%, respectively. However, the mortality associated with TB in this cohort means that there is a different risk-benefit relationship for TB screening in these children in contrast to otherwise healthy children.
There were limitations to this study. Given the retrospective nature, not all data elements were available for all children. We did not have data on the total number of children with comorbidities cared for at this medical center, precluding incidence estimates. IGRAs were not available early in the study period and were not used consistently in the last decade, precluding comparisons of IGRAs and TSTs. As Houston has higher TB incidence rates than many US cities, the findings may not be generalizable to low-burden settings. The small number of cases could have been augmented if a multicenter collaborative network could be established for domestic childhood TB.
This study shows that use of standard American Academy of Pediatrics screening guidelines for TB infection and disease would have potentially missed >40% of cases of TB disease in children with significant comorbidities. These children tolerated LTBI and TB disease therapy well, but both all-cause and TB-specific mortality rates were high in children with TB disease. This warrants further investigation in a larger cohort of children to find whether these relationships persist. If they do, then consideration should be given to integrating TB screening into the initial and ongoing evaluations of children diagnosed with cancer and other medical comorbidities.
1. Stefan DC, Kruis AL, Schaaf HS, et al. Tuberculosis
in oncology patients. Ann Trop Paediatr. 2008;28:111–116
2. Toussi SS, Pan N, Walters HM, et al. Infections in children and adolescents with juvenile idiopathic arthritis and inflammatory bowel disease
treated with tumor necrosis factor-α inhibitors: systematic review of the literature. Clin Infect Dis. 2013;57:1318–1330
3. Vecino R, Santiago B, Baquero-Artigao F, et al. Tuberculosis
in pediatric solid organ and hematopoietic stem cell transplant recipients. Pediatr Infect Dis J. 2012;31:774–777
4. Centers for Disease Control and Prevention
. . Targeted tuberculin testing and treatment of latent tuberculosis
infection. MMWR Recomm Rep. 2000;49:1–51
5. Graham SM, Ahmed T, Amanullah F, et al. Evaluation of tuberculosis
diagnostics in children: 1. Proposed clinical case definitions for classification of intrathoracic tuberculosis
disease. Consensus from an expert panel. J Infect Dis. 2012;205(suppl 2):S199–S208
6. Winston CA, Menzies HJ. Pediatric and adolescent tuberculosis
in the United States, 2008–2010. Pediatrics. 2012;130:e1425–e1432
7. Pickering LKAmerican Academy of. Pediatrics.. Tuberculosis
. 201229th Ed. Red Book: 2012 Report of the Committee on Infectious Diseases. Elk Grove Village, IL American Academy of Pediatrics:745 In:
8. Donald PR. Antituberculosis drug-induced hepatotoxicity in children. Pediatr Rep. 2011;3:e16
9. Fábrega E, Sampedro B, Cabezas J, et al. Chemoprophylaxis with isoniazid in liver transplant recipients. Liver Transpl. 2012;18:1110–1117
10. Centers for Disease Control and Prevention
. . Trends in tuberculosis
– United States, 2012. Morbid Mortal Wkly Rep.. 2013;62:201–205