Division of Internal Medicine and Hypertension, Department of Medical Sciences, University of Turin, Turin, Italy.
Department of Electronics and telecommunications, Polytechnic University of Turin, Turin, Italy.
Ann Surg. 2020 Dec;272(6):1125-1132. doi: 10.1097/SLA.0000000000003200.
To develop a prediction model for clinical outcomes after unilateral adrenalectomy for unilateral primary aldosteronism.
Unilateral primary aldosteronism is the most common surgically curable form of endocrine hypertension. Surgical resection of the dominant overactive adrenal in unilateral primary aldosteronism results in complete clinical success with resolution of hypertension without antihypertensive medication in less than half of patients with a wide between-center variability.
A linear discriminant analysis model was built using data of 380 patients treated by adrenalectomy for unilateral primary aldosteronism to classify postsurgical clinical outcomes. The total cohort was then randomly divided into training (280 patients) and test (100 patients) datasets to create and validate a score system to predict clinical outcomes. An online tool (Primary Aldosteronism Surgical Outcome predictor) was developed to facilitate the use of the predictive score.
Six presurgical factors associated with complete clinical success (known duration of hypertension, sex, antihypertensive medication dosage, body mass index, target organ damage, and size of largest nodule at imaging) were selected based on classification performance in the linear discriminant analysis model. A 25-point predictive score was built with an optimal cut-off of greater than 16 points (accuracy of prediction = 79.2%; specificity = 84.4%; sensitivity = 71.3%) with an area under the curve of 0.839.
The predictive score and the primary aldosteronism surgical outcome predictor can be used in a clinical setting to differentiate patients who are likely to be clinically cured after surgery from those who will need continuous surveillance after surgery due to persistent hypertension.
建立单侧醛固酮增多症患者单侧肾上腺切除术术后临床结局的预测模型。
单侧醛固酮增多症是最常见的可通过手术治愈的内分泌性高血压。单侧醛固酮增多症中优势活性肾上腺的手术切除可使近一半患者的高血压完全治愈，无需降压药物治疗，但降压成功率存在中心间的较大差异。
采用线性判别分析模型，对 380 例单侧醛固酮增多症患者的肾上腺切除术治疗数据进行分析，以对术后临床结局进行分类。然后，将总队列随机分为训练集（280 例患者）和测试集（100 例患者），以创建和验证预测术后临床结局的评分系统。开发了一个在线工具（原发性醛固酮增多症手术结局预测器），以方便使用预测评分。
根据线性判别分析模型中的分类性能，选择了 6 个术前因素（高血压已知持续时间、性别、降压药物剂量、体重指数、靶器官损害和影像学最大结节大小）与完全临床治愈相关。建立了一个 25 分的预测评分，最佳截断值为大于 16 分（预测准确率为 79.2%；特异性为 84.4%；敏感性为 71.3%），曲线下面积为 0.839。
该预测评分和原发性醛固酮增多症手术结局预测器可用于临床，区分术后可能临床治愈的患者和术后因持续性高血压而需要持续监测的患者。
