We present a case of tuberculosis (TB) presenting as chronic pneumonia with partial response to conventional antibiotics, complicated by pneumothorax secondary to cavity formation. A 2-year-old boy of Indian descent presented to the Children’s Hospital at Westmead, Sydney, Australia, with a 3-month history of recurrent fever and cough, not responding to oral antibiotics. He was born in Australia and had not travelled outside the country. The only neonatal complication of note was an Erb’s palsy following shoulder dystocia. He had no previous hospitalization or major illnesses and was thriving with a weight of 12 kg (weight for height on the 90th centile). His immunizations were all up to date.
On clinical examination, he displayed no signs of acute respiratory distress, although coryza was noted together with crepitations in the left lung base. His white cell count was 33.0 × 109/L (neutrophils, 20.5 × 109/L) with a C-reactive protein (CRP) level of 31 mg/L. Given the raised inflammatory markers, his clinical and chest radiograph findings were thought to be consistent with partially treated bacterial pneumonia. The patient improved with second-line intravenous antibiotics that included cefotaxime and lincomycin. No causative organism was isolated from blood culture or nasopharyngeal aspirates; mycoplasma serology was negative. He was discharged after 5 days of intravenous antibiotics, during which time he defervesced, and completed a 1-week course of oral amoxicillin–clavulanate at home.
His intermittent fever and chronic cough recurred after cessation of the oral antibiotics. On representation, 9 days after discharge, he was alert but mildly tachypneic (respiratory rate, 30–35/min) with some increased work of breathing. Minimal crepitations persisted in the left lung base, now associated with decreased air entry and dullness on percussion. A chest radiograph at this time showed persisting left lower lobe pneumonia with a possible pneumatocele, as well as a small left-sided effusion and hilar lymphadenopathy (Fig. 1). The white cell count remained elevated with a neutrophil count of 16.1 × 109/L; CRP, 46.4 mg/L. Restarting intravenous cefotaxime and lincomycin led to clinical improvement and resolution of tachypnea. A tuberculin skin test was strongly positive at 15 mm in a child who had not received Bacille Calmette–Guérin vaccine.
Early morning gastric washings were obtained. Bronchoscopy was performed to assess airway patency and to obtain a high-quality sample from the lower respiratory tract for microbiologic evaluation. Obstruction of subsegmental left lower lobe bronchi was noted, and the bronchoalveolar lavage specimen was positive for acid-fast bacilli on Ziehl–Neelsen staining. The presence of Mycobacterium tuberculosis was confirmed by polymerase chain reaction and culture. He was HIV uninfected. Although there was no history of recent TB exposure, the grandmother from Northern India cared for the patient in preceding months and returned to India before presentation. A diagnosis of complicated intrathoracic lymph node disease was made with extrinsic airway compression. Endobronchial TB could not be established, due to the lack of visible granulomas inside the bronchus.
The patient was commenced on standard TB treatment; rifampicin (100 mg daily), isoniazid (100 mg daily) and pyrazinamide (250 mg daily) during the 2-month intensive phase followed by rifampicin and isoniazid during the 4-month continuation phase. The organism was confirmed to be fully sensitive to these drugs. A 1-month course of oral prednisone (2 mg/kg/d) was added to try to reverse the associated airway obstruction. Less than 2 weeks after treatment initiation, the patient presented with acute respiratory distress, with a left-sided pneumothorax on chest radiograph (Fig. 1B). This was likely caused by spontaneous rupture of the left-sided parenchymal cavity. An intercostal chest drain was inserted which led to reexpansion of the left lung, but segmental left lower lobe collapse/consolidation persisted.
The subsequent course was characterized by recurrent fever and cough whenever amoxicillin–clavulanate was stopped, leading to a prolonged (6-week) course of oral antibiotics. Full clinical resolution occurred after 2 months of TB treatment, although follow-up bronchoscopy after 3 months of treatment revealed no improvement in the degree of endobronchial obstruction. He was treated with 3 months of oral prednisolone. Repeat bronchoscopy at the end of TB treatment (6 months) demonstrated excessive airway fibrosis and persistent obstruction of 2 subsegmental airways (Fig. 1C), accompanied by distal segmental collapse with expansion of the surrounding functional lung parenchyma. He has remained clinically well without recurrence of respiratory symptoms in the subsequent year.
TB is the most common infectious cause of death on the planet, being responsible for 1.8 million deaths in 2015.1 It is estimated that 1 million children developed TB in 2015,1 with most cases occurring in resource-limited settings where children have poor access to appropriate diagnostics and treatment. In these settings, TB is a common cause of death in young children, although these are mostly recorded as pneumonia, malnutrition or meningitis deaths. This case presented highlights several important teaching points.
First, TB is an important differential diagnosis in children with chronic pneumonia or poor response to an initial course of oral antibiotics, especially in TB endemic settings. The cooccurrence of TB disease infection with underlying TB and additional bacterial pneumonia can complicate the picture.2 A TB diagnosis was delayed by nearly 4 months from initial presentation in our patient, with ongoing disease progression contributing to the excessive fibrosis associated with the bronchial obstruction observed. After correct diagnosis and treatment, our patient still developed likely chronic bacterial pneumonia, distal to the bronchial obstruction. The abnormal white cell count (WCC) and CRP observed could have been due to TB alone but was presumed to be bacterial in origin given the clinical improvement documented with antibiotic treatment. Sputum and bronchoalveolar lavage samples did not isolate a bacteria but were performed while on antibiotics. The fact that children are more likely to develop primary M. tuberculosis infection with highly reactive regional lymph node enlargement, as well as the increased malleability and small size of their airways, increase the incidence of airway compression and other complications associated with intrathoracic lymph node disease.3 Endobronchial disease has been observed in around 40% of children with TB requiring brochoscopic evaluation for airway obstruction,4 with extrinsic lymph node compression—without endobronchial disease—in around 60%.5 Either can result in partial obstruction with distal hyperinflation or complete obstruction with distal collapse, potentially complicated by secondary bacterial infection. Endobrochial TB may also be associated with bronchial spread and TB bronchopneumonia.6 In children with airway obstruction resulting from possible TB disease, bronchoscopy has diagnostic and potential therapeutic value; careful consideration of infection control measures is essential.
The second teaching point is that unusual complications such as pneumothorax can occur. In our patient, this resulted from primary parenchymal disease with cavity formation. Pneumatocele development, due to the subsegmental ball valve effect produced by extrinsic compression, was considered to be less likely, given the thickness and irregularity of the cavity wall. Lung cavities in adults with pulmonary TB may precipitate a spontaneous pneumothorax,7 but lung cavities are uncommon in young children.6 However, young children may experience progressive parenchymal caseation with cavitation if there is uncontrolled progression of the primary Ghon focus or following complete bronchus obstruction with an expansile caseating pneumonia distal to the obstruction.8
The final teaching point is that the role of corticosteroids remains controversial in the treatment of associated airway obstruction. The role of corticosteroids in this situation is contentious9 with very little evidence of benefit in pediatric lung disease. In adults, steroid therapy has been shown to reduce mortality in patients with severe forms of TB,9 but only a small benefit was reported for pulmonary disease. In children, 1 small prospective study of airway obstruction due to hilar lymph adenopathy described faster clinical improvement and fewer complications when providing a course of oral prednisone.10 In our case, prednisone 2 mg/kg/d was initiated on diagnosis of airway obstruction. This higher dose was chosen due to concurrent induction of hepatic enzymes by rifampicin, which increases steroid metabolism by approximately 50%. The airway fibrosis and persistent obstruction—despite steroid treatment—observed on follow-up may place our patient at risk of recurrent infection and bronchiectasis in this region in the future. However, in most children, the affected lung segments collapse with expansion of neighboring healthy parenchyma leading to complete resolution without any long-term sequelae.11 Decompression of the external lymph nodes compressing the airway is a therapeutic option12 but was not considered in our patient. Despite achieving a good treatment response, our patient may have benefitted from increased TB drug dosages in accordance with revised international guidance [isoniazid 10 mg/kg (range, 7–15 mg/kg), rifampicin 15 mg/kg (range, 10–20 mg/kg), pyrazinamide 35 mg/kg (range, 30–40 mg/kg)].13
1. World Health Organization. Global Tuberculosis
Report. 2016.Geneva, Switzerland: World Health Organization;
2. Oliwa JN, Karumbi JM, Marais BJ, et al. Tuberculosis
as a cause or comorbidity of childhood pneumonia in tuberculosis
-endemic areas: a systematic review. Lancet Respir Med. 2015;3:235–243.
3. Marais BJ, Gie RP, Schaaf HS, et al. A proposed radiological classification of childhood intra-thoracic tuberculosis
. Pediatr Radiol. 2004;34:886–894.
4. Cakir E, Uyan ZS, Oktem S, et al. Flexible bronchoscopy for diagnosis and follow up of childhood endobronchial tuberculosis
. Pediatr Infect Dis J. 2008;27:783–787.
5. de Blic J, Azevedo I, Burren CP, et al. The value of flexible bronchoscopy in childhood pulmonary tuberculosis
. Chest. 1991;100:688–692.
6. Perez-Velez CM, Marais BJ. Tuberculosis
in children. N Engl J Med. 2012;367:348–361.
7. Winer-Muram HT, Rubin SA. Thoracic complications of tuberculosis
. J Thorac Imaging. 1990;5:46–63.
8. Goussard P, Gie RP, Kling S, et al. Expansile pneumonia in children caused by Mycobacterium tuberculosis
: clinical, radiological, and bronchoscopic appearances. Pediatr Pulmonol. 2004;38:451–455.
9. Critchley JA, Young F, Orton L, et al. Corticosteroids for prevention of mortality in people with tuberculosis
: a systematic review and meta-analysis. Lancet Infect Dis. 2013;13:223–237.
10. Toppet M, Malfroot A, Derde MP, et al. Corticosteroids in primary tuberculosis
with bronchial obstruction. Arch Dis Child. 1990;65:1222–1226.
11. Marais BJ, Gie RP, Schaaf HS, et al. The natural history of childhood intra-thoracic tuberculosis
: a critical review of literature from the pre-chemotherapy era. Int J Tuberc Lung Dis. 2004;8:392–402.
12. Goussard P, Gie RP, Janson JT, et al. Decompression of enlarged mediastinal lymph nodes due to mycobacterium tuberculosis
causing severe airway obstruction
in children. Ann Thorac Surg. 2015;99:1157–1163.
13. World Health Organization. WHO Guidance for National Tuberculosis
Programmes and the Management of Tuberculosis
in Children. 2014.2nd ed. Geneva, Switzerland: World Health Organization;