No need for fasting prior to doppler ultrasound of pediatric liver transplants: A self‐controlled study

Children frequently undergo routine Doppler‐ultrasound (DUS) after liver transplantation (LT) for which they are fasted, but this may cause hunger and discomfort.

In general, patients are fasted prior to an ultrasound to optimize visibility by keeping bowel gas to a minimum, and to avoid discomfort from compression on a full stomach. 5Although available literature does not specify whether children should be fasted for a routine LT DUS, to our knowledge children are fasted in most LT centers.
However, fasting may cause hunger and distress, and for children with LTs requiring frequent surveillance DUS this may add substantially to their disease burden.In children it is unknown how eating would affect a DUS.[8] Of note, adult cirrhotic livers showed less changes than healthy livers. 7Therefore, it is possible that a postprandial LT DUS produces different measurements than a pre-prandial DUS.
For assessment of significant PVAS the anastomotic PSV threshold was recently shown to be dependent on the time after LT, ranging from 171 cm/s at day 0 intraoperatively to 53 cm/s at 2 years. 9Depending on the degree of increased postprandial PV PSV these thresholds could be reached in non-significant PV anastomotic narrowing, resulting in a false positive PVAS diagnosis.Postprandial changes could also result in decreased hepatic artery (HA) PSV because of the hepatic buffer. 7,10r diagnosing PVAS, an alternative to the anastomotic PSV is the velocity ratio (VR), which is the ratio between pre-and anastomotic PV PSV. 11For this, a pre-to anastomosis VR of 1:4 corresponds to a 50% stenosis, which is considered significant. 11,12If the PV is considered as a pipe with a central narrowing (the anastomosis), according to basic physics, an increase of postprandial PSV would result in a linear increase of the anastomotic PSV. 12 Conversely, the VR should correct for pre-anastomotic increased PV PSV postprandially, and may therefore be more reliable than anastomotic PV PSV.However, this hypothesis depends on if and how PV diameters might change with increased mesenteric perfusion, for which there are conflicting reports in adult studies. 6,7,13 aimed to determine whether fasting is necessary for routine pediatric LT DUS by comparing pre-and postprandial DUS measurements.In addition, we tested the hypothesis that the PV VR is less affected by prandial changes than the PV anastomotic PSV.Last, we investigated the degree of burden of fasting as experienced by the children and their parents.

| ME THODS
This prospective single-center study was performed in the University Medical Center Groningen, which is the national pediatric LT center of the Netherlands.The study was approved by the local research ethics committee (registry number 201900869), and written informed consent was obtained from all participants and/or guardians.
All consecutive patients that underwent an LT between June 2020 and October 2022 were identified.Inclusion criteria were age at LT <18 years, and ability to fast.Primary or repeat LT was allowed, but each patient was only included once.Patients were excluded if ≥18 years old at the time of DUS.Children aged 0-4 years were asked to skip or delay one feeding session until after DUS, and children aged 4-17 years were fasted for 4 h.
To determine fasting related changes, we used a selfcontrolled study design.At 6 months after LT, patients underwent a routine pre-prandial DUS.After this routine DUS, included patients were asked to have their breakfast or lunch, followed by a second (postprandial) DUS 30-60 min after finishing this meal. 7tients and/or their parents were also asked to fill in a short questionnaire about the experienced degree of burden from having to fast.
Collected patient data included age at DUS, interval time between the end of the meal and the second DUS (min), LT donor type (heart beating, non-heart beating, living), graft type (full size, split liver), and type of PV anastomosis (end-end, preventive portoplasty, interposition graft).DUS were performed by either a pediatric radiologist or dedicated pediatric sonographer, and the first and second DUS in the same patient were always performed by the same person.Consequently, the sonographer was not blinded to the results from the first DUS.All DUS were performed on a Fujifilm ARIETTA 850 ultrasound machine (Fujifilm, Tokyo, Japan) using a 6 MHz convex transducer.
Both pre-and postprandial DUS consisted of routine analysis of the PV, HA, and hepatic vein(s).Aside from vascular patency, this includes extensive flow velocity measurements, with particular attention to the vascular anastomoses.Our DUS protocol and justification of the type of measurements are detailed in a previous publication by our group, 4 and consistent with published reviews on this topic. 3,11In addition to our standard DUS protocol, the PV diameters were measured (pre-, anastomotic, and post-anastomotic), and, if visible, flow velocities were measured in the origin of the superior mesenteric vein (SMV) and superior mesenteric artery (SMA).The sonographer was also asked to grade the degree of obscuration by bowel gas (scaled 1-5; completely obscured to completely visible) and difficulty of DUS due to distress (scaled 1-5; very easy to very hard).Other obtained variables such as parenchymal appearance, bile duct abnormality, focal liver lesions, or splenic size were assessed as part of standard practice, but not included in this study.

| Statistics
Continuous variables were summarized using mean and standard deviation (SD) when normally distributed, or using median and interquartile range (IQR) if skewed.Statistics were performed using SPSS for Windows (version 28, IBM).The level of significance was set at α < .05.
Normality of continuous variables was assessed using the Shapiro-Wilk test.The PV VR was obtained by dividing the anastomotic flow velocity by the pre-anastomotic flow velocity. 11,12The percentual postprandial change of anastomotic PV PSV and PV VR were compared using paired sample t-tests, to test our hypothesis that PV VR is less affected by prandial changes than the anastomotic PV PSV.
The chi-square test was used to compare categorical variables, the paired sample t-test for comparison of continuous variables, and the Mann-Whitney U-test for ordinal categorical variables.Intraclass correlation coefficients (ICC) were calculated for DUS measurement to test consistency and confidence intervals (CI) were determined.
ICC was classified as slight for 0.00-0.20;fair for 0.21-0.40;moderate for 0.41-0.60;good for 0.61-0.80;and excellent for 0.81-1.00. 14l children were assigned to group A (total study population).In order to specifically investigate whether the PV VR is less affected by prandial changes than the PV anastomotic PSV, a subgroup analysis (group B) was performed to further analyze only those children with actual postprandial increased anastomotic PV PSV (defined ≥1 cm/s postprandial increase of PV anastomotic PSV).

| Study population
Between June 2020 and October 2022, 47 children who underwent an LT were identified, and of these, 28 consented to participate (median age 3.5 years, IQR 1.6-10.8).Four patients were subsequently excluded because they had not fasted (n = 2) or withdrew consent for the second DUS because the child was in too much distress (n = 2).
Median age of the remaining 24 patients was 3.0 years old (IQR 1.2-10.8, Figure 1).The mean time between the end of the meal and the start of the second (postprandial) DUS was 45 min (SD 12 min).Donor type, graft type, and PV anastomosis type are detailed in Table 1.No children at 6 months after LT had previously been treated for PVAS.

| Questionnaire results
The majority of parents and/or children somewhat or strongly agreed that fasting was difficult (16/28, 57.1%), and several somewhat or strongly agreed that the child got upset when fasted (13/28, 46.4%).Almost half of parents somewhat or strongly disagreed that they themselves got upset when the child was fasted (13/28, 46.4%) (Table 2).

| Assessability of the postprandial DUS
In the majority of pre-prandial DUS, there was no or minimal bowel gas obscuration (completely visible or almost completely visible; 79.2%).During postprandial DUS this was 75.0%.There was no significant difference between pre-and postprandial obscuration by bowel gas (p = .92).In addition, on the postprandial DUS the SMV and SMA were visible in the large majority (91.7% and 87.5%, respectively), with no significant difference compared to pre-prandial DUS (p = 1.00 and p = .64,respectively) (Table 3).

| Pre-and postprandial DUS measurements
Among the total included study population (group A), 14/24 demonstrated an increase of postprandial anastomotic PV PSV compared to their pre-prandial DUS (group B).
In both group A and group B, there were no significant pre-to postprandial differences for pre-, anastomotic, or post-anastomotic PV PSV, nor for PV diameters or VR (Table 4).There were also no significant differences for the HA PSV, HA RI, HV venous pulsatility index (VPI), SMA PSV, SMA RI, SMV PSV, splenic artery RI, or splenic vein PSV (Table 4).ICC for DUS measurement consistency was overall moderate to excellent in group A, but with fair consistency for splenic vein PSV and post-anastomotic PV PSV (Table 4).In group B, ICC was overall fair to excellent, but with slight consistency for preanastomotic PV PSV and HA PSV.
For group B, there was also no significant (p = 0.16) difference between pre-to postprandial anastomotic PSV percentual increase F I G U R E 1 Study population flowchart.

| DISCUSS ION
This study assessed whether eating prior to routine DUS after pediatric LT would significantly change DUS measurements and assessability compared to fasting.We also tested the hypothesis that while anastomotic PV PSV would increase postprandially, the PV VR would remain stable, thus representing a more reliable method of assessing the PV anastomosis.
Comparison of pre-and postprandial DUS measurements in 24 children after LT found no significant differences.Importantly, obscuration by bowel gas and ease of DUS due to distress did not change postprandially either.Furthermore, our hypothesis that PV VR is more reliable than anastomotic PV PSV was disproven.We did demonstrate that the majority of children found fasting difficult and that they got upset because of the fasting.
Our findings indicate that there are no strong arguments to have children fasted prior to routine DUS of their LT.Omitting fasting will likely make it easier for children and their parents to travel to the hospital and undergo their DUS, thereby reducing their disease burden.
[8] One factor that may explain this lack of apparent increase of PSV is change in vessel diameters, which although not statistically significant, were larger in each PV segment postprandially.Because a small increase in diameter will result in an exponentially larger vessel area (according to the formula: area(mm 2 ) = π x r 2 ), this could accommodate the increased flow rate (mm 3 /s) and keep PSV reasonably stable. 12Of note, although we did attempt this, measuring small vessel diameters in the right plane on ultrasound has been reported to have substantial inter-and intra-observer variation. 5Therefore, to what extent vessel diameter impacts DUS measurements remains unclear.For the purpose of our study, the measured small diameter increases may have mitigated the increased postprandial mesenteric perfusion.The resulting non-significant DUS measurement changes justify omitting fasting in our pediatric population.
In contrast to our hypothesis, there was no significant difference between PV anastomotic PSV and PV VR before and after eating.
Based on our data, it remains uncertain why PV VR increased along with PV anastomotic PSV when there was no disproportionate change in vessel caliber to explain the relative increase of anastomotic PSV to pre-anastomotic PSV.This may represent the issue of suboptimal PV diameter measurements, or may entail suboptimal PV PSV measurements because of intravascular turbulence. 15Nevertheless, this is part of daily practice, and leads us to disprove our hypothesis.
Although postprandial DUS did not show significantly more bowel gas obscuration or distress for the liver and mesenteric vessel ultrasound, we suggest some caution to allowing children to eat whatever they want.In general (pediatric), radiology fasting for ultrasound is advised not only because of bowel gas but also because a full stomach may be uncomfortable during ultrasound. 16To avoid unnecessary distress or vomiting from abdominal compression after eating, and potential incidences of children eating a lot, in our practice we will now allow children to have a light meal (e.g., two biscuits and a non-fizzy drink) prior to their DUS.
The main limitation of this study was the non-consecutive inclusion of patients during the study period.Although the reasons to withhold consent to participate are unknown, it is likely that a part of the population had co-morbidity of their LT and therefore refused the additional burden of the postprandial DUS.This may have resulted in selection bias in which this study included relatively healthier children than our overall LT population.The resulting lack of PVAS warrants some caution to the generalizability of our results in the overall pediatric LT population.
In conclusion, in a cohort of pediatric LT patients without PVAS, measurements during routine DUS are not significantly affected by postprandial changes, and neither are assessability or obscuring bowel gas.Therefore, there are no strong arguments to fast children for these ultrasounds.Omitting fasting may decrease the disease burden.In addition, there is no preference for either VR or anastomotic PSV when assessing for PVAS in these children.

patients (N = 24), group A Only postprandially increased PV anastomotic PSV (N = 14), group B Pre-prandial, mean (SD) Postprandial, mean (SD)
a Mann-Whitney U-test, level of significance set at α < .05.TA B L E 3 Visibility, bowel gas, and difficulty of ultrasound.TA B L E 4 Pre-and postprandial DUS measurements.All included