37 patients who were diagnosed with PA were enrolled in the study (Fig. 1), with female proportion of 32.4% (12/37) and mean age of 51.3 ± 10.9 years. The median duration of hypertension was 13.3 years, ranged from 0.1 to 53 years. The median of serum potassium was 2.92 mmol/L, ranged from 1.9 mmol/L to 4.03 mmol/L. Patients with bilateral adrenal mass accounted for 54.1% (20/37). The median of the maximum diameter of adrenal lesions was 1.50 cm, ranged from 0.8cm to 3.8cm. The median of the direct renin concentration (DRC) in standing position was 3.4 mU/L (0.5mU/L to 69.8mU/L, normal range 4.4–46.1 mU/L), and the median plasma aldosterone concentration (PAC) was 29.2 ng/dL (12.2ng/dL to 160.0ng/dL, normal range 3.0–35.3 ng/dL), with the aldosterone renin ratio (ARR) of 11.82 (normal range < 3.7). Captopril challenge test was performed in 35 patients, and the median of PAC after captopril challenge was 25.2 ng/dL, with the suppression ratio of PAC < 30% in all 35 patients. The saline solution infusion test, as a second confirmation test, was conducted in 12 patients, and median PAC after saline solution infusion was 16.2 ng/dL (> 10ng/dL was considered the confirmed diagnosis of PA). The main features of the population are shown in Table 1.

Sequential AVS was conducted in all patients, and the successful cannulation in the right adrenal vein was achieved in 26 patients (70.3%). Using LI in the 26 cases of successful bilateral catheterization and LAV/IVC in the 11 cases of the catheterization of right adrenal vein failed, 20 cases (54.1%) presented with left side hypersecretion, 8 of right side hypersecretion (21.6%), and 9 of bilateral hypersecretion (24.3%). Of all 37 cases experienced AVS, no adverse events including adrenal hemorrhage and contrast-induced nephropathy were reported.

In 37 patients enrolled in this study, 54.1% (20/37) of them were presented with bilateral adrenal lesions (Fig. 2, Supplementary Figs. 2 & 3), and 11(29.7%) of them had adrenal hyperplasia or adrenal nodules ≤ 1.0 cm. The lateralization was confirmed via LI or LAV/IVC using AVS. The diagnostic accuracy based on visual analysis using 68Ga-pentixafor was 86.5%, and the sensitivity and specificity of 68Ga-pentixafor PET/CT in distinguishing lateralization by visualization were 89.3% and 77.8%, respectively. The area under the ROC curve (AUC) for detecting positive lateralization based on the value of 68Ga-pentixafor SUVmax was 0.75 (95%CI 0.57–0.92, p = 0.026, Fig. 3A). The optimum SUVmax cut-off value was 6.86, and the sensitivity was 78.6%, specificity was 66.7%, and accuracy was 78.4% at this cut-off value (Table 2). Defining SUV ratio as SUVmax of the higher side/SUVmax of contralateral adrenal gland, the AUC for identifying lateralization based on SUV ratio was 0.71 (95%CI 0.50–0.92, p = 0.061, Fig. 3B). The optimum SUV ratio cut-off was 2.40, and the sensitivity was 60.7%, specificity was 88.9%, and accuracy was 67.6% at this cut-off value (Table 2).

The performance of 68Ga-Pentixafor PET/CT imaging in PA patients. A 60-year-old woman presented with hypertension and hypokalemia of 2.45mmol/L. CT scan showed a nodule of 1.9 cm and the second nodule of 0.9 cm in the left adrenal gland, and another adrenal lesion of 1.0 cm on the right side. AVS indicated left aldosterone hypersecretion, with the lateralization index of 13.1 (> 2). (B) CT scan showed a nodule of 1.9 cm. (A, C, D) Positive findings of 68Ga-Pentixafor PET/CT demonstrated strong uptake of 68Ga-Pentixafor on the left adrenal lesion, with SUVmax of 13.9 in the nodule of 1.9 cm, which was in accordance with the findings of AVS. Left adrenalectomy was performed, and the plasma potassium increased to 4.64 mmol/L without oral supplement. PA, primary aldosteronism. AVS, adrenal vein sampling. (CT scan and 68Ga-Pentixafor PET/CT of the 0.9 cm nodule in the left adrenal gland and the 1.0 cm nodule on the right side are shown in Supplementary Material as Supplementary Fig. 2 & Fig. 3)

The ROC curve for lateralization using AVS. (A) The ROC curve for detecting positive lateralization based on the value of 68Ga-pentixafor SUVmax in 37 patients via LI or LAV/IVC. (B) The ROC curve for identifying lateralization according to the value of 68Ga-pentixafor SUV ratio in 37 patients via LI or LAV/IVC. (C) The ROC curve for detecting positive lateralization based on the value of 68Ga-pentixafor SUVmax in 26 patients via LI. (D) The ROC curve for identifying lateralization according to the value of 68Ga-pentixafor SUV ratio in 26 patients via LI. LI, lateralization index, defined as aldosterone/cortisol ratio in the dominant side to the contralateral adrenal vein. LAV/IVC, aldosterone/cortisol ratio in the left adrenal vein to intra vena cava

In the 26 patients with successful catheterization in the right adrenal vein, the lateralization was confirmed via LI. The sensitivity, specificity, and accuracy of 68Ga-pentixafor PET/CT in distinguishing lateralization by visualization were 85.7%, 100.0% and 88.5%, respectively (Table 3). The AUC for detecting positive lateralization based on the value of 68Ga-pentixafor SUVmax was 0.84 (95%CI 0.69–0.99, p = 0.018, Fig. 3C). The optimum SUVmax cut-off value was 6.86, the sensitivity was 76.2%, specificity was 100.0%, and the accuracy was 80.8% at this cut-off value (Table 3). The AUC for identifying lateralization based on the value of 68Ga-pentixafor SUV ratio was 0.76 (95%CI 0.56–0.95, p = 0.074, Fig. 3D). The optimum SUV ratio cut-off was 2.25, and the sensitivity was 52.4%, specificity was 100.0%, and accuracy was 61.5% at this cut-off value (Table 3).

Using visualization of 68Ga-pentixafor PET/CT in distinguishing lateralization in 20 cases with bilateral adrenal lesions, the consistency of 68Ga-pentixafor PET/CT with AVS was 80.0% (16/20), which showed no significant difference compared with those with unilateral lesion (consistency, 94.1%, 16/17; p = 0.737). In the 11 cases with adrenal hyperplasia or adrenal lesion of diameter ≤ 1 cm, the consistency with AVS was 81.8% (9/11), and no significant difference was discovered compared with those with adrenal lesions over 1 cm (consistency, 88.5%, 23/26; p = 0.884).