Abstract

Introduction
Tuberculosis (TB), caused by Mycobacterium tuberculosis (MTB), is encountered more commonly in developing countries, though is becoming more frequent in industrialized nations.1,2 Despite this changing landscape, the majority of recent cases detailed in the United States are in patients of foreign descent.2,3 Disease burden is highest among older individuals and lowest among children younger than 14 years, though the latter tend to have a higher propensity for more invasive disease. 3 This unique case series details the epidemiology of a local Louisiana outbreak of active tuberculosis in a Caucasian cohort with no known epidemiologic risk factors and where 3 of the patients described are younger than 2 years.
Methods
Medical records from Our Lady of the Lake Children’s Hospital (OLOLCH) and the State of Louisiana Department of Health and Hospitals (LADHH) were retrospectively reviewed for demographic, clinical, and treatment data. The study population was limited to the 3 actively infected pediatric patients treated at OLOLCH, additional active and latent TB cases uncovered by repeat testing during this outbreak, and the adult source.
Case Presentations
Case 1 was a 19-month-old Caucasian female with a past medical history significant for prematurity, surgical necrotizing enterocolitis, grade 1 intraventricular hemorrhage, and retinopathy of prematurity. She presented with weakness, mental status changes, failure to thrive, diarrhea, and intermittent fever after a recent hospitalization for pyelonephritis. Head imaging emergently obtained after acute worsening of her neurologic status noted hydrocephalus requiring urgent ventricular drain placement. Subsequent imaging noted diffuse edema, hypodensity of the brainstem, and basal ganglia with multifocal infarcts. Serial electroencephalograms were read as diffusely slow consistent with a diffuse disturbance in brain function. Her cerebrospinal fluid (CSF) studies were notable for mild pleocytosis (28 white blood cells [WBC]/mm3) with a lymphocytic predominance (86%), low glucose (38 mg/dL), and elevated spinal fluid protein (47.4 mg/dL). Her initial TB evaluation (ie, tuberculin skin test [TST], MTB polymerase chain reaction [PCR] and acid-fast bacilli [AFB] stain and culture from ventricular CSF) was unremarkable. Given her presentation, she was initially treated for bacterial meningitis. After initial mild improvement, her clinical status precipitously declined. Because of the lack of clinical response to empiric antimicrobial treatment and further head imaging suggestive of disease progression, she was started on steroids and 4-drug anti-TB therapy with rifampin (RIF), isoniazid (INH), pyrazinamide (PZA), and ethambutol. She was diagnosed with tuberculous meningitis (TBM) based on CSF pleocytosis, positive MTB PCR from a tracheal aspirate, and head imaging consistent with TBM. Additional evaluation revealed a positive Quantiferon-Gold TB In-Tube (QTF-G). She had severe neurological sequelae resulting from TBM.
Case 2 was a 7-month-old previously healthy Caucasian male who was a cousin of our index patient and presented with fever and cough. At the time of presentation, there was no known adult exposure to TB. His initial workup was notable for mild leukocytosis (17 × 103/μL) with bandemia (14%), and CSF studies notable for mild pleocytosis (15 WBC/mm3) with a lymphocytic predominance (81%). Initial imaging studies, including head computed tomography (CT) and chest radiograph (CXR), were unremarkable. A subsequent chest CT was noted to have paratracheal adenopathy with an indistinct right middle lobe (RML) infiltrate. He was diagnosed with TBM and active pulmonary tuberculosis based on CSF pleocytosis, positive TST, positive QTF-G, and abnormal chest imaging. He was subsequently started on steroids and anti-TB therapy, the latter continued for 9 months. He had no significant sequelae resulting from his active tuberculosis.
Case 3 was a 19-month-old Caucasian male, another cousin of case 1, with a past medical history significant for global developmental delay secondary to absence of the corpus callosum. He had recently been treated for a persistent pneumonia when he presented with a positive TST (10 mm induration) that had been performed due to close contact with the preceding 2 cases described. A gastric aspirate was found to be AFB-culture positive for MTB. The CXR was remarkable for a persistent RML infiltrate. CSF analysis was also performed and found to be normal. He completed 6 months of anti-TB therapy. CXR following completion of TB therapy showed resolution of the RML opacity. He had no significant sequelae from his active pulmonary tuberculosis.
Epidemiologic investigation resulted in the identification of an adult source—a Caucasian male who provided child care services for the 3 affected patients. Forty-two contacts of this adult source were evaluated for possible tuberculosis. Of these, 14 individuals (7 children and 7 adults) were found to have TB infection—10 with latent TB and 4 with active disease; all were treated.
Discussion
We present a rare clustering of invasive MTB in infants without foreign contacts. While pulmonary TB and TBM are commonly confronted in developing nations, the described cohort demonstrates that these diseases are also being encountered more frequently in the United States. 4 Progression from primary M tuberculosis to TBM is more common in children secondary to impairment in the ability to contain the primary infection in the lung. 4 The adult source for this series was identified retrospectively as a Caucasian male who was providing child care services for the 3 children described above. He had a remote history of incarceration but no foreign travel.
Tuberculous meningitis is a subacute disease process with symptoms present for a median of 10 days prior to diagnosis and an estimated incidence of 100 to 150 cases per year in the United States. 5 Previous studies have noted that the clinical outcome of TBM is directly related to early diagnosis and treatment. 1 However, diagnosis is often delayed, because of nonspecific clinical features, which include fever, malaise, headache, dizziness, and vomiting.1,2 As the disease progresses, clinical features can include altered mental status, stroke, hydrocephalus, cranial neuropathies, and seizures. 1 A delay in TBM diagnosis occurred in case 1, whose initial presentation was nonspecific—her clinical symptoms were similar to those seen in bacterial meningitis, she lacked known TB risk factors, and the initial TB evaluation was unremarkable (including ventricular CSF testing that was negative, similarly seen in prior reports 6 ).
While early diagnosis is key, diagnostic evaluation in the pediatric population remains challenging. Current diagnostic methods such TST, interferon-γ release assay (IGRA), AFB culture from respiratory or gastric samples and CSF have relatively low sensitivity for the disease. 4 Nevertheless, these were among the diagnostic methods used in this case series. The CSF results in the first pediatric patient described exhibited the findings suggestive of TBM-lymphocytic-predominant pleocytosis, elevated protein levels, and low glucose.1,4,5 Although multiple CSF and tracheal aspirate samples were obtained in case 1, only 1 resulted positive (MTB PCR from endotracheal tube aspirate), which resulted 6 weeks after initial presentation. The second pediatric patient described was diagnosed on the basis of positive IGRA, TST, mediastinal lymphadenopathy, and CSF pleocytosis, while the third pediatric patient was diagnosed on the basis of a positive TST, persistent pneumonia, and MTB – positive gastric aspirate (both presenting after case 1 was diagnosed). The gold standard for TBM diagnosis remains isolation of bacillus by culture or visualization on an acid-fast smear 1 ; however, this was seen in only one of the pediatric patients described in this case series. Repeated large volume samples increase the diagnostic yield,1,5 but are challenging to obtain in the pediatric population. 4 As illustrated in the first case, empiric therapy should ideally be started prior to confirmation of disease in suspected cases due to the delay in obtaining results and poor sensitivity of accurate diagnostic evaluation.1-5
Neuroimaging can be useful in diagnosing TBM and was key in the diagnosis of the first pediatric patient described. As magnetic resonance imaging is the imaging study of choice for visualizing abnormalities associated with TBM,1,2,5 this was the study used on the 3 pediatric cases described. Common neuroradiologic features seen in TBM are basal meningeal enhancement and hydrocephalus as noted in the first pediatric patient.1,5 Tuberculomas and vascular infarcts of the basal ganglia and midbrain are also characteristic. 1
After diagnosis of active TB or TBM, all 3 pediatric patients were started on 4-drug anti-TB therapy. The recommended treatment regimen for drug-susceptible TBM includes 2 months of daily INH, RIF, PZA, and either streptomycin or ethambutol, followed by 7 to 10 months of INH and RIF. 3 Treatment is generally administered by direct observation therapy to ensure compliance. Surgical intervention or diuretic therapy may also be required to treat hydrocephalus, the former being necessary for our index patient.
Morbidity and mortality remain high in children with TBM—almost 50% of patients have permanent neurological sequelae despite treatment.1-3,5 The first pediatric patient described was noted to have severe neurological sequelae resulting from her TBM, including tracheostomy and ventilator dependence, temperature instability, and clonus, while the other 2 pediatric patients have not suffered long-term neurological or pulmonary sequelae. This cohort, where all are Caucasian and toddlers represent most of the active disease, illustrates the invasive proclivity of active TB in toddlers and the rarity of TBM in this country.
Author Contributions
All the authors participated in the conceptualization, design, analysis and interpretation of data, drafting and revision of the manuscript, and have approved the manuscript as submitted.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
