REPLY: Science needs differences in opinions to ensure that facts are presented in a truly reproducible way. That is why we welcome the comments by Banerjee et al. (1), and we have to admit that we actually agree with most of them.
Needless to say, the results of our study must be tested for replicability—as every set of novel data must be tested. Having said that, we would like to highlight that we have never claimed that this study was a registration trial. In fact, we clearly state that these are the first results of clinical imaging of congenital hyperinsulinism with 68Ga NODAGA-exendin-4, and we did raise this issue in the discussion and conclusion sections (2).
Pertaining to the comments on retrospectively including 11 patients outside the prospective study, we do agree that there is a possibility of extending the bias. However, as the nature of the study was proof of concept and considering the expertise of the Charité Berlin Centre in treating congenital hyperinsulinism patients from all around the world, we decided to include all available patient data. We acknowledge the concern of Banerjee et al. (1), as it becomes clear from our respective remarks concerning this limitation in the introduction and discussion sections. In fact, we have concluded that exendin PET has the potential to be an improvement over 18F-DOPA PET, but that more work is needed to validate these preliminary results (2).
To allay the concerns of bias spilling into the results of our study, we would like to stress that our group did realize that PET readings could be positively biased if the reporting was left only to the inventors of the technique. That is why we chose 2 methods of reading—that is, clinical reading at the respective site where imaging had been performed and an independent, masked reading. Thereafter, to overcome any discordant results, a joint reading was performed. Although we took the joint reading as reference, knowing that the gold standard for imaging findings can only be histopathology, specifically for new tracers and new indications, we did perform histopathologic confirmation in all specimens operated on (Supplemental Table 2 in (2)). Even the surgeons, trained on using 18F-DOPA PET–directed surgical planning, clearly stated more confidence in exendin PET in interpretation of the image results than in 18F-DOPA PET. These results are presented in the supplemental materials using the validated Likert scale (2).
We are grateful for the comments from Banerjee et al. supporting the conclusions we have drawn, as the points they raised are in line with the arguments we have put forward in our paper. Also, we are grateful that Banerjee et al. raise the issues for which the answers can be found in the supplemental material, thus putting emphasis on this part of the paper as well. In addition, we hope that our remarks with respect to a potential bias have helped to allay such concerns.
- © 2022 by the Society of Nuclear Medicine and Molecular Imaging.
- Revision received December 17, 2021.
- Accepted for publication January 4, 2022.