Langerhans Cell Histiocytosis – A Mimicker of Tuberculosis of the Spine


  • Conflict of interest: None.

10 November 2011

Dear Editor,

Langerhans cell histiocytosis (LCH) and tuberculosis are mimickers in the bone.[1-3] We present an interesting case of an 8-year-old Indonesian-Chinese boy, who presented with 1 month of low back pain and thoracolumbar tenderness. Imaging studies demonstrated T10 vertebral plana with preserved disc spaces and a paraspinal soft tissue mass causing cord compression. Bone scan revealed increased uptake over T10 but skeletal survey was unremarkable. Tumour resection, posterior instrumentation and fusion of the spine were performed. Histology was consistent with eosinophilic granuloma.

He re-presented 3 years later with right neck/shoulder pain, right arm weakness and limited abduction for 8 months with no other systemic symptoms. With a positive household contact of pulmonary tuberculosis (pTB) in Indonesia 2 months prior, positive Mantoux test (18 mm) and imaging studies (Fig. 1), he was diagnosed with cervical spine tuberculosis. Blood counts, erythrocyte sedimentation rate and chest X-ray were unremarkable. He was treated with anti-tuberculosis therapy (rifampicin, isoniazid, ethambutol and pyrazinamide for 6 weeks, then isoniazid monotherapy) with no adverse side effects. He underwent laminectomy/drainage of cervical spine abscess within the first week of anti-tuberculosis therapy.

Figure 1.

Pre-operative imaging studies at re-presentation. (a) Computed tomography cervical spine showing bony destruction of C3 and C4 vertebrae (red arrow) with (R) epidural soft tissue mass, possibly malignancy or granulomatous disease. (b) MRI cervical spine demonstrated vertebral plana at C3 (red arrow), with involvement of anterior pre-vertebral soft tissue, epidural intraspinal space and (R) paraspinal muscles and possible fluid collection.

Intra-operatively, soft friable tumour tissue eroding C3 lamina, compressing the spinal cord, could not be resected completely. Histology confirmed fibrosis with reparative bone formation and focal areas of LCH, positively staining S100 and CD1A. Ziehl–Neelsen stain and tuberculosis cultures were negative. The patient was commenced on LCH-III protocol chemotherapy for multifocal bone disease. A re-evaluation magnetic resonance imaging spine at 6 months showed no disease progression.

This is an interesting case where a LCH spine recurrence was misdiagnosed as tuberculosis of spine due to a positive contact of pTB, positive Mantoux test and imaging studies. To highlight the pitfalls in diagnosis, in a patient from Indonesia where pTB is rampant, the likelihood of ‘history of exposure’ to tuberculosis in the community is very high and does not necessary equate to true tuberculosis exposure. Furthermore, the median time for reactivation of unifocal bone disease in LCH has been reported to be 1 year;[4] in this case, a second involvement of another vertebra after several years is uncommon, which might have been an added challenge.

We highlight some indicators which may help in distinguishing the two conditions:

  1. Radiological investigations. Presence of vertebral plana with preserved disc spaces is typical of LCH, while the disc spaces will be lost in tuberculosis of the spine.[5] Despite the presence of vertebral plana with preserved disc spaces, the history of pTB contact and positive Mantoux test led to the misdiagnosis of tuberculosis spine in this patient.
  2. Histology. A bone biopsy for histology and mycobacterial culture is recommended to differentiate the two conditions and to exclude dual pathology.