Identifying unilateral aldosterone-producing adenomas using published algorithms and imaging: a systematic review and meta-analysis

Objective Primary aldosteronism (PA) caused by a unilateral aldosterone-producing adrenal adenoma is surgically curable. However, radiologically visible adrenal adenomas are not necessarily functional. Adrenal vein sampling (AVS) is recommended to lateralise aldosterone excess, but is technically challenging and limited in availability. We sought to evaluate the diagnostic accuracy of published algorithms for lateralisation without AVS, and identify the top-performing algorithms with the highest specificity. Design Systematic review and meta-analysis. Methods Algorithms to predict unilateral PA and enable lateralisation without AVS, using AVS or surgical outcomes as the comparator, were systematically evaluated. Meta-analysis and meta-regression were conducted to obtain and compare aggregated estimates, respectively. Results There were 43 studies evaluating 30 unique algorithms grouped into 4 categories: (1) those combining biochemical, radiological and demographic characteristics; (2) algorithms involving confirmatory testing or adrenocorticotropic hormone stimulation; (3) anatomical imaging; and (4) functional imaging. The meta-analysis demonstrated the highest specificity (95%) in the first two categories. The algorithm with the highest specificity (98%) for unilateral PA consisted of an adrenal nodule on CT, plasma aldosterone concentration >20.0 ng/dL (554 pmol/L), hypokalaemia (≤3.5 mmol/L) and renin concentration ≤ 5 mIU/L. Anatomical imaging for subtyping had poor specificity (69%) while functional imaging had higher specificity (86%), noting variation in functional imaging modalities and criteria for lateralisation. Conclusions Algorithms may be used to identify unilateral PA without invasive testing in a modest proportion of patients. The place of other modalities, including functional imaging, remains to be formally established. Validation in different populations is required to enable the widespread use of these tools. Significance Statement Primary aldosteronism (PA) is gaining recognition as an important cause of hypertension and cardiovascular disease. The bottleneck created by adrenal vein sampling (AVS) to subtype PA leads to delays in diagnosis and treatment. Selecting patients likely to have unilateral aldosterone-producing adrenal adenoma who can bypass AVS will facilitate prompt surgical treatment and optimize the use of healthcare resources. Incorporating algorithms into the diagnostic evaluation of PA will enable streamlining of patient care and expedition of appropriate treatments that will minimise the sequelae of aldosterone excess.