Department of Orthopedics, All India Institute of Medical Sciences, Bhubaneswar, 751019, Odisha, India.
Department of Radiology, AIIMS Bhubaneswar, Bhubaneswar, 751019, Odisha, India.
Eur Spine J. 2024 Sep;33(9):3409-3419. doi: 10.1007/s00586-024-08453-w. Epub 2024 Aug 21.
The objective of this study is to evaluate the prognostic value of nutritional scores comprising the Controlling Nutritional Status (CONUT) score and the Prognostic Nutritional Index (PNI), in prediction of multilevel vertebral involvement (> 2 vertebra) in Spinal Tuberculosis (STB).
Retrospective analysis of 39 STB patients was conducted to assess nutritional indices (CONUT and PNI) and the numbers of vertebral affection. Spearman's correlation was used to examine the association between these variables. Receiver Operating Characteristic (ROC) curves were utilized to determine optimal cutoff values, with Area Under the Curve (AUROC) evaluation. Additionally, multiple logistic regression was performed as a predictive model.
There were 24 males and 15 females, with a mean BMI of 18.88 kg/m² (± 1.37). Spearman's correlation analysis revealed negative correlations between BMI, PNI (rho - 0.68, p < 0.001) and multilevel vertebra involvement, while ESR (rho 0.83, p < 0.001), CRP (rho 0.81, p < 0.001), and CONUT score (rho 0.83, p < 0.001) positively correlated with multilevel vertebral affection (> 2 vertebra). Age and comorbidities showed no correlation with the level of vertebral affection. ROC analysis revealed a CONUT Score ≥ 3 cutoff (sensitivity-95.7%, specificity-87.5%) and PNI ≤ 38.605 (sensitivity-78.3%, specificity-93.8%) for predicting multilevel STB (> 2). PNI exhibited superior specificity and positive predictive value where as CONUT score was a better parameter for sensitivity, negative predictive value and diagnostic accuracy. Both CONUT score and PNI were significant predictors of vertebral involvement in univariate analysis, with multivariate analysis identifying CONUT score as the sole predictor of multilevel vertebral affection.
Nutritional scores, including CONUT score and PNI, emerged as significant predictors of multilevel STB. CONUT score displayed superior sensitivity, negative predictive value, and overall diagnostic accuracy, while PNI served as a nutritional marker with high specificity and positive predictive value in predicting multilevel involvement in spinal tuberculosis.
本研究旨在评估包含控制营养状况（CONUT）评分和预后营养指数（PNI）的营养评分对预测脊柱结核（STB）多节段椎体受累（>2 个椎体）的预后价值。
对 39 例 STB 患者进行回顾性分析，评估营养指标（CONUT 和 PNI）和椎体受累数量。采用 Spearman 相关分析评估这些变量之间的相关性。使用受试者工作特征（ROC）曲线确定最佳截断值，并评估曲线下面积（AUROC）。此外，还进行了多元逻辑回归作为预测模型。
患者中男性 24 例，女性 15 例，平均 BMI 为 18.88kg/m²（±1.37）。Spearman 相关分析显示，BMI、PNI 与多节段椎体受累呈负相关（rho -0.68，p<0.001），而 ESR（rho 0.83，p<0.001）、CRP（rho 0.81，p<0.001）和 CONUT 评分（rho 0.83，p<0.001）与多节段椎体受累呈正相关（>2 个椎体）。年龄和合并症与椎体受累程度无相关性。ROC 分析显示，CONUT 评分≥3 分（敏感性 95.7%，特异性 87.5%）和 PNI≤38.605（敏感性 78.3%，特异性 93.8%）可预测多节段 STB（>2）。PNI 具有更高的特异性和阳性预测值，而 CONUT 评分是敏感性、阴性预测值和诊断准确性的更好参数。CONUT 评分和 PNI 在单因素分析中均为椎体受累的显著预测因素，多因素分析则确定 CONUT 评分是多节段椎体受累的唯一预测因素。
营养评分，包括 CONUT 评分和 PNI，是多节段 STB 的重要预测指标。CONUT 评分具有较高的敏感性、阴性预测值和整体诊断准确性，而 PNI 作为一种具有高特异性和阳性预测值的营养标志物，可预测脊柱结核的多节段受累。
