Mediastinal Lymphadenopathy in Children With Histoplasmosis

Background: Mediastinal lymphadenopathy (ML) in children can arise from malignancy, infection, or rheumatic illness among others, and may be found incidentally on imaging or during workup for a variety of symptoms. Our aim was to describe the clinical presentation and natural history of histoplasmosis in children who present to a tertiary care center with ML in an endemic area of the country.

Methods: After institutional review board approval, a retrospective study of all children (aged < 21 y) presenting with proven (positive serologies) or suspected histoplasmosis (negative serologies, negative tuberculosis testing, and benign outcomes in follow-up) over a 5-y period was done. Seventy-four patients were tested; those with another diagnosis (n = 6) or without ML (n = 26) were excluded, for a total cohort of 44 patients. Demographics, clinical presentation, symptoms, laboratory data, treatment course, radiography studies, and inpatient and outpatient visits were examined.

Results: Of the 44 patients with ML, 27 had proven histoplasmosis, and 19 had suspected histoplasmosis. The median follow-up by imaging or clinical examination was 6.9 mo (0.3-73.2 mo). Sixteen patients received antifungal therapy with itraconazole, and 15 patients received at least one course of steroids, nearly all for respiratory symptoms; 11 patients (24%) received both. There was no difference in readmission rate (n = 5 versus 2, P = 0.7) or recurrence of symptoms (n = 2 versus 5, P = 0.4) between patients who received an antifungal and those that did not. Receiving steroid therapy was associated with airway narrowing and a higher readmission rate (n =6 versus 2 who were not treated with steroids, P = 0.04), but not with symptom recurrence. Nine lymph node or mass biopsies were performed; however, the pathology only confirmed nonspecific infection in three and was nondiagnostic in the remaining six patients. Twenty-seven patients had at least one confirmatory laboratory test positive for histoplasmosis. Thirty-nine of the 44 patients (84%) with a diagnosis of histoplasmosis (proven or suspected) were asymptomatic by 1-2 mo follow up, with the remainder having intermittent chest pain or reactive airway disease.

Conclusions: ML because of proven or suspected histoplasmosis is usually a self-limiting disease that can be managed with treatment of the child's symptoms. Antifungals and steroids are of unclear benefit and may not alter the natural course of the disease. Biopsies are rarely diagnostic in the setting of ML, and invasive procedures should be avoided.