Endocrine Abstracts

It is recommended that all patients with PA who are suitable for surgery should undergo adrenal CT and AVS unless there is a large unilateral adenoma with a completely normal contralateral gland. We reviewed 100 patients diagnosed with PA. AVS was performed in 93. Different lateralisation criteria for AVS were assessed using ROC curve analysis to determine the optimal one for identifying an adenoma. We reviewed the outcomes of adrenalectomy evaluating which pre-operative characteristics predicted cure of hypertension.

Successful cannulation was defined as a 1.2-fold increase in cortisol from IVC to adrenal. Bilateral cannulation was successful in 67%. Using results of those with successful bilateral cannulation, a cortisol corrected aldosterone affected to unaffected (Aldo/Cort A:U) cut-off of 2.0 was the best predictor of adenoma identifying 77.2% of patients.

Our policy is to refer patients for surgery if they have clear lateralisation on AVS using our previously published criteria and concordant CT results. Using the best ROC curve criteria of Aldo/Cort A:U >2.0 together with concordant CT results, 91% of patients are correctly identified for surgery.

Forty-three of forty-six patients referred for surgery had confirmed unilateral disease, three had what appears now to be bilateral disease. Of the 43, 39 had long-term outcome data. Cure rate for blood pressure was 38.5% with improvement in another 59.0%. On univariate analysis, significant (P≤0.05) predictors of persistent post-operative hypertension were increased weight, raised creatinine, LVH and male sex. On multivariate analysis, male sex and higher pre-operative systolic BP remained predictive.

Our studies demonstrate that patients with PA should have both CT scanning and AVS. Aldo/Cort A:U >2.0 is the most accurate of previously published ratios in predicting unilateral disease but can only be used with successful bilateral cannulation. When patients were carefully selected for surgery, 97% had cure or improvement in blood pressure control.