ACR2010_分析TEMPO研究数据发现12周疗效可以预测依那西普联合甲氨蝶呤治疗RA达1年时的疗效

初始使用ETN或ETN+MTX的患者在开始治疗12周内，可以分类为可能在第52周疗效好或不好。
        

原文	译文
[1117] - Prediction of One-Year Response to Etanercept and Methotrexate in Rheumatoid Arthritis Patients in TEMPO. Jeffrey R Curtis, MD, MPH 1,Shuo Yang2,Lang Chen, MD3,Grace S Park, DrPH4,Bojena Bitman, MS Biostatistics5,Brian C. Wang, M.S.6,Iris Navarro-Millan, MD7,Arthur Kavanaugh, MD8. 1Rheumatology & Immunology, University of AL Brimingham, Birmingham, AL,2University of Alabama at Birmingham,3University of Alabama at Birmingham, Birmingham, AL,4Amgen Inc., Thousand Oaks, CA,5Amgen Inc., South San Francisco, CA,6KForce Clinical Research, South San Francisco, CA,7University of Alabama at Birmingham, Vestavia, AL,8Rheumatology/Allergy Immunology, La Jolla, CA Background: The ability to predict which rheumatoid arthritis (RA) patients (pts) will respond to biologic therapy is a challenging but important endeavor. TEMPO (Trial of Etanercept and Methotrexate with Radiographic Patient Outcomes) data were used to evaluate moderate to severe RA pts initiating etanercept (ETN) with or without methotrexate (MTX) to derive and validate a decision tree that would predict which pts would have a good response after 1 year of therapy, determine which pts could have a treatment decision made at week (wk) 12, and identify pts who needed additional time on therapy. Methods: Pts were included if they had disease activity scores (DAS28) at 52 (or 48) wks of therapy with either ETN 25 mg twice weekly with MTX (n = 193) or ETN alone (n = 172). Pts were classified as responders if they had low disease activity (LDA) at wk 52, i.e., DAS28 ≤ 3.2. Pts who dropped out for safety reasons were excluded. Classification and Regression Trees (CART) software [Salford Systems], which identifies variables affecting the outcome and partitions data accordingly, was used to develop and validate decision trees on clinical and demographic variables, stratified by TEMPO treatment arm. Tenfold cross-validation was used to guard against overfitting. Results: Sixty percent (115/193) of pts in the ETN + MTX arm (Figure 1) and 39% (67/172) of pts in the ETN arm (Figure 2) achieved LDA response at wk 52. As shown in Figure 1, LDA was predicted by DAS28 at wk 12 and DAS28 change from baseline values at wk 8. Pts were categorized into 3 groups by 12 wks: 1) high-likelihood responders, such that 63% (121/193) of pts were able to be classified as responders by wk 12 with 81% accuracy (98/121 correctly classified as responders); 2) non-responders, 25% (49/193) of pts with 88% (43/49) accuracy; and 3) uncertain likelihood of response for the remaining 12% (23/193) of pts.	分析TEMPO研究数据发现12周疗效可以预测依那西普联合甲氨蝶呤治疗RA达1年时的疗效         Curtis JR, et al. ACR 2010. Present No: 1117.      背景: 医生尽力想预测哪些RA患者对生物制剂反应好。TEMPO试验评估初始使用依那西普（ETN）合并或不合并甲氨蝶呤（MTX）的中重度RA患者，以期建立并验证决策树，预测哪些患者治疗1年后反应好，哪些患者在第12周可以确定治疗方案，哪些患者需要更长时间治疗。 方法: 患者使用ETN（25mg，每周2次）加用MTX（193例）或不加用MTX（172例）治疗52周（或48周）。如果患者在第52周为低疾病活动度（LDA），即DAS28 ≤ 3.2，则为反应者。排除因安全性原因而退出的患者。采用分类和回归树（CART）软件（Salford系统）建立并验证有关临床和人口学变量的决策树，相关变量根据TEMPO治疗组进行分层。 结果: 60%（115/193）ETN + MTX组患者（图1）和39%（67/172）ETN组患者（图2）在第52周达到LDA。如图1所示，LDA可由第12周DAS28和从基线期到第8周的DAS28变化所预测。在第12周，患者可分为3组: 1）可能性大的反应者，63%（121/193）患者在第12周分类为反应者的准确性为81%（98/121正确分类为反应者）；2）无反应者，25%（49/193），准确性为88%（43/49）；3）不能确定，12%（23/193）。
In the ETN only arm, results were similar (Figure 2). Response to therapy in this cohort was predicted by DAS28 at wk 12 and tender joint count at wk 8. By wk 12, 53% (92/172) of pts were able to be classified as responders, 39% of pts as non-responders (67/172), and the remaining 8% were not able to be classified well.	ETN组结果相似（图2）。疗效好的预测因素为第12周DAS28和第8周关节压痛数。在第12周，53%（92/172）患者分类为反应者，39%为无反应者（67/172），另外8%不能确定。
Conclusion: Most TEMPO pts initiating ETN with or without MTX could be classified within 12 wks of starting therapy as likely to have a good response or not at wk 52. However, approximately 10–15% of pts needed additional time on therapy to decide whether to continue ETN. This preliminarily validated decision tree needs replication and may help physicians in deciding whether to continue or change anti-TNF therapy at 12 wks.	结论: 初始使用ETN或ETN+MTX的患者在开始治疗12周内，可以分类为可能在第52周疗效好或不好。然而，10-15%患者需要更长时间治疗才能确定是否继续使用ETN。这种初步验证的决策树需要重复验证，可能有助于临床医生在第12周决定是否继续或更换抗TNF治疗。