Diagnostic performance of plain radiography for sacroiliitis in patients with suspected axial spondyloarthritis: a systematic review and meta-analysis

Abstract

Background

Plain radiography serves a pivotal role in diagnosing axial spondyloarthritis. However, a broad range of diagnostic performance of plain radiography has been reported.

Purpose

To perform a systematic review and meta-analysis to measure the diagnostic performance of plain radiography for sacroiliitis in patients suspected of having axial spondyloarthritis using magnetic resonance imaging (MRI) findings as the reference standard.

Material and Methods

Studies comparing radiography and MRI in the diagnosis of sacroiliitis in patients suspected of having axial spondyloarthritis were searched in PubMed and EMBASE. Additionally, studies analyzed SPondyloaArthritis Caught Early (SPACE), DEvenir des Spondylarthropathies Indifferenciées Récentes (DESIR), GErman Spondyloarthritis Inception Cohort (GESPIC), and South Swedish Arthritis Treatment Group (SSATG) cohorts were manually searched. Pooled sensitivity and specificity of radiography were calculated by using a bivariate random-effects model. Meta-regression analyses were performed to identify the sources of heterogeneity.

Results

Eight eligible studies with 1579 patients were included. The pooled sensitivity and specificity of radiography were 0.55 (95% confidence interval [CI] = 0.40–0.69) and 0.87 (95% CI = 0.72–0.95). The meta-regression analyses showed prospective study design and criteria for MRI positivity considering only active bone marrow edema were associated with lower sensitivity.

Conclusion

The plain radiography showed low sensitivity and reasonable specificity in diagnosis of sacroiliitis in patients suspected of having axial spondyloarthritis.

Keywords

Radiography magnetic resonance imaging sacroiliitis spondylarthropathies meta-analysis

Introduction

Spondyloarthritis (SpA) is a clinical entity comprising heterogeneous correlated disease groups that had been previously classified as ankylosing spondylitis (AS), psoriatic spondyloarthritis, reactive spondyloarthritis, enteropathic spondyloarthritis, juvenile-onset spondyloarthritis, and undifferentiated spondyloarthritis (1). Diagnosis of SpA has been a criteria-based diagnosis with a combination of clinical, laboratory, and imaging arms. Historically, plain radiography played an important role in the imaging arm for establishing the diagnosis of sacroiliitis and it has been integrated into the evolving diagnostic criteria of SpA from Amor (2), European Spondyloarthropathy Study Group (ESSG) (3), to Assessment of SpondyloArthritis international Society (ASAS) (4).

Low cost and wide availability of plain radiography made its role as a first-line imaging method and triaging tool for SpA (1). Widely accepted radiographic diagnosis criteria for sacroiliitis is the modified New York criteria, which interprets as positive if it is grade II or higher bilaterally or grade III and higher unilaterally (5). Since plain radiography has a wide range of inter-observer agreement (κ = 0.19–0.79) (6 –8), and subjectivity in grading grade I sacroiliitis over grade II (9), concerns are growing over the diagnostic performance of plain radiography as an initial diagnosing method for sacroiliitis (10). On the other hand, magnetic resonance imaging (MRI) is a three-dimensional multiplanar imaging technique that is sensitive for diagnosing acute inflammatory changes (11), showing good reliability and substantial inter-observer agreement (κ = 0.73) (12), and it can detect early structural changes in a manner similar to computed tomography (CT) (13).

Given the broad range of reported diagnostic performance of plain radiography for diagnosing sacroiliitis, we conducted a systematic review and meta-analysis to assess the overall diagnostic performance of plain radiography for diagnosing sacroiliitis using MRI findings as the reference standard.

Material and Methods

The present study followed the guideline of Preferred Reporting Items for a Systematic Review and Meta-analysis of Diagnostic Test Accuracy Studies (PRISMA-DTA) (14). Approval from the Institutional Review Board was not required for this study.

Search strategy

Our target population was composed of patients suspicious of inflammatory sacroiliitis due to chronic low back pain with clinical suspicion for axial spondyloarthritis (axSpA). The index test was set as plain radiography, and the target outcome was the diagnostic performance of plain radiography for sacroiliitis, with MRI being set as the imaging reference standard, as it is regarded as the most sensitive modality for detecting sacroiliitis (11 –13,15).

We conducted online literature searches of PubMed and EMBASE on 7 July 2019. The following search term was used: (sacroili* OR spondyloarth*) AND (radiography OR radiograph OR X-ray) AND (sensitivity OR specificity OR accuracy OR receiver operating characteristic OR roc curve). The literature search was not restricted to any publication date or study setting but it was limited to studies written in English.

Additionally, we manually searched articles dealing with imaging in axSpA that analyzed cohorts as follows: SPondyloArthritis Caught Early (SPACE); DEvenir des Spondylarthropathies Indifferenciées Récentes (DESIR); GErman Spondyloarthritis Inception Cohort (GESPIC); and South Swedish Arthritis Treatment Group (SSATG).

Study selection

Two investigators (Y.K. and C.G.C. with four and seven years of radiology research) independently screened the titles and abstracts for potential eligibility and reached consensus. They excluded animal studies, case reports or series, letters, editorials, conference abstract, review articles, guidelines, or consensus statements. The abstracts or full text of the remaining articles were reviewed and only the articles that had appropriate target population and information of target outcome were considered eligible.

Data extraction

The two investigators independently extracted data from the selected studies using the following standardized form: (i) study characteristics: authors, year of publication, study design and setting (type and location of investigating center), patient inclusion criteria, sample size, MRI sequence, magnetic strength of MRI machine, specialty of image reviewers, definition of sacroiliitis in plain radiography, definition of sacroiliitis in MRI, whether image reviewers were blinded to clinical information or other imaging method; (ii) demographic characteristics: mean age, sex, and proportion of human leukocyte antigen-B27 (HLA-B27); and (iii) 2 × 2 contingency table for plain radiography as an index test and MRI as reference standard test in diagnosis of sacroiliitis (number of true-positive, false-positive, false-negative, and true-negative results). The 2 × 2 contingency table was recorded per patient.

Quality assessment

Each included study was assessed using Quality Assessment of Diagnostic Accuracy Studies (QUADAS)-2 (16). Initially, there were two pre-specified criteria for evaluating the concerns regarding applicability. In the patient selection domain, they rated it as high concern when a study included only patients who were fully diagnosed as having axSpA. In the reference standard domain, they rated it as high concern when a study used MRI system with magnetic strength <1.5 T or when either T1-weighted (T1W) or fluid-sensitive images were absent in the MRI protocol, and they rated it as unclear when a study used their own criteria during the judgment in the presence of sacroiliitis on MRI. Furthermore, whether or not the oblique view of radiography was used in the study did not affect the assessment of the risk of bias in the index test domain, since it is known that addition of oblique view did not improve the diagnostic performance significantly (17).

Statistical analysis

For the 2 × 2 contingency tables, radiograph-positive and MRI-negative sacroiliitis was considered a false-positive finding and was assumed not to be sacroiliitis. We used a bivariate random-effect model to calculate the pooled per-patient sensitivity and specificity. The sensitivities and specificities of included studies were plotted in forest plot. Heterogeneity was assessed through visual inspection of the hierarchical summary receiver operating characteristic curve (HSROC) with 95% confidence and prediction regions (17). Heterogeneity was judged to be present when there was a substantial difference between the 95% prediction and confidence interval regions or when there was large deviation of studies from the fitted HSROC. Additionally, Cochran’s Q test, I² test, and Spearman’s correlation coefficient between false-positive rate and sensitivity were used to evaluate any threshold effect. Spearman’s coefficient over 0.6 was considered as a considerable threshold effect (18). For evaluation of potential publication bias, we evaluated Deeks’ funnel plots of individual study log odds ratios plotted against the one divided by square root of effective sample size (19). Positive likelihood ratio and negative likelihood ratio were calculated, and resultant Fagan’s nomogram was drawn. The pre-test probability of having sacroiliitis was assumed to be 0.17, as reported in the previous literature in patients with chronic low back pain with clinical suspicions for axSpA (Supplemental material) (20).

Subgroup analyses with meta-regression were performed to identify factors associated with heterogeneity of sensitivity and specificity. The following variables were evaluated in the subgroup analysis: (i) study design (prospective design vs. retrospective design); (ii) study institute (single center vs. multicenter); (iii) year of publication (before 2010 vs. 2010 and beyond); (iv) number of patients included (≥100 vs. <100); (v) proportion of men (≥ 50% vs. <50%); (vi) proportion of patients with HLA-B27 positivity (≥ 60% vs. <60%); (vii) whether image reviewers were blinded to clinical information or other imaging methods when interpreting plain radiography or MRI; (viii) magnetic strength of MRI machine (≥ 1.5 T vs. <1.5 T); (ix) MRI sequence (presence of both T1W image and short tau inversion recovery [STIR] sequence and at least one oblique coronal imaging plane vs. absence of either T1W image, STIR, or oblique coronal imaging plane); (x) diagnostic criteria of sacroiliitis in MRI (published criteria vs. institutional criteria); (xi) readers (purely radiologists vs. radiologists and clinicians); and (xii) whether only active bone marrow edema components were used for judging positivity or negativity in MRI. Those cases where the subgroup could not be dichotomized because the study did not give a clear statement about the covariate, or where the study consisted of a single study within a certain subgroup, were not included in the meta-regression analyses.

All statistical analyses were performed using Stata version 15.0 (StataCorp, College Station, TX, USA) with “midas” and “metandi” modules and R 3.6.1 (open source, GNU project) with “mada” module. Two-tailed P value <0.05 was considered statistically significant.

Results

Study search

From the online literature search in PubMed and EMBASE, 1522 studies were identified. Eight studies were finally included in both qualitative and quantitative synthesis. The result of study search and selection is summarized in Fig. 1.

Fig. 1.

Study search and selection.

Study characteristics

The study characteristics and demographics of the eight studies are summarized in Table 1 (10,21–27). A total of 1597 patients were included for meta-analysis. Five studies included patients with inflammatory back pain, two studies included patients with chronic back pain referred to rheumatologists for suspected axSpA, and one study included patients diagnosed with axSpA according to ASAS criteria. A total of 664 (41.6%) patients had MRI-positive sacroiliitis and 455 (28.5%) patients had radiograph-positive sacroiliitis. The mean patient age was in the range of 27–37 years. Only one study utilized both the oblique and anteroposterior views for the diagnosis of sacroiliitis in radiograph. All studies adopted modified New York criteria for radiographic diagnosis of sacroiliitis. MRI criteria for diagnosing sacroiliitis varied among studies with six studies using their institutional criteria and only two studies utilized the published criteria (ASAS/OMEARCT) for diagnosing MRI-positive sacroiliitis (28). Four studies used only an active bone marrow edema component for judging MRI as positive for sacroiliitis.

Table 1.

Study characteristics.

Study	Year	Country	Patients (n)	MRI-positive patients (n)	Inclusion criteria	Mean age (years)	Male patients (%)	HLA-B27 (+) (%)	MRI strength (T)	Use of MR contrast	Including essential sequences*	MRI slice thickness (mm)	Image interpreter
Oostveen et al.	1999	The Netherlands	25	17	Inflammatory back pain and HLA-B27 (+)	37	56	100	1.5	No	No	NA	2 radiologists
Ortolan et al.	2018	The Netherlands, Norway, Sweden, Italy	719	234	Chronic back pain referred to rheumatologists	29	38	44	1.5	No	Yes	4	2 rheumatologists
Sudol-Szopinska et al.	2016	Poland	100	14	Chronic back pain referred to rheumatologists	NA (19–71)	40	NA	1.5	Yes	Yes	NA	NA
Blachier et al.	2013	France	475	225	Inflammatory back pain	33	50	83	1–1.5	No	Yes	4	Numerous radiologists and clinicians
Blum et al.	1996	Germany	44	43	Inflammatory back pain	34	61	57	0.23	Yes	No	6	2 radiologists
Inanc et al.	2005	Turkey	54	48	Inflammatory back pain and ESSG criteria	35	28	NA	NA	NA	No	NA	1 radiologist
Poddubnyy et al.	2013	Germany	112	71	ASAS criteria	37	61	82	NA	No	No	NA	1 radiologist, 1 rheumatologist
Heuft-Dorenbosch et al.	2006	Netherlands	68	12	Inflammatory back pain	35	38	46	1.5	Yes	Yes	NA	2 radiologists

Essential MRI sequence was defined as protocol including both T1-weighted and fluid-sensitive sequences, and with at least one view of the oblique coronal plane.

ASAS, Assessment of SpondyloArthritis international Society; ESSG, The European Spondyloarthropathy Study Group; NA, not available.

Quality assessment of selected studies

Table 2 contains the results of the assessment of the reporting quality of the studies by using the QUADAS-2 scoring systems. Five studies satisfied five or more of the seven QUADAS criteria. For the majority of the studies, the applicability concerning the reference standard had high risks. Most studies used their institutional criteria for the diagnosis for sacroiliitis other than ASAS/OMERACT criteria, but it must be noted that half of the studies were published before the publication of the ASAS/OMERACT criteria. The verification bias was considered low in all our included studies since all patients with conventional radiograph underwent MRI regardless of whether the radiograph findings were positive or negative.

Table 2.

Assessment of reporting quality by QUADAS-2 scoring system.

Study	Risk of bias				Applicability concerns
	Patient selection	Index test	Reference standard	Flow and timing	Patient selection	Index test	Reference standard
Oosteven et al., 1999	☺	☺	☺	☺	☺	☺	☺
Ortolan et al., 2018	☺	☺	☺	☺	☺	☺	☺
Sudol-Szopinska et al., 2016	?	?	?	☺	☺	☺	☺
Blachier et al., 2013	☺	?	?	☺	☺	☺	☹
Blum et al., 1996	☺	☺	☺	☺	☺	☺	☹
Inanc et al., 2005	?	☺	?	☺	☹	☺	?
Poddubnyy et al., 2013	☺	☺	☺	☺	☹	☺	☹
Heuft-Dorenbosch et al., 2006	?	☺	☺	☺	☺	☺	?

☺, low risk; ☹, high risk; ?, unclear risk.

Overall diagnostic performance of plain radiography

The sensitivity and specificity of the eight included studies were in the range of 0.12–0.83 and 0.62–1.00, respectively. The pooled sensitivity and specificity were 0.55 (95% confidence interval [CI] = 0.40–0.69) and 0.87 (95% CI = 0.72–0.95). The HSROC and forest plot of sensitivity and specificity were drawn in Figs. 2 and 3, respectively. As there was substantial difference between the 95% prediction and confidence interval regions, we assumed that there was substantial heterogeneity in our included studies. The Q test revealed significant heterogeneity in both sensitivity (Q = 77.43, P <0.01) and specificity (Q = 147.27, P <0.01). Additionally, in the I² test, both sensitivity (I² = 90.96%) and specificity (I² = 95.25%) exhibited considerable heterogeneity. The overall arrangement of the coupled forest plot did not follow V or inverted-V shape and the Spearman’s correlation coefficient between false-positive rate and sensitivity was 0.29 (95% CI = −0.55–1.13), which was <0.6. Therefore, the contribution of threshold effect to heterogeneity was regraded not remarkable, and subgroup analysis for the sources of heterogeneity was warranted. The Deeks’ funnel plot and asymmetry test (P = 0.78 for the slope coefficient) both indicated no influence of publication bias on our meta-analysis (Supplemental material). Assuming the 17% prevalence of MRI-positive sacroiliitis in patients with chronic low back pain, the post-test probability of a positive test and negative test would be 0.46 and 0.10 using the Fagan’s nomogram, respectively (29) (Supplemental material).

Fig. 2.

Hierarchical summary receiver-operating characteristic curve (HSROC) of plain radiography for sacroiliitis in patients with suspected axial spondyloarthritis.

Fig. 3.

Forest plot for sensitivity and specificity of plain radiography for sacroiliitis in patients with suspected axial spondyloarthritis.

Subgroup analysis and sources of heterogeneity

The results of subgroup analysis and meta-regression for each covariate are presented in Table 3. Study design and whether considering active bone marrow edema as positive MRI were the source of heterogeneity. The two covariates incidentally yielded identical subgroups. The pooled sensitivity was lower in the prospective design study (0.41; 95% CI = 0.23–0.61) than in the retrospective study (0.71; 95% CI = 0.62–0.79) (P = 0.013) and it was lower in the subgroup which considered only active bone marrow edema as positive MRI (0.41 [95% CI = 0.23–61] vs. 0.71 [95% CI = 0.62–0.79]) (P = 0.013). Although the difference was not statistically significant (P = 0.054), the specificity was higher in the subgroups when clinical information was blinded to the readers (0.93 [95% CI = 0.79–0.98] vs 0.72 [95% CI = 0.61–0.81]).

Table 3.

Subgroup analysis and meta-regression.

	Studies (n)	Sensitivity	P value	Specificity	P value
Study design			0.013*		0.229
Prospective	4	0.41 (0.23–0.61)		0.78 (0.59–0.90)
Retrospective	4	0.71 (0.62–0.79)		0.93 (0.76–0.98)
Study institute			0.140		0.495
Single center	6	0.60 (0.39–0.78)		0.83 (0.63–0.93)
Multi center	2	0.43 (0.28–0.60)		0.91(0.69–0.98)
Year of publication			0.962		0.370
Before 2010	4	0.56 (0.27–0.81)		0.93 (0.84–0.97)
2010 and beyond	4	0.54 (0.36–0.71)		0.81 (0.59–0.93)
Included patients (n)			0.962		0.370
≥100	4	0.54 (0.36–0.71)		0.81 (0.59–0.93)
<100	4	0.56 (0.27–0.81)		0.93 (0.84–0.97)
Proportion of men (%)			0.923		0.648
≥50	4	0.53 (0.28–0.77)		0.77 (0.72–0.81)
<50	4	0.56 (0.35–0.75)		0.89 (0.70–0.97)
Proportion of HLA-B27 positivity (%)			NA		NA
≥60	4	0.44 (0.21–0.70)		0.89 (0.69–0.97)
<60	2	0.67 (0.54–0.78)		0.93 (0.83–0.97)
Unclear	2	0.63 (0.50–0.74)		0.63 (0.53–0.72)
Blinded to clinical information when interpreting radiography or MRI			0.493		0.054
Yes	5	0.53 (0.27–0.78)		0.93 (0.79–0.98)
Unstated	3	0.57 (0.51–0.62)		0.72 (0.61–0.81)
MRI strength (T)			NA		NA
≥1.5	4	0.42 (0.18–0.70)		0.92 (0.73–0.98)
<1.5	2	0.56 (0.50–0.62)		0.78 (0.72–0.82)
Unclear	2	0.73 (0.64–0.80)		0.70 (0.56–0.81)
Inclusion of essential MR sequences^†			0.355		0.965
Yes	4	0.52 (0.34–0.69)		0.87(0.68–0.96)
No	4	0.56 (0.29–0.80)		0.81(0.42–0.96)
Diagnostic criteria of MRI			0.92		0.518
Published criteria	2	0.55 (0.24–0.83)		0.89 (0.57–0.98)
Institutional criteria	6	0.55 (0.38–0.72)		0.84 (0.67–0.93)
Readers			NA		NA
Exclusively radiologists	4	0.56 (0.27–0.81)		0.93 (0.84–0.97)
Radiologist/Clinicians	3	0.55 (0.33–0.75)		0.86 (0.63–0.96)
Unclear	1	0.50 (0.26–0.74)		0.62 (0.51–0.71)
Criteria for MRI positivity			0.013*		0.229
Considered only active bone marrow edema	4	0.41 (0.23–0.61)		0.78 (0.59–0.90)
Considered chronic lesion	4	0.71 (0.62–0.79)		0.93 (0.76–0.98)
Usage of contrast enhancement			0.46		0.528
Yes	3	0.64 (0.52–0.74)		0.83 (0.52–0.95)
No	5	0.49 (0.28–0.71)		0.88 (0.72–0.96)
Inclusion criteria			0.14		0.466
Inflammatory back pain	4	0.53 (0.26–0.78)		0.89 (0.71–0.96)
Chronic low back pain	2	0.33 (0.27–0.39)		0.87 (0.48–0.98)
Axial spondyloarthritis	2	0.73 (0.64–0.80)		0.70 (0.56–0.81)

Values are given as pooled sensitivities (or specificities) with 95% confidence intervals in the parentheses.

P < 0.05.

†

Essential MRI sequence was defined as protocol including both T1-weighted and fluid sensitive sequences, and with at least one view of oblique coronal plane.

MRI, magnetic resonance imaging; NA, not available.

Discussion

We performed a meta-analysis on the diagnostic performance of plain radiography in detection of sacroiliitis when the reference standard was set to MRI findings. The plain radiography showed a reasonable pooled specificity of 0.87 but a low pooled sensitivity of 0.55. Furthermore, the included studies showed considerable heterogeneity on the diagnostic performance of radiography. The meta-regression analysis showed prospective study design rather than retrospective study design and MRI criteria considering only bone marrow edema as positive finding were associated with lower sensitivity.

Plain radiography has been considered as the initial imaging modality of choice in patients with suspected axSpA since it has advantages in that it is inexpensive and accessible. However, the pooled sensitivity of 0.55 in detecting sacroiliitis with plain radiography questions its role. The detection of active inflammation of the sacroiliac joint is important not only for diagnosing non-radiographic spondyloarthritis but also for determining whether to use biological agents such as tumor necrosis factor alpha inhibitors (30). Therefore, additional MRI scans may be needed to detect active inflammation even in the patients who were already diagnosed as axSpA. Furthermore, the proportion of patients with suspected axSpA benefiting from plain radiography seemed quite limited. When plain radiography was performed in patients with inflammatory back pain in the SPACE cohort (20), only 9/105 (8.6%) patients with inflammatory back pain and 12/157 (7.6%) patients with chronic low back pain exhibited positive findings on plain radiography. The study found that a considerable portion of patients with inconclusive outcomes or negative findings on plain radiography needed additional MRI examinations for the diagnosis of axSpA.

Although we undoubtedly diagnose a patient with positive radiographic findings as having axSpA in practice, our result suggests that considerable portion of patients might have false-positive diagnoses. The post-test probability of plain radiography was 0.46 if the radiography was positive for axSpA. Owing to high probability of false-positive diagnosis in plain radiography, we might have to consider MRI examination even in the patients with radiographic findings of grade II.

We found significant heterogeneity in the diagnostic performance of plain radiography, which was not explained by a threshold effect. The subgroup analysis showed that such heterogeneity was caused by two factors: (i) whether it is a prospective study; and (ii) whether MRI diagnosis solely included bone marrow edema as positive marker. Since the two covariates yielded identical subgroups by chance, it would be difficult to decide which covariate truly produced the heterogeneity in the present study. We assumed that both factors might have affected the heterogeneity of diagnostic performance of plain radiography in our study. First, in the retrospective design, there exists a selection bias that could not be eliminated with study design, and clinical and laboratory information would be less strictly managed compared with prospective studies. Therefore, overestimation of the sensitivity in the retrospective studies and more conservative results in the prospective studies would be a plausible consequence. Second, considering only active bone marrow edema as positive MRI may lower the sensitivity of plain radiography. Active bone marrow edema, by itself, is not detectable on plain radiography and it is the chronic structural changes that enables the diagnosis of sacroiliitis with plain radiography. By considering the chronic components of sacroiliitis on MRI, more wide range of imaging findings (i.e. subchondral sclerosis, erosions) can be interpreted as positive sacroiliitis, which would yield more true-positive results than considering only active bone marrow edema.

Although plain radiography has certain limitations mentioned above, it seems difficult to completely omit radiography in the diagnostic procedure of axSpA at this time. Other image modalities, which are suggested as alternatives to plain radiography such as CT and MRI, also have some limitations in diagnosing axSpA. CT has strength in detecting structural damage and there is an effort to detect active inflammation using dual-energy CT (31), but there are concerns for a potential radiation hazard. MRI has strength in detecting active inflammation (32), but there are disagreements as to which of the chronic structural lesions detected on MRI findings might be meaningful in diagnosing and treating axSpA (33). Further investigations of CT and MRI are warranted to diagnose axSpA more accurately.

The present study has several limitations. First, only a small number of studies were included. Nevertheless, we could find there was limited sensitivity of plain radiography in detecting sacroiliitis and reported diagnostic performance of plain radiography was heterogeneous across the included articles. Second, most of the included studies used their institutional criteria for diagnosis of sacroiliits on MRI, as the criteria for diagnosing sacroiliitis on MRI were not established at the point of publication of the included studies. Third, detailed protocol for plain radiography and MRI was often missing in the included studies, which may lower the reproducibility of the study results.

In conclusion, plain radiography showed limited diagnostic performance in detecting sacroiliitis compared with MRI. Although radiography is currently used as an initial screening test for axSpA, its role in diagnostic procedure of axSpA should be reconsidered, taking its limited diagnostic performance into account.

Supplemental Material

sj-pdf-1-acr-10.1177_0284185120930624 - Supplemental material for Diagnostic performance of plain radiography for sacroiliitis in patients with suspected axial spondyloarthritis: a systematic review and meta-analysis

Supplemental material, sj-pdf-1-acr-10.1177_0284185120930624 for Diagnostic performance of plain radiography for sacroiliitis in patients with suspected axial spondyloarthritis: a systematic review and meta-analysis by Youngjune Kim, Choong Guen Chee, Junghoon Kim, Jungheum Cho, Min A Yoon and Hye Won Chung in Acta Radiologica

Footnotes

Authors’ note

Youngjune Kim’s current affiliation is: Aerospace Medical Group, Air Force Education and Training Command, Jinju, Republic of Korea.

Acknowledgement

The authors thank Jong-Bok Lee from the Division of Biostatistics, Asan Medical Center for his assistance in statistical analysis.

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.

ORCID iDs

Youngjune Kim

Choong Guen Chee

Jungheum Cho

Supplemental material

Supplemental material for this article is available online.

References

Taurog

Chhabra

Colbert

RA.

Ankylosing spondylitis and axial spondyloarthritis. N Engl J Med 2016; 374:2563–2574.

Amor

Dougados

Mijiyawa

Criteria of the classification of spondylarthropathies. Rev Rhum Mal Osteoartic 1990; 57:85–89.

Dougados

van der Linden

Juhlin

, et al. The European Spondylarthropathy Study Group preliminary criteria for the classification of spondylarthropathy. Arthritis Rheum 1991; 34:1218–1227.

Rudwaleit

Landewe

van der Heijde

, et al. The development of Assessment of SpondyloArthritis international Society classification criteria for axial spondyloarthritis (part I): classification of paper patients by expert opinion including uncertainty appraisal. Ann Rheum Dis 2009; 68:770–776.

van der Linden

Valkenburg

Cats

Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum 1984; 27:361–368.

Hollingsworth

Cheah

Dawkins

, et al. Observer variation in grading sacroiliac radiographs in HLA-B27 positive individuals. J Rheumatol 1983; 10:247–254.

Yazici

Turunc

Ozdogan

, et al. Observer variation in grading sacroiliac radiographs might be a cause of ‘sacroiliitis’ reported in certain disease states. Ann Rheum Dis 1987; 46:139–145.

Taylor

Wardle

Beswick

, et al. The relationship of clinical and laboratory measurements to radiological change in ankylosing spondylitis. Br J Rheumatol 1991; 30:330–335.

van Tubergen

Heuft-Dorenbosch

Schulpen

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

10.

Poddubnyy

Radiographic evaluation of sacroiliac joints in axial spondyloarthritis—still worth performing?

J Rheumatol 2017; 44:1–3.

11.

Navallas

Ares

Beltran

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

12.

van den Berg

Lenczner

Thevenin

, et al. Classification of axial SpA based on positive imaging (radiographs and/or MRI of the sacroiliac joints) by local rheumatologists or radiologists versus central trained readers in the DESIR cohort. Ann Rheum Dis 2015; 74:2016–2021.

13.

Braun

Sieper

Bollow

Imaging of sacroiliitis. Clin Rheumatol 2000; 19:51–57.

14.

McInnes

MDF

Moher

Thombs

, et al. Preferred reporting items for a systematic review and meta-analysis of diagnostic test accuracy studies: the PRISMA-DTA statement. JAMA 2018; 319:388–396.

15.

Mandl

Navarro-Compán

Terslev

, et al. EULAR recommendations for the use of imaging in the diagnosis and management of spondyloarthritis in clinical practice. Ann Rheum Dis 2015; 74:1327–1339.

16.

Whiting

Rutjes

Westwood

, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011; 155:529–536.

17.

Battistone

Manaster

Reda

, et al. Radiographic diagnosis of sacroiliitis—are sacroiliac views really better? J Rheumatol 1998; 25:2395–2401.

18.

Deville

Buntinx

Bouter

, et al. Conducting systematic reviews of diagnostic studies: didactic guidelines. BMC Med Res Methodol 2002; 2:9.

19.

Deeks

Macaskill

Irwig

The performance of tests of publication bias and other sample size effects in systematic reviews of diagnostic test accuracy was assessed. J Clin Epidemiol 2005; 58:882–893.

20.

van den Berg

de Hooge

van Gaalen

, et al. Percentage of patients with spondyloarthritis in patients referred because of chronic back pain and performance of classification criteria: experience from the Spondyloarthritis Caught Early (SPACE) cohort. Rheumatology (Oxford) 2013; 52:1492–1499.

21.

Ortolan

van Lunteren

Ramiro

, et al. Are gender-specific approaches needed in diagnosing early axial spondyloarthritis? Data from the SPondyloArthritis Caught Early cohort. Arthritis Res Ther 2018; 20:218.

22.

Sudol-Szopinska

Wlodkowska-Korytkowska

Kwiatkowska

Spectrum of inflammatory changes in the SIJs on radiographs and MR images in patients with suspected axial spondyloarthritis. Pol J Radiol 2016; 81:125–133.

23.

Heuft-Dorenbosch

Landewe

Weijers

, et al. Combining information obtained from magnetic resonance imaging and conventional radiographs to detect sacroiliitis in patients with recent onset inflammatory back pain. Ann Rheum Dis 2006; 65:804–808.

24.

Inanc

Atagunduz

Sen

, et al. The investigation of sacroiliitis with different imaging techniques in spondyloarthropathies. Rheumatol Int 2005; 25:591–594.

25.

Oostveen

Prevo

den Boer

, 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.

26.

Blum

Buitrago-Tellez

Mundinger

, et al. Magnetic resonance imaging (MRI) for detection of active sacroiliitis–a prospective study comparing conventional radiography, scintigraphy, and contrast enhanced MRI. J Rheumatol 1996; 23:2107–2115.

27.

Blachier

Canoui-Poitrine

Dougados

, et al. Factors associated with radiographic lesions in early axial spondyloarthritis. Results from the DESIR cohort. Rheumatology (Oxford) 2013; 52:1686–1693.

28.

Rudwaleit

Jurik

Hermann

, et al. Defining active sacroiliitis on magnetic resonance imaging (MRI) for classification of axial spondyloarthritis: a consensual approach by the ASAS/OMERACT MRI group. Ann Rheum Dis 2009; 68:1520–1527.

29.

Dougados

d’Agostino

Benessiano

, et al. The DESIR cohort: a 10-year follow-up of early inflammatory back pain in France: study design and baseline characteristics of the 708 recruited patients. Joint Bone Spine 2011; 78:598–603.

30.

Braun

Landewe

Hermann

, et al. Major reduction in spinal inflammation in patients with ankylosing spondylitis after treatment with infliximab: results of a multicenter, randomized, double-blind, placebo-controlled magnetic resonance imaging study. Arthritis Rheum 2006; 54:1646–1652.

31.

Zhang

Shi

, et al. Axial spondyloarthritis: dual-energy virtual noncalcium CT in the detection of bone marrow edema in the sacroiliac joints. Radiology 2019; 290:157–164.

32.

Braun

Bollow

Eggens

, et al. Use of dynamic magnetic resonance imaging with fast imaging in the detection of early and advanced sacroiliitis in spondylarthropathy patients. Arthritis Rheum 1994; 37:1039–1045.

33.

Lukas

Cyteval

Dougados

, et al. MRI for diagnosis of axial spondyloarthritis: major advance with critical limitations ‘Not everything that glisters is gold (standard)’. RMD Open 2018; 4:e000586.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.25 MB