Tuberculous sacroiliitis:
A case report and review of the literature

O.D. SAVVIDOU[2], E.CH. PAPADOPOULOS[1], P.J. PAPAGELOPOULOS[1]
[1]Department of Orthopaedics, Athens University Medical School, [2]"Hygeia" Hospital Athens

ABSTRACT
A 32 year-old man with tuberculous sacroiliitis associated with tuberculosis of the greater trochanter is reported. The patient had right buttock pain and a progressively growing mass on the lateral aspect of his right proximal and mid-thigh. Imaging modalities revealed erosions of the right sacroiliac joint, destruction of the greater trochanter and a large mass of the thigh. Open biopsy, histology, cultures and polymerase chain reaction test established the diagnosis of tuberculosis. After debridement of the tuberculous abscess, the patient was treated successfully with a triple antibiotic regimen for 12 months. Five years postoperatively, the patient is disease-free and with no functional limitation.
Key words: Skeletal tuberculosis, tuberculous sacroiliitis, tuberculosis of the greater trochanter.

INTRODUCTION
Tuberculosis affects worldwide one-third of the world's population and is the most common infectious cause of death[1]. On the contrary, its incidence in the United States[2] is declining after an almost two decades interval of increased rates[6]. Despite this decline, in some urban areas in the United States, the incidence of tuberculosis is comparable to that of developing countries as a result of high-risk population segregation[40].
Due to the optimism succeeded the development of antibiotics in the 1950's, tuberculosis was often overlooked as a possible underlying diagnosis of skeletal infection[41]. The lack of high index of suspicion and the scant experience of the medical community[40] accounts for the average delay of 17 months between the initial symptoms and the definite diagnosis of skeletal tuberculosis[41].
One to 5 percent of all the recorded tuberculosis cases involves the musculoskeletal system[5,10]. Sacroiliac joint tuberculous infection is reported in 3-9.7% of the skeletal tuberculosis cases[10,29], while the reported incidence in developing countries is negligible[24].
Sacroiliac joint infection represents a diagnostic dilemma for its vague and unspecific clinical presentation[15]. Herein, the authors report a patient with chronic tuberculosis of the sacroiliac joint that remained undiagnosed for several years, until the involvement of the greater trochanter and the development of an ipsilateral thigh abscess. To the authors' knowledge, this combination of tuberculous lesions has not been reported before.

1. 2.
Figure 1. Anteroposterior radiograph of the pelvis in a 37 year-old man showing erosions of the right sacroiliac joint and demarcation of the cortex of the right ischial tuberocity.
Figure 2. Bone scintigraphy with 99mTc MDP exhibiting increased radioisotope uptake at the right sacroiliac joint, the right trochanteric area and ischial tuberocity.

3.
Figure 3. Preoperative anteroposterior radiograph of the pelvis showing destructive lesion of the right trochanter and right ischial tuberocity.

CASE REPORT
A 37-year old man, school teacher in a rural area, has been admitted at the authorsÕ institution complaining of right buttock pain and an enlarging mass over the lateral aspect of his right hip and mid-thigh.
The right buttock pain started five years ago after a fall. At that time, plain radiograph of the pelvis showed no fracture; however, there were extensive erosions of the right sacroiliac joint and faint calcifications of the ipsilateral ischial tuberocity (figure 1). Tc99-bone scintigraphy demonstrated high radioisotope uptake at the right sacroiliac joint and at the ipsilateral greater trochanter (figure 2). The sacroiliac joint lesion was attributed by his local physician to a past systemic brucellosis infection and the increased radioisotope uptake at the region of the greater trochanter was attributed to his fall. The patient refused any further investigation because his pain was adequately managed with non-steroid anti-inflammatory drugs (NSAIDs).
Twelve months before his admission to the authors' institution his buttock pain worsened and a thigh mass emerged. The pain was radiating in the ipsilateral calf especially during stair climbing and intense walking. The patient reported episodes of low-grade fever and night sweats twice a year (spring and autumn) usually of an average duration of 2 to 3 weeks that were treated symptomatically.
His past medical history included systematic brucellosis with positive agglutination test 10 years ago. This was treated successfully with the antibiotics. Interestingly, before the definite diagnosis of brucellosis, the patient had blindly received a 3-week course of intramuscular streptomycin that arrested his high fever and devastation (loss of 10 kilograms of body weight in a month). His family history was negative for tuberculosis.
The patient walked with an antalgic gait. Clinical examination revealed a painless mass over the lateral aspect of his right hip and mid-thigh. The inner thigh and the ipsilateral perineal area were also mildly swollen. There was tenderness in deep palpation over the right sacroiliac joint. Lateral pelvic compression test, Patrick and Gaenslen test provoked right buttock pain. Right hip joint motion was painful at the extreme. Right straight leg raising test was positive at 500. There was no motor or sensitivity deficit of the lower extremity.
Plain radiograph of the pelvis showed additionally to the sacroiliac joint lesion a significant erosion of the right greater trochanter (figure 3). Computed tomography of the pelvis confirmed the presence of right sacroiliac joint erosions and extensive destruction of the ipsilateral trochanter and ischium associated with scattered calcifications (figure 4A,B). In addition, magnetic resonance imaging (MRI) demonstrated a large mass at the right mid-thigh having features of an abscess (figure 5A,B). Chest radiograph was negative.

4A. 4B.
Figure 4A. Axial computer tomography image showing erosions of the articular surface of the right sacroiliac joint. B. Axial CT-image depicting destruction at the right trochanter and ischial tuberocity. Soft tissue calcifications are delineated in both sites.

Erythrocyte sedimentation rate (ESR) was 5mm/1h and C-reactive protein (CRP) was 0.9mg/dl (normal value, >0.5mg/dl). There was a mild hypochromic, microcytic anemia; hematocrit was 39.5%. Slide agglutination test with Brucella antiserum was negative. Tuberculin skin test was positive (10mm); however the patient had a history of BCG vaccination.
Open biopsy was performed through a lateral incision over the right trochanter proximal thigh. Extended debridement of the mass and greater trochanter resulted in soft, friable, whitish-gray debris resembling clumped cheese-like material. Direct smear and stain of the material showed acid-fast bacilli. Histology was consistent with a granulomatous infection. Polymerase chain reaction (PCR) showed amplification of the M. tuberculosis genome and confirmed the diagnosis of tuberculosis. Finally, cultures in Lšwestein-Jensen medium isolated M. tuberculosis.
Postoperatively, the patient started immediately chemotherapy with a triple antituberculous regimen including isoniazid (5mg/kg of body weight), rifampin (10mg/kgr of body weight), and pyrazinamide (35mg/kg of body weight).
Patient received chemotherapy for a total of 12 months. Within this period, his symptoms resolved completely. Subsequent MRI evaluation showed complete resolution of the abscess. At the latest follow-up, 5 years postoperatively, the patient was pain-free with no functional limitations. Plain radiograph of the pelvis demonstrated spontaneous ankylosis of the sacroiliac joint (figure 6).

5A. 5B.
Figure 5A. Coronal T-2 weighted MRI of the pelvis showing right hip and proximal thigh soft tissue mass, associated with destruction and significant edema of the right trochanter. B. Axial T-2 weighted MRI showing a soft tissue mass of the right proximal thigh.

Figure 6. Postoperative anteroposterior radiograph of the pelvis five years after the treatment. The sacroiliac joint is spontaneously ankylosed; there are residual changes of the right trochanter and ischial tuberocity.

DISCUSSION
The inhaled Mycobacterium tuberculosis is the responsible pathogen in most of the tuberculosis cases, as the ingested M. bovis is largely eradicated[32]. A history of tuberculous infection or exposure is not always present; nevertheless a history of local trauma is reported in 30-50% of the cases of skeletal tuberculosis[11,32,41].
Skeletal involvement occurs mainly by the hematogenous route from a primary pulmonary infection particularly in children. When it occurs later in life, hematogenous seed arises either from a dormant pulmonary infection or from another extraosseous secondary focus. Reactivation of a quiescent, healed skeletal lesion may occur in several occasions, such as general debilitation or local trauma as occurred in the present patient[32].
Tuberculous sacroiliitis is generally considered sub-acute infection[14,33] and only exceptionally acute or intermittent[31]. The present case of sacroiliac joint infection should be characterized as chronic, considering the protracted history of symptoms. The slow onset of symptoms, the mild intensity of the pain and paucity or absence of constitutional symptoms usually differentiates sub-acute from acute pyogenic infection of the sacroiliac joint[12,34]. Early recognition of an acute sacroiliitis is important, since an anterior joint capsule rupture and a subsequent retroperitoneal irritation may produce signs of atypical appendicitis[22,30], retroperitoneal abscess and urinary track infection[22].
Constitutional symptoms such as night pain, fever, night sweats, weight loss and fatigue may be absent[8,14,30,31,33,38,39], hence chronic pain may be the sole symptom of tuberculous arthritis[32]. Buttock pain is invariably present in tuberculous[8,14,19,30,31,33,39] or pyogenic[9,12,13,34] sacroiliac joint infections. Pain is often associated with a low back element or is radiating to the leg. Richter et al[33] reported that 92% of their patients had been initially treated for lumbosacral radiculopathy; misdiagnosis can lead to needless lumbar spine surgery[33]. Pain is aggravated by weight bearing during walking[9,19,30,31,33] or while performing more intensive activities[33]. Prolonged sitting[33] or bending[8] may also aggravate pain. Intensification of pain with sneezing or coughing is not always a reliable differentiating sign, as it may[22] or not[8] increase the leg pain. In the present patient, the trochanteric lesion was not associated with local pain, a finding consistent with other reported cases of isolated trochanteric tuberculosis[4,20,23].
Although sacroiliac joint pain is basically of somatic type[3], resulting from the joint destructive process, it is almost universally deep, ill-defined and poorly localized and may suggest referred type of pain. Thus differential diagnosis may include pain originated from lumbar spine, hip and lower abdominal quadrate[9,13,15,31,38]. Sacroiliac joint related pain is usually reproduced or enhanced when the sacroiliac joint is mechanically strained during weight bearing or physical exam. However, sacroiliac pain may also radiate to the lower extremity either as referred[2] or as neuralgic. In the latter condition, there is irritation of the nerve structures, that cross anteriorly the sacroiliac joint, either by the infection per se or by the adhesions or the bulging anterior capsule that may compress the nerves and cause nerve ischemia[8].
An enlarging mass was one of the main differential diagnosis issues in the present patient. Abscesses represent the commonest complication of mycobacterial[14,19,30,31,33] and bacterial[9,12,13,15,34,36] sacroiliac joint infection, ensuing from pus released after the rapture of the sacroiliac joint anterior capsule; subsequently pus follows various paths of low resistance. It may fill the iliopsoas muscle sheath and irritate the peritoneum, ascent to the lumbar spine or tracking the iliopsoas tendon descent to the hip or in the inguinal area. Similarly it may follow the piriformis muscle sheath to the buttock and to all the structures anteriorly the gluteus maximus. The purulent material may also penetrate the pelvic floor and discharge through the rectum or vagina[33]. Due to the lesion extent and the atrophic performs muscle seen in the present patient's MRI, the authors assume that a sequential invasion occurred, spreading from the sacroiliac joint, through the piriformis muscle, to the greater trochanter and its bursa. Moreover, through the loose connective tissue between the rotators and the gluteus maximus the pus spread to the ischial tuberocity and its bursa.
Sinuses when present may contribute to earlier diagnosis. Although sinuses were reported often in the past emerging either during the disease course or after the surgical drainage[38], this complication is rare in the antibiotic era[19,31,33].
A delayed diagnosis is attributed to the inaccessibility of the sacroiliac joint while the patient is examined in supine position[15] as well as to the physicians' failure to perform the sacroiliac joint pain provocation tests[9]. Erythema over the joint is never observed and should not be anticipated[3]. Tenderness over the sacroiliac joint is a universal sign[13-15,26]. Pain provocation tests[21] such as pelvic compression and distraction tests, pelvic torsion test (Gaenslen's test), Patrick test (FABER test) yield positive results in the affected site. Passive hip motion is often painfully limited in the extremes. This results from motion transmission to the sacroiliac joint; this type of pain should be differentiated from the near total painful limitation of the septic hip arthritis[13].
Additionally, hip motion may irritate muscles already inflamed by the sacroiliac pyarthrosis or a soft tissue abscess. A possible antalgic spasm of these muscles may splint the hip joint and inhibit a proper sacroiliac joint examination[30]. Iliopsoas muscle may cause of hip flexion contracture that may draw attention away from the sacroiliac joint[9]. Tenderness on rectal examination is possible[15,22]. Straight leg raising test often elicits pain[8,13,15,31,33,34], possibly due to stretching of the inflamed sacral nerves as they pass along the anterior sacroiliac joint15. In the case reported by Chen[8], sciatica was also aggravated with the extension of the hip joint, as occurred in the present patient.
Tuberculous sacroiliitis should be differentiated from degenerative arthritis, post-traumatic arthrosis, infectious or inflammatory arthritis and tumors[22]. The unilateral disease, the negative history of joint trauma and the absence of additional manifestations from other systems should raise the suspicion of infective sacroiliitis[22,32].
In the present case, ESR was normal, CRP was slightly elevated and the patient had mild anemia of chronic disease. Half of the patients with tuberculous sacroiliitis exhibit no significant laboratory findings[33]. This is often observed in skeletal tuberculosis[29]. Tuberculin skin testing should be careful interpreted. Although a negative skin test usually excludes the diagnosis[22], a positive one is of little help in the diagnosis of skeletal tuberculosis[32], indicating a previous infection but not necessarily an active one. Additionally BCG vaccinees as the present patient, become tuberculin skin test positive, which eliminates the usefulness of the test. A negative chest radiograph is not helpful in excluding skeletal tuberculosis in the adult patient[32], as only half of the patients will have concomitant pulmonary tuberculosis[11,22].
Pelvic radiograph depicted in the present patient in addition to the sacroiliac joint lesion a large destructive lesion in the greater trochanter, which had to be differentiated from a bone tumor[27]. The multiplicity of the lesions, the rounded sclerotic osteolysis of the greater trochanter and the scattered calcifications in the soft tissue allowed the authors to suspect an infectious process of low virulence, such as tuberculosis, brucella and fungal infection[11].
Early radiographic sacroiliac joint findings may be either normal or minimal[8,22,31,33]. Kim et al[19] classified tuberculous sacroilitis into four types based on the clinical and radiographic findings: widening of the joint space and blurring of the margin of the joint (type I), joint erosions (type II), and severe destruction, with cyst formation and sclerosis (type III). The sacroiliac joint lesion combined with an affected vertebra or an abscess as in the present patient was classified as type IV.
Computer tomography is considered superior to bone scintigraphy for an early diagnosis of bacterial sacroiliac joint infection[26]. For an early recognition of tuberculous sacroiliitis, these two modalities are considered of equal value[22,25]. Computer tomography findings such as joint sclerosis and bone erosions displayed in the present patient are characteristic of tuberculous infection of the joint. Additional information such as the extent of an abscess and the involvement of the adjacent structures facilitates the planning of a surgical debridement[19,22,31,39].
In the present patient, bone scintigraphy depicted a unilateral sacroiliac joint disease. Bone scintigraphy is a reliable, sensitive technique for the early detection of both pyogenic[9,15] and tuberculous[35] sacroiliac joint infection. Despite its low specificity[22], a combination of 99mTc MDP and Gallium-67 citrate scintigraphs proved adequate in establishing the diagnosis and monitoring patients with bacterial sacroiliitis[18]. There are no similar data for tuberculous sacroiliitis.
Magnetic resonance imaging provides preoperative information about the extent of the disease in multiple planes[19] and may contribute in differentiating an abscess from a soft tissue tumor. In addition, MRI may be useful in the differentiation of tuberculous arthritis from pyogenic arthritis, based on differences in bone erosion, marrow and synovial lesion signal intensity, boundaries for extraarticular infection and abscess rim enhancement[17].
Any monoarticular sacroiliac joint process, must be regarded infectious until proven otherwise[22,31]. The best way to establish the diagnosis of tuberculous sacroiliitis is to identify the mycobacterium within the joint by biopsy. Biopsy can be omitted only when tuberculous infection is already diagnosed in other body site[19]. There are several ways to obtain material, such as open biopsy that provides abundant material while necrotic tissues are removed, closed needle aspiration[16,28] that is direct joint fluid aspiration and closed needle biopsy that provides osteoarticular fragments suitable for both pathology and cultures[30,31]. Acid-fast bacilli in direct smear and stain, the growth of the bacilli in the Lšwestein-Jensen culture or the granulomatous lesion identified in the histologic specimen will confirm the diagnosis of tuberculosis. However, negative results should be anticipated in long standing cases[33]. Positive culture result and an antibiogram are highly desirable, because fungal infection and brucellosis yield similar histologic findings[22,31]. Although skeletal tuberculosis is frequently isolated with no other organ involvement, the culture or histology of other tissues and secretions, such as sputum will support the diagnosis; sputum culture can be positive despite normal chest radiography[31].
Multiple drug therapy is employed to delay or prevent emergence of resistant bacilli strains. Isoniazid and rifampicin are the cornerstone in treatment of skeletal tuberculosis. These drugs as also streptomycin, pyrazinamide, ethambutol and thiacetazone are characterized as "first line" drugs because of their efficacy and the acceptable degree of toxicity; all except the latter two are bacteriocidal. To address the problem of drug resistance, three bacteriocidal drugs should be combined and treatment has to be maintained for at least a six months[37]. Non-judicious use of antibiotics and poor patient compliance are the leading cause for the raise of multidrug-resistant organisms[40]. Kidney function and auditory acuity when prescribing streptomycin and visual acuity when prescribing ethambutol, should periodically be assessed.
Before the absolute advent of chemotherapy, sacroiliac joint arthrodesis was considered of value in hastening the end result, which regardless the treatment was the spontaneous ankylosis of the joint[38]. Kim et al[19] described two methods of surgery, the one of which is joint curettage in the early stages of the disease and the second joint curettage combined with arthrodesis. The latter is the preferred method when the disease is advanced and joint instability is anticipated. All except one of their patients were immobilized in a hip-spica cast for an average of 3 months; only one patient developed a sinus that eventually healed after sinusectomy. Curettage serves in eradicating pus and necrotic tissues and potentially augments the drug effectiveness by increasing the blood flow into the lesion[19]. Richter et al[33] operated on 44 patients with excellent results in terms of residual disease and pain relief. A sinus emerged in two patients postoperatively. In the present patient, the authors elected to operate directly on the abscess site and to treat the sacroiliac joint lesion conservatively. The greater trochanter was consciously debrided, as aggressive intervention may spread the tuberculous infection to the hip joint[25].
The present patient had an excellent outcome fulfilling the healing criteria set by Kim et al[19]. He became progressively pain-free over the lesion site, he had no pain or discomfort during walking, his abscess disappeared and his radiograph revealed clearance and sclerosis of the joint margin and fusion of the sacroiliac joint. Analogous response to the either conservative or surgical treatment is reported in the literature[8,22,33,38,39]. Recurrences are generally not anticipated[19,33]. Of the 43 patients that Richter et al[33] reported, only one, a labor, could not resume his previous activity and had to be retrained.
This case highlights the importance of continued vigilance for the early detection of a tuberculous sacroiliac joint infection. Two factors are implicated in delayed diagnosis. One is the low index of suspicion due to the scant experience of the medical community with skeletal tuberculosis[40]. The second is the vague and unspecific clinical presentation of tuberculous sacroiliitis. Although the disease is not deadly, as half century ago[38], it is a devastating infection that merits accurate recognition and treatment.

REFERENCES
1. Babhulkar S.S. Editorial comment. Clin Orthop. 2002; 398, 2-3.
2. Bernard T.N, Jr., Kirkaldy-Willis WH. Recognizing specific characteristics of nonspecific low back pain. Clin Orthop. 1987; 217, 266-80.
3. Bonica J.J. Anatomic and physiologic basis of nociception and pain. Lea & Febinger, 1990.
4. Brown N.L. Radiologic case study. Tuberculosis of the greater trochanter and its bursa. Orthopedics. 1986; 9, 9, 1276, 1279-80.
5. Center for Disease C. Tuberculosis-United States 1981. Morbidity and Mortality Weekly Report. 1981; 30, 325-6.
6. Center for Disease C. Tuberculosis-United States, 1992. Morbidity and Mortality Weekly Report 42. 1993; 36, 696-7.
7. Center for Disease C. Tuberculosis morbidity among U.S.-Born and Foreign-Born populations-United States, 2000. Morbidity and Mortality Weekly Report. 2002; 51, 05, 101-4.
8. Chen W.S. Chronic sciatica caused by tuberculous sacroiliitis. A case report. Spine. 1995; 20, 10, 1194-6.
9. Coy J.T., 3rd, Wolf C.R., Brower T.D., Winter W.G., Jr. Pyogenic arthritis of the sacro-iliac joint. Long-term follow-up. J Bone Joint Surg Am. 1976; 58, 6, 845-9.
10. Davies P.D., Humphries M.J., Byfield S.P., et al. Bone and joint tuberculosis. A survey of notifications in England and Wales. J Bone Joint Surg Br. 1984; 66, 3, 326-30.
11. Davidson P.T., Horowitz I. Skeletal tuberculosis. A review with patient presentations and discussion. Am J Med. 1970; 48, 1, 77-84.
12. Delbarre F., Rondier J., Delrieu F., et al. Pyogenic infection of the sacro-iliac joint. Report of thirteen cases. J Bone Joint Surg Am. 1975; 57, 6, 819-25.
13. Dunn E.J., Bryan D.M., Nugent J.T., Robinson R.A. Pyogenic infections of the sacro-iliac joint. Clin Orthop. 1976; 118, 113-7.
14. Feldmann J.L., Menkes C.J., Weill B., Delrieu F., Delbarre F. [Infectious sacroiliitis. Multicenter study of 214 cases]. Rev Rhum Mal Osteoartic. 1981; 48, 1, 83-91.
15. Gordon G., Kabins S.A. Pyogenic sacroiliitis. Am J Med. 1980; 69, 1, 50-6.
16. Hendrix R.W., Lin P.J., Kane W.J. Simplified aspiration or injection technique for the sacro-iliac joint. J Bone Joint Surg Am. 1982; 64, 8, 1249-52.
17. Hong S.H., Kim S.M., Ahn J.M., Chung H.W., Shin M.J., Kang H.S. Tuberculous versus pyogenic arthritis: MR imaging evaluation. Radiology. 2001; 218, 3, 848-53.
18. Horgan J.G., Walker M., Newman J.H., Watt I. Scintigraphy in the diagnosis and management of septic sacro-iliitis. Clin Radiol. 1983; 34, 3, 337-46.
19. Kim N.H., Lee H.M., Yoo J.D., Suh J.S. Sacroiliac joint tuberculosis. Classification and treatment. Clin Orthop. 1999; 358, 215-22.
20. King A.D., Griffith J., Rushton A., Metreweli C. Tuberculosis of the greater trochanter and the trochanteric bursa. J Rheumatol. 1998; 25, 2, 391-3.
21. Laslett M., Williams M. The reliability of selected pain provocation tests for sacroiliac joint pathology. Spine. 1994; 19, 11, 1243-9.
22. Ludmerer K.M., Kissane J.M., eds. Clinicopathologic conference. Severe right hip pain in a 73-year-old woman. Am J Med. 1986; 81, 1, 117-24.
23. Lynch A.F. Tuberculosis of the greater trochanter. A report of eight cases. J Bone Joint Surg Br. 1982; 64, 2, 185-8.
24. Martini M., Ouahes M. Bone and joint tuberculosis: a review of 652 cases. Orthopedics. 1988; 11, 6, 861-6.
25. McNeur J.C., Pritchard A.E. Tuberculosis of the greater trochanter. J Bone Joint Surg Br. 1955; 37, 2, 246-51.
26. Morgan G.J., Jr., Schlegelmilch J.G., Spiegel P.K. Early diagnosis of septic arthritis of the sacroiliac joint by use of computed tomography. J Rheumatol. 1981; 8, 6, 979-82.
27. Moujtahid M., Essadki B., Lamine A., et al. Tuberculous osteitis in limbs. Twenty five case-reports. Rev Rhum Engl Ed. 1996; 63, 5, 344-8.
28. Miskew D.B., Block R.A., Witt P.F. Aspiration of infected sarco-iliac joints. J Bone Joint Surg Am. 1979; 61, 7, 1071-2.
29. Nicholson R.A. Twenty years of bone and joint tuberculosis in Bradford. A comparison of the disease in the indigenous and Asian populations. J Bone Joint Surg Br. 1974; 56, 4, 760-5.
30. Osman A.A., Govender S. Septic sacroiliitis. Clin Orthop. 1995; 313, 214-9.
31. Pouchot J., Vinceneux P., Barge J., et al. Tuberculosis of the sacroiliac joint: clinical features, outcome, and evaluation of closed needle biopsy in 11 consecutive cases. Am J Med. 1988; 84, 3 Pt 2, 622-8.
32. Resnick D., Niwayama G. Osteomyelitis, septic arthritis, and soft tissue infection: organisms. In: D.R, ed. Diagnosis of bone and joint disorders. Philadelphia, PA: W.B. Saunders Co, 1995:2461-85.
33. Richter R., Nubling W., Kobler G., Iljinski A. [Tuberculosis of the iliosacral joints. Treatment, results, prognosis and differential diagnosis]. Z Orthop Ihre Grenzgeb. 1983; 121, 5, 564-70.
34. Rosenberg D., Baskies A.M., Deckers P.J., Leiter B.E., Ordia J.I., Yablon I.G. Pyogenic sacroiliitis. An absolute indication for computerized tomographic scanning. Clin Orthop. 1984; 184, 128-32.
35. Salomon C.G., Ali A., Fordham E.W. Bone scintigraphy in tuberculous sacroiliitis. Clin Nucl Med. 1986; 11, 6, 407-8.
36. Shanahan M.D., Ackroyd C.E. Pyogenic infection of the sacro-iliac joint. A report of 11 cases. J Bone Joint Surg Br. 1985; 67, 4, 605-8.
37. Shembekar A., Babhulkar S. Chemotherapy for osteoarticular tuberculosis. Clin Orthop. 2002; 398, 20-6.
38. Soholt S.T. Tuberculosis of the sacro-iliac joint. J Bone Joint Surg Am. 1951; 33, 1, 119-30.
39. Vaiopoulos G., Sfikakis P.P., Velikas E., Kittas C., Kaklamanis P. Tuberculosis of the sacroiliac joint. Eur Spine J. 1997; 6, 5, 330-1.
40. Watts H.G., Lifeso R.M. Tuberculosis of bones and joints. J Bone Joint Surg Am. 1996; 78, 2, 288-98.
41. Yao D.C., Sartoris D.J. Musculoskeletal tuberculosis. Radiol Clin North Am. 1995; 33, 4, 679-89.

Mailing address:
Panagiotis Papagellopoulos
4 Christovasili Str.
15451 Athens
Ôel./fax: 210-6721355
e-mail: pjp@hol.gr