Childhood Cancers Misdiagnosed as Tuberculosis in a High Tuberculosis Burden Setting.

BACKGROUND: Tuberculosis (TB) and childhood cancers have overlapping presentations and malignancies may be misdiagnosed as TB in high TB-burden settings. METHODS: This retrospective study investigated the diagnosis of TB in children with cancer registered in the Tygerberg Hospital Childhood Tumor Registry from 2008 to 2018. We studied children on anti-tuberculosis treatment (ATT) at cancer diagnosis or diagnosed with TB within 1 month of cancer diagnosis. We describe the circumstances and extent of this misdiagnosis, quantify the delay in therapy and document the outcomes of these children. RESULTS: Twenty-seven of 539 (5%) children in the registry started ATT before cancer diagnosis. Both pulmonary and extrapulmonary TB complicated the cancer diagnosis. Of the 27 patients on ATT at cancer diagnosis, 22 (81%) had contact with a TB case and in 6 of 12 children (50%) a tuberculin skin test was positive. At cancer diagnosis, 16/27 (59%) children had chest radiograph changes interpreted as TB with 11/27 (41%) regarded as suggestive of TB on expert review. The median diagnostic delay between TB and cancer diagnoses was 25 days (interquartile range 3.5-58). Of 539 children with cancer, 204 (38%) died of cancer, including 18/30 (60%) children on ATT at cancer diagnosis or diagnosed with TB within 1 month of cancer diagnosis (odds ratio 2.6; 95% confidence interval: 1.2-5.4; P = 0.012). CONCLUSIONS: The clinical and radiologic overlap of TB and cancer causes diagnostic confusion in a significant number of children with cancer and may contribute to increased mortality.

Bronchoscopic assessment of airway involvement in children presenting with clinically significant airway obstruction due to tuberculosis.

INTRODUCTION: The incidence of complicated lymph node disease in tuberculosis (TB) in children less than 15 years of age varies from 8% to 38%. There are few published studies on the bronchoscopic appearance and severity of airway obstruction caused by lymph node involvement of the airways resulting from Mycobacterium tuberculosis (MTB). The primary aim of the study was to describe the flexible bronchoscopic findings of lymph node involvement of the airways caused by MTB in children with severe airway obstruction. The secondary aim was to compare the degree of airway involvement in HIV negative to HIV positive children as well the airway involvement caused by drug susceptible and drug resistant MTB. PATIENTS AND METHODS: All children between 1 month and 13 years of age presenting with clinical and radiological signs of significant airway obstruction suspected of being the result of MTB infection were studied. In addition to routine examination for MTB disease a flexible bronchoscope and bronchoalveolar lavage (BAL) for MTB culture were performed on all the children. RESULTS: Two hundred fifty children (16% HIV positive) were studied. Median age was 14 months and the median weight 8.5 kg. MTB was cultured from 78% (n = 194) of children with the BAL positive in 44%. The BAL culture yield was significantly higher in children with radiological evidence of pneumonia when compared to children with airway involvement alone (P = 0.004). The bronchial tree was obstructed on the right in 85% (n = 212), the left in 66% (n = 164), and both sides in 53% (n = 132) of cases. The commonest sites of obstruction were bronchus intermedius (72%) and left main bronchus (62%). Drug resistance was present in 16% (n = 28). There was no difference in the site or severity of obstruction when comparing drug susceptible to drug resistant cases or HIV positive to HIV negative children. CONCLUSIONS: Bronchus intermedius and left main bronchus were the commonest sites of airway obstruction. The MTB culture yield from BAL was higher in children with pneumonia when compared to those with airway involvement alone. HIV positive or children with drug resistant TB did not have more severe airway obstruction.