|
Diagnostic Utility of Sacroiliac
Joint MRI in Non-Radiographic Spondyloarthritis: Validation of
Three Assessment Methods in Two Inception Cohorts
Ulrich Weber
1, Veronika Zubler2,
Susanne Juhl Pedersen3, Stanley Chan1, Kaspar
Rufibach4, Robert GW Lambert1 and Walter P.
Maksymowych1, 1University of Alberta,
Edmonton, AB, 2Balgrist University Hospital, Zurich,
Switzerland, 3Copenhagen University Hospital at
Glostrup, Copenhagen, Denmark, 4University of Zurich,
Zurich, Switzerland
Presentation
Number:
1320
Background/Purpose: A systematic evaluation of
sacroiliac joint (SIJ) MRI based on both active and structural
lesions (global evaluation) has shown high diagnostic utility in
axial spondyloarthritis (SpA) [1]. The incorporation of erosions
(ER) with bone marrow edema (BME) as defined by ASAS improved
diagnostic utility compared to BME alone.
Objective: To validate global and lesion-based
(BME, BME plus ER) approaches to evaluation of SIJ by MRI in two
inception cohorts of consecutive patients with back pain suspected
of having SpA.
Method: Cohort A comprised
88 consecutive patients, age ≤50, with acute anterior uveitis (AAU)
and back pain referred by an ophthalmology centre. Cohort B
comprised 69 consecutive patients ≤50 years referred by practising
rheumatologists and primary care physicians to a tertiary care
center for further assessment of suspected inflammatory back pain.
They were classified according to clinical, laboratory and imaging
evaluation as having non-radiographic SpA (nrSpA) (n=20; 65% HLAB27
positive, 55% male), ankylosing spondylitis (AS) by the modified
New York criteria (mNYc) (n=10), and non-specific back pain (NSBP)
(n=39). Scans from 20 local healthy controls (HC) were also
assessed. SIJ MRI were scored independently in random order by 4
readers blinded to patient identifiers.
Result: The assessment of
ER in addition to BME enhanced diagnostic utility in both cohorts
despite differences in the prevalence of SpA. For the AAU cohort,
AS by mNYc was diagnosed in 28/88 (31.8%) and 16/88 (18.2%) were
diagnosed with NSBP. Of 44/88 (50%) that had suspected SpA (68%
HLAB27 positive, 52% male), sensitivity/specificity of MRI
according to 2 readers was 32%/100% by global assessment, 27%/100%
for BME alone, and 36%/100% for BME plus ER, compared to NSBP.
Similar data was noted in cohort B. Sensitivity/specificity (mean
of 4 readers) of SIJ MRI for nrSpA compared to NSBP was 75%/96% by
global assessment, 72%/85% for BME alone, and 81%/83% for BME plus
ER. In cohort B, ER was detected in 75% of nrSpA patients. ER was
not recorded in NSBP patients and HC, but both groups showed BME
(NSBP in 23%, HC in 20%).
Table 1: Cohort B, nrSpA versus
NSBP
|
Assessment method
|
Sens
|
Spec
|
LR+
|
LR-
|
|
Global
|
0.75
|
0.96
|
20.2
|
0.3
|
|
BME alone
|
0.72
|
0.85
|
4.3
|
0.3
|
|
BME + ER
|
0.81
|
0.83
|
7.0
|
0.2
|
Conclusion: In two unselected
cohorts of consecutive patients we show that ER detected by MRI is
highly specific for nrSpA and contributes to diagnostic utility
beyond assessment of BME alone.
[1] Arthritis Rheum
2010;62:3048
|
骶髂关节MRI在放射学阴性的脊柱关节炎中的诊断作用：3种评估法在两个起始队列中的验证
Ulrich Weber, et al. Present No: 1320
背景/目的: 一项基于活动性和结构性损害(整体评估)的骶髂关节(SIJ)MRI的系统评价,显示其对中轴SpA有较高的诊断价值[1]。ASAS定义的侵蚀(ER)与骨髓水肿(BME)同时出现比仅有BME诊断价值更高。
目的:
在两个有腰背痛疑诊为SpA的起始队列中，验证以整体和损害为基础(BME、BME加ER)的方法在评价MRI检查中SIJ的作用。
方法:队列A为88个连续的患者组成, 年龄≤50岁,因急性前葡萄膜炎(AAU)和背痛由一个眼科中心转诊而来。队列B为69个年龄≤50岁的连续患者，由执业风湿病医生和初级医师转诊给一家三级医疗中心对可疑的炎症行背痛作进一步评价。根据临床、实验室及影像学检查分类为放射学阴性
SpA(nrSpA)(n =
20;65%HLAB27阳性,
55%男性),符合改良纽约标准(mNYc)的强直性脊柱炎(AS) (n=10例),和非特异性背部疼痛(NSBP)(n = 39)。同时评估当地20名健康人(HC)的影像。MRI的SIJ
由4位阅片者随机盲法独立评估。
结果:BME加上ER的评估在两个队列中均增加了诊断效力，尽管SpA的患病率有所不同。AAU队列中,符合mNYc
标准的AS患者为28/88(31.8%),NSBP患者为16/88(18.2%)。44/88(50%)例疑诊SpA患者(68%
HLAB27阳性,男性52%)，根据2位阅片者的结果，相比NSBP，整体评价时MRI灵敏度／特异性为 32%
/ 100%,单有BME为27% /
100%,BME加ER为
36% / 100%。队列B数据与此类似。nrSpA中MRI SIJ敏感性/特异性 (平均4位阅片者),整体评估为75% /
96%,单有BME为
72% / 85% 、BME 加ER为81% / 83%。队列B中，75%的nrSpA患者检测到ER。而 NSBP患者和HC中,则无ER发现，但两组均发现有BME(NSBP 23%,HC 20%)。
Table 1: Cohort B, nrSpA versus
NSBP
|
Assessment method
|
Sens
|
Spec
|
LR+
|
LR-
|
|
Global
|
0.75
|
0.96
|
20.2
|
0.3
|
|
BME alone
|
0.72
|
0.85
|
4.3
|
0.3
|
|
BME + ER
|
0.81
|
0.83
|
7.0
|
0.2
|
结论：通过两个未筛选的连续患者队列，我们证实MRI检测的ER对nrSpA高度特异，相比单有BME能更好地应用于该病的诊断。
|