Molecular Imaging Versus Adrenal Vein Sampling for the Detection of Surgically Curable Primary Aldosteronism: A Prospective Within-Patient Trial

Background: Primary aldosteronism (PA) causes hypertension and is potentially surgically curable when it is caused by a unilateral aldosterone-producing adrenal adenoma (APA). Adrenal vein sampling (AVS) is required to guide surgery, but it is invasive, is technically difficult, and has limited availability. Objective: To determine whether the accuracy of post-dexamethasone [C-11]metomidate ([C-11]MTO) positron emission tomography-computed tomography, a diagnostic test for APAs, is superior or noninferior to the accuracy of AVS in predicting outcomes from unilateral adrenalectomy, and whether [C-11]MTO is interchangeable with its longer-half-life analogue, para-chloro-2-[F-18]fluoroethyletomidate ([F-18]CETO). Design: Prospective within-patient comparison of diagnostic interventions. (ClinicalTrials.gov: NCT02945904). Setting: Three referral centers. Participants: 174 patients with PA desiring surgery if a unilateral source of PA was diagnosed. Intervention: [C-11]MTO and AVS in 169 patients, plus [F-18]CETO in the final 31. Measurements: International consensus criteria for biochemical and clinical success at 6 and 24 months after surgery; kappa statistic and Bland-Altman analyses comparing predictions of surgical outcomes by [C-11]MTO and [F-18]CETO. Results: Eighty-nine of 169 (52.7%), 78 of 169 (46.2%), and 43 of 169 (25.4%) patients were predicted to have unilateral PA by [C-11]MTO, AVS, or both, respectively. One hundred of 169 (59.1%) were assigned to adrenalectomy by the multidisciplinary team; primary outcome data were available for 156 of 169. Predictions were most accurate for complete or partial biochemical success ([C-11]MTO, 71.3% [95% CI, 61.0% to 80.1%]; AVS, 62.8% [CI, 52.2% to 72.6%]) and least accurate for complete clinical success (home blood pressure <135/85 mm Hg off treatment). The 95% CIs around differences between accuracies crossed zero, excluding superiority for [C-11]MTO, but not the prespecified lower bound of -17%, allowing [C-11]MTO to be declared noninferior to AVS. [F-18]CETO and [C-11]MTO agreed in 29 of 31 patients (kappa = 0.85 [CI, 0.68 to 1.00]). Limitation: The accuracy of [C-11]MTO could be assessed only in the surgical group. Conclusion: Molecular imaging is an accurate, noninvasive alternative to AVS.