Ann Thorac Surg 2001;71:303-308
© 2001 The Society of Thoracic Surgeons
Original article: general thoracic
Colon interposition for esophageal replacement: isoperistaltic or antiperistaltic? Experimental results
Bernhard Dreuw, MDa,b,
Jürgen Fass, PhDa,b,
Svetlana Titkova, MDa,b,
Michail Anurov, MDa,b,
Michail Polivoda, MDa,b,
Alexander P. Öttinger, MDa,b,
Volker Schumpelick, MDa,b
a Department of Surgery, Aachen University of Technology, Aachen, Germany
b Department of Digestive Physiology, University Research Center, Moscow, Russia
Accepted for publication June 5, 2000.
Address reprint requests to Dr Dreuw, Department of Surgery, Aachen University of Technology, Pauwelstrasse 30, D- 52074 Aachen, Germany
e-mail: bernhard.dreuw{at}post.rwth-aachen.de
 |
Abstract
|
|---|
Background. Isoperistaltic colon is preferred to antiperistaltic colon for esophageal replacement, but experimental data do not exist to support this practice.
Methods. In 7 dogs a 20 cm long colon loop was interposed between the skin and the small bowel, isoperistaltically in 3 dogs and antiperistaltically in 4 dogs. Three months later five strain-gauges were implanted and evacuation was investigated by motility testing, barium studies, and scintigraphy.
Results. Motility recording showed normal colon motility in the excluded loops. Quiescent states (duration 40.2 ± 13.6 minutes) were followed by contractile states (duration 7.5 ± 2.4 minutes, frequency 3.3 ± 0.6 per minute). The main peristaltic direction of isoperistaltic loops was isoperistaltic, and the main peristaltic direction of antiperistaltic loops was antiperistaltic. Evacuation took place exclusively during the contractile status. Half time emptying was more rapid in isoperistaltic loops (35 ± 11 vs 69 ± 16 minutes). The content of antiperistaltic loops was held back by antiperistaltic activity. Application of oatmeal porridge into the loops shortened the quiescent status from 40.2 to 13.2 ± 4.8 minutes.
Conclusions. The colon graft for esophageal replacement is an active system. Food is stored during the quiescent states and evacuated during the contractile states. The original peristaltic direction is preserved so that retroperistalsis in antiperistaltic loops may lead to patient discomfort and pulmonary complications.
|
Introduction
|
|---|
Functional results after esophageal resection are considered to be best if the colon is used for esophageal replacement. A colon graft is usually preferred in children, patients with benign or limited malignant disease with an expected long-term survival. Judgment of postoperative function is mainly based on clinical observations, questionnaires, evaluation of patients symptoms, and radiologic or scinthigraphic examinations [1]. Detailed motility studies are rare. If they are done the colon graft seems to have little motor activity. Cinefluoroscopic studies failed to demonstrate peristalsis in an interposed colon [2, 3]. This indicates that the colon graft mainly acts as a passive tube conducting the pharynx and the small bowel and will evacuate mainly by gravity [4–7].
However, there are some reports of peristaltic motor activity after swallowing [8, 9]. Motility may be induced by acid infusion or other stimuli [10–12]. This has led to the opinion that an isoperistaltic colon interposition should be preferred to an antiperistaltic one, although the clinical late results are the same for both types of grafts [5, 13, 14]. Until now, there has not been any experimental data supporting this practice.
Our objective in this study was to investigate the motility and evacuation function of isoperistaltic and antiperistaltic colon loops, excluded from the intestinal stream, as they are used for esophageal replacement.
|
Material and methods
|
|---|
Surgical procedure
The following surgical procedure was performed on 7 dogs each weighing 17 to 25 kg (20 ± 3 kg). After premedication with 15 mg/kg ketamine hydrochloride, 1 mL xylazine hydrochloride (xylazinum hydrochloride, Bayer AG, Leveukusen, Germany), and 0.5 mL atropine sulfate, a general anesthesia was maintained with 30 mg/kg pentobarbital sodium. The abdomen was opened by a midline incision. To minimize shortening of the bowel, 5 mL of xylocaine hydrochloride was injected into the base of the mesentery. The colon mesentery was visualized before an adequate part of the colon was selected. It consisted of a 20 cm long segment of the midcolon. Care was taken not to damage the inferior mesenteric vein to prevent necrosis of the rectum. The proximal end was marked with a suture. Then the loop was excised, pediculated by its mesentery. The proximal and distal ends of the remaining colon were sutured together end-to-end with a layer of all single stitch anastomosis using 3-0 polyglycol acid first. Then this suture line was invaginated by a second seromuscular 3-0 single stitch suture line. All following intestinal anastomoses were done with the same technique. The small bowel was cut 40 cm proximal to the ileocecal valve and the proximal end was anastomosed 20 cm proximal to the ileocecal valve. This resulted in a 20 cm long Roux-Y type of small bowel. After randomization the colon was anastomosed to this Roux-Y loop in an isoperistaltic or an antiperistaltic manner. This design was taken to prevent the colon loop from intestinal reflux. The proximal part of the colon was brought out to the skin as a colon fistula (Fig 1). Finally the abdomen was closed and the dogs were allowed 12 weeks to recover.
View larger version (12K):
[in this window]
[in a new window]
|
Fig 1. Experimental model: isoperistaltic or antiperistaltic colon segments as a Roux-Y loop with skin fistula.  1–5: Orientation of the strain gauges. A, a = proximal colon; B, b = distal colon. The antiperistaltic colon segments are arranged antiperistaltically (a = distal colon; b = proximal colon).
|
|
Ten days before performing all tests, the abdomen was opened again under general anesthesia to implant the strain gauges. The strain gauges (type SA-03–090DH-350, Measurement Group, Inc, Raleigh, NC) were adhered back-to-back on a curved beryllium copper strip to provide a half bridge circuit. After they were coated for waterproofing, both the strain gauges and the lead wires (type 336 FTE) were sheathed in silicon (Perthèse silicone sheeting, Laboratoire Perouse Implant, Bornel, France and GB silicone rubber, GE Silicones, Cheshire, UK). Finally the transducers had a size of 18 x 15 mm. Three strain gauges were spaced equal distances apart (mean 5.5 ± 0.4 cm) along the colon and 2 strain gauges on the small bowel loop using seromuscular sutures. They were orientated to record circular muscular contractions. The lead wires which had been connected to a specially designed connector were brought up to the skin surface of the dog’s abdomen to allow connection to the motility recording unit.
Motility testing
After an overnight fast, motility was recorded from the implanted strain gauges for 2 hours to determine basal fasted colonic activity in the conscious dog. After that, a 20 French Foley catheter was introduced into the stoma and blocked with 4 mL of air. Fifteen minutes later, after the distension stimulus from the balloon had resolved, 40 mL of oatmeal porridge was inserted into the colon. Motility was recorded for another 2 hours. All motility data were played to a standard personal computer by a digital analogous converter for storage and analysis. The display and analysis program for the computer was written (by Vladimir Klioutchikov) from the department of physiology at the University of Moscow (Moscow, Russia).
Combined motility and fluoroscopy
Two days later, after an overnight fast, a combined motility test and fluoroscopy were recorded again from the implanted strain gauges for 2 hours, but this time the dog was sedated with 15 mg/kg ketamine sulfate. After 2 hours a 20 French Foley catheter was introduced into the stoma and blocked with 4 mL of air. Fifteen minutes later, 40 mL of oatmeal porridge mixed with barium sulfate was inserted into the colon. This was done under continuous fluoroscopic control for 10 minutes. The fluoroscopy was stored on a standard video recorded tape after synchronization with the strain gauge motility recording unit. Fluoroscopy was performed and stored for 10 minutes during motility quietness and during the phases of colonic motility complexes. The investigation continued for 2 hours or until the colon loop was completely empty of barium. Finally the colon was filled again with 40 mL barium porridge mixture, before 2 mg of prostigmine were given intravenously. Then motility and fluoroscopy were recorded until the next regular colon motor complex occurred.
Scintigraphy
After an overnight fast the dogs were sedated with 15 mg/kg ketamine sulfate. A 20 French Foley catheter was introduced into the stoma and blocked with 4 mL of air. Fifteen minutes later the dogs were positioned under the gamma counter. From the previous fluoroscopy, the ideal body position that minimized overlay of the colon loop by small bowel loops was known, and the dogs were positioned in that way. Then 40 mL of oatmeal porridge, mixed with 100 mBq Technetium sulfate, was inserted into the colon. Using one count per second, 60 frames were done during 1 hour. If half time evacuation was not achieved after that time, measurement was continued for another half hour.
Analysis of data
All motility data were analyzed visually from the computer analysis program. As described in the literature, colonic motility was defined as bursts of contractions and quiescent periods in between. A colonic motility complex (CMC) was defined as a contractile burst of duration for more than 1 minute. A period of quiescence of more than 1-minute duration separated two individual periods of adjacent motor complexes. CMC measurements were calculated for each individual strain gauge. By visual inspection, the start and end of each CMC and each separated contraction amplitude were marked in the computer program. The analysis program then calculated the duration, frequency, mean amplitude, and motility index of each strain gauge. By combining the begin of each motor complex between the different strain gauges, a separation of peristaltic, segmental, or antiperistaltic activity was made.
For comparison of the motility measurements, statistical analysis was performed by repeated measures of analysis of variance and covariance, using SPSS for Windows version 7.52 (SPSS, Chicago, IL). For all statistical data, a value of p less than 0.05 was considered to be significantly different.
All animals received humane care in compliance with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (NIH Publication No 86-23, revised 1985).
|
Results
|
|---|
Colonic motor complexes
Fasting period
The characteristics of the CMC contractile patterns in the fasting state of all dogs were periods of contractile activity, interrupted by quiescent states. The contractile phases of the isoperistaltic colon loops had a duration of 7.5 minutes (standard deviation [SD] of 2.4) and a frequency of 3.3 contractions per minute (SD of 0.6). The resting state had a duration of 40.2 minutes (SD of 13.6) until the next phase of motor activity began (Table 1).
For the antiperistaltic loops the results were similar. A contractile status of 7.2 minutes in duration (SD 2.5) and a frequency of 3.4 contractions per minute (SD 0.6) were followed by a resting state of 39.7 minutes in duration (SD 12.8). Giant contractions were not identified in any dog during the time of measurement.
The direction of the peristaltic sequence was antiperistaltic in 35% of phases in the antiperistaltic loops (Fig 2). The remaining 39% were mixed isoperistaltic and antiperistaltic (Fig 3), peristaltic in 22% (Fig 4) and segmental in 4%.
View larger version (22K):
[in this window]
[in a new window]
|
Fig 2. Antiperistaltic colonic motility complex (CMC) in a dog with antiperistaltic colon loop. After a short quiescent state another CMC occurred, this time with mixed segmental and peristaltic direction.
|
|
View larger version (17K):
[in this window]
[in a new window]
|
Fig 3. Mixed isoperistaltic and antiperistaltic colonic motility complex in a dog with isoperistaltic colon loop.
|
|
In the peristaltic loops the main direction of the CMC was peristaltic in 77%. Mixed peristaltic sequences occurred in 3% and segmental in 20%. An antiperistaltic CMC over the whole length of the graft was not observed (Table 2).
Feeding period
After filling the colon loops with oatmeal porridge there was a marked shortening of the quiescent states to 13.2 minutes (SD of 4.8) for the isoperistaltic loops and 11.9 minutes (SD of 5.3) for the antiperistaltic loops. The frequency of contractions did not change. The duration of the contractile state enlarged: the isoperistaltic loop was 11.3 ± 1.9 minutes and the antiperistaltic loop was 11.9 ± 2.1 minutes. The quality and direction of the contractile sequences changed to more segmental sequences in the antiperistaltic grafts and to 100% peristaltic sequences in the peristaltic grafts (Tables 1 and 2).
Simultaneous motility and fluoroscopy
Any emptying of the colon loops was accompanied by contractile states. There was no passive emptying of the loops. In the quiescent states the content of the colon loops remained unchanged although the distal small bowel loops had marked peristaltic activity. During antiperistaltic sequences we noted a flow of the oatmeal porridge from distal to proximal (Fig 5). In two instances the retrograde peristaltic force propelled the blocked balloon out of the skin fistula. After 1 hour of observation all isoperistaltic loops were completely empty. In three of the four antiperistaltic loops, however, barium oatmeal was present, even after 4 hours when the study was stopped.
View larger version (50K):
[in this window]
[in a new window]
|
Fig 5. Retroperistalsis: Three images from the fluoroscopy series with a time gap of 10 seconds in a dog with an antiperistaltic colon loop. An antiperistaltic wave is squeezing against the balloon that occludes the skin fistula (arrows). During the same time a phase of antiperistaltic colonic motility complex was observed (Fig 2).
|
|
Scintigraphy
Placement of two regions of interest, one to the proximal end of the loop and one to the distal end, resulted in characteristic biphasic waves (Fig 6), indicating a retrograde flow of the radiolabeled meal.
View larger version (16K):
[in this window]
[in a new window]
|
Fig 6. Biphasic course of scinthigraphic curves in a dog with antiperistaltic colon loop. Region 1 is set to the proximal part of the colon loop; region 2 is set to the distal part of the colon loop. Initially, after filling the loop (0 minutes), activity is high in the proximal part. Then the radioactive labeled food is transported to the distal part of the loop with increasing activity distal and decreasing activity proximal (13 minutes). Instead of emptying, the food is transported back to the proximal part of the loop as indicated by an increase of radioactive activity (25 minutes).
|
|
The evacuation half time of the isoperistaltic colon loops for semisolid food (oat meal porridge) was 35 minutes (SD of 10) and was significantly longer for the antiperistaltic loops with 69 minutes (SD of 15; p < 0.001). After the follow-up period the percentage of activity left in the isoperistaltic colon graft was 11% (SD of 3) and in the antiperistaltic graft it was 39% (SD of 12; p < 0.002).
|
Comment
|
|---|
The main motor action of the colon is to slowly propel the luminal content toward the anus and rapidly move its fecal material caudad during mass movement and defecation [15]. The motor complexes described in the colon consist of alternating contractile and quiescent states. The contractile states are composed of individual phasic contractions of short or long duration. Groups of contractions are organized as colonic migrating motor complexes or colonic nonmigrating motor complexes. The frequency of short and long duration contractions is in the range of 4 to 6 minutes and 0.5 to 2 minutes, respectively. The contractions of the colon are poorly coordinated in space and generally propagate only over a fairly short distance [15].
Unlike the upper gastrointestinal tract the colon is seldom empty. When the colon is used as an esophageal substitute this will change. Then most of the time the colon will be empty and it will only fill when the patient eats. In addition, the luminal content will become different, changing from a semisolid or solid consistency to a more fluid consistency. To date, little is known about the motility of such colon interpositions used for esophageal replacement.
Miller and colleagues [8] reported pressure waves in 3 patients with a total motor activity of 37%. The main pressure type were rhythmic type II waves according to Templeton and Lawson [16]. The authors also reported a pressure wave showing propagation on swallowing water, but they could not see the waves on fluoroscopic examination.
Isolauri and associates [5] performed radionuclide transit studies in 18 antiperistaltic and seven isoperistaltic colon grafts from 5 to 175 months after the operations. They found that the emptying of the grafts was markedly slower than that of the normal esophagus. The intraabdominal third of the graft had a residual capacity of 50.5% after 20 minutes. They did not found any difference between the antiperistaltic and the isoperistaltic grafts and between grafts with short or long follow-up.
In our study, quiescent states and motor active states changed in a rhythmic manner. Quiescent states sometimes lasted for more than 60 minutes. This may be one explanation for why short time radiographic, scinthigraphic, and manometric studies sometimes fail to demonstrate motility in the colon interposition.
This study has shown that peristaltic and antiperistaltic colon loops, excluded from the intestinal stream, will keep their physiologic motility pattern. The motility pattern of the colon loops, both peristaltic and antiperistaltic, were similar to that described in the literature for normal colon motility [15, 17]. The amplitude, duration, and motility indexes were no different between the peristaltic and the antiperistaltic groups. These results indicated that a colon loop, pedunculated by its mesentery, preserves the motility pattern at least during the first postoperative month. This is not a passive pipe but an organ with active peristalsis. We did not find results indicating that the colon will function like the esophagus. A swallow will not be transmitted by active force through the colon loop. As long as there is no motility, food is stored in the relatively large organ. However, the distension stimulus by the food may induce a motor active state that will help to evacuate the colon graft if it is arranged in an isoperistaltic manner. Clearance of the colon graft will take place during phases of colonic motility complexes. This reminds one of the storage capacity of the stomach and may explain in part the clinically good functional results.
The situation in the case of antiperistaltic colon loops is more complex. The motility observed in our dogs was mainly antiperistaltic. When the colon graft recovers from the postoperative ileus, and the patient is still in the recumbant position during the early postoperative period, antiperistaltic motility is a risk factor for the development of occult aspiration and pneumonia. In the long run this antiperistaltic force is not strong enough to overcome the gravity power of a 40 to 50 cm water column that would be necessary to propel the luminal content into the pharynx in an upright sitting adult. However, antiperistalsis leads to a markedly delayed emptying of the colon loop with retained food in the distal colon and clinical sensation of regurgitation in humans [13]. Therefore, patients with antiperistaltic colon grafts should be informed that the emptying of the colon may be delayed. They should not eat late in the evening before they go to bed. It should be natural for them to sleep in an elevated bed to avoid night-time aspiration during antiperistaltic CMCs.
Our results indicate that evacuation of food from the interposed colon as esophageal replacement does not only occur by gravity but by active clearance through regular colonic motility complexes. Antiperistaltic sequences persist in antiperistaltic loops and improve the reservoir function. However, regurgitation, reflux, and aspiration may occur and need special attention.
The isoperistaltic colon has a reasonable reservoir function. The clearance is completed by peristaltic activity during colonic motor complexes that prevent regurgitation and aspiration. From a functional point of view, the isoperistaltic interposition should be preferred to the antiperistaltic whenever possible.
|
Acknowledgments
|
|---|
This study was supported in part by the German Research Society, Bonn, Germany, Grant 436 RUS 17/49/98. The authors thank Dr Alexander Gregory for performing the scintigraphy and Vladimir Klioutchikov for writing the computer display and analysis program.
|
References
|
|---|
-
DeMeester T.R., Johansson K.E., Franze I., et al. Indications, surgical technique, and long-term functional results of colon interposition or bypass. Ann Surg 1988;208:460-474.[Medline]
-
Rodgers B.M., Talbert J.L., Moazam F., Felman A.H. Functional and metabolic evaluation of colon replacement of the esophagus in children. J Pediatr Surg 1978;13:35-39.[Medline]
-
Sieber A.M., Sieber W.K. Colon transplants as esophageal replacement: cineradiographic and manometric evaluation in children. Ann Surg 1968;168:116-122.[Medline]
-
Clark J., Moraldi A., Moossa A.R., et al. Functional evaluation of the interposed colon as an esophageal substitute. Ann Surg 1976;183:93-100.[Medline]
-
Isolauri J., Koskinen M.O., Markkula H. Radionuclide transit in patients with colon interposition. J Thorac Cardiovasc Surg 1987;94:521-525.[Abstract]
-
Othersen H.B., Jr, Clatworthy H.W., Jr Functional evaluation of esophageal replacement in children. J Thorac Cardiovasc Surg 1967;53:55-63.[Medline]
-
Postlethwait R.W. Colonic interposition for esophageal substitution. Surg Gynecol Obstet 1983;156:377-383.[Medline]
-
Miller H., Lam K.H., Ong G.B. Observations of pressure waves in stomach, jejunal, and colonic loops used to replace the esophagus. Surgery 1975;78:543-551.[Medline]
-
Moreno O.E., Tomas R.M., Paris F., et al. Motor activity of esophageal substitute (stomach, jejunal, and colon segments). Ann Thorac Surg 1986;41:515-519.[Abstract/Free Full Text]
-
Hankins J.R., Cole F.N., McLaughlin J.S. Colon interposition for benign esophageal disease: experience with 23 patients. Ann Thorac Surg 1984;37:192-196.[Abstract/Free Full Text]
-
Jones E.L., Skinner D.B., DeMeester T.R., Elkins R.C., Zuidema G.D. Response of the interposed human colonic segment to an acid challenge. Ann Surg 1973;177:75-78.[Medline]
-
Kelly J.P., Shackelford G.D., Roper C.L. Esophageal replacement with colon in children: functional results and long-term growth. Ann Thorac Surg 1983;36:634-643.[Abstract/Free Full Text]
-
Isolauri J. Colonic interposition for benign esophageal disease. Long-term clinical and endoscopic results. Am J Surg 1988;155:498-502.[Medline]
-
Thomas P., Fuentes P., Giudicelli R., Reboud E. Colon interposition for esophageal replacement: current indications and long-term function. Ann Thorac Surg 1997;64:757-764.[Abstract/Free Full Text]
-
Sarna S.K. Physiology and pathophysiology of colonic motor activity (1). Dig Dis Sci 1991;36:827-862.[Medline]
-
Templeton R.D., Lawson W. Studies in the motor activity of the large intestines. I. Normal motility in the dog, recorded by the tandem balloon method. Am J Physiol 1931;96:667.[Free Full Text]
-
Sarna S.K. Physiology and pathophysiology of colonic motor activity (2). Dig Dis Sci 1991;36:998-1018.[Medline]
This article has been cited by other articles:
|
|
|
|
 
T. Grikscheit, E. R. Ochoa, A. Srinivasan, H. Gaissert, and J. P. Vacanti
Tissue-engineered esophagus: experimental substitution by onlay patch or interposition
J. Thorac. Cardiovasc. Surg.,
August 1, 2003;
126(2):
537 - 544.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
