Adrenal venous sampling in primary hyperaldosteronism: Help for surgical decision?

The localisation of aldosterone-producing adenomas (APA's) remains difficul t. Indeed, CT scan may not detect small APA's while CT detected tumours do not necessarily produce aldosterone. Objective: to evaluate the value of adrenal vein catheterization (AVC) for the diagnosis of APAs and also the rates of unsuccessful procedures and com plications. Patients: One-hundred-and-nine hypertensives with biological features of pr imary hyperaldosteronism were included. Plasma sodium, potassium, aldostero ne and renin levels were obtained after one night in lying position. Aldost erone and renin levels were also measured after a one-hour walking period a nd after a sodium expansion with saline. All patients underwent a CT scan a nd AVC. For adrenal samples, the ratio of aldosterone to cortisol concentra tions was calculated for both sides. Both ratios were divided by the opposi te one and the higher was retained as an index of laterization (IL). For la ck of a golden standard variable, we have taken as a working hypothesis tha t operated patients had an APA (n=38) and the non operated ones bilateral h ypereplasia (n=71). Results: Operated patients had a shorter history of hypertension, more freq uently a left ventricular hypertrophy on ECG, and lower serum creatinine an d potassium levels, along with a higher sodium level. None of these feature s appeared discriminant, however. Presence of a 10-mm, tumour on CT scan wa s not significantly different between the 2 groups. Operated patients had m ore than a five-fold higher IL compared with that of non-operated patients. Analysis using a ROC curve showed that the value of 12 for "IL" was an acc eptable operational criterium of lateralization, producing a specificity of 90%, while maintaining a sensitivity of 62%. AVC appeared most useful in c ase of middle-range pre-test probability of an APA being present, that is w hen when clinical, biological, or radiological features are not fully conco rdant. The rate of unsuccessful procedure and the rate of complications wer e low (1.8% and 3.6%. respectively). Conclusion: With a low iatrogenicity, AVC appears helpful in indicating sur gery mostly when clinical, biological, and CT scan features are not fully c oncordant. A value of 12 for IL appears to allow a high specificity while r etaining an acceptable sensitivity.