- Email: [email protected]
Failed gastroplasty for morbid obesity
Failed Gastroplasty for Morbid Obesity Revised Gastroplasty Versus Roux-Y Gastric Bypass
Harvey J. Sugerman, MD, Richmond, Virginia James L. Wolpew, MD, Richmond, Virginia
Gastric restriction operations have become popular for the surgical management of morbid obesity after the initial loop gastric bypass operation of Mason and Ito [I]. Several studies have shown clearly that gastric bypass is superior to intestinal bypass [2-51, with a lower rate of complications and an equivalent degree of weight loss. Several surgeons have recommended that gastroplasty, separating a small upper gastric pouch from the remainder of the stomach by a small stoma, would carry a lower morbidity than gastric bypass [6-8]. However, long-term follow-up of these gastroplasty procedures suggests a higher failure rate without a reduction in complications [9-151. Before June 1982, we performed 122 gastroplasties at the Medical College of Virginia by one of three techniques. Each of these techniques carried a high incidence of failure. Forty-three revisional procedures were performed in 39 of these patients. Ten patients underwent a gastroplasty revision, and 33 patients were converted to a Roux-Y gastric bypass. This paper describes the high incidence of complications and failure when an attempt was made to revise a failed gastroplasty. An increased complication rate was also noted in patients converted to a gastric bypass when compared to patients who underwent a primary gastric bypass. However, both gastric bypass groups had a high rate of successful weight reduction. Material and Methods Patients were considered for gastric restriction surgery when they were more than 100 pounds over ideal body weight. Dietary counseling confirmed that patients accepted for surgery had failed multiple attempts at dietary weight reduction programs for 5 years or more. Gastroplasty was the initial operation of choice. As each type of gastroplasty procedure failed, a new method was used in From the lhpcemm of supay and Medkine, Mdkal Cd@e of Vkginia. Virginia Commonwealth University, Richmond. Virginh. Requests fw reprints should be addressed to liarvey J. Superman. MD. Box 519, MCV Station, RichmoM. Virginia 23298.
vdlmm
148, sophneu
1914
the hope that the failure rate would be reduced. However, all of the procedures had a high incidence of failed weight loss or weight regain. The first operation used in 21 patients from March I979 to April 1980 was the procedure recommended by Pace et al [6], with the single application of a TA 90 stapler from which three staples had been removed. Each patient was restricted to a full liquid diet for 8 weeks after the procedure to optimize staple line healing, as recommended by these investigators [16]. From April 1980 to February 1981, 46 patients had two superimposed applications of the
TA 90 stapler from which three staples had been removed in the middle of the staple line and a 10 cm stoma reinforced with a circumferential running 2-O polypropylene suture. From February 1981 to June 1982,55 patients underwent a Gomez gastroplasty [ 71. In this procedure, two superimposed applications of the TA 90 stapler were used, leavingastomaonthegreatercmvatureutihxingtheGomez C clamp. The entire staple line was reinforced with a running through-and-through suture of 2-O polypropylene. The stoma was reinforced with a circumferential 0 chromic catgut suture over a no. 30 Maloney dilator, over which a circumferential running whipstitch of 2-O polypropylene was placed. This was further buttressed with interrupted 3-O silk vertical Lembert sutures. A Silastie ring was used to reinforce the stoma in one of these patients. An attempt was made to follow each patient 2 weeks, 1 month, and 3 months postoperatively and then semiannually. Questionnaires were mailed to patients annually. Telephone contact was used to reach those patients who failed to return or answer the questionnaire. Weight loss was considered successful when a patient lost two thirds of excess weight. Those patients who initially failed to lose weight or regained previously lost weight were asked to have an upper gastrointestinal radiographic series as well as upper endoscopy. Those patients who had evidence of staple line disruption or stomal dilatation were offered operative revision. Those patients with persistent vomiting were evaluated to rule out stomal stenosis and were endoscopically dilated if stenosis was documented. If endoscopic dilatation failed to relieve obstructive symptoms, the patient had surgical revision. Initially, gastroplasty revision was attempted. However, because of the high complication and failure rate associated with gastroplasty
331
Sugerman and Wolper
DILATED OBSTRUCTION
Y
d
OPERATED OWXRUCTION
STOMAL DISRUPTION
so-
:
60 F 5 v) 40: F 20%
SINGLE
DOUBLE
DOUBLE
STAPLE LINE Pac*,Mortin. Tetirlck, et 01
STAPLE LINE, CENTRAL REINFORCED STOMA
STAPLE LINE, LATERAL REINFORCED STOMA I Gomcz 1
I
Figure
1. Failures
I
after
three
types
of gastroplasty
for morbid
obesely.
revision, all gastroplasty failures were subsequently converted to a Roux-Y gastric bypass. Four patients who had failed gastroplasty revisions underwent conversion to a gastric bypass as their third procedure. Recently, we have been conducting a randomized study comparing a vertical banded or Gomez gastroplasty with Roux-Y gastric bypass as the primary procedures for weight reduction. Forty-one patients have had primary gastric bypass procedures performed, and 16 of these patients had this operation more then 1 year ago. The complications and success rate in these patients were compared to those in patients who underwent conversion from gastroplasty to gastric bypass. Results Of the 21 patients who had a single TA 90 application, 15 had disruption of their staple line or the development of stoma1 dilatation, for a failure rate of 71 percent (Figure 1). Forty-six patients had a double TA 90 application with a central reinforced stoma. In 13 of them, stoma1 dilatation developed, and 3 required operative revision for stomal stenosis. One patient had an upper pouch leak, for a total failure rate of 37 percent (Figure 1). A Gomez gastroplasty was performed in 55 patients, 14 of whom had development of stoma1 dilatation that, on endoscopic examination, was invariably associated with erosion of the circumferential prolene suture into the lumen. In one patient, the stoma was reinforced with a Silastic ring, and this also eroded into the lumen with subsequent stoma1 dilatation. Three patients in the Gomez group lost less than 20 percent of their excess body weight at 1 year but have not, as yet, undergone endoscopic or radiographic studies to confirm stoma1 dilatation. Three patients required operative revision for stoma1 stenosis, and in three
332
patients, a leak from the gastric pouch developed with two deaths. Thus, the Gomez gastroplasty has carried a 42 percent failure rate to date (Figure 1). As noted by Freeman and Burchett [13], the majority of patients who failed to return for follow-up visits had failed to lose or had regained lost weight when subsequently contacted. Of the 122 primary gastroplasty procedures, there were 45 weight loss failures, 6 stoma1 stenoses requiring operative revision, and 4 gastric leaks, for a total failure rate of 45 percent. Two patients died from peritonitis secondary to gastric leak, for a 1.6 percent mortality rate. Forty-four revisional operations were performed in 40 patients for either stoma1 dilatation (36 patients), stomal stenosis (6 patients), or delayed gastric emptying (2 patients). Four of these patients had had their primary gastroplasty procedure performed elsewhere by other surgeons. Early in our series, 10 patients underwent attempted revision of failed gastroplasties. In each instance, the previous stoma was completely closed, and a 10 mm diameter gastrogastrostomy was constructed above the staple line around a no. 30 Maloney dilator using a running 2-O polypropylene suture. Major problems developed in 9 of these 10 patients, including 5 patients who either failed to lose or rapidly regained lost weight, 1 patient who was lost to follow-up, 1 patient with gastric pouch leak, and 2 patients in whom severe, delayed distal gastric emptying developed, which was unrelieved with metoclopramide (Tables I and II). Minor problems included one patient with transient stoma1 stenosis treated by endoscopic stoma1 dilatation, one patient with delayed gastric emptying that responded to metoclopramide, and one incisional hernia (Table I). Four of the patients who had revision of a failed gastroplasty underwent conversion to a Roux-Y gastric bypass as a third gastric restriction procedure. Since several patients did well after conversion of their failed gastroplasty revision to Roux-Y gastric bypass and because of the unacceptable incidence of major problems with gastroplasty revision, all subsequent patients have been converted to Roux-Y gastric bypass. Of the 34 patients included in this series, 26 have been followed for more than 1 year since their gastric bypass. One of the 26 patients has failed to lose more than 40 percent of her excess body weight, but, as a group, they have lost an average of 66 f 18 percent of their excess weight (Table III). This was significantly better than the weight loss seen after gastroplasty revision (p
The American Journal of Surgery
Revision
TABLE I
Complkstlons of Revldm
of Failed Gastroplasty
of Falled Gastruplasty Gastroqastrostomv Gastroplasty Revision (n = 10)’
.
Roux-Y Gastric Bypass Conversion Primarv (n = 34) (n = 46)
Major complications 0
Subphrenlc abscess Gastric leak Failed to lose X0% excess weight 1 yr postop Lost to follow-up Delayed gastric emptying Respiratory arrest Pulmonary embolus Hypoglycemia. anemia, neuropathy Small bowel obstruction Minor ccMnplicatlcils stoma1 stenosis Stoma1 ulcer Pancreatic fistula (transient) Incisional hernia Splenectcmy
1 5
1 2 3
0 0 1
1 3+ 0 0 0 0
0 0 0 1 2 1
0 0 1 1 0 0
2 0 0
3 3 1
5+ 1 0
1 1
2 2
0 0
Four patients subsequently converted to Roux-Y gastric bypass as a third procedure. + Only one patient was successfully treated with metoclcpramida. t Patients with stoma1 stenosis after primary gastric bypass all underwent successful endoscopic l
postprandial hypoglycemia (Table I). Both of these latter patients underwent repeated endoscopic stoma1 dilatation for stoma1 stenosis and probably represent malnutrition, although they have maintained their weight at 5 percent above ideal body weight for a year. Both patients have also had prox; imal gastric resection for recurrent stenosis and ulceration. Of the patients with a gastric leak after conversion to a gastric bypass, the leak developed in one patient secondary to marked distal gastric distention that disrupted the gastrojejunostomy. A gastrostomy tube has been used in all other gastric bypass procedures. The second leak occurred when the gastrostomy tube became dislodged where it was not possible to affix the stomach to the abdominal wall drain site. There were 11 minor complications in the 34 patients converted to gastric bypass, including 3 patients with stoma1 stenosis requiring endoscopic dilatation, 3 patients with a marginal ulcer treated with cimetidine, 1 patient with a transient pancreatic fistula, 2 patients with an incisional hernia, and 2 patients with incidental splenectomy (Table I). There were significantly (p CO.01) fewer major complications in the patients with failed gastroplasties who were converted to gastric bypass than in the patients who underwent revision as a gastrogastrostomy (Table II). When the total number of major and minor complications was considered, there was no significant difference (p
balloon dilatation.
During this same period, 46 patients underwent Roux-Y gastric bypass as their primary gastric restriction procedure. Of these patients, 16 were followed for more than 1 year and lost 69 f 17 percent of their excess body weight (Table III). Only one patient failed to lose more than 50 percent of her excess weight. Major complications in this group totalled two and included a nonfatal pulmonary embolus in one patient and severe postoperative atelectasis with asphyxia and death in another patient who weighed 478 pounds (Table I). Minor problems included stoma1 stenosis in five patients, all of whom were successf@ly endoscopically dilatated (Table I). The incidences of major complications and total complications were both significantly (p
TABLE II
Comptlcatlcm of Roux-Y Gastrk Bypass GSStrOgastrostomy
Complications
Gastroplasty Revision (n = 10)
Roux-Y Gastric Bypass Primary Cor+versio+l (n = 46) (n = 34)
Major Minor
10’ 4
3 6
lo+ 11
Tote1
14
9
21’
p
333
Sugefman and Wolpw
TABLE III
Weight Loss Atter Gastric Bypass In Patients Followed
Procedure
No. of Patients
Primary gastric bypass Conversion to gastric bypass
16 26
higher in those patients who were converted to a gastric bypass after a failed gastroplasty than in those who underwent gastric bypass as a primary procedure (Table II). The patients who underwent gastric bypass as a primary operation had significantly (p
334
for More Than 1 Year
Preoperative Weight (lb) 287 f 47 330 f a3
Weight Loss (lb)
Weight Lost (“/I
99 f 31 125 f 57
34 f 7 36f 12
Excess Weight Lost (%) 69f 66f
17
ia
failure appeared to be associated with migration of the suture into the lumen and subsequent stoma1 dilatation. Mason has hypothesized that prevention of stomal dilatation by banding with a polypropylene mesh collar will reduce the high incidence of gastroplasty failure [HI. Pories et al [9] believe that gastric bypass is more effective due to an unexplained effect of distal gastric and duodenal exclusion. We are now conducting a randomized, prospective trial comparing the Roux-Y gastric bypass with the vertical banded gastroplasty. Early in our experience, we attempted revision of the gastric partition in 10 patients with failed gastroplasties. Nine of these 10 patients had a major complication (Table I). In two of these patients, severe delayed distal gastric emptying developed, which was refractory to metoclopramide. The gastroparesis was probably caused by inadvertent injury to the nerve of Laterjet, which is extremely difficult to identify in the midst of the extensive adhesions encountered on reexploration in these patients. Four of these patients have since been converted to Roux-Y gastric bypass as their third weight reduction procedure. Because of the high failure rate associated with gastroplasty revision and the improved weight loss associated with gastric bypass [9-151, all subsequent gastroplasty failures were converted to Roux-Y gastric bypass. These patients have had a much better weight reduction, with significantly fewer major complications than those who had gastroplasty revision. These patients lost, on the average, 66 f 18 percent of their excess body weight when followed for more than 1 year since revision. Although these improved results may have been due, in part, to increased surgical experience, it is more likely that gastric bypass is a better operation. This may be partially a result of a mild dumping syndrome seen in some of these patients so that the mere thought of sweets is repulsive. Certainly, there is no concern about distal gastric emptying and injury to the nerve of Laterjet after gastric bypass. It is not a perfect operation since two patients lost only 40 percent of their excess weight and claimed an increased ability to eat. Second and third operations on gastroplasty failures are technically difficult. Dense adhesions to the
lhe Amwkan Journal of Surgery
Revision of Failed Gastroplasty
liver, spleen, and pancreas are common, making it hard to identify the staple line, the stoma, or the cardioesophageal area. Major complications after conversion to gastric bypass were significantly greater in this series than when gastric bypass was performed as a primary operation. These included two gastric leaks and one subphrenic abscess. Increased morbidity after revision of failed gastric restriction operations has also been noted [22,23]. In two patients who were converted to gastric bypass, recurrent stoma1 stenosis developed that required multiple endoscopic dilatations. One of these patients had failure of the revised gastroplasty and had the gastric bypass as her third procedure. Recurrent marginal ulcers developed on the gastric side of the anastomosis, which were treated with either cimetidine or sulcrafate. They appeared to respond better to the latter drug. The ulcers may have been due to ischemia of the gastric pouch. Both patients lost 95 percent of their excess weight, which they have maintained for over 1 year, and both had serum albumin values greater than 4.2 g/dl. However, in both of these patients iron deficiency anemia developed that was refractory to oral ferrous sulfate but, in one patient, responded to intramuscular iron dextran. Osteomalacia developed in the other patient, and both have a peripheral neuropathy that has responded to supplemental oral multivitamins. Peripheral neuropathy has previously been reported in patients with severe malnutrition and hypoalbuminemia after gastroplasty [24]. Both of these patients had to undergo resection of the proximal gastric pouch. Three of our patients who were converted to gastric bypass lost less than 45 percent of their excess body weight. Each of these patients continues to ingest large quantities of nondietetic beverages and high calorie sweets. It is, therefore, important to carefully screen these patients and refuse revisional procedures, as suggested by Halverson and Koehler [25], for noncompliant patients. One of our patients is representative of the marked functional improvement that can be seen after gastric reduction surgery and exemplifies the importance of revising a failed gastroplasty. This 31 year old male taxicab driver weighed 580 pounds and had severe respiratory insufficiency due to obesity. This included a marked obstructive sleep apnea syndrome, with 60 percent of his sleep apneic on screening capneography. In addition, he had severe obesity hypoventilation syndrome, with a room air Pa02 of 36 torr and PaCOs of 57 torr, a forced vital capacity 40 percent of normal, and marked pulmonary artery hypertension (76 mm Hg systolic, 34 mm Hg diastolic). He previously had multiple auto accidents from falling asleep while driving and had become bedridden, with severe stasis ulcers on his legs. Nine months after tracheostomy and Gomez gastroplasty, he weighed 355 pounds. His sleep apnea was com-
vohkme 148, September IS84
pletely cured, and his obesity hypoventilation syndrome was markedly improved. His room air PaOs had increased to 71 torr and his PaCOs had decreased to 35 torr. His forced vital capacity was 88 percent of normal, and his pulmonary artery pressure had decreased to 32/19 mm Hg. One year later, he no longer had early satiety, his weight rapidly increased to 400 pounds, and he began to develop shortness of breath. Six months after conversion to gastric bypass, he weighed 283 pounds and was symptom-free. His Pa& was 84 torr. He now works regularly and is no longer on welfare. We have had two other patients whose respiratory insufficiency of obesity was cured with gastroplasty, only to recur after weight regain and be corrected again after conversion to gastric bypass and renewed weight loss. In conclusion, gastroplasty for morbid obesity, as performed in this and other series, carried an unacceptable failure rate that was not prevented by a 2 month liquid diet. Furthermore, gastroplasty was not associated with a lower incidence of acute complications than was gastric bypass. It remains to be proved whether the newer technique of vertical banded gastroplasty [22] will be a superior procedure to gastric bypass. Revising a failed gastroplasty also has an unacceptable failure rate, which was significantly less when the failed gastroplasty was converted to a Roux-Y gastric bypass. Conversion of a failed gastroplasty to a gastric bypass carried a significantly higher morbidity than when gastric bypass was performed as a primary procedure. Revisional procedures are technically difficult and should be considered very carefully. If the patient is aware of the risks and still desires surgical correction, gastric bypass should be the procedure of choice. Summary Forty-six percent of 122 gastroplasties for morbid obesity failed. This included a failure rate of 71 percent for a single staple line without stoma1 reinforcement, 37 percent for a double staple line and a central stoma reinforced with 2-O polypropylene, and 42 percent for the Gomez gastroplasty. Revisional procedures were performed in 44 patients. Ten underwent revision of a failed gastroplasty using a gastrogastrostomy and 34 had conversion to a Roux-Y gastric bypass. Patients who had revisional gastroplasty as a second procedure had a significantly higher failure and complication rate than those converted to gastric bypass. Four of these 10 patients were subsequently converted to gastric bypass as their third weight reduction procedure. Conversion of a failed gastroplasty to a Roux-Y gastric bypass is a difficult procedure that carried a significantly higher complication rate in our study than that of a group of 46 patients who underwent a primary gastric bypass procedure. Of 26 patients followed for more than 1 year after conversion to Roux-Y gastric by-
335
Sugefman and Wolper
pass, the average weight loss was 66 f 18 percent of their excess body weight. This was comparable to 16 patients who had undergone a primary gastric bypass more than 1 year previously and had lost 69 f 17 percent of their excess body weight. References 1. 2.
3.
4.
5.
6.
7. 8. 9.
10.
336
Mason EE.Ito C. Gastric
bypass in obesity. Surg Clin North Am 1967;47:1345-51. Griffen WO, Young VL, Stevenson CC. A prospective comparison of gastric and jejunoiieal bypass procedures for morbid obesity. Ann Surg 1977;166:500-7. Buckwalter JA. A prospective comparison of the jejunoileal and gastric bypass operations for morbid obesity. World J Surg 1977;1:757-66. Rucker RD, Horstmann J, Schneider PD, Varco RL, Buchwald t-f. Comparisons between jejunoileal and gastric bypass operations for marked obesity. Surgery 1962;92:241-9. Alden JF. Gastric and jejunoileal bypass. A comparison in the treatment of marked obesity. Arch Surg 1977;112:799606. Pace WG, Martin EW, Tetirick T, Fabri PJ, Carey LC. Gastric partioning for rnorbkf obesity. Ann Surg 1979;190:392400. Gornez CA. Gastroptasty in morbkf obesity. Surg Clin North Am 1979;59:1113-9. Cohn R, Merrell RC, Koslow A. Gastric stapling for morbid obesity. Am J Surg 1961;142:67-72. Portes WJ, Flicklnger EG, Meelheim D, VanRij AM, Thomas FT. Tha effectiveness of gastric bypass over gastric partition in morbid obesity. Consequence of distal gastric and duodenal exclusion. Ann Swg 1962;196:369-99. Lechner GW, Callender AK. Subtotal gastric exclusion and gastric partitioning: a randomized prospective comparison of one hundred patients. Surgery 1961;90:637-44.
11. Laws HL, Piantadosi S. Superior gastric reduction procedure for morbkf obesity. A prospective. randomized trial. Ann Sug 1961;193:334-6. 12. Freeman JB, Bwchett f-U. A comparison of gastric bypass and gastroplasty for morbid obesity. Surgery 1960;88:433. 13. Freeman JB, Burchett H. Failure rate with gastric partitioning for marked obesity. Am J Surg 1983;145:113-9. 14. Linner JH. Comparative effectiveness of gastric bypass and gastroplasty. Arch Surg 1982: 117:695-700. 15. MacLean LD, Rhode Bh4, Shizgal HM. Gsstroplasty fcx obesity. Surg Gynecol Obstet 1981;153:200-8. 16. Ellison EC, Martin EW, Jr., Laschinger J, et al. Prevention of early failure of stapled gastric partitions in treatment of morbid obesity. Arch Surg 1980;115:528-33. 17. Drenick EJ, Gurunanjappa SB, Seltzer F, Johnson DG. Excessive mortality and causes of death in morbidly obese men. JAMA 1980;243:443-5. 18. Johnson D, Drenick EJ. Therapeutic fasting in morbid obesity. Long-term follow-up. Arch Intern Med 1977;137:1381-2. 19. Mason EE, Printen KJ, Barron P, Lewis JW, Kealey Gp, Blornmsrs TJ. Risk reduction in gastric operations for obesity. Ann Surg 1979;190:158-65. 20. Cogbill TH, Moore EE. Perforation after gastric partitioning for morbid obesity. Surgery 1982;92:551-2. 21. Mason EE. Vertical banded gastroptasty for obesity. Arch Surg 1982;117:701-6. 22. MacArthu RI, Smith DE, Hermreck AS, Jewel1 WR. Hardin CA. Revision of gastric bypass. Am J Surg 1980;140:751-4. 23. Mason EE, Printen KJ, Lewis KW, Scott DH, Blommers TJ. Gastric bypass criteria for effectiveness. Int J Obes 1981: 5:405-l 1. 24. Feit H, Glasberg M, lreton C, Rosenberg RN, Thal E. Peripheral neuopathy and starvation after gastric partttioning for morbkf obesity. Ann Intern Med 1982;96:453-5. 25. Halverson JD, Kcehler RE. Assessment of patients with failed gastric operations for morbid obesity. Am J Surg 1983; 145~357-63.
The Anndun
Journal ol !Zurgory
Harvey J. Sugerman, MD, Richmond, Virginia James L. Wolpew, MD, Richmond, Virginia
Gastric restriction operations have become popular for the surgical management of morbid obesity after the initial loop gastric bypass operation of Mason and Ito [I]. Several studies have shown clearly that gastric bypass is superior to intestinal bypass [2-51, with a lower rate of complications and an equivalent degree of weight loss. Several surgeons have recommended that gastroplasty, separating a small upper gastric pouch from the remainder of the stomach by a small stoma, would carry a lower morbidity than gastric bypass [6-8]. However, long-term follow-up of these gastroplasty procedures suggests a higher failure rate without a reduction in complications [9-151. Before June 1982, we performed 122 gastroplasties at the Medical College of Virginia by one of three techniques. Each of these techniques carried a high incidence of failure. Forty-three revisional procedures were performed in 39 of these patients. Ten patients underwent a gastroplasty revision, and 33 patients were converted to a Roux-Y gastric bypass. This paper describes the high incidence of complications and failure when an attempt was made to revise a failed gastroplasty. An increased complication rate was also noted in patients converted to a gastric bypass when compared to patients who underwent a primary gastric bypass. However, both gastric bypass groups had a high rate of successful weight reduction. Material and Methods Patients were considered for gastric restriction surgery when they were more than 100 pounds over ideal body weight. Dietary counseling confirmed that patients accepted for surgery had failed multiple attempts at dietary weight reduction programs for 5 years or more. Gastroplasty was the initial operation of choice. As each type of gastroplasty procedure failed, a new method was used in From the lhpcemm of supay and Medkine, Mdkal Cd@e of Vkginia. Virginia Commonwealth University, Richmond. Virginh. Requests fw reprints should be addressed to liarvey J. Superman. MD. Box 519, MCV Station, RichmoM. Virginia 23298.
vdlmm
148, sophneu
1914
the hope that the failure rate would be reduced. However, all of the procedures had a high incidence of failed weight loss or weight regain. The first operation used in 21 patients from March I979 to April 1980 was the procedure recommended by Pace et al [6], with the single application of a TA 90 stapler from which three staples had been removed. Each patient was restricted to a full liquid diet for 8 weeks after the procedure to optimize staple line healing, as recommended by these investigators [16]. From April 1980 to February 1981, 46 patients had two superimposed applications of the
TA 90 stapler from which three staples had been removed in the middle of the staple line and a 10 cm stoma reinforced with a circumferential running 2-O polypropylene suture. From February 1981 to June 1982,55 patients underwent a Gomez gastroplasty [ 71. In this procedure, two superimposed applications of the TA 90 stapler were used, leavingastomaonthegreatercmvatureutihxingtheGomez C clamp. The entire staple line was reinforced with a running through-and-through suture of 2-O polypropylene. The stoma was reinforced with a circumferential 0 chromic catgut suture over a no. 30 Maloney dilator, over which a circumferential running whipstitch of 2-O polypropylene was placed. This was further buttressed with interrupted 3-O silk vertical Lembert sutures. A Silastie ring was used to reinforce the stoma in one of these patients. An attempt was made to follow each patient 2 weeks, 1 month, and 3 months postoperatively and then semiannually. Questionnaires were mailed to patients annually. Telephone contact was used to reach those patients who failed to return or answer the questionnaire. Weight loss was considered successful when a patient lost two thirds of excess weight. Those patients who initially failed to lose weight or regained previously lost weight were asked to have an upper gastrointestinal radiographic series as well as upper endoscopy. Those patients who had evidence of staple line disruption or stomal dilatation were offered operative revision. Those patients with persistent vomiting were evaluated to rule out stomal stenosis and were endoscopically dilated if stenosis was documented. If endoscopic dilatation failed to relieve obstructive symptoms, the patient had surgical revision. Initially, gastroplasty revision was attempted. However, because of the high complication and failure rate associated with gastroplasty
331
Sugerman and Wolper
DILATED OBSTRUCTION
Y
d
OPERATED OWXRUCTION
STOMAL DISRUPTION
so-
:
60 F 5 v) 40: F 20%
SINGLE
DOUBLE
DOUBLE
STAPLE LINE Pac*,Mortin. Tetirlck, et 01
STAPLE LINE, CENTRAL REINFORCED STOMA
STAPLE LINE, LATERAL REINFORCED STOMA I Gomcz 1
I
Figure
1. Failures
I
after
three
types
of gastroplasty
for morbid
obesely.
revision, all gastroplasty failures were subsequently converted to a Roux-Y gastric bypass. Four patients who had failed gastroplasty revisions underwent conversion to a gastric bypass as their third procedure. Recently, we have been conducting a randomized study comparing a vertical banded or Gomez gastroplasty with Roux-Y gastric bypass as the primary procedures for weight reduction. Forty-one patients have had primary gastric bypass procedures performed, and 16 of these patients had this operation more then 1 year ago. The complications and success rate in these patients were compared to those in patients who underwent conversion from gastroplasty to gastric bypass. Results Of the 21 patients who had a single TA 90 application, 15 had disruption of their staple line or the development of stoma1 dilatation, for a failure rate of 71 percent (Figure 1). Forty-six patients had a double TA 90 application with a central reinforced stoma. In 13 of them, stoma1 dilatation developed, and 3 required operative revision for stomal stenosis. One patient had an upper pouch leak, for a total failure rate of 37 percent (Figure 1). A Gomez gastroplasty was performed in 55 patients, 14 of whom had development of stoma1 dilatation that, on endoscopic examination, was invariably associated with erosion of the circumferential prolene suture into the lumen. In one patient, the stoma was reinforced with a Silastic ring, and this also eroded into the lumen with subsequent stoma1 dilatation. Three patients in the Gomez group lost less than 20 percent of their excess body weight at 1 year but have not, as yet, undergone endoscopic or radiographic studies to confirm stoma1 dilatation. Three patients required operative revision for stoma1 stenosis, and in three
332
patients, a leak from the gastric pouch developed with two deaths. Thus, the Gomez gastroplasty has carried a 42 percent failure rate to date (Figure 1). As noted by Freeman and Burchett [13], the majority of patients who failed to return for follow-up visits had failed to lose or had regained lost weight when subsequently contacted. Of the 122 primary gastroplasty procedures, there were 45 weight loss failures, 6 stoma1 stenoses requiring operative revision, and 4 gastric leaks, for a total failure rate of 45 percent. Two patients died from peritonitis secondary to gastric leak, for a 1.6 percent mortality rate. Forty-four revisional operations were performed in 40 patients for either stoma1 dilatation (36 patients), stomal stenosis (6 patients), or delayed gastric emptying (2 patients). Four of these patients had had their primary gastroplasty procedure performed elsewhere by other surgeons. Early in our series, 10 patients underwent attempted revision of failed gastroplasties. In each instance, the previous stoma was completely closed, and a 10 mm diameter gastrogastrostomy was constructed above the staple line around a no. 30 Maloney dilator using a running 2-O polypropylene suture. Major problems developed in 9 of these 10 patients, including 5 patients who either failed to lose or rapidly regained lost weight, 1 patient who was lost to follow-up, 1 patient with gastric pouch leak, and 2 patients in whom severe, delayed distal gastric emptying developed, which was unrelieved with metoclopramide (Tables I and II). Minor problems included one patient with transient stoma1 stenosis treated by endoscopic stoma1 dilatation, one patient with delayed gastric emptying that responded to metoclopramide, and one incisional hernia (Table I). Four of the patients who had revision of a failed gastroplasty underwent conversion to a Roux-Y gastric bypass as a third gastric restriction procedure. Since several patients did well after conversion of their failed gastroplasty revision to Roux-Y gastric bypass and because of the unacceptable incidence of major problems with gastroplasty revision, all subsequent patients have been converted to Roux-Y gastric bypass. Of the 34 patients included in this series, 26 have been followed for more than 1 year since their gastric bypass. One of the 26 patients has failed to lose more than 40 percent of her excess body weight, but, as a group, they have lost an average of 66 f 18 percent of their excess weight (Table III). This was significantly better than the weight loss seen after gastroplasty revision (p
The American Journal of Surgery
Revision
TABLE I
Complkstlons of Revldm
of Failed Gastroplasty
of Falled Gastruplasty Gastroqastrostomv Gastroplasty Revision (n = 10)’
.
Roux-Y Gastric Bypass Conversion Primarv (n = 34) (n = 46)
Major complications 0
Subphrenlc abscess Gastric leak Failed to lose X0% excess weight 1 yr postop Lost to follow-up Delayed gastric emptying Respiratory arrest Pulmonary embolus Hypoglycemia. anemia, neuropathy Small bowel obstruction Minor ccMnplicatlcils stoma1 stenosis Stoma1 ulcer Pancreatic fistula (transient) Incisional hernia Splenectcmy
1 5
1 2 3
0 0 1
1 3+ 0 0 0 0
0 0 0 1 2 1
0 0 1 1 0 0
2 0 0
3 3 1
5+ 1 0
1 1
2 2
0 0
Four patients subsequently converted to Roux-Y gastric bypass as a third procedure. + Only one patient was successfully treated with metoclcpramida. t Patients with stoma1 stenosis after primary gastric bypass all underwent successful endoscopic l
postprandial hypoglycemia (Table I). Both of these latter patients underwent repeated endoscopic stoma1 dilatation for stoma1 stenosis and probably represent malnutrition, although they have maintained their weight at 5 percent above ideal body weight for a year. Both patients have also had prox; imal gastric resection for recurrent stenosis and ulceration. Of the patients with a gastric leak after conversion to a gastric bypass, the leak developed in one patient secondary to marked distal gastric distention that disrupted the gastrojejunostomy. A gastrostomy tube has been used in all other gastric bypass procedures. The second leak occurred when the gastrostomy tube became dislodged where it was not possible to affix the stomach to the abdominal wall drain site. There were 11 minor complications in the 34 patients converted to gastric bypass, including 3 patients with stoma1 stenosis requiring endoscopic dilatation, 3 patients with a marginal ulcer treated with cimetidine, 1 patient with a transient pancreatic fistula, 2 patients with an incisional hernia, and 2 patients with incidental splenectomy (Table I). There were significantly (p CO.01) fewer major complications in the patients with failed gastroplasties who were converted to gastric bypass than in the patients who underwent revision as a gastrogastrostomy (Table II). When the total number of major and minor complications was considered, there was no significant difference (p
balloon dilatation.
During this same period, 46 patients underwent Roux-Y gastric bypass as their primary gastric restriction procedure. Of these patients, 16 were followed for more than 1 year and lost 69 f 17 percent of their excess body weight (Table III). Only one patient failed to lose more than 50 percent of her excess weight. Major complications in this group totalled two and included a nonfatal pulmonary embolus in one patient and severe postoperative atelectasis with asphyxia and death in another patient who weighed 478 pounds (Table I). Minor problems included stoma1 stenosis in five patients, all of whom were successf@ly endoscopically dilatated (Table I). The incidences of major complications and total complications were both significantly (p
TABLE II
Comptlcatlcm of Roux-Y Gastrk Bypass GSStrOgastrostomy
Complications
Gastroplasty Revision (n = 10)
Roux-Y Gastric Bypass Primary Cor+versio+l (n = 46) (n = 34)
Major Minor
10’ 4
3 6
lo+ 11
Tote1
14
9
21’
p
333
Sugefman and Wolpw
TABLE III
Weight Loss Atter Gastric Bypass In Patients Followed
Procedure
No. of Patients
Primary gastric bypass Conversion to gastric bypass
16 26
higher in those patients who were converted to a gastric bypass after a failed gastroplasty than in those who underwent gastric bypass as a primary procedure (Table II). The patients who underwent gastric bypass as a primary operation had significantly (p
334
for More Than 1 Year
Preoperative Weight (lb) 287 f 47 330 f a3
Weight Loss (lb)
Weight Lost (“/I
99 f 31 125 f 57
34 f 7 36f 12
Excess Weight Lost (%) 69f 66f
17
ia
failure appeared to be associated with migration of the suture into the lumen and subsequent stoma1 dilatation. Mason has hypothesized that prevention of stomal dilatation by banding with a polypropylene mesh collar will reduce the high incidence of gastroplasty failure [HI. Pories et al [9] believe that gastric bypass is more effective due to an unexplained effect of distal gastric and duodenal exclusion. We are now conducting a randomized, prospective trial comparing the Roux-Y gastric bypass with the vertical banded gastroplasty. Early in our experience, we attempted revision of the gastric partition in 10 patients with failed gastroplasties. Nine of these 10 patients had a major complication (Table I). In two of these patients, severe delayed distal gastric emptying developed, which was refractory to metoclopramide. The gastroparesis was probably caused by inadvertent injury to the nerve of Laterjet, which is extremely difficult to identify in the midst of the extensive adhesions encountered on reexploration in these patients. Four of these patients have since been converted to Roux-Y gastric bypass as their third weight reduction procedure. Because of the high failure rate associated with gastroplasty revision and the improved weight loss associated with gastric bypass [9-151, all subsequent gastroplasty failures were converted to Roux-Y gastric bypass. These patients have had a much better weight reduction, with significantly fewer major complications than those who had gastroplasty revision. These patients lost, on the average, 66 f 18 percent of their excess body weight when followed for more than 1 year since revision. Although these improved results may have been due, in part, to increased surgical experience, it is more likely that gastric bypass is a better operation. This may be partially a result of a mild dumping syndrome seen in some of these patients so that the mere thought of sweets is repulsive. Certainly, there is no concern about distal gastric emptying and injury to the nerve of Laterjet after gastric bypass. It is not a perfect operation since two patients lost only 40 percent of their excess weight and claimed an increased ability to eat. Second and third operations on gastroplasty failures are technically difficult. Dense adhesions to the
lhe Amwkan Journal of Surgery
Revision of Failed Gastroplasty
liver, spleen, and pancreas are common, making it hard to identify the staple line, the stoma, or the cardioesophageal area. Major complications after conversion to gastric bypass were significantly greater in this series than when gastric bypass was performed as a primary operation. These included two gastric leaks and one subphrenic abscess. Increased morbidity after revision of failed gastric restriction operations has also been noted [22,23]. In two patients who were converted to gastric bypass, recurrent stoma1 stenosis developed that required multiple endoscopic dilatations. One of these patients had failure of the revised gastroplasty and had the gastric bypass as her third procedure. Recurrent marginal ulcers developed on the gastric side of the anastomosis, which were treated with either cimetidine or sulcrafate. They appeared to respond better to the latter drug. The ulcers may have been due to ischemia of the gastric pouch. Both patients lost 95 percent of their excess weight, which they have maintained for over 1 year, and both had serum albumin values greater than 4.2 g/dl. However, in both of these patients iron deficiency anemia developed that was refractory to oral ferrous sulfate but, in one patient, responded to intramuscular iron dextran. Osteomalacia developed in the other patient, and both have a peripheral neuropathy that has responded to supplemental oral multivitamins. Peripheral neuropathy has previously been reported in patients with severe malnutrition and hypoalbuminemia after gastroplasty [24]. Both of these patients had to undergo resection of the proximal gastric pouch. Three of our patients who were converted to gastric bypass lost less than 45 percent of their excess body weight. Each of these patients continues to ingest large quantities of nondietetic beverages and high calorie sweets. It is, therefore, important to carefully screen these patients and refuse revisional procedures, as suggested by Halverson and Koehler [25], for noncompliant patients. One of our patients is representative of the marked functional improvement that can be seen after gastric reduction surgery and exemplifies the importance of revising a failed gastroplasty. This 31 year old male taxicab driver weighed 580 pounds and had severe respiratory insufficiency due to obesity. This included a marked obstructive sleep apnea syndrome, with 60 percent of his sleep apneic on screening capneography. In addition, he had severe obesity hypoventilation syndrome, with a room air Pa02 of 36 torr and PaCOs of 57 torr, a forced vital capacity 40 percent of normal, and marked pulmonary artery hypertension (76 mm Hg systolic, 34 mm Hg diastolic). He previously had multiple auto accidents from falling asleep while driving and had become bedridden, with severe stasis ulcers on his legs. Nine months after tracheostomy and Gomez gastroplasty, he weighed 355 pounds. His sleep apnea was com-
vohkme 148, September IS84
pletely cured, and his obesity hypoventilation syndrome was markedly improved. His room air PaOs had increased to 71 torr and his PaCOs had decreased to 35 torr. His forced vital capacity was 88 percent of normal, and his pulmonary artery pressure had decreased to 32/19 mm Hg. One year later, he no longer had early satiety, his weight rapidly increased to 400 pounds, and he began to develop shortness of breath. Six months after conversion to gastric bypass, he weighed 283 pounds and was symptom-free. His Pa& was 84 torr. He now works regularly and is no longer on welfare. We have had two other patients whose respiratory insufficiency of obesity was cured with gastroplasty, only to recur after weight regain and be corrected again after conversion to gastric bypass and renewed weight loss. In conclusion, gastroplasty for morbid obesity, as performed in this and other series, carried an unacceptable failure rate that was not prevented by a 2 month liquid diet. Furthermore, gastroplasty was not associated with a lower incidence of acute complications than was gastric bypass. It remains to be proved whether the newer technique of vertical banded gastroplasty [22] will be a superior procedure to gastric bypass. Revising a failed gastroplasty also has an unacceptable failure rate, which was significantly less when the failed gastroplasty was converted to a Roux-Y gastric bypass. Conversion of a failed gastroplasty to a gastric bypass carried a significantly higher morbidity than when gastric bypass was performed as a primary procedure. Revisional procedures are technically difficult and should be considered very carefully. If the patient is aware of the risks and still desires surgical correction, gastric bypass should be the procedure of choice. Summary Forty-six percent of 122 gastroplasties for morbid obesity failed. This included a failure rate of 71 percent for a single staple line without stoma1 reinforcement, 37 percent for a double staple line and a central stoma reinforced with 2-O polypropylene, and 42 percent for the Gomez gastroplasty. Revisional procedures were performed in 44 patients. Ten underwent revision of a failed gastroplasty using a gastrogastrostomy and 34 had conversion to a Roux-Y gastric bypass. Patients who had revisional gastroplasty as a second procedure had a significantly higher failure and complication rate than those converted to gastric bypass. Four of these 10 patients were subsequently converted to gastric bypass as their third weight reduction procedure. Conversion of a failed gastroplasty to a Roux-Y gastric bypass is a difficult procedure that carried a significantly higher complication rate in our study than that of a group of 46 patients who underwent a primary gastric bypass procedure. Of 26 patients followed for more than 1 year after conversion to Roux-Y gastric by-
335
Sugefman and Wolper
pass, the average weight loss was 66 f 18 percent of their excess body weight. This was comparable to 16 patients who had undergone a primary gastric bypass more than 1 year previously and had lost 69 f 17 percent of their excess body weight. References 1. 2.
3.
4.
5.
6.
7. 8. 9.
10.
336
Mason EE.Ito C. Gastric
bypass in obesity. Surg Clin North Am 1967;47:1345-51. Griffen WO, Young VL, Stevenson CC. A prospective comparison of gastric and jejunoiieal bypass procedures for morbid obesity. Ann Surg 1977;166:500-7. Buckwalter JA. A prospective comparison of the jejunoileal and gastric bypass operations for morbid obesity. World J Surg 1977;1:757-66. Rucker RD, Horstmann J, Schneider PD, Varco RL, Buchwald t-f. Comparisons between jejunoileal and gastric bypass operations for marked obesity. Surgery 1962;92:241-9. Alden JF. Gastric and jejunoileal bypass. A comparison in the treatment of marked obesity. Arch Surg 1977;112:799606. Pace WG, Martin EW, Tetirick T, Fabri PJ, Carey LC. Gastric partioning for rnorbkf obesity. Ann Surg 1979;190:392400. Gornez CA. Gastroptasty in morbkf obesity. Surg Clin North Am 1979;59:1113-9. Cohn R, Merrell RC, Koslow A. Gastric stapling for morbid obesity. Am J Surg 1961;142:67-72. Portes WJ, Flicklnger EG, Meelheim D, VanRij AM, Thomas FT. Tha effectiveness of gastric bypass over gastric partition in morbid obesity. Consequence of distal gastric and duodenal exclusion. Ann Swg 1962;196:369-99. Lechner GW, Callender AK. Subtotal gastric exclusion and gastric partitioning: a randomized prospective comparison of one hundred patients. Surgery 1961;90:637-44.
11. Laws HL, Piantadosi S. Superior gastric reduction procedure for morbkf obesity. A prospective. randomized trial. Ann Sug 1961;193:334-6. 12. Freeman JB, Bwchett f-U. A comparison of gastric bypass and gastroplasty for morbid obesity. Surgery 1960;88:433. 13. Freeman JB, Burchett H. Failure rate with gastric partitioning for marked obesity. Am J Surg 1983;145:113-9. 14. Linner JH. Comparative effectiveness of gastric bypass and gastroplasty. Arch Surg 1982: 117:695-700. 15. MacLean LD, Rhode Bh4, Shizgal HM. Gsstroplasty fcx obesity. Surg Gynecol Obstet 1981;153:200-8. 16. Ellison EC, Martin EW, Jr., Laschinger J, et al. Prevention of early failure of stapled gastric partitions in treatment of morbid obesity. Arch Surg 1980;115:528-33. 17. Drenick EJ, Gurunanjappa SB, Seltzer F, Johnson DG. Excessive mortality and causes of death in morbidly obese men. JAMA 1980;243:443-5. 18. Johnson D, Drenick EJ. Therapeutic fasting in morbid obesity. Long-term follow-up. Arch Intern Med 1977;137:1381-2. 19. Mason EE, Printen KJ, Barron P, Lewis JW, Kealey Gp, Blornmsrs TJ. Risk reduction in gastric operations for obesity. Ann Surg 1979;190:158-65. 20. Cogbill TH, Moore EE. Perforation after gastric partitioning for morbid obesity. Surgery 1982;92:551-2. 21. Mason EE. Vertical banded gastroptasty for obesity. Arch Surg 1982;117:701-6. 22. MacArthu RI, Smith DE, Hermreck AS, Jewel1 WR. Hardin CA. Revision of gastric bypass. Am J Surg 1980;140:751-4. 23. Mason EE, Printen KJ, Lewis KW, Scott DH, Blommers TJ. Gastric bypass criteria for effectiveness. Int J Obes 1981: 5:405-l 1. 24. Feit H, Glasberg M, lreton C, Rosenberg RN, Thal E. Peripheral neuopathy and starvation after gastric partttioning for morbkf obesity. Ann Intern Med 1982;96:453-5. 25. Halverson JD, Kcehler RE. Assessment of patients with failed gastric operations for morbid obesity. Am J Surg 1983; 145~357-63.
The Anndun
Journal ol !Zurgory