The ¹⁸F- isotope was produced by proton irradiation of an ¹⁸O-H₂O target in the cyclotron and passed through a QMA column to capture the ¹⁸F-. The ¹⁸F- was then eluted with 0.5 mL of saline into the reaction tube, where a mixed solution containing 10 μL of aqueous AlCl₃ (1 μg/mL), 200 μL of pH=4 acetate buffer, and 0.6 mL of precursor solution (200 μg in 0.6 mL acetonitrile) was added sequentially. The reaction mixture was heated to 105 ℃ for 15 min, and upon cooling to room temperature, it was transferred to an HLB column for purification. The mixture was rinsed with 20 mL of water for injection, eluted with 2 mL of 50% ethanol, and the final solution was filtered through a 0.22 μm sterile filter membrane. The final product was a colorless, transparent liquid with an ethanol content not exceeding 10% and a pH ranging from 5 to 8. The synthetic chemical formula is provided in Figure 1 .

Patients diagnosed with PA or NFA, aged 18-80 years, at the First Affiliated Hospital of Wenzhou Medical University between May 25^th, 2023, and April 30^th, 2024 were recruited for the study ( Figure 2 ). All participants provided written informed consent, and the study protocol was approved by the Medical Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (Approval Number: KY2024-R245).

The diagnosis of PA was based on the clinical guidelines from the subcommittee of the Endocrine Society (1). All patients underwent PA screening using the plasma aldosterone–renin ratio (ARR) and received adrenal CT scans (slice thickness: 1.25 mm). Patients who tested positive in the screening (ARR > 30 pg·mL⁻¹/pg·mL⁻¹)) proceeded to confirmatory testing with a captopril challenge test (CCT) and seated saline infusion test (SSIT). All included patients underwent a 1-mg overnight dexamethasone suppression test (ODST) to exclude cortisol co-secretion in PA patients or mild autonomous cortisol secretion in NFA subjects. Patients who had unilateral or bilateral adrenal nodule(s) with clear borders identified on CT but presenting negative CCT and SSIT results or negative screening test results, along with normal serum potassium levels, were included as NFA. P The PA group was further subtyped into UPA and BPA. PA meeting any of the following criteria were classified as UPA: 1) unilateral lesion identified in AVS; 2) adrenal cortical adenoma confirmed by postoperative pathology, and complete or partial biochemical and clinical remission after surgery based on the Primary Aldosteronism Surgery Outcome criteria (19) (follow-up time: 1-6 months). PA meeting any of the following criteria were classified as BPA: 1) bilateral lesions identified in AVS; 2) adrenal cortical hyperplasia confirmed by postoperative pathology, and partial or no biochemical and clinical remission after surgery based on the Primary Aldosteronism Surgery Outcome criteria (follow-up time: 1-6 months) (20). Subtype diagnosis was determined based on postoperative pathological and follow-up results when two criteria were contradictory. All enrolled patients underwent 18F-AlF-NOTA-Pentixafor PET/CT imaging.

Patients maintained a normal diet without any special preparation. ¹⁸F-AlF-NOTA-Pentixafor was administered intravenously at a dose of 0.1 mCi/kg. Approximately 60 minutes after the ¹⁸F-AlF-NOTA-Pentixafor injection, a local PET/CT scan of the upper abdomen was performed. Twenty-one patients diagnosed with or suspected of PA, who underwent 18F-AlF-NOTA-Pentixafor PET/CT scan between July 27^th and Oct. 19^th in 2023 at the first affiliated hospital of Wenzhou Medical University got additional PET/CT scan at 20 minutes post-injection (p.i.). Studies were performed on two dedicated PET/CT scanners (Gemini TF 64, Philips Medical Systems, Netherlands, and uMI Panorama 35S, equipped with third-generation TOF 3D acquisition technology) with the following parameters: 120 kV, 80 mA, pitch of 0.829, tube rotation time of 0.5 seconds per rotation, and reconstruction thickness and interval of 5.0 mm.

Two nuclear medicine physicians certified by the committee, who were blinded to the patient’s clinical information, evaluated the PET/CT data. Any inconsistencies in their evaluation were resolved through negotiation, and the final decision was reached through consensus. Lesions were considered positive (also referred to as hot lesions) on PET/CT based on visual analysis if the adrenal nodule(s) showed higher uptake than the normal ipsilateral and contralateral adrenal glands. Adrenal lesions with similar uptake to the contralateral and adjacent normal adrenal tissue on visual assessment were classified as warm lesions. Adrenal lesions with lower uptake than the contralateral and adjacent normal adrenal tissue on visual assessment were classified as cold lesions. Warm and cold lesions were collectively referred to as negative lesions. The main lesion was identified as the adrenal nodule with the highest ¹⁸F-AlF-NOTA-Pentixafor uptake and clear borders in patients with multifocal nodules. Lesions identified on CT, or those with no abnormalities on CT but suspected of increased tracer uptake on PET, were designated as regions of interest, and the maximal standardized uptake value (SUV_max) was measured in these regions. The liver SUV_max was defined as the average SUV_max of five round spheres, each with a diameter of 2 cm, selected from the liver. Specific uptake value ratios, such as the lesion-to-liver ratio (LLR) and the lesion-to-normal adrenal ratio (LAR), were calculated. Additionally, in 18 patients who underwent AVS, SUV_max values within each adrenal gland were recorded. The side with the higher SUV_max in both adrenal glands was considered the dominant side. The lateralization index (LI) based on SUV_max was calculated as (SUV_max of dominant side)/(SUV_max of non-dominant side).

Patients diagnosed with PA underwent AVS within 3 months of completing ¹⁸F-pentixafor PET/CT to determine the lateralization of aldosterone hypersecretion. The decision on whether to perform AVS was made by endocrinologists with over 20 years of experience based on the clinical indicators, PET/CT results, and the patient’s inclination. AVS without adrenocorticotropic hormone stimulation was performed in the morning between 8:00 AM and 12:00 PM. Successful catheterization is considered if the adrenal cortical cortisol/peripheral venous cortisol ratio (ie, selectivity index) is 2 or greater.

A diagnosis of UPA was made if LI based on AVS was 4 or greater or 2 to 4 in combination with contralateral suppression or CT showing a typical adenoma on the dominant side, while those with LI based on AVS of less than 2 or 2 to 4 without meeting the previously described criteria were diagnosed as BPA (21, 22).

The pathological results of all surgical patients were independently assessed by a pathologist who was blinded to the clinical and imaging findings.

At least one follow-up was conducted by endocrinologists within 1 to 6 months after surgery (routine follow-up was scheduled at 1, 3, and 6 months post-surgery). For patients with PA, the Postoperative Aldosterone Outcome Score (PASO) system was used to evaluate their outcomes, categorizing patients into three groups: (1) Cure: blood pressure is normal without the use of antihypertensive medication; normal serum potassium and ARR, or ARR remains elevated but aldosterone levels can be suppressed in confirmatory testing; (2) Improvement: blood pressure has decreased to normal or has remained relatively unchanged with a reduction in antihypertensive medication use; normal serum potassium levels, elevated ARR, but plasma aldosterone levels have decreased by 50% compared to pre-surgery levels, or confirmatory test results have improved compared to baseline; (3) No improvement: antihypertensive medication dosage remains the same or has increased, and blood pressure has not improved or has worsened; low serum potassium and/or elevated ARR and/or aldosterone levels are not suppressed in confirmatory testing (19).

IBM SPSS Statistics 25.0 was used for statistical analysis. The Kolmogorov-Smirnov (K-S) test was employed to assess normality initially. Normally distributed data were reported as mean ± SD (χ ± S), and non-normally distributed data were reported as median (P₂₅, P₇₅). Nonparametric tests were used for between-group comparisons. Categorical data were presented as frequencies (percentages), and between-group comparisons were conducted using χ² or Fisher’s exact test when appropriate. Correlation analysis was performed using Pearson or Spearman correlation analysis. The diagnostic value of SUV_max, LLR, LAR, and LI of ¹⁸F-AlF-NOTA-pentixafor PET/CT was evaluated using receiver operating characteristic (ROC) curve analysis. Data visualization was conducted using GraphPad Prism 9.3 software and Origin 2022. This study employed a 2-tailed test, and a P value less than 0.05 was considered statistically significant for all tests.

¹⁸F-AlF-NOTA-pentixafor was synthesized with radiochemical yields of (37 ± 8.5) %. Radiochemical purities were > 99.5% as confirmed by radio-TLC.

The biodistribution of ¹⁸F-AlF-NOTA-pentixafor in 21 patients diagnosed with or suspected of PA at 20 minutes and 60 minutes post-injection (p.i.) is shown in Figure 3 . Among these patients, six patients were later diagnosed with UPA, and one patient was diagnosed with BPA. All the organs in the upper abdomen (including adrenal gland lesions) exhibited higher mean SUV_mean at 20 minutes p.i. than it at 60 minutes p.i., except for gallbladder. Specifically, the spleen exhibited the highest uptake at 20 minutes p.i., while the gallbladder exhibited the highest uptake at 60 minutes p.i. The adrenal gland lesion exhibited the highest SUV_mean in the upper abdomen. In 6 UPA patients, LAR and LLR at 60 minutes p.i. (2.57[95%CI, 0.71-4.43] and 5.00[95%CI, 1.96-8.04]) was higher than LAR and LLR at 20 minutes p.i. (2.28[95%CI, 0.80-3.77] and 4.28[95%CI, 1.64-6.92]) but with no statistical differences (P=0.766 and 0.657, respectively) ( Figure 4 ).

Overall, there were 45 men and 43 women in our cohort (mean age, 53.52 ± 10.28 years). Among the 88 included patients, 20 underwent AVS and 65 underwent adrenalectomy. Based on clinical evaluation, pathology, and follow-up data, 76 cases were diagnosed as UPA, 4 cases as BPA, and 8 cases as NFA (UPA lesions were classified as surgically eligible lesions, BPA and NFA lesions were classified as surgically ineligible lesions). No significant demographic differences (including age, gender, and BMI) were observed among the three groups. Compared with the UPA group, the BPA group exhibited statistically lower SUVmax, while the NFA group demonstrated statistically higher serum potassium and plasma renin concentration (PRC), and lower plasma aldosterone concentration (PAC), ARR, and SUVmax ( Table 1 , Figure 5 ). No significant statistical differences were observed in semi-quantitative parameters, including SUV_max, LAR, and LLR, between the BPA group and the NFA group (as shown in Table 1 ).

Besides, we conducted a correlation analysis among SUV_max and clinical characteristics. The results showed that SUV_max was negatively correlated with blood potassium concentration (r = − 0.444; P < 0.001) and age (r = − 0.334; P = 0.001), and positively correlated with PAC (r = 0.287; P = 0.007) and ARR (r = 0.265; P = 0.013) (partly shown in Figure 6 ). There was no correlation of SUV_max with sex, BMI, duration of hypertension, systolic blood pressure, diastolic blood pressure, or lesion diameter. In PA patients, a significant correlation was observed between SUV_max and lesion diameter (r = 0.351; P = 0.001).

In 88 patients, 95 lesions exhibited abnormalities in PET or CT scans (comprising 78 cases of surgically eligible lesions and 17 cases of surgically ineligible lesions). A total of 78 UPA and 4 BPA lesions were confirmed by AVS and/or postoperative pathology, while the remaining 13 NFA lesions were confirmed through clinical evaluation and/or postoperative follow-up evidence. Among the 95 lesions, 71/78 surgically eligible lesions demonstrated positive findings (sensitivity 89.74%), and 16/17 surgically ineligible lesions showed negative findings (specificity 94.12%) in visual analysis of ¹⁸F-AlF-NOTA-pentixafor PET/CT. The median SUV_max, LAR, and LLR of the surgically eligible lesions were 9.85(5.70,16.90), 2.85(1.66,4.52), and 5.90(3.59,9.75), respectively, significantly higher than those of the surgically ineligible lesions of 3.40(2.90,3.90), 0.92(0.83,1.11), 2.06(1.71,2.38).

Figure 7 showed LAR based on ¹⁸F-pentixafor SUV_max had a higher AUROC (0.961 [95%CI, 0.899-1.023]) than SUV_max of the lesion (0.931 [95% CI, 0.841-1.021]), visual analysis (0.919 [95%CI, 0.830-1.009]), and LLR based on SUV_max (0.901 [95% CI, 0.785-1.016]) to diagnose surgically eligible lesions. Thereinto, the AUROC of LAR was statistically higher than that of LLR (P=0.008). To achieve the maximized Youden index, the optimal cutoff LAR, SUV_max, and LLR was 1.43, 5.45, and 3.20, respectively, with a sensitivity of 83.33%, 79.49%, and 80.77%. No patients with surgically ineligible lesions would be misdiagnosed as surgically eligible lesions using these cutoffs with specificity of 100% in all (shown in Table 2 ).

To further evaluate the diagnostic value of ¹⁸F-AlF-NOTA-pentixafor PET/CT for lesions with different diameters, we divided 95 lesions into two groups: lesions with diameter < 1 cm and ≥ 1 cm. Among 95 lesions, 13 lesions examined with PET/CT were < 1 cm (10 nodules, 1 nodular thickening, 1 diffuse thickening, and 1 positive lesion without morphological abnormalities), while 82 were ≥ 1 cm in diameter (76 nodules and 6 nodular thickening). The diagnostic performance of ¹⁸F-AlF-NOTA-pentixafor PET/CT in identifying UPA was further analyzed in these 2 groups of patients and no significant demographical or biochemical differences were observed between the groups ( Supplementary Table 1 ). The small lesion group (< 1 cm in diameter) has higher optimal cutoffs of metabolic parameters (SUV_max=5.75, LAR=1.43, LLR=3.65) than the big lesion group (≥ 1 cm in diameter) (SUV_max=5.15, LAR=1.47, LLR=3.20). Compared to the small lesion group, the large lesion group demonstrated higher sensitivity, specificity, and accuracy either in visual analysis or semi-quantitative parameters ( Table 3 ). For lesions < 1 cm in diameter, LAR of 1.43 identified 9 UPA out of 11 lesions with an accuracy of 84.62% (sensitivity of 81.82%, specificity of 100%) ( Table 3 ). For lesions ≥ 1 cm in diameter, visual analysis identified 61 UPA out of 67 lesions, with an accuracy of 92.68% (sensitivity of 91.04%, specificity of 100%) for identifying UPA.

Since 7 out of 76 UPA patients showed negative findings in visual analysis, we decided to analyze the misdiagnosed reason further. All 7 missing UPA were warm lesions, with patients’ clinical and biochemical characteristics shown in Table 4 .

All 69 positive main lesions (hot lesions) out of 88 patients were proved UPA while all 4 cold main lesions were proved NFA in clinical. In 15 warm lesions, 7 lesions were proved UPA, 4 lesions were proved BPA, and 4 were proved NFA. There was a high correlation between the final diagnosis and the degree of the main lesion uptake (χ2 = 48.502; P < 0.001) ( Table 5 ).

Considering that the ultimate subtyping diagnosis needs to specify the hypersecretion side of the adrenal gland rather than simply distinguishing between unilateral or bilateral cases, we further conduct an analysis of the concordance rate of AVS and PET/CT visual diagnosis in 20 patients who underwent AVS ( Table 6 ). Of 11 UPA patients identified by PET/CT visual analysis, 10 were confirmed as having ipsilateral dominant secretion by AVS. The single patient with non-matching laterality between AVS and PET/CT showed multifocal warm nodules on the dominant secretion side of AVS, while the contralateral adrenal ramus medialis showed increased uptake but no anatomical abnormalities. Additionally, 3 of 9 patients with BPA identified by PET/CT were confirmed as having BPA by AVS. Hence, the total concordance rate of AVS and PET/CT visual diagnosis was 65.00% (13/20). The clinical characteristics of 7 patients with a disagreement between PET/CT visual diagnosis and AVS were shown in Supplementary Table 2 . The ROC analysis in the 20 patients showed that the area under the curve (AUC) for SUV_max and LI based on SUV_max reached 0.667 and 0.765, respectively, with the optimal cutoff values of SUV_max=5.15 and LI=1.12 ( Figure 8 ). However, employing SUV_max=5.15 and LI based on SUV_max=1.12 for subtyping diagnosis did not result in an improvement in the total concordance rates (65.00% (13/20); 55.00% (11/20)) when compared to visual analysis.