EFFECTS OF FASTING AND INTRALUMINAL CONTRAST ENHANCEMENT

ON ULTRASONOGRAPHIC APPEARANCE OF THE EQUINE SMALL

INTESTINE

TRACY NORMAN, KEITH CHAFFIN, DAVID SCHMITZ

The equine small intestine is challenging to evaluate ultrasonographically. In humans, hydrosonography has

been used to improve ultrasonographic images of the small intestine. We hypothesized that fasting horses for

24 h would enhance the ability to image the small intestine transabdominally by separating intestinal loops and

reducing intraluminal gas, and that the administration of intragastric contrast agent would further improve that

ability. Ten healthy horses were examined ultrasonographically under three treatment conditions: (a) regular

diet, (b) after a 24-h fast, and (c) fasted plus intragastric administration of water and mineral oil. During each

phase of the study, 30-s video clips were obtained from four predetermined abdominal windows, and were

examined to determine diagnostic quality. Fasting improved the ability to obtain high-quality images of the

small intestine significantly. The addition of contrast agent resulted in qualitative improvement in image quality,

but differences did not result in statistically significant improvement. r 2010 Veterinary Radiology &

Ultrasound, Vol. 51, No. 6, 2010, pp 642–646.

Key words: equine, image enhancement, small intestine, ultrasound.

Introduction

IN ADULT HORSES, gas and ingesta in the large colon and

the length, position, tortuous course, and intraluminal

gas of the small intestine render the bowel difficult to assess

thoroughly with ultrasound. This makes critical assessment

difficult or impossible, and intramural lesions may be

missed. This is a serious limitation, as increased wall thick-

ness is one of the most common sonographic abnormalities

of the small intestine,1 and visualization of a greater pro-

portion of the small intestine increases the probability of

identifying a focal or segmental lesion. Fasting horses for

8 h increases the length of jejunum visible ultrasonograph-

ically compared with fed horses.2,3

In humans, the challenges of examining the small intes-

tine sonographically are similar to those in the horse.

Hydrosonography of the small intestine, where a hypo-

echoic contrast agent is administered per os to provide

intraluminal contrast, is used occasionally to improve vi-

sualization of the small intestine by separating individual

intestinal loops and reducing intraluminal gas.4–7 A variety

of contrast agents have been used for hydrosonography,

including water, orange juice, cellulose-based suspensions,

and isotonic polyethylene glycol solutions.4–7

We hypothesized that fasting horses for 24h would im-

prove visualization of the small intestine, and that the ad-

ministration of intragastric fluid would further improve the

quality of small intestinal images. Our objectives were to

determine whether fasting horses for 24h improves the

overall diagnostic quality of sonographic images of the

small intestine compared with fed horses, and whether an

intragastric contrast agent in fasted horses would further

enhance diagnostic quality.

Materials and Methods

Ten normal adult Quarter Horses obtained from the

university teaching herd were studied. There were nine

geldings and one mare, and age ranged from 2 to 17 years.

Each horse was evaluated ultrasonographically three times.

The first was while the horse was on a routine feeding

schedule (fed horses). Fed horses were allowed free access

to coastal Bermuda grass hay and given 1kg of commercial

sweet feed� twice daily. The second examination was per-

formed after the horses were fasted for 24h (fasted horses).

Fasted horses had free access to water while feed and hay

were withheld. The third examination was performed on

fasted horses 15min after 3 l of water and 1 l of mineral oil

were administered by a nasogastric tube (contrast horses).

Each horse acted as its own control.

This paper has been submitted for presentation at the Annual Con-vention of the American Association of Equine Practitioners.Address correspondence and reprint requests to Dr. Tracy Norman, at

the above address. E-mail: tnorman@cvm.tamu.eduReceived April 15, 2010; accepted for publication May 6, 2010.doi: 10.1111/j.1740-8261.2010.01719.x

From the Department of Large Animal Clinical Sciences, Texas A&MCollege of Veterinary Medicine and Biomedical Sciences, College Station,TX 77802.

�Neutrena Vitalitys

Ultra, Cargill Incorporated, Minneapolis, MN.

642

In preparation for sonography, the abdomen was

clipped and cleaned with soap and water. An ultrasound

machinew with a variable frequency 2.5MHz curvilinear

tranducer was used. Adjustments to depth, power, overall

gain, and time/gain compensation were made as needed.

Images were acquired over a 30-min period. In horses re-

ceiving contrast agent, image acquisition began 15min fol-

lowing contrast agent administration and ceased 30min

later. Imaging began on the right lateral abdomen to image

the duodenum, and then proceeded to right ventral, left

lateral, and left ventral areas of the abdomen. Thirty-

second ultrasonographic video clips of the small intestine

were obtained from four discrete windows. The duodenum

was imaged on the right side of the abdomen deep to the

liver; if the duodenum was not visible at this location, it

was imaged adjacent to the cranial pole of the right kidney.

The jejunum was imaged in the right ventral abdomen, left

ventral abdomen, and left lateral abdomen deep to the

spleen. At each location, a 30-s video clip was recorded for

analysis by an investigator unaware of the imaging time

relative to fasting and contrast agent administration. All

video clips were obtained by the same investigator (T.N.).

The images were labeled with the horse’s identification and

a code to indicate the treatment phase of the study; this

code was not known to the investigator evaluating the im-

ages (K.C.). Images were shuffled before evaluation.

From each of the four abdominal windows, images were

evaluated by the investigator to determine the overall ap-

pearance and clarity of the small intestine (Table 1). Vis-

ibility was defined as whether a high-quality, diagnostic

image of the small intestine was present. Visibility was used

as a baseline or overall assessment of the images of the

small intestine. Definition was defined as whether the small

intestinal wall thickness from the lumen to the serosal

margin could be measured easily. The ability to measure

wall thickness was judged to be important because of the

number of small intestinal disorders that cause inflamma-

tion or infiltration of the bowel wall.1,3,8,9 Motility was

defined as the visible presence and relative frequency of

active contractions and dilations of the small intestine. The

visible movement of fluid or ingesta through the small in-

testine allows not only observation of serial dilations and

contractions, but potentially also of adjacent segments of

the intestine. Circumferential visibility was defined as the

ability to view the entire circumference of the small intes-

tine, including the mesenteric wall. The ability to view the

entire circumference of the wall was considered to be im-

portant because solitary mural lesions may be missed

sonographically unless the entire circumference of the wall

is visible. Dilation was defined as whether the small intes-

tine was seen to dilate to a diameter of � 3.5 cm during

the video clip. Wall thickness measurement varies with

dilation and contraction of the intestinal lumen, and ac-

curacy of such measurements is improved with full dila-

tion.8 A score was assigned by the investigator to each

parameter of each video clip separately, and the sum of the

four scores (one for each of the four abdominal quadrants)

was used to determine the total ultrasound score for that

parameter.

Data were analyzed using descriptive and inferential

methods. Summary statistics included the mean, standard

deviation, median, and range (minimum and maximum

values) for each variable measured. The data were exam-

ined graphically for violation of normal distribution by

plotting the observed values vs. standard normal quantiles;

no evidence of deviation from normality was noted. For

inferential data analysis, the study was designed to test two

hypotheses. The first null hypothesis was that sonographic

parameters did not differ significantly between fed horses

and fasted horses. The second null hypothesis was that

sonographic parameters did not differ significantly between

fasted horses and horses that were fasted and then received

contrast agent. Linear mixed-effects models were fit to the

data to test these hypotheses, accounting for repeated

measurements on individual horses. Ultrasound variables

represented the dependent variable for statistical analysis,

with treatment group (fasted, fed, or contrast) treated as a

fixed effect and individual horse as a random effect. A

significance level of Po0.05 was used for analysis, which

was performed using S-PLUS statistical software.z

Results

All horses tolerated ultrasonographic examination and

the administration of contrast agent without apparent ad-

verse effects.

The mean, standard deviation, median, and range for

the five small intestinal scores during the fed and fasted

phases of the study are given in Table 2. The total

ultrasound scores for visibility, definition, circumferential

visibility, and dilation were significantly higher in

horses during the fasted phase than during the fed phase

(Table 2). Scores for motility were slightly higher for horses

in the fasted phase than in horses in the fed phase, but that

difference was not significant (P¼ 0.0750, Table 2).

Table 1. Parameters and Scoring Method

Visibility Definition MotilityCircumferentialVisibility Dilation

0¼no 0¼no 0¼ none 0¼wall not visible 0¼no1¼ yes 1¼ yes 1¼ occasionally 1¼near wall only 1¼ yes

2¼ often 2¼ entire circumferencevisible

wMyLab 70 XVision Vet, Universal Ultrasound, Bedford Hills, NY. zS-PLUS Version 8.1, TIBCO Inc., Seattle, WA.

643EQUINE SMALL INTESTINALULTRASOUND ENHANCEMENTVol. 51, No. 6

The mean, standard deviation, median, and range for

the five small intestinal scores of horses during the fasted

and contrast agent phases of the study are given in Table 3.

When the total ultrasound scores of small intestinal images

from the fasting and contrast agent phases were compared,

total ultrasound scores for visibility, definition, motility,

and circumferential visibility tended to be higher for horses

when they received contrast agent, but these differences

were not significant (Table 3). Total scores for dilation

were similar for the two groups (Table 3).

Discussion

In healthy adult horses, withholding feed for 24h sig-

nificantly improved the total scores for four clinically im-

portant parameters relating to ultrasonographic image

quality of the small intestine.

That at least 8 h of fasting improved the sonographic

visibility of the small intestine in healthy horses has been

documented.2 The mechanisms by which fasting enhances

the quality of small intestinal sonographic images are not

clearly understood, but it is likely that withholding feed

reduces the amount of ingesta and gas in the large colon

and cecum, thus allowing unobstructed penetration of

sound waves to the small intestine in the ventral abdomen.

It is also possible that emptying of the small intestine re-

duces intraluminal gas by reducing fermentable substrate,

thus improving circumferential visibility. In contrast to the

previously reported study,2 we did not find significant re-

duction in small intestinal movement in fasted horses. In

fact, although the difference was not statistically significant,

motility scores were higher during the fasted phase vs. the

fed phase. The reason for this difference between our study

and those previously reported is unknown, but may be due

to differences in the duration of fasting or the methods

employed to assess images between the two studies.

When water and mineral oil were administered intragas-

trically as a contrast agent, total ultrasound scores for each

parameter were not significantly improved when compared

with total scores of horses that were fasted alone. Never-

theless, there were some positive aspects of contrast agent

administration. For visibility, definition, motility, and

circumferential visibility, scores for the contrast agent

phase were higher than those for the fasted phase, but

not statistically significant. It is unclear why the adminis-

tration of contrast agent did not significantly improve the

ultrasound parameters. The small number of horses likely

played a role. Assessing more horses may have resulted in

statistical significance for some of the parameters.

Another possibility was that the contrast agent was not

suitable to provide sufficient intraluminal contrast com-

pared with intestinal fluid, provided endogenously or via

recent drinking of water, by the fasting horses. In humans,

considerable effort has been directed at identifying the ideal

contrast agent for intestinal sonography.4,5 The ideal con-

trast agent should be hypoechoic, nonabsorbable, non-

foaming, isotonic, and have the ability to adsorb and

displace intraluminal gas.4–7 Although water and orange

juice have been used to provide intraluminal contrast

in people,7 inconsistent results have been obtained,

Table 2. Fasted vs. Fed: Total (Summed) Scores of Ultrasonographic and Corresponding P-Values for Ultrasonographic Video Clips ObtainedTransabdominally in Fed vs. Fasted Conditions in 10 Adult Horses

Fed Fasted

P-ValueMean SD Median Range Mean SD Median Range

Visibility 1.0 0.816 1.0 0–2 2.2 1.033 2.0 0–4 0.013�

Definition 1.0 0.816 1.0 0–2 2.2 1.033 2.0 0–4 0.013�

Motility 3.3 1.494 3.5 1–5 4.5 1.716 4.5 2–8 0.075Circumferential Visibility 4.6 1.174 5.0 3–7 5.9 1.370 6.0 3–8 0.033�

Dilation 0.9 0.737 1.0 0–2 2.0 1.054 2.0 0–4 0.0115�

�Statistically significant difference.SD, standard deviation of the mean.

Table 3. Fasted vs. Contrast: Total (Summed) Scores and Corresponding P-Values for Ultrasonographic Video Clips Obtained Transabdominallyin Fasted vs. Contrast Agent Conditions in 10 Adult Horses

Fasted Contrast Agent

P-ValueMean SD Median Range Mean SD Median Range

Visibility 2.2 1.033 2 0–4 2.7 0.949 3 1–4 0.2887Definition 2.2 1.033 2 0–4 2.5 0.850 2.5 1–4 0.4679Motility 4.5 1.716 4.5 2–8 5.6 1.174 6 4–8 0.1022Circumferential Visibility 5.9 1.370 6 3–8 6.3 1.059 6.5 5–8 0.4620Dilation 2.0 1.054 2 0–4 2.0 0.816 2 1–3 1.000

SD, standard deviation of the mean.

644 NORMAN, CHAFFIN, and SCHMITZ 2010

presumably because of rapid transit, absorption, and in-

consistent displacement of intraluminal gas.4–7 Polyethyl-

ene glycol, methylcellulose, simethicone, cellulose, and

simethicone-coated cellulose have also been used as hydro-

sonography contrast agents, with variable results.4,5,7 The

ideal contrast agent for the horse remains unknown. Most

agents used in humans have not been evaluated for safety

in the horse, and some have been associated with abdom-

inal pain and diarrhea in humans.6,7 We chose to use water

and mineral oil because these substances are anechoic and

are known to be well tolerated in horses. Water is sono-

graphically anechoic, but when mixed in the intestinal

lumen, traps microbubbles of gas, resulting in a mixed

echogenicity. Other limitations of using water alone are its

rapid transit through the gastrointestinal tract and its rapid

absorption. Mineral oil is also sonographically anechoic

and is not absorbed by the intestinal tract of horses, but

tends to trap microbubbles of gas even more so than water.

Mineral oil may reduce the small intestinal absorption of

water, thus enhancing the ability of water to provide con-

trast. However, the mixture of the two liquids in the lumen

of the intestine results in multiple echoic interfaces, with a

sonographic outcome of a mixed echogenicity. A uniform,

purely echo-poor contrast agent would be preferable for

imaging the equine small intestine. Safety trials in horses

are needed to determine which agents would be safe for this

purpose. Additionally, the volume of contrast agent ad-

ministered to horses in this study may have been insuffi-

cient for image enhancement. We used a total volume of 4 l

based on clinical experience. Perhaps, a larger volume

would have led to enhanced images, but this is not known.

Another explanation for the absence of significant find-

ings following contrast agent administration is the timing of

the ultrasound examinations. We chose to examine horses in

the contrast agent phase during the time period 15–45-min

postadministration of contrast agent in an attempt to allow

sufficient time for partial gastric emptying. It is likely that

gastric emptying and small intestinal transit time varies be-

tween horses and conditions, and thus we may not have

imaged every horse at the optimal time. It is unknown what

amount of delay, if any, would be the most appropriate

following the administration of intragastric contrast agent

to horses. In humans, immediate examination following

contrast agent administration is recommended.6,7

Additionally, it is possible that administration of con-

trast agent, regardless of the agent itself, volume, or timing

of ultrasonographic examination, simply does not enhance

the quality of ultrasound images of the small intestine of

horses over that achieved following fasting alone.

A limitation of this study was that the horses were ex-

amined using the same order of treatments. Horses were

scanned for the fed treatment phase as they entered the

study, and then had food withheld for 24h and reexamined

for the fasted phase. Immediately following this fasted ex-

amination, horses received the contrast agent and then

were rescanned 15min later. This order was chosen for

convenience, eliminating the need to fast the horses mul-

tiple times. However, some bias may have been introduced

using this same order in every horse. Similarly, during each

phase of the study, ultrasound examinations were per-

formed systematically in the same order; the right lateral

abdomen was examined first, followed by the right ventral,

left lateral, and left ventral abdomen. It is unknown

whether the uniform order of examination of the abdom-

inal quadrants introduced bias.

Although the investigator scoring the images was un-

aware of the phase of the study, the presence or absence of

contrast agent was likely obvious, thus allowing the study

phase of that image to be known. We suspect the impact of

this limitation is minimal even though intraluminal intes-

tinal fluid was clearly visible in the images because similar

contents are likely seen in individual horses, regardless of

whether they are fed, fasted, or received contrast agent.

Furthermore, when intraluminal fluid is identified sono-

graphically, its source cannot be determined.

The investigator obtaining the images was aware of the

phase of the study when the images were acquired. To

minimize any bias introduced in this way, we standardized

our method of performing the examinations and acquiring

video clips before starting the study. In addition, a strict

time limit (30min) was set for the acquisition of images to

yield consistent results. We acknowledge this limitation,

but feel that minimal bias was introduced.

Finally, the video clips may not have represented the

entire examination. Ultrasound scores were based strictly

upon images available on the 30-s video clips, even though

the actual examination included 7.5min per window. In

addition, the time limitations set for image acquisition do

not necessarily represent the conditions of a clinical exam-

ination. The investigator performing the examinations at-

tempted, as best as possible, to collect video clips that

effectively represented the findings of the entire examina-

tion. Thus, we believe this limitation also had minimal

influence.

Our results of this study demonstrate that fasting horses

for 24h significantly improves the ability to attain high-

quality diagnostic ultrasonographic images of the small

intestine. Fasting resulted in improved visibility, definition,

circumferential visibility, and dilation of the small intestine.

Administration of water and mineral oil as an intraluminal

contrast agent did not improve any of the ultrasound pa-

rameters significantly, although there were higher scores

for some parameters.

ACKNOWLEDGMENTS

The authors thank Noah Cohen for providing the statistical analysisfor this project, and Heather Quiram and Michael O’Connor for theirtechnical assistance.

645EQUINE SMALL INTESTINALULTRASOUND ENHANCEMENTVol. 51, No. 6

REFERENCES

1. Whitcomb MB. Equine abdominal ultrasound. Proceedings of the24th Annual ACVIM Forum, Louisville, KY, Richmond Hill, Ontario, CN:Content Management, May 31–June 3, 2006.

2. Mitchell CF, Malone ED, Sage AM, Nikisch K. Evaluation of gas-trointestinal activity patterns in healthy horses using B mode and Dopplerultrasonography. Can Vet J 2005;46:134–140.

3. Sage AM. New ultrasonographic approaches to chronic colic. Pro-ceedings of the 24th Annual ACVIM Forum, Louisville, KY, Richmond Hill,Ontario, CN: Content Management, May 31–June 3, 2006.

4. Lev-Toaff AS, Lnager JE, Rubin DL, et al. Safety and efficacy of anew oral contrast agent for sonography: a phase II trial. Am J Roentgenol1999;173:431–436.

5. Lund PJ, Fritz TA, Unger EC, Hunt RK, Fuller E. Cellulose as agastrointestinal contrast agent. Radiology 1992;185:783–788.

6. Maconi G, Radice E, Bareggi E, Porro GB. Hydrosonography of thegastrointestinal tract. Am J Roentgenol 2009;193:700–708.

7. Nylund K, Ødegaard S, Hausken T, et al. Sonography of the smallintestine. World J Gastroenterol 2009;15:1319–1330.

8. Fontaine GL, Hanson RR, Rodgerson DH, Steiger R. Ultrasound eva-luation of equine gastrointestinal disorders. Compendium 1999;21:253–262.

9. Whitcomb MB. Ultrasonographic evaluation of chronic equine ab-dominal pain. Proceedings of the North American Veterinary Conference,Orlando, FL, Gainesville, FL, Eastern States Veterinary Association, Jan-uary 13–17, 2007.

646 NORMAN, CHAFFIN, and SCHMITZ 2010