In vitro and in vivo technique for assessing vagus nerve regeneration after parietal cell vagotomy in the rat

Journal of the Autonomic Nervous System, 9 (1983) 27-51

27

Elsevier

In vitro and in vivo technique for assessing vagus nerve regeneration after parietal cell vagotomy in the rat Stephen N. Joffe, Aysel Crocket, Mei Chen and Kim Brackett Department of Surgery, University of Cincinnati Medical Center, 231 Bethesda A venue, Cincinnati, OH 45267 (U.S.A.)

(Received June 7th, 1983) (Accepted July 20th, 1983)

K e y words: vagus nerve - reinnervation - regeneration - gastric secretion - antrec-

tomy - parietal cell - vagotomy - pancreatic polypeptide - rats insulin hypoglycemia

Abstract This study determined if the vagus nerve can regenerate a n d / o r reinnervate the gastric parietal cell mass after parietal cell vagotomy (PCV) and examines tests for assessing vagus nerve regeneration in rats. Microscopic dissection of the neurovascular bundle allowed the vagus nerve to be divided at the gastric body with preservation of the antropyloric nerve and gastric vasculature. Gastric secretory tests were performed under basal and stimulated conditions using secretagogues and insulin hypoglycemia. The candidate hormone, pancreatic polypeptide, was measured in plasma following a mixed meal, insulin hypoglycemia and i.v. secretin. Rats were killed weekly for 9 weeks and the vagal nerve distribution examined by both light and electron microscopy. Stimulated gastric acid output fell from 164 to 26 # m o l / h immediately after operation ( P < 0.001). One week following PCV, the nerves were swollen with fibroblast infiltration and collagen around axon groups showed degeneration. By the third week after PCV, the axons were more densely packed with neurofilaments and acid output had increased to 183 # m o l / h . In the fourth and fifth weeks, the enlarged Schwann cell processes had more axons and acid output rose to 2 6 2 / x m o l / h . By the seventh week, both large and small axons were identified and the acid output was 93% higher than the preoperative level ( P < 0.001). PCV and antrectomy also was followed by reinnervation of the gastric mucosa. Pancreatic polypeptide concentration in plasma was virtually unchanged 0165-1838/83/$03.00 © 1983 Elsevier Science Publishers B.V.

28 following ingestion of food, insulin hypoglycemia or secretin. In rats, following PCV, gastric secretory tests and electron microscopy seem to be the most reliable methods of assessing vagus nerve regeneration.

Introduction

Duodenal ulceration may be treated by a large number of operations, but attention is now being focused on the more 'physiological' operation of parietal cell vagotomy (PCV) which has numerous synonyms [50]. Initially conceived by Griffith and Harkins [23], the operation was introduced clinically with a pyloroplasty by Holle and Hart [28]. Independently, Johnson and Wilkinson in England [39] and Amdrup and Jensen in Denmark [3] described a PCV without drainage for the treatment of chronic duodenal ulceration. By denervating the gastric parietal cell mass and preserving the vagal supply to the antrum and pylorus (nerves of Latarjet), the operation should avoid the risks and side effects of other surgical procedures and reduce dumping and diarrhea. Assessment of any new operation for chronic duodenal ulceration requires assessment of the 3 criteria of mortality, morbidity and recurrence rate. Mortality of PCV is acceptably low at 0.3% [41] and the decreased morbidity confirms the superiority of this operation compared to vagotomy with pyloroplasty or antrectomy in several prospective randomized clinical studies [1,13,14,42,56,63]. The incidence of recurrent ulceration in many of the initial reports was low, but in reviewing 6 of these series with a subsequent follow-up period for at least 5 years in all patients, the results indicated a linear increase of recurrent ulceration of 2% per annum [4,21,27,42,43,51]. The cause of this recurrent ulceration is essentially unknown. Technical details shown to be important include adequate clearance of vagal fibers from the lower esophagus, dissection of the 'crows foot' at the antro-pyloric junction, and division of the fundic nerve [13,21,24,51,56]. Factors which may contribute include preoperative gastric hypersecretion [42,63], postoperative reduction in acid output [11,46,47], hypergastrinemia and autonomous-secreting gastrinomas [51,53,54], denervation hypersensitivity [12,35], ulcerogenic drugs [33], mucosal cell proliferation [26] and site of the primary ulcer [5]. After a PCV, a large number of patients convert from a 'negative' to a 'positive' insulin (Hollander) test [7]. Immediately after operation, only 3%-6% are positive, but this increases to between 30 and 90% at 1-4 years after operation [40,42,56]. Initially, this conversion was thought not to be important, but there now appears ,to be a correlation between a positive Hollander test and the risk of developing a recurrent peptic ulcer. Why a negative response to a Hollander test should convert to a positive response is unknown. Possible reasons suggested include regeneration of vagal fibers [40], increased sensitivity of parietal cell mass to humoral stimuli [20], or a temporary neuropraxia [20,40]. Faxen [16] showed the acid response to insulin increased more rapidly with time after a PCV than a selective vagotomy and postulated a reinnervation from the antrum to the body. Hancock et al. [25], finding

29 a significant reduction in the peak acid output using insulin discriminants [15,52] when a PCV was combined with a rotational gastropexy, suggested that this prevented vagal nerve regeneration. However, all these suggestions are purely hypothetical. We have used a previously modified technique [55] of producing duodenal ulcers in rats involving the infusion of gastric secretagogues [19,32,33]. These peptic ulcers can be prevented by various gastric antisecretory agents [31,32,34], truncal vagotomy [30,37], and PCV [35,36]. In rats we found that a neurovascular parietal cell vagotomy only temporarily prevented these experimental duodenal ulcers for 5 days due to a gradual recovery of the parietal cell mass to stimulation [35]. Gastric ischemia could have partially accounted for this recovery. Subsequently, we have modified the technique of PCV by dissecting out the neurovascular bundle and dividing only the vagal nerve fibers to the body of the stomach while preserving the gastric blood vessels - - a form of superselective PCV [38]. Postoperative tests to determine the completeness of vagotomy and possible regeneration require the collection of gastric juice. It would be advantageous to measure an immunoreactive polypeptide hormone response to stimulation in peripheral blood which could indicate the intactness of vagal nerves. Pancreatic polypeptide (PP) may fulfill this role. Since first isolated [9] as a contaminant of chicken insulin [44], there have been numerous publications dealing with its physiological and pharmacological functions [2,9,17,18,44,45,49,57,58,60-62]. The role of the vagus in the control of PP secretion after a meal is controversial. Three months after truncal vagotomy, the primary rapid increase in PP after food is eliminated and the secondary response may be reduced [57] or unchanged [2]. This led to the suggestion that a cholinergic non-vagal pathway may be of importance because patients evaluated several years after truncal vagotomy showed a normal response to food [61]. Stern [60], studying the release of PP after a protein-rich meal, found the initial PP response was greater after a PCV at 3 months postoperatively than following a truncal vagotomy with an antrectomy or pyloroplasty. Their results indicated that the release of PP by food requires both intact vagi and an intact stomach [60]. Insulin hypoglycemia produces a rapid rise in plasma PP in normal and duodenal ulcer patients, but does not occur after truncal vagotomy [2] and is impaired after selective vagotomy [57,58]. Cephalic stimulation of PP by sham feeding is virtually eliminated by truncal vagotomy [61] but is unaffected by PCV [58]. The aim of this study was to examine some of the in vitro and in vivo techniques for assessing vagus nerve regeneration after a parietal cell vagotomy (PCV) in the rat. In vivo studies will include the measurement of gastric secretion and plasma levels of immunoreactive pancreatic polypeptide (PP) under basal and following various stimulatory tests. The vagus nerve and its distribution in the stomach were examined by light microscopy using the methylene blue squash preparation technique and by electron microscopy at regular intervals after the PCV. These results will be of considerable importance regarding the surgical treatment of peptic ulceration by PCV and will form the basis of future clinical studies in man.

3o

Materials and Methods

Parietal cell vagotomy (PCV)

Operative technique Male albino Wistar rats (weight 150-250 g), following a 12-h fast were anesthetized with ether and a 2 cm midline abdominal incision made from the xiphisternum. The stomach was then delivered into the wound and the esophago-gastric junction and anterior neurovascular bundle exposed using a x 16 magnification (Mediscopy Mark II). By careful dissection, the anterior nerves to the fundus and body of the stomach were divided and tied, leaving the gastric blood vessels and vagus nerve to the antrum and pylorus intact. Reflecting the greater curvature of the stomach superiorly exposed the posterior neurovascular bundle and a similar procedure completed the denervation of the parietal cell mass (PCV). The neuroanatomy of the rat reveals only one major vagus nerve to the body of the stomach anteriorly and posteriorly and occasionally an accessory branch at the lower end of the esophagus [6]. This is different from man, dog and rabbit. The abdomen was then closed using 000 catgut and autoclips for skin. We have now performed over 300 superselective PCVs and 250 neurovascular PCVs in rats with no mortality [6,35,37,48]. For comparisons of insulin hypoglycemia data, a bilateral subdiaphragmatic truncal vagotomy was performed [30,37,48,59].

In vitro tests Light microscopy Ten male Wistar rats were starved overnight, anesthetized and the stomachs perfused with 0.05% solution of methylene blue in Ringer's lactate for 10 min. The stomach was then excised and, overstretching being avoided, was carefully pinned out. The specimen was kept in the methylene blue solution and oxygen passed over the stomach at a rate of 4 1/min for 1 h. The stomach was then fixed overnight in 8% ammonium molybdate at a temperature of 4°C. After the specimen had been washed with distilled water, squash preparations were made [8]. The slides were examined under light microscopy and photographed. Electron microscopy A parietal cell vagotomy was performed in 70 male Wistar rats (weight 150-2~0 g) and at each week postoperatively, groups of 5 rats were killed. Under ether anesthesia, a segment was removed of stomach wall distal to the site of nerve division (Fig. 1). A minimum of 7 pieces of gastric fundus from each rat was fixed in 2% glutaraldehyde ( m / 1 5 Sorensen's phosphate buffer, pH 7.35-5) and stained with 2% osmium tetroxide and lead citrate and processed for electron microscopy. Sections were examined using a TEM-JEOL JEM-100C electron microscope. To determine normal morphology, sham-operated and unoperated rats were starved for

31

I! "',,\;

Fig. 1. Diagrammatic representation of a parietal cell vagotomy,showing site of vagus nerve division and an area of gastric wall excised for electron microscopy. 12 h, anesthetized and a similar area of gastric fundus processed and examined by electron microscopy. In vivo tests Gastric secretory tests

Basal and stimulated gastric output were measured on sham-operated rats and after PCV. This technique has been successfully performed for over 6 years [32]. This was performed immediately after operation and at weekly intervals for 12 weeks in 5 rats at each time interval. Under light ether anesthesia, a no. 8 French infant feeding catheter with side holes was passed i n t o the stomach and secured with pursestring suture in the rumen. Contents of the stomach were sucked out and washed with 0.9% saline. Animals were discarded if the stomach contained food residue or feces, or if fluid could not be easily recovered. The abdominal wall was then closed and the catheter sutured to the anterior abdominal wall. A butterfly needle (Abbott R-19) was sutured into the dorsal s.c. tissue and the rat placed in an individual restraining cage. On recovery from anesthesia, the stomach juice was then aspirated and discarded and a 1-h basal period of gastric juice was then collected. Following this, a combination of two gastric secretagogues, pentagastrin (4 # g / k g / m i n ) and carbachol (0.8 # g / k g / m i n ) were continuously infused subcutaneously using a Braun Unita pump. Gastric juice was collected at 15-min intervals by gentle suction into 5 ml syringes for a further 4 h. Insulin hypoglycemia test (Hollander) was performed in sham-operated rats and immediately after recovery from the PCV. Following a 1-h basal collection of gastric juice, soluble insulin was

32 injected in a dose ranging from 0.4 to 0.8 units/kg and the gastric juice was collected for a further 2 h at 15-min intervals. Hypoglycemia was confirmed during the experiments by sampling venous blood from the tail. The volume of gastric juice collected each hour was measured and the acidity determined by titrating against 0.1 N NaOH with phenol red as an indicator. Total acid output was obtained by multiplying the volume and titratable acidity and expressed in ~ m o l / h . Statistical analysis was performed using the Student's t-test for paired and unpaired observations and Mann-Whitney U-test.

Plasma immunoreactive pancreatic polypeptide (PP) Rats were fasted for 12 h and under pentobarbitol anesthesia (36 m g / k g intraperitoneally) a jugular cannula was inserted. The rats were then placed in modified restraining cages and 3 h after recovery of consciousness the following 3 stimulatory tests known to release PP in man were performed. Insulin hypoglycemia. Basal blood samples were drawn at - 3 0 and - 1 5 min. Insulin (0.4 U / k g ) was then administered as an i.v. bolus infusion and further blood samples drawn at 15, 30 and 60 min. At each time interval, 0.4 ml of blood was collected in heparinized tubes mixed with aprotinin (1000 units/ml blood). The specimens were cold centrifuged and snap frozen and stored at - 7 0 ° C . Standard mixed meal. Basal blood samples were drawn at - 30 and - 15 min and a standard meal given. This consisted of 2 g of Rodent Laboratory Chow (Ralston Purina, St. Louis, MO 63188, U.S.A.). This contained not less than 23% protein, 4.5% carbohydrate and added ground corn, soyabean, oat meal and wheat. Blood samples were collected at 15, 30 and 60 min with 0 min being the time the rats started eating. At each time interval, 0.4 ml of blood was collected in heparinized tubes, mixed with aprotinin (1000 U / m l blood), cold centrifuged, snap frozen and stored at - 70°C. Secretin stimulation test. Basal blood samples were drawn at - 30 and - 15 min. G I H secretin (Karolinska Institute, Sweden) was given as an i.v. bolus (2 C U / k g / 3 0 s) at 0 min and 0.4 ml of blood was drawn at 2.5, 5, 7.5 and 10 minutes. This was collected in heparinized tubes, mixed with aprotinin (1000 U/ml), cold centrifuged, snap frozen and stored at - 7 0 ° C . As there was a 7-day delay between each of these tests to allow recovery from the withdrawal of 4.5 ml blood, rats were randomly allocated to the tests. Radioimmunoassay of pancreatic polypeptide (PP). A specific and sensitive radioimmunoassay described by Chance [10] for pancreatic polypeptide is in use in our laboratory. Antiserum and pure pancreatic polypeptide were kindly donated by Dr. A.E. Chance of Eli Lilly Research Laboratories, Minneapolis, MN, U.S.A. The hormone was radiolabeled with 12~I using chloramine T [24] and purified by ion exchange chromatography on SP Sephadex C25 equilibrated with 0.05 M NaH2PO 4 containing 0.2% human serum albumin (Behringwerke, Marburg, F.R.G.) and irrigated with a gradient of 0-0.5 M NaC1 in the equilibration buffer. All assays are done in duplicate which requires only 0.2 ml plasma. The assay shows no cross-reactivity with gastrin 17 and 34, VIP, GIP, highly purified porcine cholecystokinin, monocomponent human insulin, N- and C-terminal glucagon, motilin or secretin.

33 The experimental detection limit is 5 f m o l / m l with an intrassay variation of 5% and an interassay variation of 11%.

Parietal cell vagotomy with antrectomy A second experiment was designed to test the theory that reinnervation of the parietal cell mass may be the result of ingrowth of nerve fibers originating in the antrum. Eighteen rats underwent PCV as previously described, combined with antrectomy. Groups of 3 rats were sacrificed at 3 days, and at intervals of 3 weeks postoperatively up to 15 weeks. A 1-cm square portion of the fundic stomach was removed at necropsy, cut up into small cubes and processed for electron microscopy.

Results Parietal cell vagotomy

In vitro Light microscopy On light microscopy, nerves were identified in the muscle coat in all the unoperated and sham-operated rats studied, in the muscularis mucosae in 7 animals, and in the mucosae in 3. In the muscle coat, the nerves were clearly visible showing a plexiform arrangement of large-diameter beaded axons. The principal directions of the axons were parallel to the muscle fibers, with branches running between the muscle fibers. The axons passed close to the interphase between the muscle coat and the connective tissue on the epithelial side of the muscle coat, then left the muscle coat to ramify in the connective tissue plane. In the connective tissue spaces on both sides of the muscularis mucosae there were also beaded and unbeaded axons of equal or even slightly greater diameter, which usually ran in pairs beneath the muscularis mucosae. The variability of the beading was striking. Fine nerve fibers were barely discernible in the deepest part of the epithelium (Fig. 2). An additional type of fiber was present in the connective tissue between the muscularis mucosae and the muscle coat which passed at right angles to the muscle coats. These fibers also crossed between the epithelial glands in company with small blood vessels. Between the epithelium and the muscularis mucosae, nerve axons of a fairly uniform diameter were found which were clearly distinguishable from the markedly beaded fibers. Occasionally, from the superficial mucosa, small beaded axons were seen to be terminating on the surfaces of cells or ramifying over their surfaces (Fig. 3). At 1 and 2 weeks following PCV, these nerve fibers could no longer be identified in the muscle coat, muscularis mucosae or mucosae. In view of the greater specificity of electron microscopy, no further studies were done.

34

Electron microscopy E l e c t r o n m i c r o s c o p y s h o w e d that in all u n o p e r a t e d a n d s h a m - o p e r a t e d rats, the n e r v e fibers in the m u c o s a w e r e u n m y e l i n a t e d with c o n t a c t p o i n t s b e i n g f o u n d

Fig. 2. Axons in the connective tissue spaces including submucosa. This shows axons running with blood vessels and indicates the variability of the beaded outlines and unbeaded axons ( × 140).

35

between axons and parietal cells (Figs. 4 and 5). At one week after PCV most of the axons were swollen, evacuated and contained an increased number of fibroblasts. Diameters of the axons were much more variable with a large amount of collagen surrounding Schwann cells. These changes represented degeneration of the nerve

Fig. 3. Fine beaded axons between epithefial cells ( × 288).

36

Fig. 4. A small nerve bundle running in the connective tissue between gastric epithelial cells. Three axons are contained in the cytoplasm of one Schwann cell ( × 16,800).

Fig. 5. Zones of contact (arrowed) between axon and an epithelial cell ( x 60,000).

37

Fig. 6. One week after PCV. A Schwann cell containing swollen axons and fibroblasts indicating degeneration ( x 40,000).

fibers. Also present were isolated and randomly arranged but normal-appearing axons, possibly indicating rapid regeneration (Fig. 6). At 2 and 3 weeks after PCV, the axons were more densely packed with neurofilaments and less swollen. Plenty of collagen was still present around the axon groups. By the fourth and fifth weeks, a similar pattern was seen. There were a greater number of smaller and more normal-looking axons which were more densely packed with neurofilaments. Some swollen axons were still present but contained less fibroblasts in the Schwann cell processes. Enlarged Schwann cell processes were packed with axons in greater numbers than the controls, and these axons appeared squashed rather than the normal circular shape (Fig. 7). In the sixth postoperative week, the Schwann cell processes contained numerous axons. Very few swollen axons or collagen fibers could be found. However, two types of axons of differing diameter were now identified, a large axon which was less densely packed with neurotubules and a smaller axon containing two types of neurovesicles. One was a densely centered neurovesicle. The small axons were densely packed with neurotubules and filaments (Fig. 8). By the seventh, eighth and ninth weeks after PCV, the axons appeared similar to those found in control animals. The Schwann cell processes looked smaller and contained a smaller number of axons. Small nerve bundles containing both types of axons were found close to

38

Fig. 7. Four weeks after PCV. Axons are more densely

packed with neurofilamentsand are more normal

in appearance ( x 9000).

the gastric epithelial cells and the remaining collagen was more mature (Fig. 9). In the tenth and eleventh weeks postoperatively, the axons were still randomly arranged, and there was an increase in the number of vesicles contained. Specimens in the twelfth, thirteenth and fourteenth weeks showed similar features. Both naked axons were found, as were axons complete with a Schwann cell covering. In vivo Gastric secretion (Fig. 10a, b) In sham-operated rats, the mean basal acid output (BAO) was 75 /amol/h and after secretagogue stimulation the acid output had significantly increased to a mean of 164 # m o l / h ( P < 0.001). On recovery from a PCV (day 0), the gastric juice ,was achlorhydric with a basal volume of 0.3 ml/h. Stimulation with gastric secretagogues caused an increase in both volume (0.57 m l / h ) and acidity of gastric juice and the acid output was 26 /xmol/h. On the seventh day after a PCV, the basal and stimulated volume and acidity of gastric juice following secretagogue stimulation increased when compared with the early postoperative results ( P < 0.001). By the second and third weeks after PCV, the stimulated gastric acid output increased

39

Fig. 8. Six weeks after PCV. There are two types of axons, small and large in diameter. Large axons are less denselypacked with neurotubules ( × 7000).

further and continued to increase with each week of testing. By the seventh week, the stimulated acid output was 93% higher than the preoperative level ( P < 0.001) and by the ninth week was 12 times greater than that immediately after the PCV (Table I). On the first postoperative day, the BAO was 4.2 # m o l / h and after stimulation by insulin hypoglycemia was unchanged at 4.4 # m o l / h and was similar to our previously obtained results following a bilateral subdiaphragmatic truncal vagotomy (Fig. 11).

Pancreatic polypeptide (Fig. 12) Basal PP levels in the plasma were similar in unoperated and sham-operated rats on both the day of the insulin hypoglycemia test (mean 27 + 2.2 pg/ml), meal test (mean 26 + 2.4 pg/ml), and after secretin challenge (mean 24 ___3.2 pg/ml). In these same unoperated and sham-operated rats following insulin hypoglycemia and a meal test, there was virtually no rise in plasma PP levels. Following the secretin challenge, there was a mean A% increase of the immunoreactive plasma PP levels by 58% at 2.5 min after the bolus injection of secretin which rapidly returned to normal by 7.5 min. This increment was accounted for by a dramatic rise in plasma PP in only 1 rat, whereas in the remaining 7 rats, there was virtually no

40

Fig. 9. Nine weeks after PCV. A small nerve bundle, normal in appearance containing both small and large axons ( × 15,000).

ioo

BASAL ACID OUTPUT

STIMULATED ACID OUTPUT

~oo[-

-~

TTT

,

T 5c

8

k

Sham ~2y 0~1

50

0 2

3

4 ~ 6 WEEKS

7

8

9 h

b

Sham Do 0

I

2

3

4

.5 6 WEEKS

7

8

9

,

Fig. 10. a: Basal and stimulated acid output after pentagastrin (4 / ~ g / k g / m i n ) with carbachol (0.8 # g / k g / m i n ) infusion in rats after a sham operation, b: immediately after recovery from a PCV (day O) and at weekly intervals up to 9 weeks postoperatively.

58.4 140.1 56.6 5

46.3 26.4 7.5 5

65.6

164.1

65.6 10

( t ~ m o l / h / 1 0 0 g) N u m b e r of rats

74.0 25.4

0 0

75.2 30.7 2.4

41.1

0

50.1

2.5

1.8

0.3

1.5

0.57

1

55.7 5

161.0

64.4

2.5

69.1 24.1

42.2

1.65

2

59.2 5

183.6

68

2.7

70.0 26.0

40.0

1.75

3

Parietal cell vagotomy (postoperative weeks) Day 0

Basal volume ( m l / h ) acid concentration (mmol/1) acid output ( ~ mol/h) (t~ m o l / h / 1 0 0 g) Stimulated volume ( m l / h ) acid concentration (mmol/l) acid output (~tmol/h)

Sham operation

Stimulation with pentagastrin (4 t ~ g / k g / h ) + c a r b a c h o l (0.8 t~g/kg/h).

G A S T R I C S E C R E T O R Y TEST A F T E R P A R I E T A L C E L L V A G O T O M Y IN RATS

TABLE I

87.2 5

242.5

84.8

2.9

78.1 28.1

62.5

1.25

4

85.1 5

262.6

81.3

3.2

79.4 26.8

48.1

1.65

5

89.4 5

316.2

98.2

3.2

84.4 30.3

52.7

1.6

6

79.8 5

317.2

101.7

3.1

88.6 34.3

39.4

2.25

86.4 5

302.2

92.1

3.3

86.3 36.5

41.1

2.1

88.8 5

320.7

94.9

3.4

91.1 39.7

39.6

2.3

42 BASAL

STIMULATED (Insulin hypoglycemia)

6-

5

150

100

3-

2~ii!i!i~i!i E 1

i8

_=

>° o

o

-HSV

Sham

[ ]

Volume

m TV

~-]

~o ~ Sham

Acid Concentration

HSV

B

TV

Acid Output

Fig. 11. Basal and stimulated acid output following insulin hypoglycemia (0.2-0.4 units/kg insulin subcutaneously in sham-operated and after parietal cell vagotomy (HSV) and truncal vagotomy (TV).

change. With these negative results obtained following 3 stimulatory tests prior to a PCV, it was not considered worthwhile to repeat the experiments after the vagotomy.

PCV with antrectomy At 3 days after PCV plus antrectomy, populations of normal and swollen axons were present in mucosa and submucosa. Swollen axons contained relatively few neurotubules, attenuated cytoplasm and phospholipid membranous debris. These axons were interpreted as degenerative. Many Schwann cells ensheathed a mixture of both normal and degenerating axons (Fig. 13). At 3 weeks the number of swollen axons appeared to be decreased. The division of axons into those which were large, less densely packed with neurotubules and those which were small and mo~e densely packed (as seen at 6 weeks in the previous PCV study) was already becoming apparent (Fig. 14). At each interval of 3 weeks, an increasing number of normal appearing axon bundles was noted in the mucosa. These bundles were contained within Schwann cell sheaths. Naked axons, as found in the previous PCV study, were rarely seen in the first 9 weeks of healing. Synaptic contact between nerve

43 PLASMA PANCREATIC POLYPEPTIDE RESPONSE TO PROVOCATION TESTS I00

-

60-

x

I~"

II

SECRETIN

~

.

.

.

~

O / -

O-Jill I

-30

ai I

-15

° I

I

I

0

2.5

5

!

- " - -

I

7.5 IO MINUTE

- " -

O

~

l

l

INSULIN A MEAL

I

I

J

15

30

60

Fig. 12. Plasma immunoreactivepancreatic polypeptide under basal conditions and following stimulation by a mixed meal (A), insulin hypoglycemia(e), and i.v. secretin (0). A% is the percentage increase over basal levels following the provocation test.

fibers and parietal cells was occasionally seen. Synaptic vesicles were clear and similar in morphology to cholinergic transmitter vesicles (Fig. 15).

Discussion This study has shown the sequential neuropathological changes of nerve degeneration, regeneration and functional reinnervation of the gastric mucosa with increased sensitivity of the parietal cell mass after parietal cell vagotomy. The insulin hypoglycemia test confirmed the completeness of vagotomy. The most accurate technique of assessing vagus nerve regeneration in the rat appears to be the response to gastric stimulation by the gastric secretagogues, pentagastrin and carbachol and the serial ultrastructural studies of electron microscopy. It was disappointing to note the failure of measurement of release of immunoreactive pancreatic polypeptide by the 3 provocation tests of insulin hypoglycemia, a mixed meal and i.v. secretin. Either rodents do not respond to these stimuli or the radioimmunoassay is not sensitive enough to detect changes in rodent

44

Fig. 13. Degeneration of axons (*) is evident at 3 days after PCV with antrectomy. Other axons within this Schwann cell appear swollen. Bar represents 1 #m.

45

PP level in the plasma. The assay does, however, measure changes in plasma PP in man, dog and the guinea pig. A PCV is currently the preferred treatment of chronic duodenal ulceration by many centers because of its efficiency in reducing gastric

Fig. 14. The division of axons into small, densely packed with neurotubules and large, less densely packed axons is becoming apparent. Three weeks after PCV with antrectomy. Bar represents 1 pro.

46 secretion without the sequelae of other operative procedures [1,13,56,63]. However, there is an unexplained and increasing incidence of recurrent peptic ulceration with each year of follow-up [4,27,43,51]. D u o d e n a l ulcers can be produced in rats by the infusion of gastric secretagogues, and are prevented by various gastric antisecretory

Fig. 15. Synaptic contact between nerve fibers and basal portion of parietal cell (P). Synaptic vesicles (v) are mostly small and clear-centered. PCV + antrectomy, 9 weeks. Bar represents 0.5 #m.

47

agents, truncal vagotomy and PCV [32,33,35]. In rats, a neurovascular PCV only temporarily prevented experimental duodenal ulcers for 5 days due to a gradual recovery of the parietal cell mass to stimulation [35]. Gastric ischemia could have partially accounted for this recovery. In these experiments, a technique of performing a supraselective vagotomy was developed which only divided vagal nerve fibers to the parietal cell mass while leaving the blood vessels intact. Recovery of gastric secretion has been more gradual with this supraselective PCV and has shown a close relationship to the electron microscopic features of vagus nerve regeneration seen in the gastric mucosa and submucosa just distal to the site of transection of the vagus nerves [38]. After PCV, it has been shown that a number of patients convert from a negative to a positive insulin hypoglycemia test [40,42]. This conversion to a positive test was initially thought not to correlate with increased risk of developing a recurrent peptic ulcer, but more recent evidence suggests that there may be an association [50]. Possibilities for this insulin conversion include vagal nerve regeneration, increased sensitivity of the parietal cells, neuropraxia or reinnervation [20,40]. This study has shown the sequential changes of vagal nerve degeneration and regeneration following a PCV and has attempted to correlate these morphological changes with gastric secretion. If these changes are confirmed in the clinical situation, then it may explain both insulin conversion from a negative to a positive test and the increased incidence of recurrent peptic ulceration. Furthermore, it casts serious doubt on the possible long-term effects of PCV in the treatment of chronic duodenal ulcers. Following PCV and antrectomy, there was an increase in the number of normalappearing axons present in the mucosa, indicating extensive reinnervation. Further investigations are necessary to determine the exact nature and source of these nerve fibers, but it appears that they do not arise as an ingrowth from the antrum.

Acknowledgements Thanks to Dr. K. Tabata and Mr. M. Vorhees for technical assistance.

References 1 Adami, H.O., Enander, L.K., Ingvaar, C. and Rydberg, B., Clinical results of 299 patients with duodenal ulcer 1-6 years after highly selective vagotomy, Brit. J. Surg., 67 (1980) 29-32. 2 Adrian, T.E., Bloom, S.R., Besterman, H.S., Barnes, A.J., Cooke, T.J.C., Russel, R.C.G. and Faber, R.G., Mechanism of pancreatic polypeptide release in man, Lancet, i (1977) 161-163. 3 Amdrup, E. and Jensen, H.E., Selective vagotomy of the parietal cell mass preserving innervation of the undrained antrum, Gastroent., 59 (1970) 522-527. 4 Amdrup, E., Andersen, D. and Hostrup, H., The Aarhus county vagotomy trial. I. An interim report on primary results and incidence of sequelae following parietal cell vagotomy and selective gastric vagotomy in 748 patients, World J. Surg., 2 (1978) 85-90.

48 5 Andersen, D., Hostrup, H. and Amdrup, E., The Aarhus county vagotomy trial. II. An interim report on reduction in acid secretion and ulcer recurrence following parietal cell vagotomy and selective gastric vagntomy, World J. Surg., 2 (1978) 91-100. 6 Bapat, R.D., Ferric, M.M. and Joffe, S.N., A simple technique for gastric parietal cell vagotomy in the rat, Experientia, 33 (1977) 1254-1255. 7 Baron, J.H., Clinical Tests of Gastric Secretion. History, Methodology and Interpretation, MacMillan, London, 1978, 277 pp. 8 Carleton, H.M. and Drury, R.A.B., Histological Technique, Oxford University Press, London, 1957. 9 Chance, R.E. and Jones, W.E., Polypeptide from Bovine, Ovine, Human and Porcine Pancreas, U.S. Patent No. 3,842,063. 10 Chance, R.E., Moon, N.E. and Johnson, M.G., Human pancreatic polypeptide (HPP) and bovine pancreatic polypeptide (BPP). In B.M. Jaffe and H.R. Behrman (Eds.), Methods of Hormone Radioimmunoassay, Academic Press, New York, 2rid Edn., 1979, pp. 657-672. 11 Christiansen, J., Jensen, H.E., Egby-Paulsen, P., Bardram, L. and Henriksen, F.W., Primary results, sequelae, acid secretion, and recurrence rates two to five years after operation, Ann. Surg., 193 (1981) 49-55. 12 Csendes, A., Ornsholt, J., Venturelli, A. and Henriquez, A., Dose response studies of acid secretion after administration of tetragastrin, Amer. J. Surg., 139 (1980) 832-837. 13 Doricott, N.J., McNeish, A.R., Alexander-Williams, J., Royston, C.M.S., Cooke, W.M., Spencer, J., DeVries, B.C. and Muller, H., Prospective randomized multicentre trial of proximal gastric vagotomy or trancal vagotomy and antrectomy for chronic duodenal ulcer: interim results. Brit. J. Surg., 65 (1978) 152-154. 14 Dunn, D.C., Thomas, W.E.G. and Hunter, J.O., An evaluation of highly selective vagotomy in the treatment of chronic duodenal ulcer, Surg. Gynec. Obstet., 150 (1980) 845-849. 15 Faber, R.G., Russel, R.C.G., Parkin, T.V., Whitfield, P. and Hobsley, M., The predictive accuracy of the post vagotomy insulin test, Gut, 16 (1975) 337-342. 16 Faxen, A., Parietal cell vagotomy and selective vagotomy with pyloroplasty, Scand. J. Gastroent., 13, Suppl. 51 (1978) 7-32. 17 Floyd, J.C., Fajans, S.S., Pek, S. and Chance, R.E., A newly recognized pancreatic polypeptide, plasma levels in health and disease, Recent Progr. Horm. Res., 23 (1977) 519-576. 18 Floyd, J.C., Pancreatic polypeptide, Clin. Endocr. Metab., 3 (1979) 279-299. 19 Gaskin, R.J., Gad, A., Barros D'Sa, A.A.J., Joffe, S.N. and Barron, J.H., Natural history and morphology of secretagogue-induced duodenal ulcers in rats, Gastroent., 69 (1976) 903-910. 20 Gillespie, G., Elder, J.B., Gilliespie, I.E., Kay, A.W. and Campbell, E.H.G, The longterm stability of the insulin test, Gastroent., 58 (1970) 625-632. 21 Goligher, J.C., Hill, G.L., Kenny, T.E. and Nutter, E., Proximal gastric vagotomy without drainage for duodenal ulcer: results after 5-8 years, Brit. J. Surg., 65 (1978) 145-151. 22 Grassi, G., Orrechia, C. and Sbuelz, B., Early results of the treatment of duodenal ulcer by ultraselective vagotomy without drainage, Surg. Gynec. Obstet., 136 (1973) 726-728. 23 Griffith, C.A. and Harkins, H.N., Selective gastric vagotomy: physiological basis and technique, Surg. Clin. North Amer., 42 (1962) 1431-1441. 24 Hallenbeck, G.A., Gleysteen, J.J., Aldrete, J.S. and Slaughter, R.L., Proximal gastric vagotomy: effects of two operative techniques on clinical and gastric secretory results, Ann. Surg., 184 (1976) 435-442. 25 Hancock, D.M., Sankar, M.Y., Old, J.M., Bose, A.A.K., Punnen, P.C., Mishra, S.M., Lobo, F.X. and Trinder, P., The combination of proximal gastric vagotomy with a rotational gastropexy for duodenal ulcer, Brit. J. Surg., 65 (1978) 706-711. 26 Hansen, O.H., Larsen, J.K. and Svendsen, L.B., Changes in gastric mucosal cell proliferation bafter antrectomy or vagotomy in man, Scand. J. Gastroent., 13 (1979) 947-952. 27 Hedenstedt, S. and Lundquist, G., Selective proximal vagotomy (SPV) as an emergency and definitive operation for massive ulcerous bleeding, Acta Chit. Scand., 144 (1978) 241-248. 28 Holle, F. and Hart, W., Nene wege der Chirurgie des gastroduodenalukus, Med. Kin., 62 (1967) 441-450. 29 Hunter, W.M. and Greenwood, F.C., Preparation of iodine -131, human growth hormone of high specific activity, Nature (Lond.), 194 (1962) 495-496.

49 30 Joffe, S.N. and Ferrie, M.M., Vagotomy and experimental duodenal ulcers, Acta Hepato-gastroenterol., 23 (1976) 435-440. 31 Joffe, S.N., Gaskin, g.J., Barros D'Sa, A.A.J. and Baron, J.H., Prevention of secretagogue-induced ulcers in the rat, Gastroent., 70 (1976) 751-752. 32 Joffe, S.N., Polak, J.M., Pallone, F., Bloom, S.R., Gaskin, R.J., Barros D'Sa, A.A.J. and Baron, J.H., Gastrin, antral G cells, and gastric acid in secretagngne-induced and antihistamine-inhibited duodenal ulcers, Scand. J. Gastroent., 12 (1977) 27-32. 33 Joffe, S.N., Gaskin, R.J., Barros D'Sa, A.A.J. and Baron, J.H., Secretagogne-produced duodenal ulcers in the rat, Brit. J. Surg., 64 (1977) 218-220. 34 Joffe, S.N. and Fettle, M.M., The effect of GIH and Boots secretin and cholecystokinin on duodenal ulcers in the rat, Acta Hepato-gastroenterol., 24 (1977) 458-461. 35 Joffe, S.N. and Bapat, R.D., The temporary effect of proximal gastric vagotomy on experimental duodenal ulcers and gastric secretion, Brit. J. Surg., 66 (1979) 234-237. 36 Joffe, S.N. and Primrose, J.N., A prospective study evaluating preoperative gastric secretion and choice of an operation for duodenal ulcer. Surg. Gynec. Obstet., 152 (1981) 421-423. 37 Joffe, S.N. and Crocker, A., The effect of parietal cell vagntomy and truncal vagntomy with and without drainage on duodenal ulcers, Digestion, 22 (1981) 47-65. 38 Joffe, S.N., Crocket, A. and Doyle, D., Morphological and functional evidence of reinnervation of the gastric parietal cell mass following a parietal cell vagotomy, Amer. J. Surg., 143 (1982) 80-85. 39 Johnston, D. and Wilkinson, A.R., Highly selective vagotomy without a drainage procedure in the treatment of duodenal ulcer, Brit. J. Surg., 57 (1970) 289-295. 40 Johnston, D., Wilkinson, A.R., Humphrey, C.S., Smith, M.B., Goligher, J.C., Kragelund, E. and Am&up, E., Serial studies of gastric secretion in patients after highly selective (parietal cell) vagotomy without a drainage procedure for duodenal ulcer: II. The insulin test after highly selective vagotomy, Gastroenterology, 64 (1973) 12-21. 41 Johnston, D., Operative mortality and postoperative morbidity of highly selective vagotomy, Brit. Med. J., 4 (1975) 545-547. 42 Jordan, P.H., Our first 35 patients studied five years after parietal cell vagotomy, Arch. Surg., 114 (1979) 528-535. 43 Kennedy, T. and Spencer A., Cimetidine for recurrent ulcer after vagotomy or gastrectomy; a randomised controlled trial, Brit. Med. J., 1 (1978) 1242-1243. 44 Kimmel, J.R., Pollock, H.G. and Hazelwood, R.L., Isolation and characterization of chicken insulin, Endocrinology, 83 (1968) 1323-1330. 45 Kimmel, J.R., Hayden, L.J. and Pollock, H.G., Isolation and characterization of a new pancreatic polypeptide hormone, J. biol. Chem., 250 (1975) 9369-9376. 46 Kjaergaard, J., Jensen, H.E. and Allermand, H., Inadequately reduced gastric secretion after vagotomy for duodenal ulcer, Ann. Surg., 192 (1980) 711-715. 47 Kronborg, O., Jorgensen, P.M. and Holt-Christensen, J., Influence of different techniques of proximal gastric vagotomy upon risk of recurrent duodenal ulcer and gastric acid secretion, Acta Chir. Scand., 143 (1977) 53-56. 48 Larmi, T.K.I. and Kairaluoma, M.I., Phenylbutazone ulcer and vagotomy in the rat, Scand. J. Gastroent., 8 (1973) 307-311. 49 Lin, T.M., Pancreatic polypeptide: isolation, chemistry and biological function. In G.B. Jerzy Glass (Ed.), Gastrointestinal Hormones, Raven Press, New York, 1980, pp. 275-306. 50 MacLeod, R.S. and Cohen, Z., Highly selective vagotomy and truncal vagotomy and pyloroplasty for duodenal ulcer: a clinical review, Canad. J. Surg., 22 (1979) 113-120. 51 Madsen, P. and Kronborg, O., Recurrent ulcer 5½-8 years after highly selective vagotomy without drainage and selective vagotomy with pyloroplasty, Seand. J. Gastroent., 15 (1980) 193-199. 52 Maybury, N.K., Faber, R.G. and Hobsley, M., Postvagotomy insulin test: improved predictability of ulcer recurrence after correction for height and collection errors, Gut, 18 (1977) 449-456. 53 Pedrazzoli, S., Brandsetter, S. and Dod, G., Peptic ulcer, what can we expect from various diagnostic tests, Chir. Triveneta, 19 (1979) 198-202. 54 Primrose, J.N., Ratcliffe, J.G. and Joffe, S.N., Assessment of the secretin provocation test in the diagnosis of gastrinoma, Brit. J. Surg., 67 (1980) 744-746.

50 55 Robert, A., Stowe, D.F. and Nezamis, J.E., Prevention of duodenal ulcers by administration of prostaglandin E2, Scand. J. Gastroent., 6 (1971) 303-305. 56 Sawyers,J.L., Herrington, J.L. and Burney, D.P., Proximal gastric vagotomy compared with vagotomy and antrectomy and selectivegastric vagotomy and pyloroplasty, Ann. Surg., 186 (1977) 510-517. 57 Schwartz, T.W., Rehfeld, J.F., Stadil, F., Larson, L.I., Chance, R.E. and Moon, N., Pancreatic polypeptide response to food in duodenal ulcer patients before and after vagotomy, Lancet, i (1976) 1102-1105. 58 Schwartz, T.W., Hoist, J.J., Fahrenkrug, J., Jensen, S.L., Nielsen, O.V., Rehfeld, J.F., MuckadeL O.B.S., Stadil, F., Vagal, cholinergic regulation of pancreatic polypeptide secretion, J. clin. Invest., 61 (1978) 781-789. 59 Shay, H., Kormarox, S.A. and Gruenstein, M., Effect of vagotomy in the rat, Arch. Surg., 59 (1949) 210-226. 60 Stem, A.I., Hansky, J., Korman, M.G., Coupland, G. and Waugh, J., Pancreatic polypeptide release following surgery for duodenal ulcer disease, Dig. Dis. Sci., 25 (1980) 485-488. 61 Taylor, I.L., Feldman,: M., Richardson, C.T. and Walsh, J.H., Gastric and cephalic stimulation of human pancreatic polypeptide release, Gastroenterology, 75 (1978) 432-437. 62 Valenzuela, J.E., Taylor, I.L. and Walsh, J.H., Pancreatic polypeptide response in patients with chronic pancreatitis, Dig. Dis. Sci., 24 (1979) 862-864. 63 Wastell, C., Colin, J., Wilson, T., Walker, E., Gleeson, J. and Zeegen, R., Prospectively randomised trial of proximal gastric vagotomy either with or without pyloroplasty in treatment of uncomplicated duodenal ulcer, Brit. Med. J., 2 (1977) 851-853.

Discussion

Dr. Kral: W h y a r e n ' t the recurrence rates of ulcers in m a n over 5 - 9 years more than 15% if nerve regeneration occurs?

Dr. Joffe: I n m a n I d o n ' t think the regeneration process takes place as rapidly as in other m a m m a l s . Dr. Sclafani: Do you think regeneration is responsible for the return of acid secretion i n the second week? Dr. Joffe: I d o n ' t know. I think it is unlikely, b u t it seems as if these two processes are going o n simultaneously. Dr. Norgren: D e g e n e r a t i o n has not occurred completely in that period of time. Even in the b r a i n it d o e s n ' t work that fast. Dr. Nyhus: Have you d o n e any tests on PCV patients with recurrence who have been reoperated? Dr. Joffe: W e have looked at the mucosa of a few patients when they have had antrectomies following recurrences. There are nerves present, b u t I c a n ' t say whether others have come a n d how m a n y m a y have gone. Dr. Novin: W o u l d n ' t a more direct test of stimulated acid o u t p u t be to put electrodes o n the vagus nerve above the cut a n d see if acid o u t p u t follows the same time-course? Have you tried that? Dr. Joffe: I agree, b u t we h a v e n ' t done it. Dr. Ahlman: H o w can you prove that you are studying vagal fibers? It m a y be i m p o r t a n t that you have intact sympathetic fibers. After vagotomy it is a well-known fact that i n various organs, one gets a c o m p e n s a t o r y h y p e r t r o p h y of sympathetic nerves.

51

Dr. Nyhus: But adrenergic reinnervation would inhibit the acid output. Dr. Koopmans: Would you expect that the adrenergic nerves would make contact with parietal cells? Dr. Joffe: I ' m sure they can make contact with parietal cells, but in the vagotomy and antrectomy the nerves studied have had the structural features of cholinergic fibers. Dr. Sawchenko: Does the rate of functional recovery depend on the extent or the level of your vagotomy? Dr. Joffe: I don't know, since the vagotomy is just done by dividing the fibers distal to the nerve of Latarjet. Dr. Brooks: You spoke about the 15% recurrence rate of PCV. Is that what you expect in a regenerating nerve, that one will get only a functional junction in 15% of the cases? This needs to be answered. Dr. Joffe: Maybe the sympathetic cell regeneration is converted to cholinergic by the environmental proximity of the right receptors. So it is not the vagus but the sympathetic fibers which haven't been cut that are having the effect via a recruitment of function. A regenerating cell is very much like a cell in tissue culture - - it can be modified. Dr. Schwartz: I can relieve you of your doubts about pancreatic polypeptide assay: it cannot be done adequately in the rat. Dr. Olbe: I was puzzled about the quick return of acid secretion. What is the effect of truncal vagotomy? Do you have this increased acid response to microstimulation in the truncal vagotomies too? Dr. Joffe: We don't see an increase in the stomach, but you get it in other organs, denervation hypersensitivity is a well-known phenomenon. Dr. P. Smith: We tried to denervate the pancreas of the dog by cutting the nerves following the vascular supply. I found that there were still plenty of nerve fibers present - - the problem probably was that we simply could not get all the fibers. H o w confident are you that there are not not many smaller nerve branches coming off the vagus that you just can't identify with a dissecting microscope and that kind of operation? Dr. Joffe: We're not too certain but in this operation one totally frees the nerve of Latarjet. Also, a specimen of gastric wall is taken just distal to where that main trunk comes and is divided.