Reversible truncal vagotomy in conscious dogs

GASTROENTEROLOGY

Reversible Truncal Conscious Dogs

Vagotomy in

J. J. GLEYSTEEN,

M. J. ESSER,

Surgical Section,

Administration

Veterans

1983;85:578-83

and A. L. MYRVIK

Medical

Prior attempts at in vivo reversible vagal denervation of the gastrointestinal tract have been limited to cervical cooling techniques that also denervate both sympathetic and vagal pulmonary and cardiac branches. Denervation of vagal efferent fibers at this level has produced results that are inconsistent with those obtained after surgical truncal vagotomy. We have, therefore, developed a technique to provide reversible vagal denervation below the pulmonary and cardiac branches for the study of gastric motility. Five dogs, previously equipped with gastric strain gages and electrodes, underwent implantation of a tubular cooling jacket around a distal thoracic vagal trunk with contralateral vagotomy (4 dogs), or around both vagal trunks (1 dog). The jacket was made of stainless steel tubing in a “I” design. Its inside channel was lined with a sterling silver sheet, and a thermistor was attached to recol d temperature change. Silicone tubing coursed externally to a pump and flask to which 95% ethanol at -70°C was circulated at variable speeds. Thoracic vagal cooling, extending up to 5 h, reversibly blocked gastric contractions induced by insulin hypoglycemia. Contractile waves were terminated at device temperatures of 2%6°C but promptly returned with warming. Dogs were tranquil during denervation, and enclosed nerves remained functional for >4O days. Acute and reversible vagotomy in the animal has experimental application to both cardiopulmonary and gastrointestinal physiology. Fabrication of a Received September 3, 1982. Accepted March 15, 1983. Address requests for reprints to: John J. Gleysteen, M.D., Surgical Service, 5000 West Wisconsin Avenue, Wood, Wisconsin 53193. This work was supported by the Veterans Administration. This study was presented at the Federation of American Societies of Experimental Biology Annual Meeting, Atlanta, Georgia, April 14, 1981. The authors appreciate the technical and interpretive assistance of Hongyung Choi, M.D., in the preparation of the vagal nerve histologic sections. 0 1983 by the American Gastroenterological Association 0016-5085/83/$3.00

Center,

Wood,

Wisconsin

thermal device for vagal cooling at the diaphragm level was designed to simulate surgical truncal vagotomy. The reversibility of the technique permitted repeated comparison of vagal intact and denervated responses in the same animal. The feasibility of reversible vagal blockade was first demonstrated in vivo in anesthetized cats by thermode cooling in the neck (1,2). Electrical stimulation above the cooled segment demonstrated reduced conduction velocity within a short temperature transition zone and terminated impulses within nerve cooled to 7”-8°C (l-3). Fishman et al. (4) first described a technique for vagal cooling in awake animals by displacing cervical vagosympathetic trunks of dogs into skin tunnels that were enclosed in copper radiators. Coolant circulated at temperatures of 3”-5°C produced responses that simulated cervical vagal transection (4). The technique was designed to study pulmonary responses to vagal denervation. Others, however, have since extended its application, specifically to report the effect of cervical blockade upon esophageal peristalsis and lower esophageal sphincter function (5,6), and also on spontaneous migrating myoelectric complexes (MMCs) of the stomach and small intestine (7,8). These studies have produced results that are discordant not only with observations after bilateral cervical vagesection (5,6), but also after diaphragmatic vagotomy (7,9,10). The following study describes the reproducibility of reversible vagal blockade (of 1-5-h duration) using cooling jackets implanted in dogs at the distal thoracic-diaphragm level. The location of the cooling jacket is the same level at which truncal vagectomy is accomplished clinically. An abstract of this material has been published elsewhere (11).

Methods Dog Preparation Five mongrel dogs, 15-22 kg, underwent my, during which two strain gages were attached

laparoto1 cm and

Septrmher

KEVEKSIHLE TKlIN(:Al> VAGOTOXIY

1983

3 cm proximal to the pylorus on the anterior surface of the gastric antrum. A bipolar silver electrode was also sutured next to the gages. Teflon-coated silver wires extended from the gages and electrode through a subcutaneous tunnel to an epoxy-molded plug implanted in the dorsal interscapular midline. This apparatus, with connections to a poly(12.13). graph recorder, has been previously described

Denervation

Device

Construction

The design of the nerve-cooling jacket is shown in Figure 1. Stainless steel tubing (OD 2.4 mm: ID 1.8 mm) was initially bent 180” to form a “U” with the arms parallel and 1 cm apart. The “U” curve was then molded back on itself to form a “J” design M.ith a s-mm channel. Small notches at l-mm intervals were filed along the inside radius of each bend to prevent crimped tubing. The inside surface of the jacket was lined with a sheet of 30-gauge sterling silver, soldered into place and filed flush with the tubing. A Yellow Springs 44004 thermistor (Yellow\ Springs Instrument Co., Yellow Springs, Ohio) was attached with epoxy to the external surface of the silver plate (Figure 1). Teflon-coated silver wire from the thermistor and thick-lvalled (0.164-cm) Is-gauge silicone tubing cemented to the tubular ends ot the jacket both extended 12 cm to connections within another epoxy plug. The exter-

579

nal surface of the jacket was covered with an insulating coat of epoxy and then with reinforced Silastic sheeting. A silicone tubing circuit [Figure 2) led from the epoxy plug to a metal coolant container placed in a Dewar insulating flask, then through a Masterflex variable speed pump (Cole-Parmer Instrument Co., Chicago. Ill.] and back to the plug. The coolant solution was 95% ethanol: its container was surrounded by a dry ice-ethanol slurry at --XX. The jacket thermistor was connected through the epoxy plug to an external telethermometer and a polygraph recorder to provide a permanent record of temperature during cooling. The device was tested for adequate insulated cooling in a 40°C water bath before implantation. A separate thermistor sealed inside the jacket channel demonstrated a lo-2°C temperature differential between the internal and external thermistors.

Device

Implantation

Each animal, previously equipped with gastric strain gages and electrodes, underwent implantation of a cooling jacket through a right (seventh interspace) thoracotomy. In order to provide a single vagal pathway to the stomach for one coolant device only, the distal 10 cm of the left nerve was resected. The right distal main nerve

580

GLEYSTEEN

Right Vagal

Trunk

ET AL

GASTROENTEROLOGY

I Cooling (0.4OC)

-

Jacket

Thertnistor~~

85, No. 3

Figure 2. Cooling and recording diagram (see text for descrip-

I

Vol.

Me

tion). The telethermometer is also connected to the polygraph recorder.

1

M

y Pump

Ethanol Solution

-,’ r-

Diaphragm

Ethanol, (-7OOC)

Antral Stram Gages and Bipolar Electrode

Dry Ice Slurry

Polygraph Recorder

trunk, with any small branches converging from the left side, was positioned inside the cooling jacket immediately above the right hemidiaphragm [Figure 1). The Silastic coverlet of the jacket was sutured to enclose the nerve and to fix itself to adjacent parietal pleura and esophagus. Silicone tubing and sheathed thermistor wires passed through the thoracic wall and subcutaneous tissue to another epoxy plug implanted caudad to the interscapular plug in the dorsal midline. In the last dog in this series, the left nerve was not transected, but was displaced behind the esophagus so that both vagal trunks at diaphragm level were enclosed in a single cooling jacket. The remainder of the operative procedure was the same.

gastric contractility after insulin injection with adequate hypoglycemia (~40 mg%). The dog was then killed, and the cooling jacket with a segment of vagal nerve extending above and below was removed. Serial sections of vagal nerve were prepared with hematoxylin-eosin stain for histologic study.

Results jacket

Within 1-2 wk after thoracotomy for cooling implantation, each dog had returned to its

normal preoperative eating habits and weight. Gastric contractile patterns were compared with those before

thoracotomy.

quiescence Motility

Recording

Gastric motility recordings were resumed after 1 wk of convalescence. Trials were begun during phase 1 contractile quiescence with intravenous injection of regular insulin (bolus 0.6 U/kg or continuous 0.2 U/kg h). The insulin response consisted of antral contractions >lO g force (Figure 31, which regularly followed each gastric pacesetter potential. Blood glucose was determined 30-40 min after insulin injection. Cooling was initiated during insulin-induced contractions. The flow was gradually increased until the device temperature had dropped to 2%6°C and antral contractions had ceased. Vagal blockade was maintained for a variable interval; flow was then slowed to permit warming of the device and to observe return of antral contractions. Heart rate was monitored during cooling and the clinical response of the dog was observed. Recordings were accomplished intermittently after the first week until the vagal trunk ceased to function. Failure of vagal function was documented by absence of induced

terns, induced injection Each

Average

or regularity,

and

promptness

duration

amplitude of onset

of phase

of phase and

1

3 pat-

amplitude

of

contractions after intravenous insulin were unchanged for each animal.

bolus

dog

trials

underwent

during insulin-induced was regularly increased

a series

of

cooling

hypoglycemia. 10-30 beatsimin

Heart rate by insulin

hypoglycemia, but was not altered again by initiation or duration of a cooling trial. Respiratory rate and volume were unchanged. The dogs remained tranquil cooling.

and

did

not

retch

or move

about

during

The effect of coolant circulation through the jacket device on insulin-induced gastric contractility is illustrated in Figure 3. As device temperature dropped to denervation levels, both electrical spike potentials (SP) and their accompanying contractions ceased. Gastric contractions returned promptly with warming after each trial. The jacket temperature necessary to terminate all

September

REVERSIBLE

1983

SP SP SP

P PP COOLANT ON I

BIPOLAR ELECTRODES ANTRUM --H

1 MIN I-

COOLANT OFF I

-

ANTRAL STRAIN GAGE 1 100 -\

40°C \

THERMISTOR TEMPERATURE

FORCE I/-----

2%

PPPP

BIPOLAR ELECTRODES ANTRUM

TACHVGASTRIA 1 MIN w

COOLANT ON I

COOLANT OFF I

-+I-

~~

ANTRAL STRAIN GAGE

I;igure

3. A. Gastric antral myoelectric and contractile response to cooling (2°C) for 5 min under insulin hypoglycemia. Gastric pacesetter potentials (P) remain, and spike potentials (SP) with their antral contractions are terminated during cooling. B. Denervation response effective at 4°C under insulin hypoglycemia. Contractions return with brief warming to 8°C. Tachygastria (rapid pacemaker) is seen near end of cooling phase.

contractions varied between animals from 2°C to 6”C, but remained effective at one temperature for each animal throughout the study period. A cooling temperature was maintained just below reproducible denervation to avoid more frigid temperature which would irreversibly damage the nerve. Exemplary of this denervation temperature threshold is the contractile response recorded in Figure 3 (B), which showed resumption of activity within 4°C of brief warming and termination again with return to denervation temperatures. Frigid destruction of nerve function was documented in 1 dog after purposely cooling for 1 h at a temperature 4°C below the known denervation threshold. Insulin-induced gastric contractions did not return upon warming. Vagal trunks in the first 3 dogs were cooled repeatedly for
TRIJNCAL

VACOTOMY

581

injections of pentagastrin (250 pg) and bethanechol HCl (2 mg) were given just before cooling. Contractile responses to these agents were not altered by nerve cooling [Figure 4). In 5 dogs, vagal trunks enclosed in a cooling jacket remained responsive to insulin hypoglycemia from 47 to 79 days. Eventual nerve failure was abruptly apparent in 4 dogs by onset of vomiting after feeding, and was then documented by failure of induced contractile activity with insulin hypoglycemia. The cooling jacket with enclosed nerve was removed 221 days later, and serial histologic sections of the nerve were obtained. Each nerve was encased in fibrous tissue which filled the jacket lumen, and which microscopically also contained inflammatory cells (Figure 5). The nerve bundle was intact without inflammatory cell infiltrate, but red blood cells were present within the nerve bundle in some sections. 6 days Serial sections of nerve in 1 dog, removed after failure, were also stained with Wright’s reticulum stain and showed no evidence of nerve cell degeneration within or distal to the enclosed segment. One nerve removed 21 days after insulin failure, however, could not be identified; fibrous scar had replaced the nerve in all sections.

Discussion The purpose of this project was to fabricate a convectional cooling device, suitable for intrathoracic implantation, which could be used chronically and could reversibly terminate vagal efferent impulses. The cooling jacket achieved vagal blockade at temperatures between 2°C and 6°C when gastric contractions, centrally induced by insulin hypoglycemia, were terminated. Unchanged contractile stimulation by agents acting at peripheral sites (pentagastrin and bethanechol HCl) during cooling confirmed that efferent vagal interruption occurred at the level of the implanted cooling jacket. Trials with BIPOLAR ELECTRODES ANTRUM

COOIANT

1 MIN.

ON

ANTRAL STRAIN GAGE 1

40%

COOlANT

OFF

10 GM FORCE

THERMISTOR

TEMPERATURE 3%

Figure

4. Gastric antral and myoelectriL response to 250 pg pentagastrin subcutaneously administered 14 min before vagal cooling. Contractions and spike potentials are not altered by vagal denervation at 3°C.

Figure

5. A. Histologic section (96x magnification) of canine vagal nerve cephalad Histologic section (63x magnification) of vagal nerve enclosed in jacket inflammation surrounds the nerve.

to jacket, showing for 79 days. Dense

normal fibrous

surrounding tissue with

tissue. 5. peripheral

September

REVERSIBLE

1983

prolonged cooling beyond 1 h, and up to 5 h, at denervation threshold temperatures were designed to make other studies of gastrointestinal myoelectric and contractile activity feasible without insulin. Reversible vagotomy with techniques of bilateral cervical vagal cooling has been reported by several investigators (5-8,141. This technique appeared adequate to assess the effect of acute vagotomy on insulin-stimulated gastric secretory responses (14), but our purpose was to observe myoelectric and contractile activity of the stomach. Inhibited relaxation of the lower esophageal sphincter with primary (swallowing) peristalsis (5,6) and elimination of spontaneous MMCs in the lower esophageal sphincter and stomach (7,8), both observed with vagal cooling in the neck, are results discordant with vagal transection in the distal thorax or diaphragm level (9,15,16). Induction of vagal blockade at the diaphragm rather than cervical level avoided simultaneous denervation of proximal intrathoracic branches (to lungs and heart), which significantly alter cardiopulmonary dynamics (5,7), and of sympathetic fibers, which course with the vagus in the neck. Our conclusion, corroborated by some preliminary observations, suggests that vagal blockade at the diaphragm simulates the myoelectric and contractile changes induced bv abdominal truncal vagectomy and has the additional advantage of reversibility. IJnilateral vagal innervation was used initially as a physiologic model of complete innervation to assess contractile activity of the stomach. Its use was justified by the crisscrossing of posterior and anterior gastric nerve axons (17). Daniel and Sarna (17) demonstrated that electrical stimulation of one gastric nerve produced prompt contractions on the stimulated side, followed rapidly by contractions on the opposite side. Stimulation through the posterior vagal trunk for anterior gastric surface recordings in our dogs produced no detectable change in promptness of contractile onset before and after unilateral vagotom):. In our last 2 dogs, however, both vagal trunks were enclosed in a single cooling jacket. One of these dogs was studied briefly before developing pneumonia 20 days after implant. The other dog was studied extensively. and histology of its nerve trunks after eventual failure documented both to be intact without evidence of cellular degeneration by reticulum stain. Enclosure of both nerves in one cooling jacket was surgically feasible, and provided the same gastric contractile responses to insulin and cooling as were seen after unilateral vagotomy. Failure of the insulin-stimulated trunk(s) to produce continuous gastric contractions coincided with symptomatic awareness of vagal denervation. Cause for nerve failure other than destruction by profoundly cold temperatures has not been ascertained from these studies. Impending nerve failure could not be

TRI!N(:AI.

VAGOTOMY

583

predicted; the denervation threshold temperature did not change for each dog throughout the course of the coolant trials. Histology of nerves removed within 6 days after failure demonstrated that nerve bundles were uniformly intact without nerve cell degeneration. Focal hemorrhage in the bundle perhaps arose from acute changes in perineural blood supply that just preceded nerve failure. One nerve removed and sectioned 21 days after failure was completely replaced by fibrous tissue. which implied that nerve bundle destruction was a delayed event after loss of vagal transmission.

References 1. Franz

DN. Iggo A. Conduction failure of mvrlinated and nonI Physiol 1968: myelinated axons at low temperatures. 199:319-45. 2. Paintal AS. Block of conduction in mammalian myelinated nerve fibers at low temperatures. J Physiol 1965:180:1-19. on conduction in single 3. Paintal AS. Effects of temperature vagal and saphenous myelinatrd nerve fibers of the cat. J Physiol 19fi5:180:20-49. 4. Fishman NH, Phillipson EA. Nadel IA. Effect of differential vagal cold blockade on breathing patterns in conscious dogs. J Appl Physiol 1973;34:754-8. 5. Ryan JP, Snape WJ. Cohen S. Influence of vagal cooling on esophageal function. Am J Physiol 1977:2j2:El59-64. 6. Price LM, El-Sharkawy TY. Mui HY. Diamant NE. Effect of bilateral cervical vagotomy on balloon-induc:Kd lower esophageal sphincter relaxation in the dog. Gastroenterology 1979: 77:324-g. 7. Hall KE. El-Sharkawy TY, Diamant NE. Vagal control of migrating motor complex in the dog. Am J Physiol 1982: 243:G276-84. 8. Hall KE, Mui H. Greenberg GR. El-Sharka\vy TY. Diamant NE. Migrating motor complexes (MMCl and plasma pancreatic polypeptide (PP) in the dog: effect of vagal blockade (abstr). Gastroenterology 1980;78:1177. 9. Kravitz JJ, Snape LVJ, Cohen S. Effer t of ihoracic. vagotomy and vagal stimulation on esophageal funciion. Am J Physiol 1978:234:E359-64. 10. Lind JF, Cotton DJ. Blanchard K. Crispin IS. Dimopolos GE. Effect of thoracic displacement and vagotomy on the canine gastroesophageal junctional zone. Gastroenterology 1969: 56:1078-85. 11. Gleysteen JJ. Esser MJ. Experimental drl~ic.e for acute and reversible truncal vagotomy in chronic studies of gastrointestinal motility (abstr.) Fed Proc 1981:40:57;‘. 12. Cowles V, Condon RE, Sclrulte WJ. W’oods JH. Sillin LF. A quarter wheatstone bridge strain gage force transducer for recording gut motility. Am J Dig Dis 1978:23:936-41. 13. Gleysteen JJ. Gohlke EG. The antrum can control gastric emptying of liquid meals. J Surg Res 1979.26:381-91. 14. Cairns D. Deveney CW. Way LW’ Mechaiiism of release of gastrin by insulin hypoglycemia. Surg Forum 1974;25:325-7. 15. Marik F, Code CF. Control of the interdigestive myoelectric activity in dogs by the vagus nerves and pentagastrin. Gastro1975:69:387-95. enterology 16. Ruckebusch Y, Bueno L Migrating complex of the small intestine: an intrinsic activity mediated 1)~.the vagus. Gastroenterology 1977;73:1309-14. 17. Daniel EE. Sarna SK. Distribution of excitatory vagal fibers in canine gastric wall to control mo!ilit\-. Gastroenterology 197fi: 71.608-13.