TEMPO研究第一年影像学数据: 骨侵蚀修复几乎只出现在无关节肿胀或肿胀改善组

D. van der Heijde ¹, C. Lukas ¹, S. Fatenejad ², R. Landewe ¹.

¹Rheumatology, University Hospital, Maastrict, Netherlands, ²Research and Development, Wyeth, Collegeville, United States

背景：双盲试验中vdHSharp评分变化呈负数提示有效的治疗可以修复关节。在单关节水平进行研究能帮助进一步理解关节修复过程。

目的：如果真的存在关节修复(repair)，判断修复是否偏好发生于无肿胀或肿胀有改善的关节。

方法：TEMPO试验第1年MTX单用组(M)、Etanercept+MTX组(M+E)患者手/腕和足部摄片后，对所有单关节的判读结 果进行评估。采用vdHSharp评分系统，对治疗和摄片顺序均不知晓的两位读片师各自对所有平片重复判读两次。计算单关节骨侵蚀评分变化，并与单关节肿 胀评分变化相关联。单关节修复的评判：四次判读中至少有一次为负数变化而其它判读变化结果均为零(即无进展)。每关节骨侵蚀变化均数，是通过减去基线均数 而得。

结果：共计11159个单关节中，判读为有修复的为557个，其中553个同时有肿胀评分。下表显示各种肿胀评分变化在"修复关节"组、 "无修复关节"组中的分布。修复组无肿胀关节共计234个，其中12个有残余肿胀，222个无肿胀。肿胀改善即评分变化为负数的关节共计318个，仅36 个有残余肿胀。与无修复相比，修复与肿胀改善显著相关(p<0.0001)。

持续肿胀组骨侵蚀变化均数(95%可信区间)如下，基线无破坏组为0.03 [0.01，0.04]，基线有破坏组为0.06 [–0.02，0.14])，而无肿胀或肿胀改善组患者的更低，基线无破坏组为0.01[0.00，0.01]，基线有破坏组为-0.09[-0.11，-0.06]。基线有骨侵蚀时，骨侵蚀变化均数只在无肿胀或肿胀改善亚组呈显著负数变化。

结论：骨侵蚀修复几乎只出现在肿胀改善或肿胀消失组。持续肿胀关节中的骨破坏仍在进展，尤其是基线已有骨损害者。这项观察研究进一步确证了骨侵蚀负数变化是骨修复的反映。

[2007] [OP0011] REPAIR OF EROSIONS OCCURS ALMOST EXCLUSIVELY IN DAMAGED JOINTS WITHOUT SWELLING: POST HOC ANALYSIS OF RADIOGRAPHIC DATA FROM YEAR 1 OF THE TEMPO STUDY

D. van der Heijde ¹, C. Lukas ¹, S. Fatenejad ², R. Landewe ¹

¹Rheumatology, University Hospital, Maastrict, Netherlands, ²Research and Development, Wyeth, Collegeville, United States

Objectives: To determine whether repair – if it truly exists – preferentially occurs in joints with no swelling or improvement in swelling.

Methods: Single-joint readings of radiographic images of the hands/wrists and feet from patients in year 1 of the TEMPO trial (the methotrexate-only group [M] and the methotrexate+etanercept group [M+E]) were evaluated. Using the SvdH scoring, 2 readers blinded to treatment and true-time sequence independently assessed each of the radiographs twice. Single-joint change scores in erosions were calculated and coupled with change in single-joint swelling scores. Repair in a joint was considered to have occurred if there was a negative erosion change score in at least 1 of the 4 potential readings with the remaining readings showing zero, ie, no progression. Mean erosion change scores per joint were calculated by taking the mean score from the first reading by each reader.

Results: Of the 11,159 single joints, 557 showed repair. For 553 of these, swelling scores were also available. The table shows the distribution of change in swelling in joints showing "repair" versus "no repair". Of the 234 joints without change in swelling in the repair group, 12 had residual swelling and 222 had no swelling. Of the 318 joints with improvement in swelling (ie negative change), only 36 had residual swelling. Repair was significantly more associated with improvement in swelling than no repair (p<0.0001).

Mean change in erosion scores (95% confidence interval [CI]) were lower in patients with no swelling or improvement in swelling (group without baseline damage 0.01 [0.00, 0.01]; group with baseline damage –0.09 [–0.11; -0.06]) compared with patients with persistent swelling (group without baseline damage 0.03 [0.01, 0.04]; group with baseline damage 0.06 [–0.02, 0.14]). The mean change in erosion score was statistically significantly negative only in the subgroup of joints with absent or improving swelling, while erosions were present at baseline.

Conclusion: Repair of erosions occurs almost exclusively in damaged joints that show either improvement of swelling, or that have no swelling at all. Progression occurs in joints with persistent swelling, preferably if there is already damage present. This observation adds to the validity that negative joint scores are a reflection of repair.