68Ga-DOTATOC-PET/MRI and 11C-5-HTP-PET/MRI are superior to 68Ga-DOTATOC-PET/CT for neuroendocrine tumor imaging

This is a comprehensive prospective comparison of Ga-68 DOTATOC PET/CT, Ga-68 DOTATOC PET/MR, 11C-5-HTP PET/MR, ceCT and ce+DWI MRI in the evaluation of grade I & II NETs. Ga-68 DOTATOC PET/MR detected the most lesions regardless of site. MRI detected the most liver lesions while Ga-68 DOTATOC PET/MR, Ga-68 DOTATOC PET/CT, and 11C-5-HTP PET had a similar sensitivity for the detection of lymph node metastases.

The study is interesting, and the data presented in this manuscript may contribute to shaping imaging of well- to moderately-differentiated NETs in routine clinical practice. The limitations of the study include the modest study population and the lack of a true gold-standard against which findings from each imaging modality is benchmarked.

Specific comments
Methods: Imaging analysis – “The examinations were initially interpreted separately by two readers, one with basic radiological training and a senior consultant radiologist with 27 years' experience. In a second joint session, image analysis was performed by the two readers together in consensus.” There appear to be a lot of disparity in the experience of the two readers with a concern that the final consensus read will be biased in favour of the opinion of the more experienced reader. Could the authors provide information on the individual readers assessment with data on inter-reader agreement?
Results: Bone metastases – Radionuclide imaging techniques and MRI are known to have high sensitivity for the detection of marrow metastases than CT, while CT is more sensitive for cortical metastases. In view of this differential sensitivity of the imaging techniques for different types of bone metastases, it may be instructive to include information on the proportion of bone metastases that were limited to the marrow and those that involved the bone cortex.
Results: Peritoneal metastases – Similar to the comment above, MRI whether as stand-alone or as hybrid imaging with PET is expected to have better sensitivity for peritoneal metastases due to its exquisite soft tissue resolution. It may be necessary here as well to give the average sizes of the peritoneal lesion. Also, the lack of significant physiologic uptake of 11C-5-HTP compared with DOTA-peptide with known physiologic intra-abdominal uptake especially in the gut would suggest that the former will perform better than the latter in the detection of peritoneal lesions. It will be necessary for the authors to provide their thoughts on what could have been responsible for their findings i.e. better detection rate and reader’s confidence on CT than MRI. That lesions depicted on PET showed morphologic correlation on CT as advanced by the authors in lines 434 and 435 seem insufficient to explain these findings.
Results: Quantitative analysis – There are many factors that influence SUVmax, which makes comparison between studies somewhat challenging. For example, two different scanners with different imaging criteria and different sensitivity were used for imaging patients in this study. In addition, certain technical factors including tracer mis-injection could also affect the quantitative parameters. The patient in figure 1 has tracer mis-injection in the right cubital fossa for the 11C-5-HTP PET study which is significant enough to make comparison of SUVmax between the two studies unreliable. Please include this as another limitation of the study.
Table 1: Please consider adding the total number of lesions detected by each of the imaging modality for each patient in this table.
Tables: Include the definitions of the abbreviations used in the tables as footnotes to each table.