生物标志物联合临床参数预测中轴型脊柱关节炎放射学进展的增益价值
重复交叉验证发现瘦素+高分子量脂联素+VEGF组合有潜在增益,联合模型AUC为0.768,但较临床模型AUC 0.665的差异未达统计学显著(P=0.051);在高特异性区域部分AUC改善显著(P=0.01)。结论为生物标志物有潜在增益但幅度较小。
要点
- 临床模型(基线韧带骨赘、CRP升高、吸烟、男性)预测2年放射学进展AUC为0.665。
- 瘦素+高分子量脂联素+VEGF联合临床参数AUC升至0.768,但差异未达显著(P=0.051)。
- 在高特异性(≥75%)区域,联合模型部分AUC改善显著(0.119 vs 0.053,P=0.01)。
核心结论
- 生物标志物联合临床参数对中轴型脊柱关节炎放射学脊柱进展有潜在预测增益,但增益幅度较小。
研究设计与样本
- 基线用ELISA检测MMP3、VEGF、钙卫蛋白、瘦素、高分子量脂联素、骨保护素、硬骨素、N端端肽、II型前胶原N端前肽和血清淀粉样蛋白A。
- 放射学脊柱进展定义为2年后改良Stoke强直性脊柱炎脊柱评分(mSASSS)恶化≥2分。
主要发现
- 重复交叉验证显示瘦素+高分子量脂联素+VEGF组合具有潜在附加预测价值。
- 生物标志物模型AUC为0.731(95%CI:0.614,0.848),数值上优于临床模型AUC 0.665(95%CI:0.553,0.776)。
- 临床+生物标志物联合模型AUC为0.768(95%CI:0.666,0.871),较临床模型的改善未达统计学显著(P=0.051)。
- 以特异度≥75%计算部分AUC,联合模型pAUC为0.119,临床模型为0.053,差异显著(P=0.01)。
关键数值与推荐
- 放射学脊柱进展定义:2年后mSASSS恶化≥2分。
- 临床预测模型组成:基线韧带骨赘、CRP升高、吸烟、男性。
- 潜在增益生物标志物组合:瘦素+高分子量脂联素+VEGF。
- 联合模型AUC 0.768(95%CI:0.666,0.871),临床模型AUC 0.665(95%CI:0.553,0.776),P=0.051。
- 高特异性区域pAUC:联合模型0.119 vs 临床模型0.053,P=0.01。
临床怎么用
- 在常规临床参数基础上,可考虑检测瘦素、高分子量脂联素和VEGF以辅助预测放射学脊柱进展。
- 联合模型在高特异性区域表现更优,可能适用于需高置信度识别进展风险患者的场景。
- 因总体增益较小且未达传统显著性,生物标志物尚不能替代临床参数用于常规预测。
局限性
- 联合模型较临床模型的AUC改善未达统计学显著(P=0.051)。
- 生物标志物附加预测价值幅度较小。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
OBJECTIVES: The objective of this study was to examine whether adding biomarkers to routine clinical parameters improves prediction of radiographic spinal progression in axial spondyloarthritis. METHODS: One hundred and seventeen patients with ankylosing spondylitis who completed the Effects of NSAIDs on RAdiographic Damage in Ankylosing Spondylitis (ENRADAS) trial were included. Radiographic spinal progression was defined as worsening of the modified Stoke Ankylosing Spondylitis Spine Score by ⩾2 points after 2 years. A clinical prediction model was constructed out of baseline syndesmophytes, elevated CRP, cigarette smoking and male sex. The following serum biomarkers were measured at baseline by ELISA: MMP3, VEGF, calprotectin, leptin, high molecular weight adiponectin, osteoprotegerin, sclerostin, N-terminal telopeptide, procollagen type II N-terminal propeptide and serum amyloid A. RESULTS: Repeated cross-validation analyses revealed one biomarker combination with potential added predictive value in addition to the clinical model: leptin + high molecular weight adiponectin + VEGF. This biomarker combination showed an area under the curve (AUC)Biomarkers = 0.731 (95% CI: 0.614, 0.848), which was numerically superior to the clinical model [AUCClinical = 0.665 (95% CI: 0.553, 0.776)]. A combination of clinical parameters + biomarkers showed an improved predictive value compared with the clinical model reflected by AUCClinical+Biomarkers = 0.768 (95% CI: 0.666, 0.871), though not statistically significant (P = 0.051). However, by considering the part of the receiver operating characteristic curve with a specificity ⩾75% resulting in partial AUC (pAUC), the improvement becomes significant (pAUCClinical+Biomarkers = 0.119; pAUCClinical = 0.053; P = 0.01). CONCLUSION: Biomarkers show potential to improve the prediction of radiographic spinal progression in axial spondyloarthritis when used in addition to the clinical parameters, though the added value seems to be rather small.
引用
- Rademacher J, Tietz LM, Le L, et al. Added value of biomarkers compared with clinical parameters for the prediction of radiographic spinal progression in axial spondyloarthritis. Rheumatology (Oxford). 2019. PMID: 30830164. doi:10.1093/rheumatology/kez025
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。