[The analysis of related factors on the unilateral aldosterone secretion in the functional diagnosis of primary aldosteronism and establishment of a predictive model].

ABSTRACT

Objective: To explore related factors associated with unilateral aldosterone secretion of patients with primary aldosteronism (PA) and construct a predictive model. Methods: The clinical data of patients who were diagnosed as PA in West China Hospital from April 2016 to September 2020 was analyzed retrospectively. According to the results of the adrenal enhancement CT, patients were divided into 3 groups, namely non-nodular group with no bilateral adrenal glands lesions, only with unilateral adrenal hyperplasia or bilateral adrenal hyperplasia, unilateral nodule group with unilateral adrenal nodules and the contralateral adrenal glands with hyperplasia or not, and bilateral nodules group with nodules in both adrenal glands. Regarding the related factors of dominant side of aldosterone secretion, univariate analysis and binary logistic regression were used. Receiver operating characteristic curve and nomogram were used to evaluate the diagnostic performance of regression models. Results: A total of 237 patients with PA were included, of which, 118 males and 119 females, the median age was 39 years, and the body mass index (BMI) was (25.2±3.5) kg/m2. There were 157 (66.2%) of 237 patients with typical imaging findings. There were 32 cases in no-nodular group, 183 cases in unilateral nodule group, and 22 cases in bilateral nodules group. Multivariate analysis showed that age (OR=0.876, P<0.001), blood potassium concentration (OR=0.430, P=0.004), and typical imaging findings (OR=2.202, P=0.035) were associated with unilateral aldosterone secretion. As for unilateral nodule group, multivariate analysis showed that age (OR=0.900, P<0.001), plasma aldosteronism concentration (PAC) (OR=1.050, P=0.018), and typical imaging findings (OR=2.637, P=0.025) were associated with unilateral aldosterone secretion. The agreement rate between the dominant side of the adrenal CT and AVS was only 50.2%. Multivariate analysis showed that age (OR=0.954, P=0.001), BMI (OR=0.893, P=0.024) and PAC (OR=1.043, P=0.011) were independently associated with concordance between AVS and CT. The cut-off value of the ROC curve was 0.43; the model sensitivity was 56.3%; the specificity was 86.7% and the area under the ROC curve was 0.742. Conclusions: Age is an important predictor in the diagnosis of PA subtypes. It is recommended to refer to subgroup based on imaging results for clinical decision. For patients with no obvious lesions or bilateral lesions on CT, AVS should be performed as far as possible to confirm the subtypes in diagnosis of PA.