Proximal Gastric Vagotomy Without Drainage for Treatment of Perforated Duodenal Ulcer

GASTROENTEROLOGY 1982;179-83 Proximal Gastric Vagotomy Without Drainage for Treatment of Perforated Duodenal Ulcer PAUL H. JORDAN, Jr. Surgical Services of the Cora and Webb Mading Department of Surgery, Baylor College of Medicine, and the Veterans Administration Medical Center, Houston, Texas One hundred and nine patients with perforated duodenal ulcer were treated by operation between 1973 and 1980. The operations performed included simple closure in 37 patients, vagotomy and drainage or gastric resection in 12 patients, and proximal gastric vagotomy without drainage and with omental patch of the perforation in 60 patients. Patients who were treated by proximal gastric vagotomy have been observed for 1-8 yr and form the basis of this study. There was no operative mortality. One patient with a postoperative infection required secondary drainage and a second patient with intestinal obstruction required lysis of adhesions. There were no other important complications. Persistent mild dumping occurred in 1 patient. Diarrhea was not a complication for any patient. One patient developed a recurrent ulcer and underwent truncal vagotomy and pyloroplasty. All patients except the patient with a recurrent ulcer had a Visick grading of lor II. Proximal gastric vagotomy, omental patch of the ulcer, and no drainage procedure is the ideal operation for patients who are candidates for definitive treatment of a perforated duodenal ulcer. In 1976 (1) we reported excellent, early results of proximal gastric vagotomy (PGV) and omental patch for treatment of perforated duodenal ulcers. In our opinion, this procedure was ideal to protect patients with a perforated ulcer against further ulcer complications. This report extends the previous study by presenting the 1-8 yr follow-up study results of this method performed upon 60 patients with a perforated duodenal ulcer. It was emphasized in the previous report that the opinion regarding the relative Received September 30, 1981. Accepted January 25, 1982. Address request for reprints to: Paul H. Jordan, [r., M.D., 1200 Moursund, Houston, Texas 77030. 1982 by the American Gastroenterological Association 0016-5085/82/070179-05$02.50 merits of definitive surgery or simple closure of a perforated ulcer did not rest entirely on the operative risk of the two methods of therapy. If poor risk patients, either because of their specific medical problems or because of complications of peritonitis, are relegated to simple closure, the remaining patients can undergo either definitive surgery or simple closure with virtually no mortality. The problem is the identification of patients who should undergo definitive surgery in order to avoid further ulcer complications. The major concern regarding the use of definitive surgery rather than simple closure is the possibility that as a consequence of a definitive operation, gastric symptoms might unnecessarily be inflicted upon a patient who would not have had further ulcer-related symptoms after simple closure. Two-thirds to three-fourths of the patients with perforated duodenal ulcer will have further significant ulcer symptoms after simple closure. Yet, a certain number of patients who because of their age, sex, and chronic history might be expected to have further ulcer symptoms do not. In other patients the need for a definitive operation is not recognized because they fail to reveal their chronic peptic ulcer history until after recovery from operation. These patients fail to give an accurate history because the pain of perforation is so great that their previous symptoms fade into insignificance by comparison. Failure to perform definitive operation in these patients partially accounts for subsequent peptic ulcer symptoms after simple closure in 14%-25% of patients thought to have an acute ulcer (2-5). It is difficult to recommend truncal vagotomy and antrectomy or pyloroplasty for all patients with perforated duodenal ulcer because these procedures, associated with identifiable postoperative sequelae, are unnecessary in approximately 15%-20% of patients with chronic ulcers and 75% of patients with acute ulcers. The specific patients who are at risk of

180 JORDAN GASTROENTEROLOGY Vol. 83, No.1, Part 2 having an unnecessary definitive operation cannot be identified at time of operation. Our objectives for the ideal definitive operation for perforated duodenal ulcer patients are a negligible mortality rate, no postoperative gastric sequelae, and good protection against recurrent ulcers. During the past 8 yr, PGV without drainage and with omental patch of the perforation was performed in 60 patients with perforated duodenal ulcer in whom it was considered safe and technically possible to perform a definitive procedure. In our opinion PGV provides excellent definitive treatment of a perforated duodenal ulcer without the fear that patients might have incapacitating gastric symptoms develop as the result of an operation that they might not have needed. Patients and Methods Between August 1973 and November 1980, 109 patients were operated upon for perforated duodenal ulcer. Treatment consisted of simple closure of the ulcer using an omental patch in 37 patients, PGV without drainage and closure of the ulcer with an omental patch in 60 patients, and other definitive methods including vagotomy, antrectomy, and Billroth I or II in 6 patients, vagotomy and drainage in 4 patients, and antrectomy and Billroth II in 2 patients. No patient during this period was treated knowingly by a nonoperative approach. This paper is based primarily upon the 60 patients who were observed 1-8 yr after treatment of their perforated duodenal ulcer by PGV. Their mean age was 47 yr (range, 20-66 yr). The mean time from perforation to operation was 9 h (range, 4-24 h). The length of ulcer history of the patient before perforation was no history in 3 patients, less than 3 mo in 5 patients, 3-12 mo in 4 patients, more than 1 yr in 48 patients, and more than 5 yr in 33 patients. The length of ulcer history was ascertained after operation in many patients because they denied having ulcer symptoms preoperatively. Ten patients who had definitive surgery other than PGV and a few of the patients in whom simple closure was performed would have been candidates for PGV had the author been available to supervise the operations. Two patients underwent definitive operations other than PGV because of a second duodenal ulcer that was bleeding. Three patients had a simple closure rather than PGV because the inflammatory response obscured the nerves of Latarjet so that they could not be identified satisfactorily. Ancillary procedures at the time of PGV included an antireflux procedure in 2 patients and appendectomy in 3 patients. Splenectomy was required as a consequence of splenic injury in 3 patients. The ulcers were not closed until after the vagotomy in order to prevent dislodgement of the omental patch from the ulcer during the manipulation required by the vagotomy. When the adequacy of the pyloric lumen was questioned, a 40F dilator was passed through the mouth and into the stomach. Passage of the dilator through the pylo- rus and into the duodenum without the application of force insured against postoperative obstruction. The abdomens were liberally irrigated with saline and antibiotics begun preoperatively were continued for an appropriate period postoperatively. After operation, efforts were made to interview patients at 6 and 12 mo and every 12 mo thereafter. All patients were not, however, studied at each time interval. The patients who were treated for perforation were more recalcitrant and less cooperative to undergo postoperative studies than those who had undergone elective operations. Many of the follow-up evaluations were done by telephone. Basal acid output (BAO) and maximal acid output (MAO) in response to histolog were tested at these time intervals if the patients permitted. The clinical results were graded at each time period according to the Visick classification. Results The results of treatment of those patients undergoing simple closure are seen in Table 1. Factors contributing to the 20 deaths in the 37 patients who underwent simple closure included head and neck and pulmonary cancer in 6 patients, chronic obstructive pulmonary disease treated with steroids in 5, over the age of 70 yr and emaciation in 3, severe cardiac pathologic features in 4, renal failure and on dialysis in 2, surgery delayed by 2 wk in 2, ischemic colitis in 1, and reoperation required for bleeding from a second ulcer in 1 patient. The results of definitive therapy by means other than PGV involved too few patients to make an analysis meaningful. There was no operative mortality in this group. One patient who underwent antrecto my and Billroth II anastomosis without vagotomy had a recurrent ulcer develop. During the period of this study, simple closure with an omental patch combined with PGV without drainage was used in 55% of the patients with a perforated duodenal ulcer. On the basis of our experience, it is estimated that 75% of the patients with a perforated ulcer would be considered candidates for PGV. This estimate is the same as the actual percentage of patients being treated in this way by Johnston (6). The operative mortality was zero. Of 5 Table 1. Thirty-seven Patients Treated by Simple Closure Group Postoperative deaths Late deaths Asymptomatic Reoperated for ulcer Remain symptomatic No. of patients 15 5 10 3 4

July 1982 PGV AND PERFORATED DUODENAL ULCER 181 Table 2. Results of the Last Gastric Analysis Performed on Each Patient Treated by Proximal Gastric Vagotomy Time BAO (rneq/h) MAO (meq /h) postoperative (yr) n 0-2 2.1-4 > 4 0-5 5.1-10 10.1-20 > 20 0.5 11 10 1 5 5 1 1 9 9 2 4 2 1 2 13 10 3 5 4 4 3 9 5 2 2 4 4 1 4 2 1 1 1 1 5 4 1 2 1 1 3 6 2 2 2 7 2 2 2 patients who had late deaths, 3 were due to squamous cell carcinoma, 1 to chronic pulmonary disease, and 1 to unknown causes. There were 2 postoperative complications. One was a right subhepatic abscess that was drained extraperitoneally. The other complication was intestinal obstruction that required lysis of adhesions. All patients were observed for at least 1 yr except for 2 patients with carcinoma of the lung who died between 2 and 6 mo after operation. Eighty-two percent of the patients were observed 2 yr or longer. Dumping was reported at some time during the follow-up period by 6 patients. It was persistent in only 1 patient (2%). Dumping was mild and was readily controlled by minor diet modifications consisting of either avoidance of overeating or the elimination or reduction of liquids with meals. Intolerance of sugar was no problem. These restrictions were usually unconsciously invoked by the patients and may account for the low rate of persistent dumping. Intolerance to milk was persistent in 1 patient which was less frequent than occurred after elective PGV. Diarrhea was reported during the study by only 3 patients (5%) and in no patient was it persistent and in no patient did it ever cause a problem. Dysphagia was reported in the early postoperative period in 10 patients but it was not persistent in any patient. Reflux was noted late in the follow-up period by 3 patients but has not been a persistent problem for any patient. The weight of each patient at the time of their last follow-up examination when compared with their preoperative weight was unchanged in 32 patients, increased more than 5 lb in 13 patients, and decreased more than 5 lb in 15 patients. Three of the 15 patients who lost weight died of cancer. The results of the last gastric analysis performed on the patients are summarized in Table 2. We have postoperative studies on 52 patients. No studies were performed on 8 patients and many patients did not permit repeat studies on subsequent visits. The BAO was :52 meqlh in 71% and the MAO was <10 meqlh in 54% of the patients who were studied. Twenty-seven patients had repeat gastric studies that spanned a 2-yr period. The BAO remained the same over this period in 24 patients and increased in 3 patients. The MAO remained the same or decreased in 22 patients and increased in 5 patients. Five patients complained of symptoms at some time after operation that were of concern because of the possibility that they might represent recurrent peptic ulcer disease. One patient did have a recurrent ulcer 2 yr after operation (2% rate of recurrence). After 1 yr of medical treatment he underwent truncal vagotomy and pyloroplasty. For 6 mo after the second operation this patient complained of dumping, diarrhea, and intolerance to milk and sugar. He had experienced none of these symptoms following PGV. After 1 yr all of these symptoms except diarrhea subsided. Of the 4 remaining patients with worrisome symptoms, 1 patient had pancreatitis, 1 had esophageal reflux, and 2 patients had symptoms of unknown causes that completely subsided. Recurrent ulcer was never documented by endoscopy in these patients. The clinical results according to the Visick grading system are tabulated in Table 3. A grade is given to each patient for each follow-up period. A total of 49 patients were observed for 2 yr or longer. The clinical results achieved by these patients at their last follow-up visit were Visick I in42 patients, Visick II in 6 patients, and Visick IV in the1 patient who had a recurrent ulcer develop. Discussion There are basically three methods of treating a patient with perforated duodenal ulcer. They are: (a) nonoperative, which relies primarily on nasogastric suction, antibiotics, and supportive treatment, (b) simple closure of the perforation and cleansing of the abdominal cavity, and (c) definitive surgery. The mortality of the patients who underwent sim-

182 JORDAN GASTROENTEROLOGY Vol. 83, No.1, Part 2 Table 3. The Visick Gradings at Each Follow-up Period after Proximal Gastric Vagotomy Time postoperative (yr) n II III IV 0.5 56 54 1 1 1 47 41 6 2 38 35 3 3 31 27 3 4 18 17 1 5 15 13 2 6 8 7 1 7 7 7 8 1 1 patients 49 42 6 1 followed ~2 yr ple closure was 41 %. This is higher than 21% and 29% reported in other series in which the poor risk patient constituted the majority of patients who underwent simple closure (7,8). Even if all our patients upon whom a definitive operation was performed had undergone simple closure, the mortality rate for simple closure would not have been better than 14%. This is higher than the 5%-7% reported by those who perform simple closure as their principal operative procedure. These observations emphasize that certain patients who were too ill to survive even simple closure might have been better treated by the nonoperative methods outlined by Berne and Rosoff (9) and Donovan (8). We were much more perceptive in selecting patients who could successfully withstand definitive surgery than we were those who could withstand simple closure. There were no deaths in the 72 patients who were treated by the various definitive methods used. If simple closure and definitive surgery were performed upon patients of equal risk, one might conclude that the operative mortality after definitive operation would be greater than the mortality of simple closure because the operation is longer and more traumatic. But just how much greater the risk, is difficult to say. The patients who undergo simple closure have been exposed to the same trauma of perforation, anesthesia, and laparotomy as those patients who undergo definitive operation. In addition, death from a complication of d~odenal ulcer that would have been prevented if definitive therapy rather than simple closure had been performed originally should be attributed to simple closure. For example, 1 patient in our study (3%) died after simple closure because of a bleeding ulcer. If this factor were considered in the mortality analysis, definitive treatment would almost certainly be as safe if not more so than simple closure in most reports. No matter how patients with perforated duodenal ulcer are categorized, if simple closure were performed upon all patients there would be many in each category who would ultimately need additional surgery. It is equally evident that a definitive operation performed upon on all patients capable of undergoing operation successfully would result in some patients having such an operation unnecessarily. Based on duration of the history or the pathological findings at operation, there were in our study approximately 8 patients who might not have needed a definitive operation, but we are not sure. It is recognized that a good gastric surgeon exercising good surgical judgment can perform definitive surgery in carefully selected patients at the time of perforation with a mortality of 0%-2.2% (7,10). Our original reservations about performing definitive surgery were based upon the possibility that gastric sequelae might occur after truncal vagotomy combined with antrectomy or a drainage procedure in patients who would not have subsequently required a definitive operation. A previous prospective, randomized study (11) demonstrated that PGV can be performed with no mortality and without postoperative gastric sequelae for the elective treatment of duodenal ulcer. The present study demonstrates that PGV can be safely applied also in patients with perforated duodenal ulcer without morbidity and mortality. Our total experience consists of PGV in 63 patients with perforated duodenal ulcer. The mortality remains zero. Of the 60 patients who have been observed 1 yr or longer, the morbidity is inconsequential and the recurrence rate is 1. 7%. Although this patient is considered a failure, the consequences of the ulcer in this patient are no different than if a persistent ulcer after simple closure had developed. Our results confirm those of others (12,13) who suggest that PGV is an excellent choice for the treatment of perforated duodenal ulcer in patients who by virtue of their age, fitness, and status of the peritoneal cavity are candidates for definitive surgery. It inflicts virtually none of the morbidity that occurs with other forms of definitive treatment upon patients who might not have needed a definitive operation. At the same time it provides definitive therapy for the larger number of patients who would have subsequently required a second operation for continued ulcer disease if simple closure alone were performed. References 1. Jordan PH Ir. Korompai FL. Evolvement of a new treatment for perforated duodenal ulcer. Surg. Gynecol Obstet 1976 ; 142 :391-5.

July 1982 PGV AND PERFORATED DUODENAL ULCER 183 2. Steiger E, Cooperman AM. Considerations in the nianagement of perforated peptic ulcers. Surg Clin N Am 1976;56:1395 401. 3. Drury JK, MacKay AI. Hutchinson JSF, Joffe SN. Natural history of perforated duodenal ulcers treated by suture closure. Lancet 1978;2:749-50. 4. Kay PH, Moore KTH, Clark RG. The treatment of perforated duodenal ulcer. Br J Surg 1978;65:801-3. 5. Kirkpatrick JR, Bowman DL. A logical solution to the perforated ulcer controversy. Surg Gynecol Obstet 1980;150:683 6. 6. Johnston D. Treatment of peptic ulcer and its complications. In: Taylor S, ed. Recent advances in surgery, No. 10, Chap. 16. London: Churchill-Livingstone, 1980:335-409. 7. Jordan GL [r, DeBakey ME, Duncan JM Jr. Surgical management of perforated peptic ulcer. Ann Surg 1974;179:628-33. 8. Donovan AI. Vinson TL, Maulsby GO, Gewin JR. Selective treatment of duodenal ulcer with perforation. Ann Surg 1979;189:627-36. 9. Berne C], Rosoff L Sr. Acute perforation of peptic ulcer. In: Nyhus LM, Waste II C, eds. Surgery of the stomach and duodenum. 3rd ed. Boston: Little, Brown & Co., 1977:441-57. 10. Lowdon AGR. The treatment of acute perforated peptic ulcer by primary partial gastrectomy. Lancet 1952;1:1270-4. 11. Jordan PH Jr. An interim report on parietal cell vagotomy versus selective vagotomy and antrectomy for treatment of duodenal ulcer. Ann Surg 1979;189:643-53. 12. Johnston D, Lyndon PI. Smith RB, Humphrey CS. Highly selective vagotomy without a drainage procedure in the treatment of hemorrhage, perforation and pyloric stenosis due to peptic ulcer. Br J Surg 1973;60:790-7. 13. Sawyers JL, Herrington JL. Perforated duodenal ulcer managed by proximal gastric vagotomy and suture plication. Ann Surg 1977;185:656-60.