Destructive Diskovertebral Lesions in Ankylosing Spondylitis: Appearance on Magnetic Resonance Imaging.

ABSTRACT: We report magnetic resonance imaging findings of diskovertebral lesions in a case of ankylosing spondylitis mimicking metastatic and/or infectious disease. Multiple hypointense areas were seen on T1-weighted images corresponding to hyperintense areas on T2-weighted images in dorsal, lumbar, and sacral vertebral bodies and the manubriosternal joint, with accompanying soft tissue masses. Diagnosis was achieved through biopsy, regression of the paravertebral soft tissue masses, later detection of bilateral sacroiliitis on computed tomography, and presence of histocompatibility antigen HLA-B27.

DEFINITIVE CLINICAL and radiographic signs of bilateral inflammatory involvement of sacroiliac (SI) joints is the standard for diagnosis of ankylosing spondylitis (AS). [1] Many articles in the English-language radiology literature have verified or compared the diagnostic value of different radiologic modalities in the detection of the pathologic changes in SI joints. However, to the best of our knowledge, magnetic resonance imaging (MRI) features of diskovertebral lesions in AS have received little attention. Andersson [2] and Edstrom [3] in 1937 and 1940 first delineated radiologic abnormalities of diskovertebral lesions in AS. Later, Cawley et al [4] classified these erosive and destructive vertebral lesions into three types. In 10% to 15% of cases, patients may have destructive diskovertebral lesions before the changes of bilateral sacroiliitis settle down. [1] At this stage, there is a diagnostic dilemma in differentiating these cases from metastatic or infectious processes solely on the basis of radiolo gic findings. We report such a case.

CASE REPORT

In September 1998, an 18-year-old woman came to the orthopedic clinic with back and chest pain of 6 months' duration. On physical examination, tenderness was present in the dorsal vertebral region over the spinal processes and paraspinal muscles with pressure. The patient had restriction in motion and stiffness. No neurologic symptoms were found. She had lost 4 kg in the previous 2 months and had low-grade fever (37.2[degrees]C). A dorsolumbar spine radiograph revealed paravertebral soft tissue masses that caused prominent erosive lesions in the left halves of the mid-dorsal vertebral bodies, including the articular faces (Fig 1). The erythrocyte sedimentation rate was elevated (35 mm/hr). There was no leukocytosis (white cell count, 8,400/[mm.sup.3]). All other blood chemistry values were within normal limits.

To better appreciate the extent of disease, an MRI examination was planned. Sagittal T1- and T2-weighted images and axial T2- and coronal T1-weighted MRI scans were obtained using a 1-T imager (Somatom Impact, Siemens, Erlangen, Germany), which revealed erosive-destructive vertebral lesions at the mid-lower dorsal, lumbar (L3 through 5), and sacral (S1 and S2) vertebral bodies, with soft tissue masses in paravertebral and epidural spaces. There was also a manubriosternal mass with adjacent osseous erosion (Fig 2). Contrast-enhanced images were not obtained at this time. Fine needle aspiration biopsy of the sixth dorsal vertebral body and the accompanying soft tissue mass was done using computed tomography (CT) guidance. The pathologic finding was necrosis with inflammatory cell infiltration. Microbiologic as well as blood cultures for tuberculosis, Brucella, and Staphylococcus aureus all yielded negative results. The patient then had CT of the thorax to rule out tuberculosis. Neither pulmonary parenchymal in filtration nor cavitary lesions were detected. The differential diagnosis included metastatic and infectious disease, but there was neither clinical nor laboratory evidence to support these conditions. The patient was discharged from the hospital and seen on an outpatient basis.

Six months later, dorsolumbar MRI was repeated with the same protocol as before. All the dorsal, lumbar, sacral, and manubriosternal soft tissue masses were seen to have completely disappeared, while the pathologic intensities (hypointense on T1-weighted images and hyperintense on T2-weighted images) in the vertebral bodies remained (Figs 3 and 4). On radiography, the destructive pattern in the end plates of the vertebral bodies were seen, but there was no facet joint ankylosis or syndesmophyte formation (Fig 4). A slight thoracal scoliosis had developed. At this time, the patient's only complaint was back pain. She had continued her usual activities. A nuclear scintigraphy examination revealed increased activity throughout the spinal column and in SI regions bilaterally. The latter drew the attention of the clinician, and the patient was again referred to the radiology department for CT examination, which showed bilateral narrowing in the SI joint spaces with erosions, sclerosis, and a small, right-sided oss eous bridge (Fig 5). After the detection of bilateral sacroiliitis, testing for histocompatibility antigen HLA-B27 was done and results were positive.

DISCUSSION

Ankylosing spondylitis is a chronic inflammatory disorder of unknown cause that affects principally the axial skeleton, though the appendicular skeleton may also be significantly involved. Alterations occur in synovial and cartilaginous articulations and in sites of tendon and ligament attachment to bone. [1] Ankylosing spondylitis is a common cause of back pain and disability, especially in young men. However, the latest studies suggest that the ratio of men to women with bilateral sacroiliitis is approximately 1:1 in those individuals who have the histocompatibility antigen HLA-B27. The onset generally occurs between the ages of 15 and 35 years.

Clinical manifestations related to the spine and the SI joints are characteristic of AS. Typically, changes initially are seen in SI joints and next appear in the thoracolumbar spine in an ascending fashion. Transient, aching pain and stiffness of variable intensity in the low back are observed, which may subsequently become persistent. Peripheral articular manifestations are apparent initially in approximately 10% to 20% of the patients. The sternoclavicular and manubriosternal joints may also be involved. [1]

The New York criteria for clinical diagnosis of AS were established in 1968 and were later modified. [5,6] Bilateral sacroiliitis is the hallmark of AS. Definitive AS is present if bilateral sacroiliitis is associated with at least one of the clinical criteria. In general, inflammatory cell infiltration and necrosis, which are nonspecific, are the histopathologic expressions of the disease. The diagnosis is established by history and clinical and laboratory findings, with findings of bilateral sacroiliitis using radiologic modalities.

In the radiology literature, most investigators are interested in showing the pathologic changes in the SI joints. Many prospective studies have compared the different radiologic modalities. [7-14] However, imaging features of the diskovertebral lesions in AS have received little attention. [15-17] To our knowledge, the appearance on MRI of the destructive type of diskovertebral lesions in AS has not been described.

Andersson [2] in 1937 and Edstrom [3] in 1940 delineated the radiologic abnormalities of diskovertebral lesions in AS. Later, Cawley et al [4] classified these erosive and destructive vertebral lesions into three types: those that involve the central portions of the diskovertebral junction, which are covered by the cartilaginous end plate (type 1); those that involve the peripheral portions of the diskovertebral junction, which are not covered by the cartilaginous end plate (type 2); and those that involve both peripheral and central portions of the diskovertebral junction (type 3). However, pathologic correlation was not established. In our case, all three types of diskovertebral lesions were observed (Figs 1 and 4). Histopathology of the destructive lesions revealed nonspecific inflammatory cell infiltration with necrosis.

These destructive changes are widespread and may involve many vertebrae at the same time. At this stage, radiologic workups, such as MRI, can result in a diagnostic dilemma or misdiagnosis. On MRI examination, hyperintense areas on T2-weighted images, corresponding to hypointense areas on T1-weighted images in many vertebrae, can be seen with or without accompanying soft tissue masses. A radiologist would probably diagnose metastatic or infectious disease (probably tuberculosis). When the diskovertebral lesions are unaccompanied by SI joint changes, the presence of AS is difficult to establish. Later, with the appearance of sacroiliitis and more typical spinal abnormalities, the diagnostic dilemma is resolved, as happened in our case.

Multiple vertebral destructions with paravertebral and epidural soft tissue masses were present in our patient. Biopsy of these vertebral lesions and the soft tissue masses revealed inflammatory cell infiltration and necrosis. There were no malignant cells, nor any microorganism overgrowth in culture plates. Clinically stable, long-standing disease in this healthy-looking patient (who complained only of back pain), with negative culture results and no malignant cells found on biopsies, made the diagnosis of metastatic or infectious disease suspect. It was on nuclear scintigraphy that bilateral increased uptake in the SI areas alerted the clinician. A CT examination of the SI joints showed the classic findings of AS, such as joint space narrowing, erosions, sclerosis of the iliac and sacral bones, and osseous bridging. Later, histocompatibility antigen HLA-B27 was found. According to modified New York criteria, [5] our case was accepted to be AS.

CONCLUSION

Bilateral sacroiliitis with back pain are the hallmarks of AS. It is easy to make the diagnosis by showing sacroiliitis with different radiologic modalities in such a situation. However, back pain is increasingly primarily investigated by MRI, and SI joint examination is overlooked. Patients who have undiagnosed AS may be found to have destructive diskovertebral lesions. Awareness of this possibility is needed to avoid misdiagnosis. In cases such as ours, we suggest inclusion of AS in the differential diagnosis, as well as examination of the SI joints.

From the Department of Radiology, Istanbul University, and the Department of Physical Medicine, Istanbul SSK Hospital, Istanbul, Turkey.

Reprint requests to Sebuh Kurugoglu, MD, Istanbul University, Cerrahpasa Medical Faculty, Department of Radiology, 34300 Istanbul, Turkey.

References

(1.) Resnick D, Niwayama G: Ankylosing spondylitis. Diagnosis of Bone and Joint Disorders. Philadelphia, WB Saunders Co, Vol 5, 2nd Ed, 1988, pp 1103-1169

(2.) Andersson O: Roentgenbilden vid spondylarthritis ankylopoetica. Nord Med Tidskr 1937; 14:2000

(3.) Edstrom G: Is spondylarthritis ankylopoetica an independent disease or a rheumatic syndrome? Acta Med Scand 1940; 14:396

(4.) Cawley MID, Chalmers TM, Kellgren JH, et al: Destructive lesions of vertebral bodies in ankylosing spondylitis. Ann Rheum Dis 1972; 31:345

(5.) Bennet PH, Wood PHN: Population studies of the rheumatic diseases. Proceedings of the Third International Symposium. New York, Excerpta Medica Foundation, 1968, p 456

(6.) van der Linden SJ, Valkenburg HA, Cats A: Evaluation of diagnostic criteria for ankylosing spondylitis: a proposal for modification of the New York criteria. Arthritis Rheum 1984; 27:361-368

(7.) Tu W, Feng F, Dion E, et al: Comparison of radiography, computed tomography and magnetic resonance imaging in the detection of sacroiliitis accompanying ankylosing spondylitis. Skeletal Radiol 1998; 27:311-320

(8.) Konig H, Sauer R, Deimling M, et al: Cartilage disorders: a comparison of spin echo, CHESS and FLASH sequence MR images. Radiology 1987; 164:753-758

(9.) Bollow M, Braun J, Taupitz M, et al: CT-guided intraarticular corticosteroid injection into the sacroiliac joints in patients with spondylarthropathy: indication and follow-up with contrast enhanced MRI. J Comput Assist Tomogr 1996; 20:512-521

(10.) Docherty P, Mitchell MJ, Mac Millan L, et al: Magnetic resonance imaging in the detection of sacroiliitis. J Rheumatol 1992; 19:393-401

(11.) Battafarano DF, West SG, Rak KM, et al: Comparison of bone scan, computed tomography and magnetic resonance imaging in the diagnosis of active sacroiliitis. Semin Arthritis Rheum 1993; 23:161-176

(12.) Blum U, Buitrago-Tellez C, Mundinger A, et al: MRI for the detection of active sacroiliitis: a prospective study comparing conventional radiography, scintigraphy and contrast enhanced MRI. J Rheumatol 1996; 23:2107-2115

(13.) Bollow M, Braun J, Biedermann T, et al: Use of contrast enhanced MR imaging to detect sacroiliitis in children. Skeletal Radiol 1998; 27:606-616

(14.) Oostveen J, Prevo R, den Boer J, et al: Early detection of sacroiliitis on magnetic resonance imaging and subsequent development of sacroiliitis on plain radiography, a prospective, longitudinal study. J Rheumatol 1999; 26:1953-1958

(15.) Tyrell PN, Davies AM, Evans N, et al: Signal changes in the vertebral discs on MRI of the thoracolumbar spine in ankylosing spondylitis. Clin Radiol 1995; 50:377-383

(16.) Marc V, Dromer C, Le Guennec P, et al: Magnetic resonance imaging and axial involvement in spondylarthropathies. delineation of the spinal enthesis. Rev Rheum Engl Ed 1997; 64:465-473

(17.) Remedios D, Natali C, Saufuddin A: A case report: MRI of vertebral osteitis in early ankylosing spondylitis. Clin Radiol 1998; 53:534-536

KEY POINTS

* Magnetic resonance imaging features of the diskovertebral lesions in ankylosing spondylitis have received little attention.

* When destructive changes involve many vertebrae, magnetic resonance imaging can result in a diagnostic dilemma.

* A radiologist will probably diagnose metastatic or infectious disease when diskovertebral lesions are unaccompanied by sacroiliac joint changes.

* Awareness of this possibility is needed to avoid misdiagnosis.