LIMPING CHILD: HIP TUBERCULOSIS IN THE DIFFERENTIAL
DIAGNOSIS
Erbil Ünsal*, Uluç Yiş**,
*Division of
Pediatric Immunology-Rheumatology, **Department of Pediatrics, ***Department of
Radiology, ****Department of Orthopaedics
and Traumatology, *****Department of Pathology
Corresponding
authors address:
Erbil Ünsal, M.D.
Associate Professor in Paediatrics
Department of Paediatric Rheumatology and Immunology,
Inciraltı, Izmir-TURKEY
Telephone: 00.90.232.4123674
e-mail: erbil.unsal@deu.edu.tr
Running title: Childhood Hip
Tuberculosis
Key words: arthritis, hip, children, tuberculosis.
Abstract
Osteoarticular tuberculosis is defined as direct infection of the bone and/or
joints by the tubercle bacillus. After the spine, the hip joint is the most
common site of involvement. Osteoarticular tuberculosis has become rare in
industrialized countries, but it is still a common problem in developing
countries. Since it has an insidious onset, the diagnosis is usually delayed or
it is not included in the differential diagnosis of juvenile idiopathic
arthritis. Delay in the diagnosis is an important cause of morbidity because of
the nature of the hip joint in a growing child. Although the definite diagnosis
is provided by identification of the bacillus in the specimens, this is not
always possible in childhood. Furthermore, the tubercle bacillus grows in only
46% of the obtained specimens. Tuberculosis of the hip must be in the
differential diagnosis of a limping child. A detailed family and environmental
history for pulmonary tuberculosis, history of tuberculous vaccination, and a
tuberculin skin test should be performed.
Introduction
Osteoarticular tuberculosis is
defined as direct infection of the bone and/or joints by the tubercle bacillus.
Osteoarticular tuberculosis is still an important cause of morbidity in
developing countries and the cases in industrialized countries are increasing
because of HIV, human migration, poverty and drug resistance. Although the
spine is a common target for osteoarticular tuberculosis, the peripheral bones
and joints also can be involved. Typically, arthritis arises on a background of
pulmonary tuberculosis as indolent, chronic monarthritis, often of the knee or
hip, that eventually results in extreme destruction of the joint and
surrounding bones. Rarely, the case presents as acute arthritis. (1)
In the differential diagnosis
of juvenile idiopathic arthritis, osteoarticular tuberculosis must always be
considered because clinical and radiological findings may be very similar.
Although synovial membrane biopsy and culture are the preferred methods to
confirm the diagnosis, both a family or environmental history of pulmonary
tuberculosis and a positive tuberculin skin test should suggest the possibility
of tuberculous arthritis (2). The Bacille Calmette Guerin vaccine (BCG) is
utilized in many countries and is usually, but not always, protective. In this
setting, a positive Mantoux may indicate either exposure to the vaccine or
actual TB disease. Because of these reasons, a detailed family history and high
clinical suspicion of the disease are key points for the diagnosis. Two cases
of tuberculous arthritis with hip involvement are presented here in which the
diagnosis was made on clinical, pathological and radiological findings.
Case 1
A four year old boy presented
with limping, pain on the posterior side of left knee and excessive sweating at
night for six months. The intensity of the pain gradually increased over time,
and he was not able to walk at presentation. He had no history of fever, joint
swelling, trauma or upper respiratory tract infection preceding the symptoms.
He had lost three kilograms of weight during this period.
On physical examination, he was
not able to walk and he appeared ill.
There was a leg length discrepancy and the left leg was 2.5 cm longer
then the right one. Left hip joint movements were restricted on flexion and abduction
and there was quadriceps atrophy. He had no BCG (Bacille Calmette Guerin)
vaccine scar. The remainder of the
physical examination did not reveal any pathological findings.
His white blood cell count was
9500/mm3 and the sedimentation rate was 133 mm/hr. He had mild
anemia (hemoglobin 11.2 g/dl, hematocrit 34.4% and MCV 70.2 fL). Peripheral
blood smear, platelet count, urinalysis, Brucella agglutination, liver and
renal function tests were unremarkable. Antinuclear antibodies and rheumatoid
factor were negative.
A radiograph of the left hip
joint showed destruction of the head of femur and a widened joint space due to
effusion (Figure 1). A T2 weighted axial MR image
revealed a fluid with high protein content in the coxafemoral joint that could
be consistent with the ongoing inflammation (Figure
2a). Contrast enhanced fat suppressed T1 weighted coronal image
of the left hip joint demonstrated abnormal thick synovial enhancement (Figure 2b). In a chest radiograph, a calcified
lymph node was noted in the left hilum (Figure 3).
Tuberculin skin test revealed a 15x13 mm induration. The detailed family
history revealed that his uncle who was in close contact with the patient had
received anti-tuberculous therapy for pulmonary tuberculosis five years
earlier.
An open drainage of the left
coxafemoral joint was deemed preferable to simple joint aspiration. The open
drainage was performed and pathological examination of the specimen showed a
granulomatous inflammation consistent with tuberculosis. Neither mycobacterium
tuberculosis nor any other pathogen could be grown in the cultures of surgical
specimens.
Based on history, radiological
and pathological findings, tuberculous arthritis was diagnosed and he was
treated with rifampin, isoniazid and pyrazinamide for nine months. There was a
remarkable improvement in the gait of the patient and in the radiological findings
at the follow up examination.
Case 2
A two year old boy presented
with pain on the posterior aspect of right knee and limping. His symptoms
started one month before and he had no history of fever, joint swelling, trauma
or upper respiratory tract infection. On physical examination, movements of the
right hip joint were painful and restricted on flexion and abduction. He had no
BCG vaccine scar. The remainder of the physical examination was unremarkable.
The sedimentation rate was 15
mm/hr and the white blood cell count was 12,900/mm3. Other
laboratory tests including hemoglobin, platelet count, peripheral blood smear,
urinalysis, Brucella agglutination, liver and renal function tests were in
normal range. Antinuclear antibodies and rheumatoid factor were negative.
Radiographs of the hip joints
showed a cystic lesion in the right proximal metaphysis. A periarticular
osseous structure was also radiolucent (Figure 4).
T2 weighted axial image of the right hip showed increased
heterogeneous signal intensity on the metaphysis of the femur. There was no
irregularity or pathological fracture in the bone and the appearance of the
epiphyseal bone was normal. There was a minimal capsular effusion (Figure 5a). There was decreased signal in the
same region of the femur on T1 weighted coronal image (Figure 5b). The chest radiograph was
unremarkable. The tuberculin skin test revealed a 13x13 mm induration.
A detailed family history
revealed that a grandfather who was in close contact with the patient had
received anti-tuberculous therapy for pulmonary tuberculosis two years ago.
Rather than joint aspiration, definitive diagnosis and treatment began with a
surgical curettage. Pathological
examination revealed a granulomatous inflammation consistent with tuberculosis.
There was no growth of tubercle bacilli or any other pathogens in the cultures
of surgical specimens. He was treated with rifampin, isoniazid and pyrazinamide
for nine months. At follow up examination, there was a distinct improvement in
his gait and in the radiological findings.
Discussion
Peripheral osteoarticular
tuberculosis represents 1-5% of cases of tuberculosis at any site (3). The
incidence of tuberculosis is rising, even in industrialized countries (4). Poor
socioeconomic conditions constitute a major risk factor (5). Under these
conditions, contact with adult tuberculous patients and failure to receive the
BCG vaccine are the most important risk factors. Although dozens of BCG trials
have been reported in various human populations, the most useful data have come
from several controlled trials. Some demonstrated a great deal of protection
from BCG vaccines, but others showed no efficacy at all. A recent meta-analysis
of published BCG vaccination trials suggested that BCG is 50% effective in
preventing pulmonary tuberculosis in adults and children. The protective effect
for disseminated and meningeal tuberculosis appears to be slightly higher, with
BCG preventing 50-80% of cases (6). In our cases, neither received a BCG
vaccine and there were adult patients in their families who received
anti-tuberculous treatment.
The tubercle bacillus can reach
the joint from an adjacent tuberculous osteomyelitis or by hematogenous spread from a
pulmonary lesion (7). Most of the cases with osteoarticular tuberculosis are
reported to have tuberculosis at other sites, usually the lungs (1, 8). In
childhood hip tuberculosis, the chest radiographs are positive for active or
healed tuberculosis in 42% of the patients (9).
The disease has an insidious
onset. Pain, restriction of motion in the region of the hip and limping are the
usual presenting symptoms. Sometimes the child complains of pain in the knee.
This is referred pain and is often misleading, as in both of our cases. There
may be constitutional symptoms such as loss of appetite, loss of weight,
low-grade fever and easy fatiguibility during physical activities.
Hip and knee joints are the
most common sites of involvement. Teklali and Ellis et al. reported hip or knee
involvement in 63% and 87% of their patients respectively (1, 10). Elbow and
ankle involvement were next most common in their series. In our cases, both
children had hip involvement. In 1983,
Shanmugasundaram classified hip tuberculosis into the following
clinico-radiological types; 1) normal type, 2) Perthes type, 3) dislocating
type, 4) arophic type, 5) aetabular protrusion type, 6) mortar and pestle type.
This classification is of prognostic value and helps in the choice of
management. Since there was no finding of atrophy or dislocation in the hips of
our two cases, the radiological appearance of both cases was classified as the
normal hip type.
The
radiological appearance of the tuberculous hip at presentation almost
invariably predicts the outcome. Hips of normal, Perthes and dislocating types
have a good result but hips of atrophic, protrusion-acetabular and
mortar-pestle types have a poor result (9). Both of our patients had a normal
clinico-radiologic type of hip disease and the results with early, aggressive
anti-tuberculous treatment appeared to be excellent.
Identification of the tubercle
bacillus in the specimens is the gold standard method in diagnosis (11), but
this is not always possible in children. Histopathological findings, such as
granulomatous inflammation, combined with clinical findings should suggest
tuberculosis in synovial biopsy specimens. In the study of Teklali et al., the diagnosis
was obtained by identification of tubercle bacillus in only 46% of patients and
by histopathological findings in 77% of cases (1). There was no growth of
tubercle bacilli or any other pathogens in the cultures of our cases. Although
we could not use PCR for mycobacterium and ELISA testing of gamma interferon,
the diagnosis of our cases were supported by the clinical, histopathological
and radiological findings.
Plain radiographs are very
important for diagnosis, but they can be normal early in the disease (12-13).
Since it is difficult to distinguish tuberculosis from other diseases, such as
inflammatory disease or neoplasia, CT or MRI should be useful for these cases.
In both of our cases, MRI findings were consistent with ongoing inflammation and
these MRI findings resolved with anti-tuberculous treatment.
The differential diagnosis of
tuberculous arthritis includes transient (or toxic) synovitis, traumatic
synovitis of the hip, septic arthritis, Perthes disease, oligoarticular
juvenile idiopathic arthritis (JIA), enthesitis-related JIA, and diseases that
cause a granulomatous inflammation.
In both our cases, transient
synovitis was not considered because there was no history of a preceding upper
respiratory tract infection and duration of the symptoms was longer. In
addition, radiological examination was not consistent with transient synovitis,
which usually causes only a minimal widening of the joint space. Neither of our
patients had a history of a recent trauma.
Bacterial septic arthritis was excluded mainly
by the absence of toxic appearance of the children and the lack of findings of
acute inflammation on histological examination of tissue from the affected
joint. It should be noted that when TB appears to be the mostly likely cause of
an osteoarthritis, joint aspiration should be bypassed in favor of the more
definitive surgical curettage and drainage.
Juvenile idiopathic arthritis
is one of the most important diseases that should be considered in differential
diagnosis of these cases. All the symptoms and radiological findings of the
patients could be seen in the oligoarticular type of juvenile idiopathic
arthritis, (including spondyloarthopathy), but the pathological examination of
the specimens were consistent with granulomatous inflammation and the response
to anti-tuberculous treatment was excellent. Malignant or benign lesions were
also excluded by pathological examination of the specimens. Although the
clinical presentation of case one was similar to the Perthes disease,
radiological and pathological findings were not suggestive. In addition to tuberculosis, granulomatous inflammation may also be
found in a few other pathological states, including infections, foreign body
reaction, silicosis and sarcoidosis. Among these
conditions, tuberculosis should be the first to be considered in differential
diagnosis in children.
In conclusion, tuberculous
arthritis is by no means rare in the world, especially in developing countries.
The diagnosis of hip tuberculosis is not easy, and if not diagnosed in the
early stages of the disease, it is an important cause of morbidity in growing
children because of the nature of hip joint. In the differential diagnosis of
oligoarticular type juvenile idiopathic arthritis, hip tuberculosis must always
be considered because of its clinical and radiological similarity and different
treatment implications. Synovial membrane biopsy and culture are the gold
standard to confirm the diagnosis of joint TB.
However, it is not always possible in young children to obtain this
confirmation because tubercle bacillus can only be grown in less than 50% of
childhood cases of tuberculous arthritis. The history of receiving a BCG
vaccine and supposed protection from TB can be misleading. Therefore, in the
evaluation of a limping child, if there is the history of an adult with
tuberculosis in the family and the child has a positive tuberculin skin test
(whether or not the child has received BCG), the clinician should be aware of
the possibility of hip tuberculosis and appropriate assessment should be
immediately performed, especially radiologic studies and tissue attainment.
References
1.
Teklali Y, El Alami Z, El Madhi T et al. Peripheral osteoarticular tuberculosis in
children: 106 case-reports. Joint Bone Spine 2003; 70:282-86.
2.
Zahraa J, Johnson D, Lim- Dunham JE, et al. Unusual features of
osteoarticular tuberculosis in children. J Pediatr. 1996; 129:597-602.
3. Perronne C, Bernard L. La
tuberculose osteoarticulaire aujourdhui. Presse Medicale 1997; 26:308-10.
4.
5. Benbouazza K, Allali F, Bezza
A. Losteoarthrite pubbienne tuberculeuse a propos de deux cas. Rev Chir Ortho
1997; 83:670-2.
6. Starke JR, Munoz F.
Mycobacterial Infections, Tuberculosis.
Behrman ER, Kliegman MR, Jenson HB (Eds), Nelson Textbook of Pediatrics,
16th edition, WB
Saunders
Company,
Philadelphia 2000, pp 885-97.
7. Gonzalez H, Farrington DM, Angulo G. Peripheral
osteoarticular tuberculosis in children: tumor-like bone lesion. J Pediatr
Orthop B 1997; 6: 274-82.
8. Hugosson C, Nyman RS, Brismar J. Imaging of
tuberculosis. Acta Radiol 1996; 37:512-16.
9.
10. Ellis
ME, Ramahi KM, Dalaan AN. Tuberculosis of peripheral joints: a dilemma in diagnosis. Tuber Lung Dis
1993; 74 (6):399-404.
11. Lemaitre F, Damade R, Pouchot J et al.
Tuberculose osteo articulaire des membres. App Loc 1979; 141:10885 A.
12. Hsu SH, Sun JS, Chen IH. Reapprasial of skeletal tuberculosis: role of radiological imaging. J
Formos Med Assoc 1993; 92:34-41.
13. Haygood
J and Williamson W: Radiographic findings of extremity tuberculosis in
childhood: back to the future? Radiographics 1994; 14:561-
570
Figures:
Figure
1: A radiograph of the left hip
joint demonstrates destruction of the head of femur and a widened joint space
due to effusion.
Figure 2a: An MRI with a T2 weighted axial image reveals a left
coxafemoral joint effusion.
Figure 2b: An MRI showing an abnormal thickened synovium utilizing contrast
enhancement of the left hip joint on T1 weighted image.
Figure 3: An anteroposterior chest radiograph shows a left hilar calcified lymph
node.
Figure 4: An anteroposterior right hip radiograph shows a radiolucent, well defined
cystic lesion on the right femur metaphysis.
Figure 5a: T2 weighted axial MRI image of the
hip joints demonstrates an increased signal on the right femur metaphysis.
Figure 5b: T1 weighted coronal image shows hypointense
signal of the same region.
