Impaired gastric emptying in children with repaired esophageal atresia: A controlled study

Impaired By Madeleine

Gastric Montgomery,

Emptying Atresia:

in Children With Repaired A Controlled Study

Ruby Escobar-Billing,

Per M. Hellstrem,

Stockholm,

6ackground; Scintigraphy is considered the “gold standard” for investigating gastric emptying. The lack of standards regarding registration technique and meal composition has been a problem especially in pediatric patients. Methods: In this study, gastric emptying of a solid meal was assessed by scintigraphy in 10 patients with repaired esophageal atresia (5 to 10 years old), and the results were compared with those in 11 healthy control children (5 to 11 years old). The meal consisted of pancakes with a fixed energy composition labeled with Tc-99m. Fractional meal retention values were plotted as a function of time. /?esu/&; Half-emptying time and lag phase values were longer in the patient group, whereas the emptying rate was slower and the retention values at 60 and 90 minutes were higher than in the control group. Extremely long lag phase and slow emptying rates were seen in two patients with

T

HERE IS A HIGH incidence of gastroesophageal reflux (GER) and motility disorders of the esophagus in patients who have undergone surgical repair of esophageal atresia (EA).lW3Esophageal dysmotility leads to an inability to clear the esophagus of acid gastric contents and can be associated with GER and slow gastric emptying4-6 consequently leading to aspiration, anastomotic stricture, and esophagitis. Jolley et al1 showed a high incidence of slow gastric emptying of liquids in patients with repaired EA and GER. However, the emptying of solid foods was not studied in this group of patients. Because the emptying of solids is considered a more sensitive indicator of abnormal gastric emptying, we have performed a series of gastric emptying tests in children with repaired EA. and compared the results with those in age-matched healthy volunteers, using a scintigraphic gastric emptying test for solids. Radionuclide scintigraphy, first introduced by Griffith From the Departments of Pediatric Surgery Pediatrzc Ra&ology, Hospital Physics, and Department of Medccme, St G&an’s/Karolinska Hospital, Karolinska Institute, Stockholm, Sweden. This study was supported by the Swedish Medical Council (7916), the Crown Princess Lovisa, the Ruth & Richard hthlin and “Mjtilkdroppen”foundations. Address reprmt requests to Madeleine Montgomev, MD. Department of Pediatric Surgev, St G&an WKarolinska Hospztal, S-11281 Stockholm, Sweden. Copyright o 1998 by WB. Saunders Company 0022.3468/98/3303-0013$03.00/0 476

K. Anders

Esophageal

Karlsson,

and Bj6rn

Frenckner

Sweden

reflux symptoms and abdominal complaints. Gastric emptying in healthy children has not previously been studied by scintigraphy. The results of this study show that values for gastric emptying of solids in healthy children correspond well to those reported in healthy adults. Conclusion:

Scintigraphy

gastric emptying

is an easy

and

studies in children.

reliable

method

The radioactive

for

dose

can be kept very low, which makes it a safe method even for pediatric patients. Delayed gastric emptying can occur in patients who have repaired esophageal atresia, and may be associated with reflux symptoms and abdominal complaints. J Pediatr Surg 33:476-480. Copyright 0 1998 by W.B. Saunders Company. INDEX subject,

WORDS: Esophageal scintigraphy.

atresia,

gastric

emptying,

healthy

et al7 being noninvasive, quantitative, and physiologic, has emerged as the method of choice for evaluating gastric emptying. Scintigraphy can be used for solids as well as liquids. The test meal must be tasty, easy to prepare, and it should have a high affinity for the radioactive marker. However, the lack of standards regarding registration technique and meal composition has been a problem, especially in studies in children, because no reference data for healthy children are available. A recent study has shown that Tc-99m-MAA (macroalbumin aggregate)-labeled pancakes provide good marker stability and homogeneous distribution in the meal.* Such a meal is suitable for pediatric patients, because most children like pancakes. The aim of this study was to evaluate gastric emptying of solids in children with repaired EA and to compare the results with those of gastric emptying in age- and sex-matched healthy controls, thereby obtaining normal values for pediatric subjects. MATERIALS

AND

METHODS

Eleven patients who had undergone repair of esophageal atresia and tracheoesophageal fistula in the newborn period in Stockholm (St. G&an’s/Karolinska Hospital) were asked to participate in the study. All of the patients had undergone a primary esophageal anastomosls. In no case was the anastomosis performed under excessive tension, and none of the patients had undergone myotomy of the esophageal segments. Patients with mental retardation or other severe concomitant malformations were excluded. Ten patients (age 5 to 10 years; median, 7.5 years) Journal

of Pecjiatric

Surgery,

Vol33,

No 3 (March),

1998: pp 476-480

GASTRIC

EMPTYING

AFTER

ESOPHAGEAL

ATRESIA

agreed to enter the study. The control group consisted of 11 healthy children (age 5 to 11 years; median, 7 years), who had no symptoms of reflux, dyspha~la. or abdominal complaints. The patients were asked to answer a questionnaire regarding dysphagia, reflux, abdominal, or respiratoq symptoms. Three patlen& who complained of GER underwent esophageal pH momtormg to determine the degree of reflux. The study was approved by the Ethics CommIttee at the hos@al. Written consent was obtained from the parents of all children.

marker remamed in the stomach. Retention values were taken from the measured data points at 60 and 90 min after the start of data acqmsition. A “corrected’. half-emptying time value was calculated as the actual half-emptying time value derived from the Iinear regressIon analysis (T50) minus the time-pcnnt of 90% retention (lag phase value). The results were expressed as median? and range. Nonparametric statistics (Mann-Whitney U test) was used for the statistical analysis. p < .05 was considered significant.

RESULTS

Symptoms Scintigraphy, Preparation of the Meal Each subject had a light breakfast and then fasted for 4 hours before the examination at lunch time (12:OO). We used Swechsh pancake for the test meal, as described by Hermansson and Sivertson.7 A milk substitute (Prosobee. Mead Johnson, Nijmegen, The Netherlands) was used instead of milk, which thereby enabled children with milk allergy to tolerate the test meal. The pancake batter was labeled before frying with 0.5 MBq (children 5 5 years old) or 1 MBq (children > 5 yeas old) Tc-99m-macroalbumin aggregates (Tc-99m-MAA. Amerscan pulmonate II Amersham, UK). The size of the meal was 3 g pancake per bodyweight served with Jam, 0.5 g/kg bodyweight and water, 4 mL/kg bodyweight. The energy content of the admmistered meal was 8 kcaL’kg bodyweight with a fixed energy composition of 59% carbohydrate, 17% protein, and 24% fat. The pancake was cut into standardized pieces (1.5 X 1.5 cm) before it was given, and the meal was eaten v&in 10 min. A separate in vitro test was carried out to determine the stab&y of the Tc-99m-labeling of the solid phase of the test meal (pancake). The pancake was cut into small pieces (4 X 4 mm), weighed, and mixed with an equal volume of gastric Juice. Seven &fferent specimens were prepared and shaken m a 37’C water bath for 2 hours. The liquid phase was then decanted and filtered through a gauze pad, permitting only particles smaller than 1 mm to pass through. Both solids and filtrates were counted in a well counter (Compucount. Beckman, Fullerton, CA).

Registration of activity started when the first lmagc was taken, within 10 minutes after intake of the meal. Static Images (128 X 128 word matrix) of 2 mim~te duration were obtained in the anterior and posterior supine projections with a gamma camera (General Electric 400 AT, Denmxk) fitted with a low-energy, all-puvose collimator. In the period between the images, the child was seated or was allowed to move freely about in the room. Images were recorded at: 0.10.20.30.45,60,75.90, and 120 minutes after the meal. The geometric mean value was calculated for each tlrne interval and corrected for radioactive decay. A computer program (FORTRAN-77, gamma 11) was developed to estabhsh automatically a region of interest (ROI) sound the stomach, at a fixed isocount level relative to the maximum count inside each ROI, ROI included areas in whch the pixel count exceeded 15% of the maximal count. For the noise reduction of the image data, a 9.pcnnt smootlung filter was used. Fractional meal retention values were plotted as a function of time. The emptying rate (% of initial value per hour) and half-emptying time values (T50) were calculated by hnear regression analysis of the emptying phase at the linea part of the retention curve. The hnear part was determined subjectively and in most cases included values recorded between 30 and 90 minutes. In a few cases with extremely long lag phases, the calculations were based on values recorded between 60 and 120 minutes. Lag phase was defined as the time-point when 90% of the

The symptoms of the 10 patients with repaired esophageal atresia are shown in Table 1. Dysphagia and symptoms of gastroesophageal reflux, such as regurgitation and heartburn, were common and occurred in five and six patients, respectively. Four patients complained of asthma as well as of reflux symptoms like regurgitation. Abdominal pain or abdominal distension after the meal occurred in four patients. Three of the patients with reflux symptoms underwent pH monitoring; two of these showed reflux.

Gastric Emptying The pattern of the gastric emptying curve is seen as an initial delay (lag phase) followed by a phase of linear emptying (Fig 1). The number of counts within one ROI ranged between 7.000 and 10,000 in the first image (0 minutes). Figure 2 shows the A/P projections of the stomach in one of the healthy conuols. The results of the gastric emptying studies are presented in Table 2. T50 values, lag phase, and corrected T.50 values were significantly longer in the patient group than in the control group. The patient group also showed significantly higher retention values at 60 and 90 minutes and slower emptying rates than the control group. Retention values at 120 minutes were very low or close to zero in many of the healthy controls. This may have been because of the low amount of isotope used. The 120-minute values could not be further used in the analysis because of the difficulties of estimating correct values. Table

1. Symptoms

in 10 Patients

,%w Patlent

lvr)

sex

Dvsphaga

With

Repaired

F&flux Svmptoms

Esophageal

Atresia

Abdom!nal Svmptoms

Asthma

1

8

Male

Mild

Yes

2

7

Male

Mild

Yes

Yes No

Abdominal No

3 4

5 10

Female Male

No Mild

No Yes

No Yes

No Abdominal

Yes

Abdominal distension Abdominal

5*

5

6+

Male

Mild

Yes

10

Female

Mild

Yes

No

7

8

Female

No

Yes

Yes

No

8 9.x

6 6

Male Male

No No

No No

Female

No

No

No No No

No No No

10 *Patients

10

5< 6, and 9 showed

delayed

gastric

distension

emptying.

pain

pain

478

MONTGOMERY

Table 2. Results

- Lag phase

Healthy

i

Children

of Scintigraphic

Gastric

and 10 Patients

With

, Vanable

Q,5

78 (62-l IO) 21 (9-47)

100 (78-173) 34 (20-68)

53 (44-76)

62 (48-105)

Emptying Retention

rate (%/hour) 60 min (%}

46 G-55) 63 (47-82)

40 (23-50) 75 (63-97)

Retention

90 min (%t

40 (26-66)

56 (42-83)

NOTE.

0,2

Control

years. Patient are expressed

O,l

in 11 Atresia

Patient Group h = IO)

Lag phase (min) Corr. T50 (T50-lag)

3 % 04' rc 0,3

of Solids

Esophageal

Control Group (n = II)

T50 (min)

2

Emptying

Repaired

ET AL

group

patients

were

age 5 to 11 years;

median,

7

group was age 5 to 10 years; median, 7.5 years. Values as medians and ranges and are statistically significant.

0 0

10

20

30

40

50

Time

60

70

80

90

100

110

120

Stability of Radioactive Labeling

[mini

Fig 1. Fractional meal retention curve from one of the control subjects. Linear regression analysis was performed between 30 and 90 minutes (r = 0.99371. The curve through the measured values were obtained by a modified power exponential formula.

Figure 3 shows the retention curve from one of the patients with reflux symptoms (regurgitation and heartburn) and abdominal complaints compared with the means (22 SD) of the controls. The curve for the means of the controls was calculated by a modified power exponential formula.9 Three of the patients had values of retention at 60 and 90 minutes outside 2 SD of the controls. Individual emptying rates, half-emptying time values and retention values at 90 minutes are shown in Fig 4. Individual values of T50 are also shown in Fig 5.

Counting of the solids and filtrates from the in vitro experiments indicated firm binding of the radioactive marker to the solid phase of the test meal. with 97% (96 to 99) remaining after 2 hours in gastric juice (pH 1.4). DISCUSSION

Radionuclide scintigraphy is an easy and reliable noninvasive method for gastric emptying studies. The Tc-99m-MAA-labeled pancake meal used in this study is easily prepared and suitable for gastric emptying studies in children. All children in our study found the meal tasty and cooperated well during the examination. Gastric emptying in healthy children has not been studied by scintigraphy before. For ethical reasons, the activity of the isotope was kept very low (0.5 to 1.0 MBq). The effective dose and the mean dose absorbed by the intestinal mucosa was estimated at 0.06 to 0.13 mSv and 0.33 to 0.66 mGy, respectively.lO To achieve sufficient number of counts, the acquisition time for each recording was set at 2 minutes. Although an error is introduced in the geometric mean value at each time interval, because of a delay between the recordings of the 12,,,,,,,,

,

06- 1 ‘.., “...,.. 1 I '..,

0706

....

,

,

-’-h ...... .

'.....,,

-

"Q.....

0.5 -

".. ........ " ...

04-

i A; I

0.3 -

oL2Al 0 ,O

,

20

30

40

50

60

1 70

60

1 90

-...,.

".....,,

'. .... I

, + 1 100 110 120130

Tune [min]

Fig 2. A/P projections of the stomach at the beginning acquisition and at 30, 60, and 90 minutes. Note increasing the intestine.

of the data activity in

Fig 3. A graphic presentation of the process of gastric emptying in one of the patients with slow gastric emptying and symptoms of GER and abdominal complaints, compared with the means of the control group (error bars correspond to 2 SD!.

GASTRIC

EMPTYING

AFTER

ESOPHAGEAL

ATRESIA

479

anterior and posterior images, this error was believed to be of no significance in the calculations. The computer program used automatically defines the stomach at a fixed isocount level, which provides repeatable results at each calculation of the retention data. This method is also less time consuming and less influenced by intraobserver error than a manual outlining technique. The process of gastric emptying is affected by a number of factors like gastrointestinal electrical activity, hormones, and the composition of the meal. For example. J

I 0,.

50

100

,,

xi&

mjn!

Tim Fig 5 Kaplan-Meier diagram, which shows the individual values in the control group compared with the group of patients.

‘1:

Conkok

IUents

160

Mients

90 -~

8070 60z 60-. 8 40=

3020

-

. . . 8 :

.

l

0.

ControIs

Patients

Fig 4. individual emptying rates, half-emptying and retention values at 90 minutes in the control groups.

time values (T50) and in the patient

T50

a test meal with a high energy content delays gastric emptying compared with a low-energy mea1.l’ To reduce this source of error, the volume/energy content of the meal was adjusted to the body weight. The pancake was cut into standardized pieces, because the food particle size also affects the rate of gastric emptying.1z.13 Solid food particles are first processed into small particles in the stomach during the lag phase, before they pass out of the stomach. This means that the way one eats may affect the duration of the lag phase, because la.rger food particles remain in the stomach longer and prolong the lag phase.1zJ3 Because children operated on for EA usually chew their food thoroughly to prevent food getting stuck in the esophagus, this could not explain the long lag phase seen in some of the patients. The values for gastric emptying in the control children in our study correspond well to those in healthy adults.13 A recent study published by Maes et all4 presents values for gastric emptying in healthy children obtained by the C-13 octanoic acid breath test. These values accord well with those for gastric emptying of solids in healthy adults. They also resemble the values of the healthy control children in this study. The group of children with repaired EA showed slower gastric emptying than did the healthy controls. All parameters (T50, lag phase, and corrected T50) were significantly longer in the patient group. The retention values at 60 and 90 minutes were significantly higher and the emptying rates (percent per hour) were lower. Three of the patients were considered to have delayed gastric emptying, based on their results of retention at 60 and 90 minutes, that were outside 2 SD of the controls (Fig 3). However, as shown in Fig 4, there was a considerable overlap between the individual data. The lag phase was extremely long in a few cases reaching values of 60 to 68 minutes. Two of the patients with reflux symptoms and abdominal complaints showed long lag phases and slow

480

MONTGOMERY

gastric emptying. However, there was no difference in gastric emptying rate between patients with symptoms (reflux or abdominal) and patients without symptoms. Delayed gastric emptying may be an important factor contributing to the reflux symptoms seen in some patients with repaired EA. The etiology of the impaired gastric emptying is not known; for instance it may be caused by operative trauma or postoperative complications (vagal injury, anastomotic leak, or infection). Another explanation is that it is present at birth and has a congenital

ET AL

origin. Some data indicate that dysmotility in the esophagus exists preoperatively in patients with EA, and it is therefore considered a congenital malfunction. ls Gastric emptying studies can be easily and reliably carried out in children by the use of scintigraphy and a Tc-99m-labeled pancake meal. The radioactive dose can be kept very low, which makes it a safe method for children. Delayed gastric emptying of solids can occur in patients with repaired EA, and may be associated with symptoms of reflux and abdominal complaints.

REFERENCES 1. Jolley Xi, Johnson DG. Roberts CC, et al: Patterns of gastroesophageal reflux in children following repair of esophageal atresia and distal tracheoesophageal fistula. J Pediatr Surg 15:857-862, 1980 2. Duranceau A, Fisher SR, Flye MW, et al: Motor function of the esophagus after repair of esophageal atresia and tracheoesophageal fistula. Surg 82:116-123. 1977 3. Tovar JA. Diez Pardo JA, Murcia J. et al: Ambulatory 24.hour manometric and pH metrrc evidence of permanent impairment of clearance capacity in patients with esophageal atresia. .l Pediatr Surg 30:1224-1231, 1995 4. Papaila JG. Wilmot D, Grosfeld JL, et al: Increased mcidence of delayed gastric emptying in children with gastroesophageal remix. Arch Surg 124:933-936, 1989 5. Rosen PR, Treves S: The relationship of gastroesophageal reflux and gastric emptying m infants and children. J Nucl Med 25:571-574, 1984 6. Fonkalsrud EW, Berquist W. Vargas J, et al: Surgical treatment of the gastroesophageal syndrome m infants and children. Am J Surg 154:11-17, 1987

7. Griffith GH, Owen GM, Kirkman S. et al: Measurement of rate of gastric emptying using chromium-51. Lancet 4:1244-1245, 1966 8. Hermansson G, Sivertson R: Tc-labelled pancake for studies of gastric emptying of solids. Nucl Med Comm 12:973-981, 1991 9. Siegel JA, Urbain JL, Adler LP. et al: Biphasic nature of gastric emptying Gut 29:85-89, 1988 10. ICRP Publication 53: Radiation dose to patients from radiopharmaceuncals. Oxford, UK, Pergamon, 1987 11. Velchik MG, Reynolds JC. Abass A: The effect of meal energy content on gastric emptying. J Nucl Med 30:1106-1110. 1989 12. Holt S, Reid J, Taylor TV et al: Gastric emptying of solids in man. Gut 23:292-296, 1982 13. Smout AJPM, Akkermans LMA: Motility of the gastrointestinal tract. Wrightson Biomedical Publishing LTD 1992, pp 96-101 14. Maes BD, Ghoos YF, Geypens BJ, et al: Relation between gastric emptying rate and energy intake in children compared with adults. Gut 36:183-188, 1995 15. Romeo G, Zuccarello B, Proietto F, et al: Disorders of the esophageal motor activity in atresia of the esophagus. J Pediatr Surg 22:120-124. 1987