INTRODUCTION

Tuberculosis (TB) is a serious health problem in poor countries and a leading cause of death. Approximately eight millions new cases of TB and three millions deaths are reported annually.¹ However, TB of central nervous system (CNS) is the most severe and life-threatening form of disease in children. Early diagnosis and prompt antituberculous therapy is important for optimal outcome, but diagnosing tuberculosis in young patients remains difficult.²The incidence of CNS TB is closely related to the prevalence of tuberculous infection in general. Initial focus of infection is usually in the lungs and through haematogeneous route tubercle bacilli spread to extra pulmonary sites. Tuberculous infection of CNS usually presents over weeks or months and because of insidious onset of symptoms, diagnosis may be delayed. Common clinical features of tuberculosis are fever, anorexia, weight loss, pallor, cough and night sweats. As the disease progresses, symptoms attributable to CNS appear. In case of tuberculomas features are headache, vomiting, seizures, cranial nerve palsies, weakness of one or more limbs and coma. Typical features of caries spine are backache, walking difficulty, weakness of one or more limbs and urinary and bowel problems. Although, TB is a widespread problem, data about intracranial tuberculomas and spinal spondylitis in children is limited.³ this study was undertaken to highlight the presence of relatively uncommon forms of CNS tuberculosis in young patients.

MATERIAL AND METHODS

Case records of children admitted in Children’s Hospital, Pakistan Institute of Medical Sciences (PIMS), Islamabad between January 1994 and August 2002 and diagnosed as cerebral tuberculoma or caries spine were reviewed. Those children who in addition also had tuberculous meningitis were excluded from the study.

Diagnostic criteria required for the diagnosis of tuberculomas was: consistent signs and symptoms for ³ 4 weeks and cranial CT scan showing single or multiple post contrast lesions with enhancing ring or uniform density with or without surrounding cerebral oedema and one of the following:

a. History of TB in the family

b. Chest x-ray (CXR) consistent with tuberculous infection of lungs

c. Induration of > 10 mm with Tuberculin skin testing after 48 to 72 hours irrespective of prior Bacillus-Calemette-Guerin (BCG).

e. Response to antituberculous therapy OR Mycobacterium tuberculosis (MTB) detected on polymerase chain reaction (PCR) OR Isolation of MTB from CSF or gastric aspirate on acid-fast stain or culture.

Criteria for the diagnosis of caries spine was:

Signs and symptoms consistent with tuberculosis
Clinical features of spinal cord compression,
Complete or partial collapse of one or more vertebrae on X-ray spine and
One of the criteria (a to e) used as above and
Magnetic resonance imaging (MRI) or CT spine showing destruction of intervertebral disc and adjoining vertebrae causing compression of spinal cord or myelography with obstruction at the level of vertebral collapse as seen on plain X-ray of the spine. Children younger than 12 years were included.

Laboratory evaluation included complete blood picture with erythrocyte segmentation rate (ESR), CSF analysis and culture, CXR, tuberculin skin test, serum electrolytes, liver and renal functions. Pre and post contrast CT scan of brain were performed on patients with suspected intracranial pathology. In children with vertebral involvement, spine was X-rayed in frontal and lateral views followed by either MRI/CT or myelography. Gastric aspirate for acid-fast stain and culture and blood or CSF PCR studies for MTB were obtained whenever possible. Gastric aspirates were obtained by either aspiration or lavage with 15 to 20 ml of distilled water for 3 consecutive mornings. Gastric aspirate specimens were sent for acid-fast staining and cultures. Sibs were screened in Outpatients with Mantoux test, ESR and CXR. Parents and other adult members of family had CXR whenever indicated.

Treatment was initiated with antituberculous therapy (ATT) and dexamethasone. Isoniazid (10-20 mg/kg), rifampicin (10-20 mg/kg) and pyrazinamide (30-40 mg/kg) was given orally in single daily dose and continued for twelve months. Streptomycin (20-30 mg/kg) was administered as intramuscular injection for the initial two months. Dexamethasone was administered intravenously or orally in full doses for three months and then tapered over the next 2 weeks. No surgical procedure was performed on children with intracranial tuberculomas irrespective of size, number or location and none of the Para vertebral abscess was drained.

RESULTS

17 children were diagnosed as having tuberculous infection of CNS during 9 years. Nine had cerebral tuberculomas and eight Pott’s disease. 11 were boys and age of the patients ranged from 10 months to 12 years with a mean of 5.5 years. Duration of symptoms before admission ranged from 1 to 12 months. Family history of TB was positive in 7 (41%) family members, mostly parents or grandparents. In patients with cerebral tuberculomas, common features were headache, vomiting, motor weakness of one or more limbs, cranial nerve palsies (3, 6 and 7^th) and abnormal movements. Presenting symptoms in case of caries spine were backache, neck pain, walking difficulty and frequent falls. On examination had gibbus, restricted movements of spine and pyramidal signs in lower limbs. Two patients had loss of sensations and a sensory level. Three children had been vaccinated against tuberculosis. Majority of children received ATT in inappropriate doses and duration with poor compliance before coming to the hospital. Hospital stay ranged from 2 to 5 weeks. Ten (58%) were malnourished and lived in overcrowded slums with unhygienic living environment.

CXRs were abnormal in 9 children and included diffuse infiltrates, hilar lymphadenopathy, consolidation and calcification. Eight patients had involvement of vertebrae, thoracic 4, cervical and lumbar region 2 each. Five of these had Para vertebral abscess, 3 thoracic and 2 cervical region. Some patients had more than one abnormality on CXR (Fig-1). ESR varied from 10 to 55 mm and in 5 (27%) patients Tuberculin skin test had induration of > 10 mm after 48 hours. Seven out of nine patients had positive PCR for MTB. Pre and post contrast CT scan brain were performed on 10 children. All but one scan were abnormal: 6 showed more than one tuberculomas and in 5 were bilateral (Fig. 2AB). Three had single tuberculoma and in two appeared as uniform density (Fig. 2C). Four patients with vertebral spondylitis had MRI showing cord compression with vertebral collapse (Fig. 3A) and myelography in two showed obstruction (Fig. 3B) and 2 CT scan spine also showed destruction of vertebrae and paravertebral abscess (Fig. 3C). Gastric aspirate results (twelve patients) were negative

Fig-1: Chest X-Ray showing large paravertebral abscess in cervicothoracic region and consolidation in left upper lobe. Lower vertebrae are affected as well (not shown)

Eight children, 3 with cerebral tuberculomas and five with vertebral spondylitis recovered completely or with minimal neurological deficit. Five children had paravertebral abscess, three of them quite large, and depending upon the size these abscesses resolved 6 to 18 months after the initiation of ATT.

Three patients were left with lower limb weakness and two of them also had kyphosis. Five children with tuberculomas were left with one or more of the neurological sequelae such as paraparesis, hemiparesis, seizures, blindness and cranial nerve palsies. One child died with cerebral tuberculomas.

DISCUSSION

Childhood tuberculosis is a serious health problem and an important cause of morbidity and mortality.⁴ Despite WHO’s universally recommended Directly Observed Treatment Strategy, Pakistan is one of the several developing countries where tuberculosis remains an highly endemic disease.⁵ Despite the fact that disease is widespread in Pakistan, culture-positive cases of TB in children are rare.³ Therefore, most cases are diagnosed on the basis of clinical features, family history and radiological abnormalities.⁶ In present series diagnosis was based upon epidemiological and clinical criteria. An effort was made to exclude other lesions, which can mimic cerebral tuberculomas and caries spine. In the presence of consistent clinical picture, one or more of the epidemiological evidence of tuberculosis such as family history of tuberculosis, abnormal CXR, neuroimaging abnormalities consistent with tuberculous infection of CNS, positive Tuberculin test and response to ATT was prerequisite to make the diagnosis of CNS tuberculosis.

Several reports of pulmonary tuberculosis and tuberculous meningitis in adults have appeared in recent years.^{7, 8} however; data about tuberculous infection of CNS in general and intracranial tuberculomas and vertebral spondylitis in particular in children is limited. To authors’ knowledge, this is the first case series about cerebral tuberculomas and caries spine in Pakistani children. In present series 9 patients had cerebral tuberculomas, in majority they were multiple and bilateral. CT scan brain with and without contrast was a very useful investigation. Intracranial tuberculomas are rare and their diagnosis is often delayed or overlooked. They are solid avascular nodular lesions with central caseation and on contrast CT brain are characterized by intense nodular or ring enhancement.⁹ The “target sign” defined as central calcification is nonspecific finding and is not pathgnomonic of CNS tuberculomas.¹⁰ Histologically central core is caseation necrosis surrounded by Langhans giant cells and epitheloid cells. Outer layer is capsule and consists of collangenous fibres. Histological findings correlate well with Gd-DTPA enhanced MRI abnormalities.¹¹ High index of suspicion is important, however, current neuroimaging techniques can differentiate other lesions which may simulate intracranial tuberculomas.¹²Intracranial tuberculomas usually do not require surgery and respond well to ATT and dexamethasone. If diagnosed early and treated promptly, are potentially curable tumours of CNS.¹³If for some reasons neuroimaging studies cannot be performed urgently, ATT and steroids may be initiated on empirical basis and final decision may be made when all relevant investigations are available.

Tuberculous osteomyelitis occurs in only 1 to 6 % of children with tuberculosis.¹⁴ Vertebrae are the most commonly affected bones,¹⁵ though, rarely cervical spine.¹⁶ Infection starts in the cancellous bone or anterior part of the vertebral body. Destruction of vertebral body leads to collapse and wedging anteriorly. Spinal cord compression is usually caused by the surrounding paraspinal abscess.¹³ Collapsed vertebral bodies, intervening discs and large paravertebral abscess are usually visible on plain spinal x-rays. Myelography, CT scan and now MRI can easily identify the exact level and extent of vertebral destruction, spinal cord compression or para vertebral abscess. Surgical treatment of vertebral tuberculosis is controversial. Surgery may be indicated for large abscesses, spinal stabilization, protection of vital structures and surgical decompression along with ATT and steroids are treatment of choice in patients with paraplegia.¹³ However, in present series none of the patient under went surgical procedure even with large para vertebral abscess and in almost two third of the cases recovery was optimal. Three of the 8 patients with spinal involvement were left with paraparesis and in addition two with some degree of kyphosis. All these children had severe vertebral destruction at the time of diagnosis. Reasons for poor outcome could be that in majority of patients lesion was quite extensive and none of the paravertebral abscess was drained. However, it seemed that size of the paravertebral abscess did not affect the outcome adversely, at least in three patients who had large paravertebral abscess and recovered. In present series children with tuberculous spondylitis presented with relatively shorter duration of illness and were of younger age group. Others have reported longer duration of illness, older age group and less common residual paraplegia (16%) in children with tuberculous spondylitis.¹⁷Clinical features of spinal tuberculosis may be apparent at the time of presentation, nevertheless, CXR as well as x-ray spine should be part of initial evaluation in a child with suspected CNS tuberculosis and look for evidence of pulmonary TB and vertebral spondylitis especially in younger patients. Occasionally spinal tumors may mimic caries spine, however, epidemiological evidence of TB and neuroimaging studies would help in differentiating other spinal pathologies. Raised ESR and Tuberculin skin reaction were positive in small number of children. We feel that these laboratory investigations were not helpful in the diagnosis of CNS TB. However, neuroimaging studies including myelography and PCR though very helpful in the diagnosis are expensive investigations.

Major limitation in our study was lack of laboratory evidence of tuberculous infection on acid-fast staining and culture. None of patient had gastric aspirates acid-fast staining or culture positive. Others have also reported low sensitivity of smears and cultures in gastric aspirate samples of children with TB.¹⁸ Children with extrapulmonary tuberculosis are less likely to be smear-positive but cultures are positive in one-third of the patients.¹⁹ Reasons for very low identification of mycobacteria on smear or culture are not clear. PCR is relatively new technique used for rapid diagnosis of TB. In present series PCR was a useful investigation and seven of out nine were positive for MTB. In a rather small series of intracranial tuberculomas PCR was found to be potentially useful approach for early and rapid diagnosis of cerebral tuberculosis even without meningitis.²⁰ Sensitivity of PCR is very variable and ranges from low to high .^2,21 PCR though a rapid and relatively new diagnostic technique is expensive and not readily available, therefore cannot be performed routinely. For these reasons tuberculosis is one of the infections where diagnosis often depends upon clinical judgment and more emphasis should be given to epidemiological evidence such as family history and abnormal CXR.

Very few children had been vaccinated with BCG in present study. In a meta analysis of large number of published data preventive value of BCG against all types of TB was approximately 50% and 64% in TBM.²² The protective efficacy varies from 0 to 80%.²³ Several case control studies have shown BCG efficacy against TBM between 85 and 100.^24,25 Neonatal BCG vaccination provides good protection against TBM and other disseminated forms of tuberculosis. Prevention of TB by widespread neonatal BCG immunization is a practical strategy that may be effective as well as cost effective. Every child with TB should have regular follow-up and administration of ATT in adequate doses and with good compliance. This must be explained to the parents and reinforced on every follow-up visit. Siblings and adult family members should be screened for TB.

CONCLUSION

Cerebral tuberculomas and spinal caries though not as common as pulmonary TB or TMB, are not rare in children. Therefore, it is important that treating paediatrician should have high index of suspicion. Positive family history and CXR may provide important clues towards the diagnosis of TB. Regular follow-up and strict treatment compliance are very important.

REFERENCES

1. Dolin PJ, Raviglione MC, Kochi A. Global tuberculosis incidence and mortality during 1990-2000. Bull World Health Organ 1994; 72:213-20.

2. Natalie Neu, Saiman L, Gabriel PS, Whittier S, Knirsch C, Ruzal-Shapiro C, et el. Diagnosis of tuberculosis in the modren era. Pediatr Infect Dis J 1999;18:122-6.

3. Qazi S, Khan S, Khan MA. Epidemiology of childhood tuberculosis in a hospital setting. JPak Med Assoc 1998;48:90-3.

4. Khan MA, Khan MM, Rehman GN. Tuberculosis in children. Pakistan Drug Information MOH, 1996 (July):19-22.

5. Ibrahim KM, Khan S, Laaser U. Tuberculosis control: current status, challenges and barriers ahead in 22 high endemic countries. J Ayub Med Coll 2002;14(3):11-5.

6. Starke JR, Taylor Watts KT. Tuberculosis in the pediatric population of Houston, Texas. Pediatrics 1998;84:28-35.

7. Aziz R, Khan AR, Qayum I, Ul Mannan M, Khan MT, Khan N. Presentation of pulmonary tuberculosis at Ayub Teaching Hospital Abbottabad. J Ayub Med Coll 2002;14(3):6-9.

8. Qureshi HU, Merwat SN, Nawaz SA, Rana AA, Malik A, Mahmud MK, et al. Predictors of inpatients mortality in 190 adult patients with tuberculous meningitis. J Pak Med Assoc 2002;52:159-63.

9. Tartaglione T, Di Lella GM, Cerase A, Leone A, Moschini M, Colosimo C. Diagnostic imaging of neurotuberculosis. Rays 1998;23:164-80.

10. Bargallo J, Berenguer J, Garcia-Barrionuevo J, Ubeda B, Bargallo N, Cardenal C, et al. The “Target Sign”: is it a specific sign of CNS tuberculoma? Neuroradiology 1996;38:547-50.

11. Inoue T, Nishino A, Uenohara H, Imaizumi S, Suzuki S, Suzuki H, et al. MR imaging of a case of cerebral tuberculoma-correlation between MRI and histological findings. No To Shinkei 1994;46:677-81.

12. Gupta RK, Kathuria MK, Pradhan S. Magnetization Transfer MR Imaging in CNS Tuberculosis. AJNR 1999;20:867-75.

13. Garg RK. Tuberculosis of central nervous system. Postgrad Med J 1999;75:133-40.

14. Davidson PT, Horowitz I. Skeletal tuberculosis. Am J Med 1970;48:77-84.

15. Bavadekar AV. Osteoarticular tuberculosis in children. Prog Pediatr Surg 1982;15:1131-51.

16. Hsu LCS, Leong JCY. Tuberculosis of lower cervical spine. J. Bone Joint Surg Br 1984;66:1-5.

17. Bailey HL, Gabriel M, Hodgson AR, Shin JS. Tuberculosis of spine in children. J Bone Joint Surg 1972;54:1633-57.

18. Kumar R, Singh SN, Kohli N. A diagnostic rule for tuberculous meningitis. Arch Dis Child 1999;81:221-4.

19. Vallejo JG, Ong LT, Starke JR. Clinical features, diagnosis and treatment of tuberculosis in infants. Pediatrics 1994;941:1-7.

20. Monno L, Angarano G, Romanelli C, Giannelli A, Appice A, Carbonara S, et al. Polymerase chain reaction for non-invansive diagnosis of brain mass lesions caused by Mycobacterium tuberculosis.. Tuber Lung Dis 1996;77:280-4.

21. Okutan O, Kartalogu Z, Ilvan A, Cerrahoglu K, Kunter E, Aydilek R. Polymerase chain reaction in pulmonary tuberculosis. East J Med 2000;5:7-9.

22. Fine PE. Variation in protection by BCG: implication of and for heterologous immunity. Lancet 1995; 346:1339-45.

23. Fine PE. The BCG story: lessons from the past and im-plications for the future. Rev Infet Dis 1989;11:353-9.

24. Miceli I, Kantor I, Colaiacovo D. Evaluation of effect-iveness of BCG vaccination using case-control method in Buenos Aires. Int J Epidemiol 1988;17(3):629-34.

25. Filho VW, de Castillo EA, Rodriques LC, Huttly SRA. Effectiveness of BCG vaccination against tuberculous meningitis: a case-control study in Sao Paulo, Brazil. Bull World Health Organ 1990; 68(2):69-74.