Purpose: Accurate and precise quantification of resting intestinal blood flow and flow changes induced by digestion or in response to gut hormones could be highly useful for both clinical and research purposes. For example, examining intestinal blood flow could help the diagnosis and monitoring of patients with conditions like occlusive chronic mesenteric ischemia (CMI) [1,2]. Attention has been brought to the underestimated prevalence of CMI [2] - typically diagnosed based on clinical symptoms and screening for stenosis - and the need for an improved diagnostic process [1]. Furthermore, current revascularization treatment has a high success rate for sustained symptom relief making the successful screening and diagnosis for CMI even more pertinent [3]. The application of two-dimensional phase-contrast magnetic resonance imaging (2D PC-MRI) to measure resting and postprandial mesenteric blood flow has been proposed for the diagnosis and monitoring of CMI [1]. However, literature is sparse on the performance and reproducibility of 2D PC-MRI measures of intestinal blood flow including day to day intrasubject variation, measure to measure variation and observer to observer post processing variation. These sources of variation are relevant to the ability of 2D PC-MRI to detect the reduced flow and to detect the reduced flow response of individual patients. The aim of this study is to examine the feasibility of accurately measuring the resting intestinal blood flow and postprandial blood flow changes using 2D PC-MR. This is achieved by evaluating the sources of variation in blood flow measured in relevant blood vessels and comparing measurement uncertainty with the magnitude of change induced by glucose consumption.

Methods: This prospective study included ten healthy men who were scanned on eight different days collecting at least three repeated 2D PC-MRI measures in an imaging plane intersecting the arteria mesenterica superior (AMS) and truncus coeliacus interceded with three repeated measures in a plane intersecting the arteria hepatica and the vena portae. On two of the days, nine additional repeated scans were acquired after the participants were given an oral glucose tolerance test (75 grams of glucose in 250 ml water) or 250 ml water, respectively.All MRI-scans were performed on a 3 Tesla Siemens Biograph mMR PET-MRI scanner using a surface body coil combined with coils integrated into the table. The sequence used was a 2D gradient echo sequence with retrospective cardiac triggering (1 slice, Field of view (FOV) = 240x340 mm2, voxel size = 0.71x0.71x5 mm3, repetition time (TR)= 11.0 ms, echo time (TE)= 3.17 ms, flip angle = 15o, 25 samples per cardiac cycle, GRAPPA acceleration factor = 2). For measurements of blood flows in AMS and truncus coeliacus a velocity encoding factor (VENC) of 200 cm/s was used and for flow in arteria hepatica and vena portae a VENC of 100 cm/s. Region of interests (ROI) delineating the vessels were manually drawn by two separate observers using in-house developed software ((https://github.com/MarkVestergaard/PCMCalculator/). The blood flow was then calculated as mean blood velocity times the cross-sectional areal of the delineated vessel.

Results: Mean blood flow values and reproducibility results in coefficients of variations (CoV) for variance between observers, day to day variance, and variance in repeated measures are presented in Figure 1. All vessels demonstrated low CoV (6%-9%) for within sessions repeated consecutive measures which was fair for between day measurements (13-15%) with the exception of large variance in arteria hepatica flow (30%). Consumption of glucose induced an average peak flow increase of 75% (+/-45) in the AMS and 60% (+/- 30%) flow increase in the vena portae, but no significant change in the arteria hepatica or truncus coeliacus (Figure 2). Individual flow measures for the prolonged postprandial and placebo skanning days are presented in Figure 3.

Conclusions: The low variability of repeated flow measures in the AMS and vena portae compared to the high change in flow after glucose consumption makes flow changes in these vessels feasible as a future diagnostic test to identify CMI patients among patients with a high clinical suspicion of CMI.

Acknowledgements

No acknowledgement found.