long-term results of selective vagotomy plus pyloroplasty: 12 to 17 year follow-up

SCIENTIFIC PAPERS

Long-Term Results of Selective Vagotomy Plus Pyloroplasty

12 to 17 Year Follow-Up

Charles A. Griffith, MD, MSc, FACS, Bellevue, Washington

We previously reported one recurrence among 103 patients followed up for 4 to 9 years after selective vagotomy plus pyloroplasty for duodenal ulcer [I]. This low recurrence rate of 0.97 percent was attrib- uted to consistently successful results of complete or adequate vagotomy provided by the anatomic accuracy of the selective technique, and also to preservation of the inhibition of gastric secretion by the hepatic and celiac vagi. However, after a longer follow-up of 12 to 17 years, five patients with negative insulin tests developed ulcerations (one stoma1 ulcer, two gastric ulcers and two cases of hemorrhagic gastritis). The present report describes these ulcerations that come to light only in the long term from causes other than incomplete vagotomy. Also, in view of the issue of the formation of gallstones after vagotomy, the results of cholecystography are presented.

Material and Methods

In the 8 years since the previous report, nine patients died of causes unrelated to ulcer and seven could not be traced. None of these patients had any signs or symptoms of recurrence before dying or being lost to follow-up 8 to 15 years after operation. The remaining 87 patients have been followed up 12 to 17 years.

Selective gastric vagotomy was done by the technique illustrated in Figure 1. Ligation and transection of the descending branch of the left gastric artery are essential to assure completeness of vagotomy. Failure to adhere to

From the Experimental Laboratory of the University of Washington School of Medicine, Seattle; and the Surgical Department of Overlake Hospital, Bellevue, Washington.

Requests for reprints should be addressed to Charles A. Griffith, MD, 12806 N.E. 34th Place, Bellevue, Washington 98005.

Presented at the 66th Annual Meeting of the North Pacific Surgical As- sociation, Portland, Oregon, November 9-10, 1979.

608

this detail by an associate not then convinced of its importance resulted in the previously reported recurrent ulcer. Pyloroplasty was done by Finney’s method [4].

Pre- and postoperative basal secretion and insulin tests were performed as previously described [5]. Basal acid output was expressed as milliequivalents of free acid per hour.

Preoperative cholecystography was performed in 73 patients. Gallstones were found in six, each of whom underwent cholecystectomy in addition to selective vagotomy plus pyloroplasty. Cholecystography was repeated during the 4th to 7th postoperative years in 53 of the remaining 67 patients. An additional 13 postoperative cholecystograms were obtained in patients who did not undergo preoperative cholecystography.

Results

Basal Secretions

Basal acid output was determined in 90 of the 103 patients. The degree of reduction and the postoperative amount of basal acid output varied. Using basal acid output of 1.5 to 2 mEq as the upper limit of normal secretion, the patients were divided into three groups.

Group 1 (83 patients): Hypersecretion usually associated with duodenal ulcer was not found in 12 patients. In the other 71 patients the preoperative basal acid output ranged from 2.6 to 15.4 mEq. Va- gotomy resulted in no free acid in 52 patients and normal levels of free acid less than 1.5 mEq in 31 patients.

Group 2 (4 patients): Basal acid output ranging

from 4.2 to 7.8 mEq was reduced to high normal

levels of 1.7 to 1.9 mEq.

Group 3 (3 patients): Basal acid output was not reduced. Preoperative hypersecretion of 2.7,3.2 and

The American Journal of Surgery

Selective Vagotomy Plus Pyloroplasty

8.9 mEq persisted after vagotomy. The vagotomy in Interpretation: Insulin tests were performed in the patient with persistent hypersecretion of 8.9 mEq 78 patients. One test was discontinued in a patient was deemed complete at operation by the Burge test who had a severe hypoglycemic reaction that re- [6]. This test was not performed in the other two quired cardiopulmonary resuscitation for recovery. patients.* The results of insulin tests were nega- Nine other tests were vitiated by bile in the gastric tive. specimens or inadequate hypoglycemia. Valid results

Insulin Tests

Vagotomy: all-or-none or segmental dener- vation: Interpretation by Hollander’s criteria for a positive result is often borderline and sometimes meaningless, primarily because Hollander based his criteria on the fallacy that vagotomy is an all-or-none phenomenon. Dragstedt et al [7] once wrote that any small vagal fiber “appears to be able to activate the entire glandular apparatus,” presumably v,ia Auer- bath’s and Meissner’s plexuses, and Hollander and his co-workers [8] stated that there are “no degrees of vagality.” We challanged their premise with the idea of parietal cell vagotomy in the mid-1950s,* but could not refute it until the mid-1960s when we proved that vagal innervation of the stomach is segmental by the method of electric vagal stimulation in the presence of circulating neutral red [9].

Degrees of vagality and adequacy of vagotomy: The method using neutral red provided accurate delineation of various-sized areas of residually innervated gastric mucosa and, for the first time, en- abled study of the secretory and ulcerogenic potentials of known amounts of residually innervated mucosa. The various anatomic types of incomplete vagotomy, which can and do occur, resulted in various-sized areas of residually innervated mucosa. The size of the innervated area of mucosa was directly correlated with its secretory and ulcerogenic potential (Figure 2). Furthermore, this correlation was distinctly manifested by the gastric secretory response to insulin (Figure 3). Thus, contrary to Dragstedt’s and Hollander’s premise of an all-or- none phenomenon with no degrees of vagality, our findings showed that there are indeed degrees of vagality. These various degrees of vagality were found to support the concept of the adequacy of vagotomy proposed by Ross and Kay [15].

* We started evaluation of the Burge test in 1962. when selective vagotomy plus antrectomy was the preferred operation. The test was continued into the early part of this study but then abandoned because, when vagotcmy was deemed complete on the basis of the accuracy of the anatomic dissection, the Burge test was also negative Only six patients in the study underwent this test.

+ After we submitted our experimental results of parietal vagotomy to Dragstedt and Hollander, both expressed disbelief that anything short of complete vagotomy of the stomach could succeed. Their replies were a chief deterrent to our clinical application of parietal vagotomy 25 years ago.

Figure 1. Completed dissection for selective vagofomy. The anterior gastric vagal division has been transected just below the or&in of the hepafic vagi, which are shown wlfhin the lesser omentum. The posterior trunk continues as the celiac vagal division, and is shown posteriorly to the pa- fienf ‘s right of the esophagus. Adjacent to the lesser curve are the ligated ends of the descending branch of the left gastric artery, which has been divided in order to transect those gastric vagi that arise from the distal posterior trunk-celiac division in 10 to 15 percent of bodies and ac- company the left gastric artery to innervate the posterior wall of the stomach [ 31. Other posterior gastric vagi have been divided by ligation and transection of the proximal lesser omenturn, so that the posterior trunk-celiac division is separated completely from the lesser curve and distal esophagus. The longitudinal muscle -of the distal esophagus is bared in ifs circumference to insure transection of any remaining small fibers to the proximal stomach (such as the so-called criminal nerve of Grassi), which originate independently from the anterior and posterior gasfric truncal divisions or the trunks or the esophageal plexus above this field of dissection. The esophageal hiatus is purposely omifted to emphasize that, in contrast to Dragstedf’s hlatal technique of fatal abdominal vagofomy, the selective fechnique is done af the lower level of the distal esophagus and gastric cardia. (Reproduced from Griffith CA 121, wifh permission of the publisher. )

Volume 139, May 1960 609

Griffith

in the remaining 68 patients were interpreted by criteria derived from Figures 2 and 3.

Only one large and early response indicative of inadequate vagotomy occurred, and this was in the previously reported patient with recurrent ulcer. Six other patients had very small and late responses which, rather than appearing in the 2nd hour as in Figure 3, did not appear until the 3rd and 4th hours after insulin. These responses were interpreted as either adequate vagotomy or, more probably, complete vagotomy with an adrenergic release of gastrin consequent to hypoglycemia [16]. The remaining 61 patients had negative insulin tests. Included among these’ patients with negative tests were the four group 2 patients with postoperative basal acid output in the upper limits of normal and the three group 3 patients with persistent hypersecretion. The group 3 patients had two insulin tests each, and both tests were negative in all three patients.

Recurrences

Recurrences included a stoma1 ulcer in one patient and gastric ulceration in four others; all five patients

Figure 2. Correlation of anatomic fype of incomplete vagofomy with adequacy. Leff, intact vagal trunk; vagofomy inadequate. With one intact vagai trunk the ipsiiaferai wall of fhe stomach remains innervafed and continues to secrefe (dark shading). The vagai release of gasfrin causes a delayed and lesser secretion from the denervafed confraiaf- erai wail ( iighf shading). Secrefion remains high and ulcers develop in 75 percent of Shay rats [ IO]. This incomplete vagofomy affords inadequafe protection against recurrenf ulcer. The same secretion occurs affer incomplete vagofomy of an intact branch of fhe esophageal plexus confrib- ufing to a vagai frunk, as in Figure 4. Righf, intact terminal branch; vagofomy adequate. With an intact terminal branch from a nerve of Lafarjef to the fundus, only a small area of gastric mucosa remains innervated. The amount of secre- f/on is only slighfly greater fhan with complete vagofomy, and no ulcers deveiop in Shay rats [ 701. This incomplete vagofomy affords adequate profecfion against ulcer. Affer 1 year the area of innervated mucosa did nof increase in size by reinnervafion due fo sprouting, and the amounf of secrefion remained the same [ 1 I- 131. Thus this adequate incomplete vagofomy is permanently adequate. in addifion to confirming the concept of adequacy of vagofomy, fhese experimenfs provided the physiologic basis of pariefal vagofomy. (Reproduced from Pritchard et al [ 91, with permission of fhe publisher. )

had negative insulin tests (Table I). Patient I was one of the three persistent hypersecretors in group 3. The other two hypersecretors have had no signs or symptoms of recurrence after 14 and 17 years. Case II was associated with gastric stasis due to narrowing of the stoma from marked angulation by adhesions between the pyloroplasty and liver. Cases IV and V

60

50

40

30

20

IO

0

Time in hours

z(4 weeks post op.)

Time in hours

0 I 2 3 4

Time in hours

3 (5 weeks post op.)

t \

\ \

\ .

Lb!%

i

I

-I 0 I 2 3 4 Time in hours

g 3 (3 weeks post op.) 3 (4 weeks post op)

c

Time in hours

-\ \ \ \ \ I i L

‘/

l--zLLh -I 0 I 2 3 4

-Time in-hours

Figure 3. Acid secretory responses to insulin in dogs [ 141 with inadequate and adequate incomplete vagofomy. Top, the large and early response wifhfn fhe 1st hour affer insulin injection before vagofomy (two fesfs). Middle, fhe large and early response persists after inadequate vagofomy of one infacf trunk (compare with Figure 2, /eff). Bottom, fhe small and delayed response in the 2nd hour after incomplete adequate vagofomy of an infacf proximal branch from the nerve of Lafarjef to fhe fundus (compare with Figure 2, righf ). (Reproduced with permission from: Griffith CA. Surgery of the stomach and duodenum, 3rd ed, chap 12. Bosfon: Little, Brown, 1977.)

610 The American Journal of Surgery

had excessive amounts of bile in the stomach seen by

gastroscopy and at operation. Biliary reflux was not

demonstrated in case III, but no other cause of gastric ulcer was apparent. Antrectomy was done in all patients except patient IV, whose hemorrhagic gastritis from biliary reflux was successfully managed by cholestyramine. Antrectomy was successful in three

of the four patients. The fourth patient (case V) bled from hemorrhagic gastritis seen by gastroscopy just proximal to the Billroth I anastomosis. The stomach

again contained excessive bile. Bleeding was con- t,rolled by cholestyramine. This is the only patient in

our entire experience who had clinically significant hemorrhagic gastritis after antrectomy complete by microscopic examination and vagotomy complete by

negative insulin testing.

Cholecystography

Comparison of the pre- and postoperative cholecystograms showed no change in the size or con- tractility of the gallbladders. Asymptomatic gall-

stones were found on 2 of the 66 postoperative cholecystograms. Another 2 of these 66 patients later

developed symptomatic gallstones found by repeat

cholecystography. One of the 37 patients who did not undergo cholecystography developed acute chole- cystitis due to stones. All five patients underwent

cholecystectomy and had an uneventful recovery. No other patient has had any signs or symptoms of disease of the biliary tract.

Dumping and Diarrhea

Statistical analysis of these sequelae was not at-

tempted because of their many degrees of severity and also because of the subjectivity of the patients and the surgeon (author). In general, the spectrum

ranged from a few completely satisfied patients with no symptoms to a few totally dissatisfied patients with symptoms more distressing and disabling than their preoperative symptoms of ulcer. In between


these extremes was the majority of patients with

symptoms that varied from mild to moderately se-

vere. Some moderate and severe episodes of dumping with intestinal hurry culminated in one or more ex- plosive diarrhea1 stools. This diarrhea associated with

dumping was distinguished from so-called postvag-

otomy diarrhea, which is characteristically episodic and, unlike diarrhea with dumping, is neither post-

cibal nor related to the type of food or liquid con- sumed. Furthermore, postvagotomy diarrhea occurs only after total abdominal vagotomy and not after

preservation of the hepatic and celiac vagi by selec-

tive or parietal vagotomy. Thus, posthepatic-celiac diarrhea did not occur in this series.

Comments Secretory Tests

The variable effect of complete vagotomy on basal acid output: Complete vagotomy reduces basal acid output effectively in the great majority of patients (such as group 1) but not in a small minority

(such as groups 2 and 3). Of particular note is group 3 with persistent hypersecretion, which was first described by Dragstedt’s group [17]: “There is a small

group of patients who although they have persis- tently negative reactions to insulin tests, maintain

a continued high night secretion. Three patients in this series of 70 demonstrated this phenomenon. . . a matter which is not as yet entirely clear.”

Gelb et al [18] and Gillespie and Kay [19] confirmed the occurrence of this type of hypersecretion and found that it was abolished by antrectomy. They

therefore proposed some type of antral hyperfunc-

tion. More recently it has been shown that this antral hyperfunction may occur with or without G-cell hy-

perplasia and that it is an entity apart from the Zol- linger-Ellison syndrome [20]. Thus, contrary to previous belief, persistent hypersecretion from antral hyperfunction disproves the thesis that all ulcers will

heal and remain healed if the vagotomy is complete and if the drainage procedure drains.

TABLE I Recurrences in Long-Term Follow-Up (All Insulin Tests Negative)

Case No. Free BAO (mEq/hour)

Preop Postop Time (yr) Recurrence

Type Cause

I 2.7 2.8 6 II 1.8 0.1 a III 3.8 1.3 9 IV 3.1 1.4 IO V 5.8 0:o 12

* See text. + Probably biliary reflux. BAO = basal acid output; postop = postoperative; preop = preoperative.

Stoma1 Hyperantrum’ Gastric Stasis Gastric Uncertain+ Bleeding gastritis Biliary reflux Bleeding gastritis Biliary reflux

Volume 139, May 1980 611

Griffith

Prognostic value: The preoperative basal acid output in several group 1 patients was significantly greater than in groups 2 and 3. The finding that basal acid output was decreased to anacidity or hypose- cretion in group 1, but not in groups 2 and 3, adds to the evidence that preoperative basal acid output is of limited value in predicting the postoperative value and the need for the addition of antrectomy. The same holds true for maximum or peak acid output of stimulated secretion. Nevertheless, on the basis that ulcers have recurred only when stimulated preoperative output is above a certain level, antrectomy has been recommended if output exceeds this level. However, most patients with an output above the critical level have not had recurrence, and in these patients antrectomy has proved unnecessary.

The postoperative basal acid output and insulin test are also of limited value in predicting recurrence. The reason for the absence of recurrence in two of the group 3 patients with persistent hypersecretion and the recurrence in the other patient (case I) is not known. On the other hand, low or normal basal acid output and negative insulin tests did not prevent gastric ulceration in cases III, IV and V.

The insulin test is valuable in grading the efficacy of the surgeon and also in determining the type of operation for recurrence. A large early response in- dicates completion of the vagotomy, whereas a small late response does not. However, insulin tests are hazardous. The near-fatal hypoglycemic reaction in this study occurred in an apparently physically fit patient. The insulin infusion test is less hazardous but may also be contraindicated in poor-risk patients. A safer and perhaps more informative method is seeing through a gastroscope the areas that secrete neutral red, as done but not published by Pritchard [9] and Jones and Griffith [II], and 1aterlreported:by Kusakari’s group with Congo red [21].

Vagotomy at the Hiatus or at the Distal Esophagus

The only recurrence due to incomplete vagotomy in this study was in the patient previously reported who, as aforementioned, did not undergo the technique illustrated in Figure 1 (the descending branch of the left gastric artery was not divided). This low rate of recurrence, plus the 61 negative insulin tests and 6 negative or adequate tests, supports our re- peatedly emphasized contention that the cardinal advantage of the technique at the distal esophagus over Dragstedt’s technique at the hiatus is the anatomic assurance that vagotomy is complete. Full appreciation of this advantage requires recognition of the different anatomic patterns and positions of

the vagi at the hiatus compared with the distal esophagus.

Weinberg [22] modified Dragstedt’s hiatal technique as follows: “The approach is made through an incision in the hiatal membrane just anterior to the esophagus instead of through the diaphragm anterior to the hiatal ring. . . . The vagus nerves are divided several centimeters above the junction of the esophagus and stomach where the nerves are compactly arranged and accessible for complete inter- ruption.” Herein lies the anatomic fallacy and the surgical flaw: the nerves are not compactly arranged and accessible at the hiatus. Rather than two and only two trunks, the vagi at the hiatus may exist as multiple fibers of either the esophageal plexus or the peripheral truncal divisions. Of more importance, one or more of these fibers may lie in variable positions closer to the margins of the hiatus and aorta than to the esophagus. In a dissection confined to the area of the hiatus, the surgeon does not know how many nerves there are or where they are. Under these cir- cumstances the occurrence of incomplete vagotomy is inevitable (Figure 4).

In a follow-up study 10 to 18 years after this hiatal technique of total vagotomy, the recurrence rate was 27.3 percent [23]. Dragstedt’s long-term recurrent rate has never been reported, but at 10 years or less was 10.6 percent [24]. Dragstedt et al [25] wrote: “The percentage of incomplete vagotomies steadily decreases as the surgeon gains experience, but is not eliminated by present techniques, and this consti- tutes the chief weakness in this method of surgical therapy.”

The “chief weakness not eliminated by present techniques” was eliminated in this study by a different technique which, in 1964, was called “a new anatomic approach to the problem of incomplete vagotomy” [26]. This approach takes advantage of the fact that the area at the most distal esophagus is the one and only place where all gastric vagi gather to innervate the stomach. Therefore, all gastric vagi may be encircled with anatomic certainty at the distal esophagus (Figures 1 and 4). This certain en- circlement of all gastric vagi is the key to eliminating inadequate incomplete vagotomy. Application of this basic surgical anatomic fact, rather than any particular surgical skill, accounted for the consistently successful results of complete or adequate vagotomy in this series.

McVay’s group [27] recently reported a 4.1 percent rate of proven plus suspected recurrence after total vagotomy. Although the follow-up period averaged only 8.5 years and only 40 percent of the patients were observed 10 years or longer, McVay’s surgical



anatomic mastery may well account for a continued

low recurrence rate. If so, his results cast doubt upon

the clinical significance of any gastrone liberated by the hepatic and celiac vagi. At any rate, the vague and usually undocumented recurrent rate of 10 percent generally quoted could be lowered if surgeons lowered their sights from the hiatus to the distal

esophagus, where either selective or total vagotomy may be performed and equally low rates of incom-

plete vagotomy obtained. Therefore, the issue of completeness of vagotomy is of no concern in the controversy of selective versus total vagotomy. What

is of concern is eliminating inadequate vagotomy by

operating at the distal esophagus instead of the hiatus.

Reflux Gastritis and Gastric Ulcer

ln addition to Dragstedt’s thesis of gastric stasis as a cause of gastric ulcer (case II), there is now ample

evidence that duodenal reflux into the stomach also causes severe gastritis and ulcer. The most convincing demonstration of duodenal reflux through an

incompetent pylorus is the cineradiographic test by Ki’lby of Capper’s group [28]. Bile unquestionably causes gastric ulcer in animals [29,30]. Silen [31] has summarized the evidence for breakdown of the so-

called mucosal barrier by bile with consequent gastritis or ulcer. No cause other than biliary reflux seems feasible for explaining the hemorrhagic gas-

trit is in cases IV and V and probably the gastric ulcer in case III. Biliary reflux accounted for more gastric complications than any other cause in this study and was seen only after long-term observation.

Gailstones After Vagotomy

Two changes account for the increased formation

of gallstones after total vagotomy. The first is stasis of bile in the dilated gallbladder, which was first described by Johnson and Boyden [32]. This dilata- tion of the gallbladder does not occur after selective vagotomy [33]. Second, a lithogenic change occurs in the biochemical composition of bile. The evidence from studies in dogs is conflicting, but the lithogenic change in monkeys is convincing [34].

In 1952 Johnson and Boyden first warned of gallstones after total vagotomy. Their warning was not generally heeded but, has since been confirmed by several investigators [35-381 who have found increased rates of gallstones ranging from 16.3 to 22.8

percent. These rates are 4 to 6 times the rate expected from the Framingham study [39]. In contrast, the rate of gallstones does not increase after selective VagotcJmy.

Volume 139, May 1980

Figure 4. Exclusion of vagal fibers at the hiatus, the most common cause of inadequate incomplete vagotomy. Left, the anterior vagal trunk or a large contributing branch of the esophageal plexus may lie well away from the esophagus adjacent to the lett hiatat margin. ftight, the posterior trunk or a large contributing branch of the esophageal plexus may lie well away from the esophagus adjacent to the right di- aphragmatic crus and aorta. The encircling finger at the hiatus ( arrow) palpates one of the two branches of the esophageal plexus, assumes it is the trunk, and excludes the other esophageal plexus branch from the encircled esophagus. This technical error is eliminated by using three constant anatomic landmarks at the distal esophagus: the card&esophageal angle of His, the celiac vagal division and the hepatic vagi. The encircling finger (arrow) includes all gastric vagi because none lie to the patient’s left of the angle of His or posterior to the celiac division or to the patient’s right of the avascular area in the lesser omentum below the hepatic vagi. (Reproduced with permission from: Griffith CA. Abdominal surgery, 6th ed, chap 15. New York: Appleton-Century-Crofts, Hall, 1974. )

Publishing Division of Prentice-

Dumping and Diarrhea

Dumping with and without associated diarrhea occurred as frequently and as severely as previously experienced with antrectomy. To prevent dumping and also to decrease basal acid output more effec-

tively, pyloroplasty was abandoned in favor of Maki’s pylorus preserving antrectomy to complement selective vagotomy in 1968. Initial experience with this procedure, with and without pylorotomy, has been reported [40]: No ulcer has recurred to date, but pylorotomy has resulted in some dumping. However,

the frequency and severity of dumping is significantly less with partial preservation of pyloric function by pylorotomy than with complete destruction of pyloric function by pyloroplasty or antrectomy. These results differ from the results of parietai cell vagotomy, which has virtually eliminated dumping but not recurrent ulcer. A small but significant rate of recurrence is inevitable because of antral hyperfunction and, in patients with extremely small antra, incomplete vagotomy of the parietal cell mass distally de-

613

Griffith

spite complete vagotomy proximally. Thus the problem of curing the diathesis of duodenal ulcer without creating the dumping syndrome remains unsolved.

Summary

Among an initial series of 103 patients with selective vagotomy plus pyloroplasty for duodenal ulcer, 9 patients died of causes unrelated to ulcer and 7 were lost to follow-up without signs or symptoms of ulcer 8 to 15 years after operation; the remaining 87 patients were followed up for 12 to 17 years. Insulin testing revealed only one inadequate vagotomy in a patient who had a recurrence in the short term. In- sulin tests were negative in 61 and negative or adequate in 6 other patients. Complete vagotomy reduced basal secretion effectively in the great majority of patients but not in a small minority. Three patients had antral hyperfunction with persistent hypersecretion despite complete vagotomy as indicated by two negative insulin tests in each patient. Inex- plicably, only one of these patients had a stoma1 ulcer recurrence.

Long-term follow-up revealed the development of gastric ulcer in one patient with stasis from a pyloroplasty stenosed by angulation from adhesions, Three other patients, one with ulcer and two with hemorrhagic gastritis, developed gastric ulceration in the long term despite low acid output and negative insulin tests. Biliary reflux was demonstrated in two of these three patients and was probably the cause of gastric ulcer in the third.

Pre- and postoperative cholecystograms in 66 patients showed the formation of gallstones in 4 patients after vagotomy. Another patient who did not undergo cholecystography developed acute chole- cystitis from stone. This rate of gallstone formation was the normal expected rate and was not increased as in some series of total vagotomy. Dumping with and without associated diarrhea was the most fre- quent and troublesome sequela. Postvagotomy diarrhea did not occur. To prevent dumping, and also to decrease acid secretion more effectively, pyloroplasty was abandoned in favor of Maki’s pylorus- preserving antrectomy to complement selective vagotomy in 1968.

22. Weinberg JA. Vagotomy and pyloroplasty in the treatment of duodenal ulcer. Am J Surg 1963;105:347.

23. Stempien SJ, Dagradi AE, Lee ER, Simonton JH. Status of duodenal ulcer patients 10 years or more after vagotomy-pyloroplasty. Am J Gastroenterol 1971;56:99.

24. Dragstedt LR. Cause of peptic ulcer. JAMA 1959; 169:203. 25. Dragstedt LR, Oberhelman HA Jr, Woodward ER. Physiology

of gastric secretion and its relation to the ulcer problem. JAMA 1951;147:1615.

References

1. Griffith CA, Leyse RM, Davis DR. Magoon CC. Mortality and recurrent ulcer after selective vagotomy plus pyloroplasty. Am Surg 1972;38:504.

26. Griffith CA. A new anatomic approach to the problem of incomplete vagotomy. Surg Clin North Am 1964;44:1239.

27. McGregor DB, Savage LE, McVay CB. Vagotomy and drainage for elective treatment of peptic ulcers. Surg Gynecol Obstet 1978;146:349.

2. Griffith CA. Selective gastric vagotomy. Surg Techniques ll- 28. Capper WM, Airth GR, Kilby JO. A test for pyloric regurgitation. lustrated 1977;2:49. Lancet 1966;2:621.

3. Jackson RG. Anatomic study of the vagus nerves, with a 29. Kirk RM. Experimental gastric ulcers in the rat: the separate

technique of transabdominal selective gastric vagus re- section. Arch Surg 1948;57:333.

4. Finney JMT. A new method of pyloroplasty. Trans Am Surg Assoc 1902;20:165.

5. Griffith CA. Completeness of gastric vagotomy by the selective technic. Am J Dig Dis 1967;12:333.

6. Burge H, Frohn MJN. The technique of bilateral selective vagotomy with the electrical stimulation test. Br J Surg 1969;56:452.

7. Dragstedt LR, Harper PV, Tovee EB, Woodward ER. Section of the nerves to the stomach in the treatment of peptic ulcer. Ann Surg 1947;126:687.

8. Jemerin EE, Hollander F, Weinstein VA. A comparison of insulin and food as stimuli for the differentiation of vagal and non- vagal gastric pouches. Gastroenterology 1943;1:500.

9. Pritchard GR, Griffith CA, Harkins HN. A physiologic demonstration of the anatomic distribution of the vagal system to the stomach. Surg Gynecol Obstet 1968; 126:79 1.

10. Legros G, Griffith CA. The anatomic basis for the variable adequacy of incomplete vagotomy. I. The various secretory and ulcerogenic potentials of various anatomic types of incomplete vagotomy in Shay rats. Ann Surg 1968;168: 1030.

11. Jones WM. Griffith CA. On the question of vagal reinnervation of the stomach. I. The permanence of the amount of the residually innervated gastric mucosa. Ann Surg 1970; 171:365.

12. Jones WM, Griffith CA. On the question of vagal reinnervation of the stomach. II. The unchanging secretory and ulcerogenic potential. Ann Surg 1970;171:369.

13. Legros G, Griffith CA. The permanence of adequate incomplete vagotomy. Surgery 1970;67:654.

14. Legros G, Griffith CA. The anatomic basis for the variable adequacy of incomplete vagotomy. II. The various responses to insulin of various anatomic types of incomplete vagotomy in dogs. Ann Surg 1968;168:1035.

15. Ross B, Kay AW. The insulin test after vagotomy. Gastroen- terology 1964;46:379.

16. Stadil F. Gastrin and insulin hypoglycemia. Stand J Gastroen- terol (Suppl) 1974;9:23.

17. Woodward ER, Harper PV Jr, Tovee EB, Dragstedt LR. Effect of vagotomy on gastric secretion in man and experimental animals. Arch Surg 1949;59:1191.

18. Gelb M, Baronofsky ID, Janowitz HD. The effect of vagotomy on the maximal acid response to histamine. Gut 1961;2: 240.

19. Gillespie IE, Kay AW. Effect of medical and surgical vagotomy on the augmented histamine test in man. Br Med J 1961;l: 1557.

20. Ganguli PC, Polak JM, Pearse AGE, Elder JB, Hegarty M. An- tral-gastrin-cell hyperplasia in peptic-ulcer disease. Lancet 1974;1:583.

21. Kusarkari K, Nyhus LM, Gillison EW, Bombech CT. An endo- scopic test for completeness of vagotomy. Arch Surg 1972;105:386.



and combined effects of vagotomy and bileduct implantation into the stomach. Br J Surg 1970;57:521.

30 Rokkjaer M, Segaard H. Kruse A, Amdrup E. Bile-induced gastric ulcer in swine. World J Surg 1977;1:377.

31 Silen W. New concepts of the gastric mucosa barriers. Am J Surg 1977;133:8.

32 Johnson FE, Boyden EA. The effect of double vagotomy on the motor activity of the human gallbladder. Surgery 1952;32: 591.

33 Rudick J, Hutchison JSF. Evaluation of vagotomy and biliary function by combined oral cholecystography and intravenous cholangiography. Ann Surg 1965; 162:234.

34. Deveney CW, Gardiner BN, Way LW. Effect of truncal vayotomy on bile kinetics in Rhesus monkeys. Surg Forum 1974;25: 402.

35 Nobles ER Jr. Vagotomy and gastroenterostomy: 15-year fol-

low-up of 175 patients. Am Surg 1966;32: 177. 36. Clave RA, Gaspar MR. Incidence of gallbladder disease after

vagotomy. Am J Surg 1969;118:169. 37. Tompkins RK, Draft AR, Zimmerman E. Lichtenstein JE, Zol-

linger RM. Clinical and biochemical evidence of increased gallstone formation after complete vagotomy. Surgery 1972;71:196.

38. Sapala MA, Sapala JA. Resto Soto AD, Bouwman DL. Chole- lithiasis following subtotal gastric with truncal vagotomy. Surg Gynecol Obstet 1979; 148:36.

39. Friedman GD, Kannel WB, Dawber TR. The epidemiology of gallbladder disease: observations in the Framingham study. J Chronic Dis 1966;19:273.

40. Griffith CA. Selective vagotomy plus suprapyloric antrectomy with and without pylorotomy for duodenal ulcer. Ann Surg 1974;179:516.

voiurne 139, May 1980 615

long-term results of selective vagotomy plus pyloroplasty: 12 to 17 year follow-up

Documents