18F-FDOPA PET/CT Imaging for Congenital Hyperinsulinism: What is the optimal imaging time?

Introduction: 18F-FDOPA PET imaging has become the diagnostic imaging tool used for the localization of a focal lesion of congenital hyperinsulinism in patients unresponsive to medical therapy or have a genetic mutation associated with a focal lesion. This study aims to determine whether the current recommended imaging protocol with dynamic imaging over 50 minutes is necessary for the identification and localization of a focal lesion.

Methods: Consecutive PET images were reviewed from 9/8/21 to 12/22/22. Dynamic images were obtained starting 10 minutes post-injection for 50 minutes. 3D-MIP images were reviewed at 10-minute (min) increments, creating 5 maximum intensity projection (MIP) images per subject, for example, 10 min, 20 min, 30 min, 40 min, and 50 min. The images were reviewed simultaneously to determine the first image that is considered diagnostic. The likert score used was 1 – non-diagnostic, 2- probably diagnostic, and 3- diagnostic. Uptake in the adjacent kidneys was also evaluated.

Results: There were 31 cases with surgical pathology confirmation. In 25 of 31 (81 %) cases, the first image at 10 minutes post-injection was considered diagnostic. In 2 cases, the second data set at 20 minutes was considered diagnostic, and in 4 cases, the 3rd data set at 30 minutes was considered diagnostic. The 30-minute image was diagnostic in 30 of 31 (97%) cases and had a decrease in kidney activity in 16 of 31 (52%). The 50-minute image added additional information in 1 case, which looked focal at 30 minutes (Likert 2) and diffused at 50 minutes (Likert 3). The focal result was used in this analysis. The sensitivity and specificity of the read at 30 minutes was 95% sensitivity and 91% specificity with 94% accuracy.

Conclusions: This retrospective review shows that the diagnosis is evident visually in the first image in most cases, and confidence increased in the second and third images. There were 2 discrepancies; one was correct on the original clinical interpretation, and the other was not. The interesting discrepancy looked focal at 30 minutes with a Likert score of 2 and looked diffuse at 50 minutes. The clinical read was correctly read as diffuse. Further evaluation with fused SPECT/CT may have been helpful in this case. The findings suggest that images acquired at 10-20 minutes are diagnostic in most of the cases. However, images at 30 min were diagnostic in 97% of cases and had equal or improved images compared to earlier images and just over 50% with a decrease in renal activity. The decrease in renal activity is a favorable finding as it can obscure a lesion in the tail. One protocol option would suggest reviewing the first data set, and if confident of a focal lesion, the study can be stopped. Otherwise, imaging can be restarted and reviewed at 30 minutes. An alternative imaging strategy would be to start imaging at 30 minutes and stop if considered diagnostic for a focal lesion. If not, the imaging can continue through 50 minutes for a total of 30 minutes of imaging. Both options would allow for a decrease in the use of anesthesia and consider sedation and a decrease in the scan time from the current protocol. The processing would need to allow for a review of the MIP while the patient is on the table. The limitations of this pilot study are the evaluation by one radiologist and the small sample size. Further study, discussion, and collaboration are welcome.