Prevalence and awareness of sacroiliac joint alterations on lumbar spine CT in low back pain patients younger than 40 years

Abstract

Background

Computed tomography (CT) examinations of the lumbar spine are commonly performed in patients aged ≤40 years due to low back pain (LBP).

Purpose

To investigate the prevalence and awareness of radiologists for the presence of structural post-inflammatory/other sacroiliac joint (SIJ) alterations on lumbar spine CTs of young patients with LBP.

Material and Methods

A total of 484 lumbar spine CT examinations (272 men, 212 women; average age, 31 years; age range, 18–40 years) of patients with LBP in which the entire SIJs were visualized were retrospectively reviewed. SIJs were scored (consensus) by two senior radiologists (study reading) for the presence of post-inflammatory structural SIJ findings or other SIJs alterations. The original reports were compared to the study reading. Fifty CT examinations were re-evaluated for reliability assessment (intra-class correlation coefficient [ICC]).

Results

A total of 150 (31%) abnormal SIJ examinations were registered (ICC: r = 0.7–0.8; P < 0.0001): suspected sacroiliitis = 50 (10.2%); definite sacroiliitis = 16 (3.3%); osteitis-condensans-ilii = 38 (7.8%); diffuse idiopathic skeletal hyperostosis = 24 (5%); degenerative changes = 22 (4.5%); accessory SIJ = 22 (4.5%); and tumor = 1. The SIJs were referenced 39 times (8.0%) in the original readings: pathological findings (n = 15); and normal SIJ (n = 24). Total diagnostic accuracy for these reports only and for the entire readings were 49% and 69%, respectively, and 13% and 1.3%, respectively, for the pathological findings.

Conclusion

Sacroiliitis and other SIJ alterations are prevalent in young individuals with LBP, albeit, the majority of these alterations are not recognized nor reported by senior radiologists thus may delay efficacious treatment.

Keywords

Spondyloarthritis computed tomography (CT)sacroiliitis sacroiliac joints low back pain skeletal-axial

Introduction

Inflammation of the sacroiliac joints (SIJs), or sacroiliitis, is the hallmark of the spondyloarthropathies (SpA), a group of diseases characterized by back pain, spinal inflammation, HLA-B27 positivity, and peripheral findings such as dactylitis, enthesitis, and uveitis (1,2). Sacroiliitis is the most common clinical finding and the presumed first manifestation of SpA (3). Clinically it manifests as inflammatory low back pain (LBP) which occurs during the second half of the night and improves with exercise rather than with rest (4,5). Chronic LBP is one of the commonest health problems afflicting 15–35% of adults aged 20–39 years (6) and one-quarter of primary care patients with chronic LBP in this age group are ultimately diagnosed with axial SpA (7). Notwithstanding, degenerative disc disease is much more common than sacroiliitis in the general population and these patients, though magnetic resonance imaging (MRI) may be preferable, may be sent for a computed tomography (CT) study of the lumbar spine as a primary diagnostic tool by their family practitioners and orthopedic surgeons. Although MRI is the modality of choice for the diagnosis of early sacroiliitis (8), structural changes such as erosions and subchondral sclerosis may also be detected on CT (9), facilitating the diagnosis of SpA. As the SIJs tend to be encompassed on routine CT examinations of the lumbar spine, they are readily available for assessment, although our impression was that they tend to be overlooked.

The primary purpose of the current study is to investigate the awareness of reading radiologists of the presence of SIJ aberrations and pathologies in general and for structural post-inflammatory pathologies in particular in patients aged ≤40 years referred to lumbar spine CT for LBP.

Material and Methods

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Institutional review board approval for the retrospective review of lumbar spine CT examinations was obtained. Informed consent was waived by the local ethic committee.

Consecutive ambulatory, non-hospitalized, CT examinations of the lumbar spine of patients aged 18–40 years performed due to LBP in our institution (a tertiary, 1200-bed, university medical center) between August 2012 and January 2014 were identified using the institutional computerized archive.

All CT examinations were carried out on two 64-slice CT scanners (ICT 956, Brilliance, Philips Medical Systems, Eindhoven, The Netherlands; VCT LightSpeed, GE Healthcare, Milwaukee, WI, USA). Slice thickness was in the range of 0.6–2.5 mm. Images were reconstructed in a bone and soft tissue algorithm, reformatted in multiple planes from the raw data and evaluated in bone window. No oral or intravenous contrast material had been administered. Original reports for these CT examinations are all performed by board certified radiologists with different subspecialties such as neuroradiologists and uroradiologists.

Studies that did not include the entire SIJs or that were technically inadequate were excluded.

SIJ findings reported in the original reports (primary reading) were compared to an independent SIJ evaluation performed in the current study (study reading).

The same PACS system was used by the primary readers as well as by the study reviewers.

SIJ evaluation: primary reading

The presence of any references to the SIJ made by the reading radiologist in the original reports, including normal anatomy, pathologies, and anatomical variants was documented in a separate session.

During that session, demographic data (age, gender) and subspecialty of the primary reading radiologist (e.g. neuro-radiologist, invasive radiologist, etc.) were also registered.

The SIJ findings in the primary reading report (primary reading) were compared and analyzed against the consensus findings of the study reviewers (study reading) considered the standard of reference.

SIJ evaluation: study reading

All CT examinations were reviewed together and in consensus by one musculoskeletal radiologist and one board-certified radiologist. This reading was used as standard of reference for statistical analysis.

Examinations were evaluated for the presence of structural SIJ findings (subchondral sclerosis, erosions, ankylosis) that may indicate the presence of inflammatory sacroiliitis. Reviewers were aware of the patient's sex, age, as well as the referrals' reported additional clinical parameters. Examinations were mainly evaluated using the axial original slices as well as coronal multiplanar reformations (MPR). Coronal oblique (MPR) paralleling the SIJ were performed and evaluated, when needed. The two reviewers evaluated the CT examinations together for any right and left SIJ abnormal/variant finding. When an abnormality was detected, reviewers discussed and agreed upon its nature.

Since the modified New York criteria designed for plain radiographs (7) were not validated on CT, we used a suggested classification for SIJ alteration for CT examinations as follows: 0, normal SIJ; 1, suspicious for inflammatory sacroiliitis (bone irregularity, or subchondral sclerosis, or mildly abnormal suspected erosions or sclerosis); 2, definite sacroiliitis (moderately abnormal with erosions, sclerosis, partial ankylosis, or severely abnormal with complete ankyloses) (10) (Fig. 1).

Fig. 1.

SIJ CT scoring: (a) 0, normal SIJ; (b) 1, suspicious for sacroiliitis with suspected small erosions (arrows); (c) 2, definite sacroiliitis with erosions (arrows) and sclerosis (arrowheads).

Both SIJs were scored in unison, that is, a single score was given per patient based on the highest SIJ score. Other SIJs findings not compatible with inflammatory sacroiliitis but that could be confused with sacroiliitis and anatomic variants were also recorded in the same reading session and in consensus by both reviewers.

Major SIJ alterations evaluated in addition to sacroiliitis were the following:

Osteitis condensans ilii (OCI): triangular-shaped sclerosis appearing mostly on the iliac side of the SIJ, sometimes bilaterally with an intact (non-eroded) SIJ cortex (Fig. 2) (11,12);

Diffuse idiopathic skeletal hyperostosis (DISH): bridging osteophytes located on the anterior part of the SIJ without joint ankylosis (Fig. 3) (13,14), often seen in patients with confluent bridging osteophytes of the spine associated with DISH;

Degenerative joint related changes: joint space narrowing, osteophytes, subchondral cysts, and subchondral sclerosis as well as vacuum phenomenon;

Space occupying bone lesion: any lytic or sclerotic lesion located in the ilium or sacrum, not confined to the subchondral area and suspected to be a tumor;

Accessory sacroiliac/facet joint: a false joint occurring at the S2 level, usually dorsal to the true synovial portion of the SIJ (Fig. 4) (15,16).

Fig. 2.

Osteitis condensans ilii in a 36-year-old woman. A triangular-shaped sclerosis is detected on the iliac side of the right sacroiliac joint (arrowheads) and to a lesser extent on the sacral side of the right sacroiliac joint and the iliac side of the left sacroiliac joint. Cortex is preserved without erosions.

Fig. 3.

Diffuse idiopathic skeletal hyperostosis in a 39-year-old man. Bridging osteophytes can be seen on the anterior part of both sacroiliac joints (arrows) with preserved sacroiliac joints.

Fig. 4.

Accessory sacroiliac joint in a 24-year-old woman: a false joint (arrows) with mild degenerative changes is detected dorsal to the true synovial portion of the right sacroiliac joint.

For the evaluation of intra-reader reliability, one of the reviewers re-evaluated twice a subset of 50 randomly selected CT examinations from the original cohort. This subset was reviewed by this same reader with an interval of 3 months after the consensus study reading and with a 2-week interval between the two reliability readings. Intra-reader reliability was assessed using the intra-class correlation coefficient (ICCs) analysis calculated by two-way mixed analysis for absolute agreement. ICC values were interpreted as follows: 0–0.2, poor agreement; 0.3–0.4, fair agreement; 0.5–0.6, moderate agreement; 0.7–0.8, strong agreement; and >0.8, almost perfect agreement. P values < 0.05 were considered to be statistically significant.

Results

A total of 484 CT examinations were evaluated by the reviewers.

A total of 499 CT examinations of patients aged ≤40 years were performed during the study period on account of LBP. Fifteen examinations did not include the entire SIJ and were thus excluded. The remaining 484 examination comprised the study cohort: 272 men, 212 women; average age, 31 years; age range, 18–40.

The original CT examination readings were performed by 15 senior (board-certified) non-musculoskeletal radiologists (general radiologists = 5, neuroradiologists = 5, pediatric radiologists = 3, invasive radiologist = 1, chest radiologist = 1) with an average of 68 ± 29 reports per radiologist.

Study reading

SIJ findings were recorded by the study reviewers in 150 (31.0%) CT examinations.

Structural changes indicating inflammatory sacroiliitis were registered in 66 (13.6%) examinations: Grade 1 (suspected sacroiliitis) = 50 (10.2%); and Grade 2 (definite sacroiliitis) = 16 (3.3%).

Other SIJ findings and anatomic variants were found in 107 (22.1%) examinations (Table 1). In 23 cases, findings appeared concomitantly.

Table 1.

Non inflammatory-related SIJ findings registered by the study readers (study cohort, n = 484).

SIJ finding	n (%)
Osteitis condensans ilii	38 (7.8%)
Diffuse idiopathic skeletal hyperostosis	24 (5.0%)
Degenerative changes	22 (4.5%)
SIJ accessory joints	22 (4.5%)
Space occupying bone lesion	1 (0.2%)

SIJs were unremarkable in the remaining 334 examinations.

Intra-observer agreement between the two additional readings performed within an interval of 3 months from the consensus study reading and with the 2-week interval was strong–almost perfect (ICC: sacroiliitis, 0.9; OCI, 0.8; DISH, 0.7; DJD, 0.7; accessory facet, 0.8; P < 0.0001).

Primary reading

The SIJs were referenced 39 times (8.0%) in the original reports. The primary reading radiologists reported 15 (10%) SIJ alterations out of the 150 detected by the study reviewers while in 24 cases a normal SIJ was documented (Table 2). In the rest of the reports, no reference to the SIJ was made. In these cases, for the study purpose the SIJ were assumed to be read as normal. The terms OCI or DISH were not used at all in the original reports. Anatomic variants were not described.

Table 2.

SIJ findings in the original reports compared to study reading (n = 39).

Original reading	n (%*)	Study reading	n (%*)	Accuracy^†
Normal	24 (4.9)	Normal	18 (3.7)	75%
		Structural sacroiliitis (grade 2)	3 (0.6)
		OCI	3 (0.6)
Structural sacroiliitis	1 (0.2)	OCI	1 (0.2)	0%
Sclerotic SIJ changes	2 (0.4)	OCI	1 (0.2)	50%
		DISH^‡	1 (0.2)
Degenerative changes	12 (2.4)	Normal	2 (0.2)	8%
		Structural sacroiliitis (Grade 2)	2 (0.4)
		OCI	7 (1.4)
		Degenerative changes	1 (0.2)

Percentage out of the entire study cohort (n = 484).

†

Calculated as correct diagnosis/description compared to study reading.

‡

Anterior bony bridges without subchondral sclerosis.

DISH, diffuse idiopathic skeletal hyperostosis; OCI, osteitis condensans ilii.

Comparison of readings

One of the original readers diagnosed inflammatory sacroiliitis in a patient with triangular subchondral iliac-sided sclerosis but otherwise regular SIJ diagnosed as OCI by the study radiologists. Two other patients were described with iliac-sided subchondral sclerosis in the body of the report but not interpreted in the summary; one had OCI according to the study reading and one had DISH and no apparent sclerosis.

Considering only the 39 reports which referenced the SIJs in the primary reading and comparing them to the 39 study reading findings, diagnostic accuracy was 49% (study reading = standard of reference) while evaluating diagnostic accuracy for all the detected pathological SIJs registered by the study reading resulted in a diagnostic accuracy of 13%.

When comparing all 484 original readings to the study readings, assuming that the non-reported SIJs in the original reports were considered normal by the primary reader, then total accuracy was 69% and accuracy for pathological SIJ reports was 1.3%.

Discussion

Sacroiliitis is an important clinical and diagnostic feature of SpA, yet it may present with an ambiguous physical examination rendering imaging a paramount role in its diagnosis (17). MRI is considered the best imaging modality for early diagnosis of SpA, yet in many instances, symptoms are mistaken for mechanical LBP and an average of 8.5 years may elapse from the onset of clinical symptoms until the diagnosis of SpA is established (18). However, undiagnosed SpA patients may undergo a CT examination as part of their chronic back pain workup—an examination, which in our institution usually encompasses the entire SIJs. CT is certainly not the diagnostic modality of choice for young patients with suspected sacroiliitis and one would not send them to perform a CT during a rheumatological workup due to excessive radiation exposure of an approximate 20 mSv (19); in addition, there is no standardized protocol for lumbar spinal CTs and not in all CTs would the SIJs be included. Still, when a CT of a patient aged under 40 years with LBP already exists and the SIJs are included in that CT examination, it is advisable and fundamental to assess the sacroiliac joints for the presence of structural post-inflammatory findings in order to facilitate earlier diagnosis. This is especially important in the face of new emerging techniques such as low radiation CT that may cause a surge in the use of CT (20) and due to the emerging role of structural changes such as erosions in the evaluation of SpA (21,22). In the current study, senior non-musculoskeletal radiologists were shown to have low awareness of SIJ pathologies generally and specifically sacroiliitis depicted on lumbar spine CT examinations of young patients with LBP. Reference to the SIJs was made in a mere 8% of the original reports and pathology was detected in only 10% of the examinations in which it was present. Moreover, the diagnostic accuracy for the pathologies detected in the original reports was extremely low (1.3%), meaning that the majority of the reported SIJ alterations by the original radiologists were inaccurate, e.g. triangular subchondral sclerosis in a woman diagnosed as inflammatory sacroiliitis instead of OCI and conversely, changes compatible with grade 2 sacroiliitis interpreted as degenerative changes.

It may be argued that the reduced accuracy may result from compromised radiological skills of the original reports' reading radiologists. However, the study was performed in an established, large tertiary university hospital and all the reading radiologists were board-certified (the majority with many years of experience in their subspecialty). Thus, regardless of its root, this reduced accuracy is a snapshot of the everyday scenario of lumbar CT radiological report.

This lack of awareness of expert radiologists for potential causes of back pain in the SIJ and for the presence of sacroiliitis or other SIJ alterations suggests that basic knowledge regarding the radiographic/CT findings for suspected sacroiliitis as well as their differential diagnosis is lacking and that training on this subject is warranted. This may result from the focus on MRI of the SIJs in recent years that may have shifted attention (and knowledge) away from CT.

Having said that, it should be noted that individual training did not improve the diagnostic ability of either radiologists nor rheumatologists for sacroiliitis on plain radiographs (23). However, the effect of individual training on CT readings of sacroiliitis has not yet been tested in the general radiologist population. As CT is a more sensitive modality for the detection of early sacroiliitis than plain radiographs (24,25), it is expected that following appropriate training, better understanding of the different structural findings as well as the differential diagnosis to sacroiliitis will be better familiarized and thus the sacroiliac joint alterations shall be more easily recognized and reported taking full advantage of the diagnostic utility of this technique.

Presently, we documented some degree of sacroiliac changes in 13.6% of this young LBP cohort and unequivocal sacroiliitis in 3.3% of the examinations.

Slobodin et al. reported structural sacroiliitis in 3.7% of abdominal CTs performed for various indications in a group patients aged 18–55 years (26), comparable with the results of the current study.

It may be argued that the study readers in the current study had an increased awareness for changes of the SIJs due to the study nature, with potentially over-diagnosis of SIJ pathologies. Still, the 3.3% prevalence seen in the current study is much lower compared to 8% reported in patients with radiographic axial SpA in a recent prospective Dutch cohort of 364 patients (age range, 20–45 years) with chronic LBP (7).

In a recently published study on 1037 patients aged < 40 years with LBP who performed MRI, SIJ erosions were found in 8% (27). It is not yet clear whether MRI is as sensitive as CT in detecting structural SIJ lesions (21), potentially explaining the higher prevalence of erosions seen in this MRI study.

The clinical significance of misdiagnosing the other, non-inflammatory SIJ alterations overlooked by the original reading radiologists can be questioned. It has been reported though that many such alterations can cause low back symptomatology and should thus not be overlooked (11,16,28). Recognizing the different features and characteristics of alternative diagnoses to sacroiliitis such as OCI, DISH, and degenerative changes in accessory SIJs, described to be identified by different imaging modalities (29 –31), is therefore of importance in facilitating correct and targeted diagnosis and for tailoring adequate treatment.

The lack of clinical correlation or follow-up is an inherent major limitation of the current study. However, the purpose of the study was not to evaluate the true prevalence of sacroiliitis, but rather to evaluate the level of awareness for SIJ alterations among the everyday reading radiologists. Another limitation is the consensus reading of the CT examinations that precludes inter-reader reliability evaluation.

In conclusion, routine assessment of the SIJs, which are sometimes fully visualized on regular CT examinations of the lumbar spine, may allow for earlier diagnosis and prompt institution of therapy in axial SpA patients leading to improved quality of life and deterring irreversible changes.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Haroon

Inman

. Ankylosing spondylitis - new criteria, new treatments. Bull NYU Hosp Jt Dis 2010; 68: 171–174.

van den Berg

van der Heijde

. How should we diagnose spondyloarthritis according to the ASAS classification criteria: a guide for practicing physicians. Pol Arch Med Wewn 2010; 120: 452–457.

Brandt

Bollow

Haberle

. Studying patients with inflammatory back pain and arthritis of the lower limbs clinically and by magnetic resonance imaging: many, but not all patients with sacroiliitis have spondyloarthropathy. Rheumatology (Oxford) 1999; 38: 831–836.

Braun

Sieper

. The sacroiliac joint in the spondyloarthropathies. Curr Opin Rheumatol 1996; 8: 275–287.

Weisman

. Inflammatory back pain: the United States perspective. Rheum Dis Clin North Am 2012; 38: 501–512.

Hoy

Bain

Williams

. A systematic review of the global prevalence of low back pain. Arthritis Rheum 2012; 64: 2028–2037.

van Hoeven

Luime

Han

. Identifying axial spondyloarthritis in Dutch primary care patients, ages 20-45 years, with chronic low back pain. Arthritis Care Res (Hoboken) 2014; 66: 446–453.

Navallas

Ares

Beltran

. Sacroiliitis associated with axial spondyloarthropathy: new concepts and latest trends. Radiographics 2013; 33: 933–956.

Lee

Hong

Park

. Value of multidetector computed tomography for the radiologic grading of sacroiliitis in ankylosing spondylitis. Rheumatol Int 2013; 33: 1005–1011.

10.

Geijer

Gadeholt Gothlin

Gothlin

. The validity of the New York radiological grading criteria in diagnosing sacroiliitis by computed tomography. Acta Radiol 2009; 50: 664–673.

11.

Mitra

. Osteitis condensans ilii. Rheumatol Int 2010; 30: 293–296.

12.

Olivieri

Ferri

Barozzi

. Osteitis condensans ilii. Br J Rheumatol 1996; 35: 295–297.

13.

Olivieri

D'Angelo

Palazzi

. Diffuse idiopathic skeletal hyperostosis: differentiation from ankylosing spondylitis. Curr Rheumatol Rep 2009; 11: 321–328.

14.

Taljanovic

Hunter

Wisneski

. Imaging characteristics of diffuse idiopathic skeletal hyperostosis with an emphasis on acute spinal fractures: review. Am J Roentgenol 2009; 193: S10–19. quiz S20–14.

15.

Ehara

el-Khoury

Bergman

. The accessory sacroiliac joint: a common anatomic variant. Am J Roentgenol 1988; 150: 857–859.

16.

Fortin

Ballard

. The frequency of accessory sacroiliac joints. Clin Anat 2009; 22: 876–877.

17.

Oostveen

Prevo

den Boer

. 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.

18.

Feldtkeller

Khan

van der Heijde

. Age at disease onset and diagnosis delay in HLA-B27 negative vs. positive patients with ankylosing spondylitis. Rheumatol Int 2003; 23: 61–66.

19.

Biswas

Bible

Bohan

. Radiation exposure from musculoskeletal computerized tomographic scans. J Bone Joint Surg Am 2009; 91: 1882–1889.

20.

Onishi

Kockelkoren

Kim

. Low-dose pelvic computed tomography using adaptive iterative dose reduction 3-dimensional algorithm: a phantom study. J Comput Assist Tomogr 2015; 39: 629–634.

21.

Lambert

Bakker

van der Heijde

. Defining active sacroiliitis on MRI for classification of axial spondyloarthritis: update by the ASAS MRI working group. Ann Rheum Dis 2016. DOI: 10.1136/annrheumdis-2015-208642.

22.

Wick

Weiss

Jaschke

. Erosions are the most relevant magnetic resonance imaging features in quantification of sacroiliac joints in ankylosing spondylitis. J Rheumatol 2010; 37: 622–627.

23.

van Tubergen

Heuft-Dorenbosch

Schulpen

. Radiographic assessment of sacroiliitis by radiologists and rheumatologists: does training improve quality?. Ann Rheum Dis 2003; 62: 519–525.

24.

Battafarano

West

Rak

. Comparison of bone scan, computed tomography, and magnetic resonance imaging in the diagnosis of active sacroiliitis. Semin Arthritis Rheum 1993; 23: 161–176.

25.

Geijer

Gothlin

. Observer variation in computed tomography of the sacroiliac joints: a retrospective analysis of 1383 cases. Acta Radiol 2007; 48: 665–671.

26.

Slobodin

Croitoru

Starikov

. Incidental computed tomography sacroiliitis: clinical significance and inappropriateness of the New York radiological grading criteria for the diagnosis. Clin Rheumatol 2012; 31: 425–428.

27.

Arnbak

Jensen

Egund

. Prevalence of degenerative and spondyloarthritis-related magnetic resonance imaging findings in the spine and sacroiliac joints in patients with persistent low back pain. Eur Radiol 2016; 26: 1191–1203.

28.

Vadivelu

Green

Bhatt

. An uncommon cause of back pain in pregnancy. Postgrad Med J 2005; 81: 65–67.

29.

Hajialilo

Ghorbanihaghjo

Khabbazi

. Ankylosing spondylitis in iran; late diagnosis and its causes. Iran Red Crescent Med J 2014; 16: e11798.

30.

Jans

Van Praet

Elewaut

. MRI of the SI joints commonly shows non-inflammatory disease in patients clinically suspected of sacroiliitis. Eur J Radiol 2014; 83: 179–184.

31.

Gleeson

O'Connell

Duke

. Coronal oblique turbo STIR imaging of the sacrum and sacroiliac joints at routine MR imaging of the lumbar spine. Emerg Radiol 2005; 12: 38–43.