Differential Diagnosis of Intraspinal and Extraspinal Non-discogenic Sciatica

Available online at www.sciencedirect.com

www.elsevier.com/locate/jocn

Journal of Clinical Neuroscience 15 (2008) 1246–1252

Clinical Study

Differential diagnosis of intraspinal and extraspinalnon-discogenic sciatica

Duygu Geler Kulcu a,*, Sait Naderi b

a Department of Physical Medicine and Rehabilitation, Yeditepe University School of Medicine, Yeditepe University Hospital,

Devlet Yolu Ankara Caddesi No. 102/104 Kozyatagı, Istanbul, Turkeyb Department of Neurosurgery, Yeditepe University School of Medicine, Istanbul, Turkey

Received 10 December 2007; accepted 5 January 2008

Abstract

The aim of this study is to present a series of 11 patients with non-discogenic sciatica (NDS), and to review the diagnostic techniquesof careful clinical and radiological examination. The cases include lumbar radicular herpes zoster, lumbar nerve root schwannoma,lumbar instability, facet hypertrophy, ankylosing spondylitis, sacroiliitis, sciatic neuritis, piriformis syndrome, intrapelvic mass andcoxarthrosis. The pain pattern and accompanying symptoms were the major factors suggesting a non-discogenic etiology. PelvicMRI and CT scans, and sciatic nerve magnetic resonance neurography were the main diagnostic tools for diagnosis of NDS. The treat-ment of choice depended on the primary diagnosis. Detailed physical examinations with special attention paid to the extraspinal causesof sciatica and to pain characteristics are the major components of differential diagnosis of NDS.� 2008 Elsevier Ltd. All rights reserved.

Keywords: Magnetic resonance neurography; Non-discogenic sciatica; Piriformis syndrome; Sacroiliitis; Sciatica

1. Introduction

Sciatica is common, and is frequently caused by lumbardisk herniation.1 However, some intraspinal or extraspinalpathologic processes along the sciatic nerve may also causesciatica. Whereas lumbar spine imaging reveals the causesof intraspinal non-discogenic sciatica (NDS), extraspinalsciatica is often misdiagnosed because routine diagnostictests focus on the lumbar spine.2 Extrapelvic causes affectthe nerve as it progresses distally from the sciatic notch.A careful patient history and clinical examination areimportant in identifying extraspinal sciatica. Further diag-nostic imaging may clarify the diagnosis.

In general, studies of NDS tend to focus on only one dis-order.1–7 This study reports a series of 11 patients with

0967-5868/$ - see front matter � 2008 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jocn.2008.01.017

* Corresponding author. Tel.: +90 216 5784100; Mob.: +90 5058575178.

E-mail address: [email protected] (D.G. Kulcu).

NDS, and overviews the different causes of intraspinaland extraspinal sciatica.

2. Patients

The patients, 6 female and 5 male, were aged between 25and 65 years old. There were 4 patients with extraspinalNDS, 4 with intraspinal NDS, and 3 with sciatica second-ary to both spinal and extraspinal processes.

2.1. Intraspinal non-discogenic sciatica

2.1.1. Lumbar radicular herpes zoster (Patient 1)

A 64-year-old woman was admitted with a 1-week his-tory of left leg pain that was not responding to analgesics.A neurological examination revealed no abnormal findings.A lumbar MRI showed degenerative changes (Fig. 1a). Aphysical examination revealed typical skin lesions alongthe L3 and L4 dermatomes (Fig. 1b). The patient was re-ferred to the dermatology department, where the lesions

mailto:[email protected]

Fig. 1. (A) Sagittal T1-weighted lumbar MRI showing degenerative changes. (B) Skin lesions along the trajectories of the L3 and L4 nerve roots.

Fig. 2. (A) Sagittal postcontrast T1-weighted lumbar MRI, and (B) axial fat-saturated T1-weighted lumbar MRI showing multiple schwannomatosis.

D.G. Kulcu, S. Naderi / Journal of Clinical Neuroscience 15 (2008) 1246–1252 1247

were determined to be caused by herpes zoster. Afterreceiving medication for herpes zoster, the patient’s symp-toms resolved.

2.1.2. Schwannomatosis (Patient 2)

A 27-year-old female patient was admitted with a 2-month history of low back pain (LBP) and left leg pain.Physical examination revealed a positive straight leg raising(SLR) test at 45�, and hypoesthesia at the left S1 derma-tome. Lumbar MRI showed multiple schwannomatosis ofthe bilateral L5 and left S1 nerve roots (Fig. 2). The symp-toms improved after medical therapy. Because of the lackof neurological deficit and the small size of the schwanno-mas, surgical treatment was not indicated.

2.1.3. Facet syndrome and lumbar disk herniation

(Patient 3)A 63-year-old woman was admitted with a history of

LBP, left leg pain and numbness. The SLR test was nega-

tive. The patient had pain with extension of the trunk.There was no neurological deficit. A lumbosacral antero-posterior (AP) radiograph showed sclerotic and degenera-tive changes in the L5–S1 facet joint on the left side. Alumbar MRI showed a L4–L5 and L5–S1 central disc pro-trusion. After physical therapy and flexion exercises, thesymptoms resolved partially.

2.1.4. Lumbar instability (Patient 4)A 31-year-old male was admitted with a 12-year history

of LBP and a 1-month history of left leg pain. Physicalexamination showed a positive SLR test at 60�. The patientwas mistakenly diagnosed as having a lumbar disc hernia(LDH) at another center and was treated with physiother-apy, which was not beneficial. Sagittal lumbar MRIshowed no disc herniation, and axial MRI showed pars de-fects of L5. Oblique radiographs showed bilateral pars int-erarticularis defects. The LBP was attributed to the lumbarinstabilty related to the isthmic defects.

1248 D.G. Kulcu, S. Naderi / Journal of Clinical Neuroscience 15 (2008) 1246–1252

2.2. Extraspinal cases

2.2.1. Sciatic neuritis (Patient 5)

A 64-year-old woman was admitted to hospital with leftbuttock pain. The patient had undergone surgery for lum-bar spondylolisthesis 10 years previously. The results of aphysical examination were unremarkable. A lumbar MRIshowed left sciatic neuritis between the piriformis muscleand the proximal femur (Fig. 3).

2.2.2. Bilateral sacroiliitis (Patient 6)

A 25-year-old female was admitted with a 2-month his-tory of LBP, left-side buttock and posterior thigh pain. Thepatient had no lower limb weakness or paresthesias andhad no history of morning stiffness or night pain. On phys-ical examination the SLR test was limited to 60� on the leftside. There was no other neurologic deficit in the lowerlimbs. Her range of motion of the trunk was not limitedand was not painful. Deep palpation of the left buttockover the sciatic nerve course was painful. Gaenslen’s testwas positive on the left side. The sacroiliac compressiontest was positive on the left side. MRI of the lumbar spineand sacroiliac joint showed L4–L5 disc protrusion with no

Fig. 3. Axial fat saturated T2-weighted lumbar MRI showing hyperin-tensity at the left sciatic nerve.

Fig. 4. (A) Coronal STIR-weighted and (B) axial postcontrast fat-saturated T

nerve root compression and bilateral sacroiliitis. Thesciatica was attributed to the sacroiliitis. The patient reci-eved indomethacin at 150 mg/day for 15 days. Thepatient’s symptoms resolved. The visual analogue scale(VAS) pain score decreased from 7 to 2. Further analyseswere performed to identify the etiology of the sacroiliitisand the patient was diagnosed with seronegativespondyloarthropathy.

2.2.3. Sacroiliitis (Patient 7)

A 53-year-old woman was admitted with left leg pain.Physical and neurological examinations yielded no abnor-mal findings. Lumbar MRI showed no abnormality. Sacro-iliac MRI showed sacroiliitis and edema in the jointcompressing the sciatic nerve (Fig. 4). The erythrocyte sed-imentation rate (ESR) was 88 mm/h. After further exami-nation, the patient was diagnosed with seronegativespondyloarthropathy. The patient started to recieve indo-methacin at 300 mg/day, salazosulphapyridine at 400 mg/day and prednisolone acetate at 7.5 mg/day. Pain de-creased and the ESR was 68 mm/h after 2 weeks. Symp-toms completely resolved after completion of the medicaltreatment.

2.2.4. Soft tissue tumor (Patient 8)

A 55-year-old male was admitted with a 1-month historyof right buttock pain. The patient had undergone L4–L5discectomies five and three years previously. Neurologicalexamination revealed no deficit. The buttock was painfulupon palpation. Lumbar spine MRI showed modicchanges at the L4–L5 level. A pelvic MRI showed a 6 cmsoft tissue mass anterior and right to the sacrum. Examina-tion of a pelvic CT scan revealed destruction of the anteriorsurface of the sacrum (Fig. 5). A CT-guided needle biopsyrevealed an angiosarcoma.

2.2.5. Piriformis syndrome and hamstring tendinopathy

(Patient 9)A 59-year-old woman was admitted with a 2-month his-

tory of left-side buttock and posterior thigh pain. The pa-tient had an antalgic gait with knee flexed and hipadducted and was not able to walk further than 10 m

1-weighted sacroiliac joint MRIs showing inflammation on the left side.

Fig. 5. (A) Axial postcontrast T1-weighted and (B) sagittal T2-weighted pelvic MRIs revealing a 4 cm � 5 cm soft tissue tumor right and anterior to thesacrum.

Fig. 6. (A) Axial T1-weighted pelvic magnetic resonance neurography(MRN) showing left piriformis muscle asymmetry and atrophy, and (B)axial postcontrast fat-saturated T1-weighted pelvic MRN showing leftischial bone marrow edema and hamstring tendinopathy. Note that theleft arrow in (A) shows an atrophic and asymmetric piriformis muscle, andthe arrow in (B) shows bone marrow edema and tendinopathy.


because of increasing symptom intensity. At physical exam-ination the SLR test was negative. Freiberg’s sign was posi-tive. The range of trunk motion was not limited. The left legwas paresthetic. Tenderness in the piriformis muscle andischium pubis was noted on deep palpation. MRI of thelumbar spine yielded no abnormal findings. Magnetic reso-nance neurography (MRN) revealed piriformis asymmetryon the left side, high signal at the sciatic nerve at this loca-tion and bone marrow edema at the ischium pubis at thehamstring muscle insertion site (Fig. 6). A physical therapyprotocol for piriformis syndrome (PS) was applied.8 At theend of the physical therapy, the VAS score decreased from10 to 3 points. The antalgic gate was corrected andFreiberg’s sign became negative. Tenderness of the ischiumpubis remained due to hamstring tendinopathy.

2.2.6. Lumbar disc hernia and piriformis syndrome (Patient

10)

A 48-year-old male was admitted with a 6-week historyof LBP and left leg pain. MRI of the lumbar spineshowed L5–S1 disc herniation compressing the left S1nerve root. Electromyography showed S1 radiculopathy.He underwent L5 hemipartial laminectomy, and left L5–S1 microdiscectomy. After surgery, the symptoms resolved

partially but the patient still complained of left thigh painafter prolonged sitting and hip internal rotation. MRIneurography showed no abnormal signal at the sciaticnerve. The patient underwent a tetracosactrin injectionprotocol but received no benefit. A new examination re-vealed negative SLR. Freiberg’s sign was positive at theleft side. Although MRN showed no abnormality, giventhe results of a physical examination, the patient wasdiagnosed as having PS and the physical therapy protocolfor PS was applied.8 The patient received 15 sessions ofphysical therapy. At the end of the physical therapyregimen, the VAS score decreased from 9 to 2 points.The patient was comfortable after prolonged sitting andinternal rotation of the hip.

2.2.7. Degenarative lumbar spine and coxarthrosis (Patient

11)

A 65-year-old male was admitted with bilateral buttockpain and neurogenic claudication. The FABER test wasalso positive at the left side. A lumbar MRI showed L3–L4 disc herniation, lumbar spinal stenosis, and L4–L5spondylolisthesis. Pelvic radiographs showed grade IVosteoarthritis of the left hip. After discussion with thepatient regarding a treatment plan, the patient underwentL3–4–5 decompression and instrumentation, and inter-transverse fusion. Two months after spinal interventionthe patient underwent total hip arthroplasty. The symp-toms resolved after surgery.

3. Discussion

Many intraspinal and/or extraspinal pathologic pro-cesses along the lumbar nerve roots and sciatic nerve cancause sciatica. In 20% of cases, the sciatica is of both disco-genic and non-discogenic origin.2 However, in practice,causes of NDS are often overlooked, partly due to the highsensitivity of lumbar spine MRI in asymptomaticpatients.9,10

The causes of NDS may be classified into two major cat-egories: intraspinal and extraspinal. Differential diagnosis

Fig. 7. An algorithm for the diagnosis of discogenic and non-discogenic sciatica. Sacroiliitis may cause posterior pelvic and sacroiliac pain during theFABER test, whereas coxarthrosis may cause groin pain during that test. In a few cases there is coincidence of spinal and extraspinal processes. In patientswhose spinal pathology has additional findings, extra examinations for extraspinal processes should be carried out. AP = anteroposterior radiography,DPP = deep piriformis palpation, SICT = sacroiliac compression test, SLR = straight leg raise test.


requires a careful and detailed physical and radiologicalexamination based on a schema (Fig. 7).

3.1. Intraspinal causes of sciatica

Many intraspinal disorders can cause sciatica. Theremay also be multiple processes, for instance intradural orextradural cysts and tumors (mainly schwannomas), adulttethered cord syndrome, spinal epidural abscesses andhematomas, facet syndrome, lumbosacral deformities andinstabilities.5,7,11–14

The pain pattern, and the presence or abscence ofaccompanying symptoms, are the most important compo-nents of the differential diagnosis.

The tumor-related pains commonly cause patients towake in the night. Whereas small schwannomas lead toradicular pain, larger tumors lead to accompanyingsymptoms due to multiple nerve root and spinal cordcompressions. Epidural abscesses and hematomas maypresent with symptoms similar to intradural tumors.Adult tethered cord syndrome may lead to sciatica asso-ciated with a stretched thigh, resulting in gait distur-bance. The pain of facet origin is located commonly inthe low back, buttock and thigh. It rarely extends tothe lower levels. The pain secondary to deformities andinstabilities is commonly associated with LBP and in-creases in spine loading.

Fortunately, because lumbar MRI shows anatomical de-tails of the lumbar spine, and other neural structures andsoft tissues, making a diagnosis of spinal NDS is relativelyeasy. Coincidental spinal and extraspinal disorders, how-

ever, may complicate the symptomatology, and requirecareful evaluation.

3.2. Extraspinal causes of sciatica

Detection of extraspinal causes of sciatica is much moredifficult than detection of intraspinal ones, and requires ahigh degree of caution. The main causes of extraspinal sci-atica include sacroiliitis,6,15 PS,5,16 intrapelvic processes,1,4

and hip arthrosis.17,18

3.3. Sacroiliitis and seronegative spondyloarthropathies

Sacroiliitis, as one of the major causes of sciatica,6,15

should be considered during the differential diagnosisof LDHs, particularly when there is posterior thighpain.

There are two potential mechanisms by which sacroiliitiscan generate sciatica: (i) referred pain (e.g. patient 6), and(ii) direct involvement of the nerve by inflammatory medi-ators released from the sacroiliac joint (e.g. patient 7).6

Sacroiliitis is a common feature of seronegative spond-yloarthropathies; the others include spondylitis, morningstiffness, LBP, decreased mobility of the lumbar spine,and peripheral arthritis.

The history of the patient and characteristics of the painallow diagnosis of sacroiliitis and its etiology. The pain insacroiliitis has an insidous onset, is commonly localizedin the deep gluteal region and may refer to the posteriorthigh. The pain decreases with activity and increases inthe late evening.


The presence of LBP in patients with sacroiliitis makesdifferential diagnosis complex. This is because LBP mayindicate both ankylosing spondylitis and degenerative lum-bar spine. The patient needs to be questioned carefully todetermine the symptoms that indicate the disease. Disc her-niation-related LBP increases after activity and decreasesafter rest, whereas spondylitis-related LBP decreases afteractivity, and increases during the latter half of the night.Morning stiffness is another characteristic of spondylitis-related LBP.19

Physical examinations should include a Sacroiliac Com-pression Test, Gaenslen’s test and a FABER test. The loca-tion of the aggrevating pain after the FABER test mayreflect the pathology. Posterior pain in the pelvis may re-flect sacroiliitis, whereas pain in the groin may reflectcoxarthrosis.19

The early clinical diagnosis of sacroiliitis may be difficultand the diagnosis should be proven radiologically. Both CTand MRI are sensitive methods of demonstrating sacroiliitis.The early CT scan findings include cortical erosions and sub-chondral sclerosis of the sacroiliac joints. The later CT scanfindings include sacroiliac joint narrowing and ankylosis.20

MRI has a similar capacity to CT for detecting the disease;however, MRI can detect the early stages of sacroiliitisbecause it can reveal bone marrow edema before morpho-logic changes can be determined by a CT scan.21,22

3.4. Intrapelvic compressive processes

Intrapelvic compressive processes may affect the nerveas it passes from the neural foramina to the greater sciaticnotch. Reported intrapelvic compressing processes includetumors,4,23 hematoma,24 endometriosis,25 tubo-ovarian ab-scess,26 presacral abscess,27 and aneurysms.28

Pelvic and femoral bone tumors may compress the sci-atic nerve. Bickel et al. analyzed a surgical database of 32patients with bone tumors that caused sciatica. Accordingto their analysis, the characteristics of the patient’s painare very important in the differential diagnosis of bone tu-mors. Sciatica due to bone tumors generally has an insi-dous onset, causes constant pain, awakening at night,and is progressive and unresponsive to position changes.1

Local compression may reveal local tenderness. As mostbone tumors occur in the pelvis or proximal femur, initialpelvic radiography is recommended for patients with atyp-ical sciatica. A simple pelvic radiograph may show tumor-related bony destruction, as in patient 8. Three-phase bonescintigraphy, CT scanning, and MRI seem to be sensitive indetecting bone tumors. MRI and MRN may show the rela-tionship between the tumor and the sciatic nerve in detail.

3.5. Piriformis syndrome

Sciatica is caused by PS in 6% of patients.29 Hypertro-phy, inflammation, anatomic variations, myositis ossificansand traumatic injury of the piriformis muscle may com-press the sciatic nerve.16,30 Any increase in the piriformis

muscle tone not only produces local muscle pain but alsomay result in sciatica.

Patients with PS exhibit significantly different symptomsand results of physical examination to those of patientswith discogenic sciatica.31 Pain is the predominant symp-tom in both, but there are some differences in pain patterns.There is usually sciatic notch tenderness or pain at the but-tock in both patient groups. The pain of discogenic originrefers to the buttock and posterior thigh without any spe-cific focal tenderness in the buttock. Unlike patients withLDH, patients with PS typically experience symptoms inmultiple dermatomes rather than either the lateral (S1radiculopathy) or the medial (L5 radiculopathy) derma-tome alone. Numbness or weakness is quite uncommonin PS. Unlike in LDH, the SLR test is generally negativein patients with PS. Most patients with PS report that pro-longed sitting and walking exacerbate their pain and thattheir symptoms reaggravate after internal rotation of thehip.

The differential diagnosis of PS requires special testsincluding Pace’s sign, Freiberg’s sign and deep digital pal-pation of the piriformis muscle. If some of these tests arepositive, advanced radiological tests such as pelvic MRIand MRN should be performed since the pelvic MRI alonemay fail to show the extraspinal parts of the sciatic nerve.MRN may identify the sciatic nerve and its relationshipwith the surrounding structures.

In one study, MRN had 93% specificity and 64% sensi-tivity for distinguishing patients with PS from those withsimilar symptoms by presenting piriformis muscle asymme-try and sciatic nevre hyperintensity at the sciatic notch.27

Although piriformis muscle hypertrophy has been demon-strated by MRI,32 Filler et al. also observed ipsilateral mus-cle atrophy in some patients.31 This finding may besecondary to disuse of the muscle in the chronic stage. Inthe present study, MRN of patient 9 showed piriformismuscle atrophy and asymmetry, and sciatic nerve hyperin-tensity. However, the MRI of patient 11 (who had PS),similar to Lee et al.,33 showed that the piriformis musclewas not compressing the sciatic nerve. The absence of thepositive finding in the neutral (static) position can be ex-plained by the dynamic nature of the piriformis muscle:as the pain is positive in Freiberg’s position, MRI andMRN should be performed in both static and dynamicpositions (i.e. Freiberg’s position).

3.6. Hip disorders

The pattern of pain distribution in hip disorders maycreate difficulty in determining the source of pain (spinevs. hip).17 Swezey et al. reviewed patients who were treatedfor lumbar spinal stenosis while the primary source of thepain was osteoarthrosis of the hip, and patients who weretreated for osteoarthrosis of the hip while the primarysource of pain was lumbar spinal stenosis. Symptoms forboth disorders are frequently present in the same pa-tients.18 The presence of limb or groin pain, and limited


internal rotation of the hips is predictive of a hip disorderrather than a spine disorder. The occurence of groin painduring the FABER test may suggest a hip disorder. Pelvicradiography may show hip osteoarthritis.

However, both hip and lumbar spine pathologies mayexist together, as in patient 12. The radicular pain associ-ated with progressive neurogenic claudication and neuro-logic deficit may dictate a lumbar decompressionprocedure before a hip procedure.

LDH is not the only cause of sciatica. A detailed patienthistory, especially focused on pain characteristics, is animportant component of patient evaluation. Physicalexamination of patients with sciatica should include inspec-tion, palpation, and all physical tests to exclude NDS.

References

1. Bickels J, Kahanovitz N, Rubert CK, et al. Extraspinal bone and soft-tissue tumours as a cause of sciatica. Spine 1999;24:1611–6.

2. Dudeney S, O’Farrell D, Bouchier-Hayes D, et al. Extraspinal causesof sciatica: a case report. Spine 1998;23:494–6.

3. Kralick F, Koenigsberg R. Sciatica in a patient with unusualperipheral nerve sheath tumors. Surg Neurol 2006;66:634–7.

4. Nishikawa T, Iguchi T, Honda H, et al. Primary bone tumors of thefemur presenting with spinal symptoms: a report of two cases andreview of the literature. J Spinal Disord 2000;13:360–4.

5. Ortolan EG, Sola CA, Gruenberg MF, et al. Giant sacral schwan-noma. A case report. Spine 1996;21:522–6.

6. Wong M, Vijayanathan S, Kirkham B. Sacroiliitis presenting assciatica. Rheumatology 2005;44:1323–4.

7. Yuksel KZ, Senoglu M, Yuksel M, et al. Brucellar spondylo-discitiswith rapidly progressive spinal epidural abscess presenting withsciatica. Spinal Cord 2006;44:805–8.

8. Fishman LM, Zybert PA. Electrophysiologic evidence of pirıformissyndrome. Arch Phys Med Rehabil 1992;73:359–64.

9. Jackson RP, Cain JE, Jacobs RR, et al. The neuroradiographicdiagnosis of lumbar herniated nucleus pulposus: II. A comparison ofcomputed tomography (CT), myelography, CT-myelography, andmagnetic resonance imaging. Spine 1989;14:1362–7.

10. Wiesel SW, Tsourmas N, Feffer HL, et al. A study of computerassisted tomography: I. The incidence of positive CAT scans in anasymptomatic group of patients. Spine 1984;9:549–51.

11. Akay KM, Ersahin Y, Cakir Y. Tethered cord syndrome in adults.Acta Neurochir (Wien) 2000;142:1111–5.

12. Naderi S, Aks�an O, Men S, et al. Postoperative lumbosacralinstability: proposal of a new classification based on etiology. World

Spine Journal 2007;2:16–21.13. Naderi S, Yucesoy K, Aks�an O, et al. Epidural brucella abscess.

Report of a spinal brucellosis manifesting itself with epidural abscess.Neuro-orthopedics 2000;28:11–5.

14. Naderi S, Manisali M, Acar F, et al. Factors affecting reduction inlow-grade lumbosacral spondylolisthesis. J Neurosurg: Spine

2003;99:151–6.15. Chen WS. Chronic sciatica caused by tuberculous sacroiliitis. A case

report. Spine 1995;20:1194–6.16. Parziale JR, Hudgins TH, Fishman LM. The PS. Am J Orthop.

1996;25:819–23.17. Brown MD, Gomez-Marin O, Brookfield KF, et al. Differential

diagnosis of hip disease versus spine disease. Clin Orthop Relat Res

2004;419:280–4.18. Swezey RL. Overdiagnosed sciatica and stenosis, underdiagnosed hip

arthritis. Orthopedics 2003;26:173–4.19. van der Linden S, van der Heijde D. Ankylosing spondylitis. In: Haris

CB, Sledge CB, editors. Kelly’s Textbook of Rheumatology. 6thed. Philadelphia: WB Saunders; 2001. p. 1039–53.

20. Maksymowych WP, Landewe R. Imaging in ankylosing spondylitis.Best Pract Res Clin Rheumatol 2006;20:507–19.

21. Battafarano DF, West SG, Rak KM, et al. Comparison of bone scan,computed tomography and magnetic resonance imaging in thediagnosis of active sacroiliitis. Semin Arthritis Rheum 1993;23:161–76.

22. Geıjer M, Gothın GG, Gothın JH. The clinical utility of computedtomography compared to conventional radiography in diagnosingsacroiliitis. a retrospective study on 910 patients and literature review.J Rheumatol 2007;34:1561–5.

23. Odell RT, Key JA. Lumbar disc syndrome caused by malignanttumors of bone. JAMA 1955;157:213–6.

24. Richardson RR, Hahn YS, Siqueria EB. Intraneural hematoma of thesciatic nerve: case report. J Neurosurg 1978;49:298–300.

25. Koga K, Osuga Y, Harada M, et al. Sciatic endometriosis diagnosedby computerized tomography-guided biopsy and CD10 immunohis-tochemical staining. Fertil Steril 2005;84:1508.

26. Andrews DW, Friedman NB, Heier L, et al. Tuboovarian abscesspresenting as sciatic pain: case report. Neurosurgery 1987;21:100–3.

27. Macfarlane R, Pollard SG. Iliac and gluteal artery aneurysmspresenting as sciatica. J Roy Soc Med 1988;81:551–3.

28. Aguilera V, Calvo F, Nos P, et al. Sciatica secondary to a presacralabscess as the first manifestation of Crohn’s disease. Gastroenterol

Hepatol 2002;25:505–7.29. Lewis AM, Layzer R, Engstrom JW, et al. Magnetic resonance

neurography in extraspinal sciatica. Arch Neurol 2006;63:1469–72.30. Ozaki S, Hamabe T, Muro T. Piriformis syndrome resulting from an

anomalous relationship between the sciatic nerve and piriformismuscle. Orthopedics 1999;22:771–2.

31. Filler AG, Haynes J, Jordan SE, et al. Sciatica of nondisc origin andpiriformis syndrome: Diagnosis by magnetic resonance neurographyand interventional magnetic resonance imaging with outcome study ofresulting treatment. J Neurosurg Spine 2005;2:99–115.

32. Jankiewicz JJ, Hennrikus WL, Houkom JA. The appearance of thepiriformis muscle syndrome in computed tomography and magneticresonance imaging. A case report and review of the literature. Clin

Orthop Relat Res 1991;262:205–9.33. Lee EY, Margherita AJ, Gierada DS, et al. MRI of PS. AJR Am J

Roentgenol 2004;183:63–4.

Documents

Differential Diagnosis of Intraspinal and Extraspinal Non-discogenic Sciatica