Adrenal tracer uptake by (18)F-FDOPA PET/CT in patients with pheochromocytoma and controls

CONTEXT: (18)F-FDOPA PET/CT accurately localizes pheochromocytoma in patients with an established biochemical diagnosis. However, cut-off (18)F-FDOPA levels of standardized uptake values (SUV(max)) for both normal adrenal glands and pheochromocytoma are lacking. OBJECTIVE: Objectives of this study were to determine (1) reference maximum standardized uptake values (SUVmax) for normal adrenal (18)F-DOPA tracer uptake and (2) the optimal diagnostic approach for pheochromocytoma localization by using (18)F-DOPA SUVmax across a series of cut-off points: the affected adrenal gland (inter-individual analysis), the difference in SUVmax between the affected adrenal gland and the contralateral normal adrenal gland (intra-individual analysis), or a combination of these two. PATIENTS AND METHODS: All patients with histologically confirmed pheochromocytoma diagnosed at our center between November 2009 and December 2017 were retrospectively analysed. Only those patients who underwent an (18)F-FDOPA PET/CT-scan for localization purposes before adrenalectomy were included for further analysis. The control group consisted of patients who underwent (18)F-FDOPA PET/CT for other indications and who had no genetic susceptibility for developing a pheochromocytoma. SUV(max) of the volume of interest surrounding the adrenal glands was determined on EARL reconstructed images. Receiver operating characteristic (ROC) analysis was performed for adrenal gland SUV(max) and intra-individual difference in SUV(max) between affected and normal adrenal gland. In addition, binary logistic regression was performed for ROC analysis of the combined parameters. RESULTS: In total, 47 histologically confirmed pheochromocytomas were diagnosed in 45 patients, and 245 disease control patients were identified. In the control group, no statistical differences between the SUV(max) of left and right adrenal glands were observed, and uptake values in both adrenal glands correlated significantly with each other (r = 0.865, p < 0.001). Median (range) adrenal gland SUV(max) in pheochromocytomas and in the control group was 12 (2.6–50) and 2.9 (1.1–6.6), respectively (p < 0.001). ROC analysis revealed 93% sensitivity and 85% specificity at an SUV(max) cut-off value of 4.1 (area under the curve (AUC) = 0.951), and 93% sensitivity and 96% specificity at an intra-individual SUV(max) difference between the affected and normal adrenal gland of 1.0 (AUC = 0.992). The combination of both variables increased the AUC to 0.995. CONCLUSIONS: (18)F-FDOPA PET/CT distinguishes pheochromocytoma from normal adrenal glands with the highest diagnostic accuracy when combining the SUVmax of the affected adrenal gland with the difference in SUV(max) between affected and normal adrenal gland.