Band Erosion: Incidence, Etiology, Management and Outcome after Banded Vertical Gastric Bypass

Transcription

1 Obesity Surgery, 11, pp-pp (Reproduced with permission from OBESITY SURGERY) Band Erosion: Incidence, Etiology, Management and Outcome after Banded Vertical Gastric Bypass Mal Fobi, MD; Hoil Lee, MD; Daniel Igwe, MD; Basil Felahy, MD; Elaine James, MD; Malgorzata Stanczyk, MD; Nicole Fobi, MD Center for Surgical treatment of Obesity, Hawaiian Gardens; Tri-City Regional Hospital, Hawaiian Gardens; Bellwood General Hospital, Bellflower; and Cedars Sinai Medical Center, Los Angeles, CA, USA Background: Prosthetic devices have been used in bariatric operations to control the outlet of the gastric pouch and thus maintain weight loss. A complication of these prostheses is erosion or migration into the gastric lumen.the transected banded vertical gastric bypass (TBVGBP) is one of the modifications of gastric bypass. This modification has a silastic ring placed around the pouch to form the stoma. Method:The records of patients with band erosion (BE) after this operation were reviewed, to determine the incidence, etiology, management and outcome during a 9-year period. Results: From May 1992 through May 2001, 2,949 primary and secondary TBVGBP were performed through the Center for Surgical Treatment of Obesity, utilizing 3 hospitals. 48 patients (1.63%) were documented to have BE: 40 documented by us and 8 by subsequent treating surgeons or at other facilities. Presenting symptoms were weight regain (18), stenosis or obstruction (17), pain (9), bleeding (7), and 5 were incidental findings. Some patients presented with more than one symptom. 8 were treated expectantly with spontaneous extrusion of the band. 16 bands have been removed endoscopically in 14 patients.26 patients had open surgical revision,with 12 having band removal only and 14 band removal and revision of either the gastroenterostomy with or without band replacement or conversion to a distal Roux-en-Y gastric bypass (DRYGBP). Two patients who had revision to DRYGBP were re-revised to a longer common limb because of protein malnutrition. Three patients who had revision of the gastroenterostomy with band removal and replacement developed leaks that were managed non-surgically. Two of these re-eroded and the band was removed R e p rint requests to: M A L Fobi, MD, Center for Surgical Treatment of Obesity, S. Pioneer Blvd., Suite 204, Hawaiian Gardens, CA 90716, USA. Fax: (562) ; fobimal@aol.com e n d o s c o p i c a l ly with a subsequent revision to a DRYGBP. There was no death due to BE. Conclusion: BE is an uncommon complication of TBVGBP. Infection,previous bariatric operations and surgical technique play a role in BE. BE is best managed by endoscopic removal but can be treated expectantly or by open surgical intervention. Band removal without replacement or revision to DRYGBP may result in weight regain. Key words: Morbid obesity, bariatric surgery, gastric bypass, device, weight loss, endoscopy, surgical complication Introduction The use of prosthetic devices around the stomach in weight loss operations includes the Wilkinson band (Figure 1) and gastric wrap, 1 the Marlex mesh band by Mason (Figure 2), 2 the Silastic ring band by Laws (Figure 3), the dual Mesh band by Molina (Figure 4), the silicone band and more recently the Lap-Band by Kuzmak (Figures 5 and 6), 5,6 the Dacron and Gore-tex band by Kolle, 7 the silastic ring and fascial band by Linner (Figure 7), 8 the S wedish Band (Obtech) (Fi g u re 8), 9 and the Silastic ring band by Fobi (Figures 9 and 10) One of the complications of these pro s t h e t i c devices is intra-luminal migration or intrusion, usually called band erosion (BE). There have been reports of erosion of various devices This paper examines the incidence of BE of the silastic band which is used to support the stoma in the transected banded (silastic ring) vertical gastric bypass FD-Communications Inc. Obesity Surgery, 11,

2 Fobi et al Marlex Mesh Band Marlex Mesh (Polypropylene) Band Figure 1. Wilkinson s Band, Figure 2. Vertical banded gastroplasty (Mason, 1980). Dual Mesh Band Figure 3. Silastic ring vertical gastroplasty (Laws, 1981). Figure 4. Molina s band, Figure 5. Inflatable gastric band (Kuzmak). Band Figure 6. A d j u s t a ble gastric band (Lap-Band, BioEnterics). (Figures 1, 3, 4, 5, and 6 reproduced with permission from Deitel M. Update Surgery for the Morbidly Obese Pa t i e n t. To r o n t o : F D Communications 2000) 2 Obesity Surgery, 11, 2001

3 Erosion of Silastic Ring in Transected Gastric Bypass Silastic ring or Fascial Band Figure 7. Linner s band, Figure 8. Swedish adjustable gastric band (Obtech). Silastic Ring Band Silastic Ring Marker Figure 9. Silastic ring vertical banded gastric bypass (Fobi, 1986). Figure 10. Transected banded (Silastic ring) vertical gastric bypass with jejunal interposition (Fobi, 1992). (TBVGBP). 11,12 Causes of BE, presenting symptoms, management and outcome are analyzed. Methods Between May 1992 and May 2001, 2,949 patients had the TBVGBP through the Center for Surgical Treatment of Obesity. This operation involved the use of a silastic tubing made by Bentec Medical as a band to control the stoma of this modification of gastric bypass (Figure 10) The data on patients having TBVGBP have been collected prospectively as part of the study of this modification in the management of morbid obesity and also for our submissions to the International Bariatric Surgery R egi s t ry. The re c o rds of these patients we re reviewed to determine the incidence of BE, presenting symptoms, m a n age m e n t, outcome and causes. Results T B VGBP was perfo rmed on 2,949 pat i e n t s between May 1992 and May 2001: 2,386 as a primary operation, 380 as a secondary operation (i.e. revision to the TBVGBP) and 183 as a revision of Obesity Surgery, 11,

4 Fobi et al Transected Banded Vertical Gastric Bypass 2,949 (100%) Primary Secondary Revision 2,386 (80.9%) 380 (12.9%) 183(6.2%) Band Band Band Erosion Erosion Erosion 22 (0.9%) 21 (5.53%) 7 (3.80%) Figure 11. Incidence of band erosion May May the TBVGBP (Figure 11). Band erosion was documented in 48 patients (1.63%), 40 by us and eight by other subsequent treating doctors. Eighteen patients presented with a history of being able to eat more than usual and weight regain (Table 1). Seventeen presented with symptoms of outlet stenosis or obstruction, with solid food intolerance and vomiting. Nine presented with pain and dyspepsia. Seven presented with bleeding, and five were incidental findings in the routine upper GI endoscopy and x-rays that we perform at 5 year follow-up intervals or as part of the pre-operative evaluation before reconstructive procedures. Some patients presented with more than one symptom. Eight patients were treated expectantly, with spontaneous extrusion of the band into the lumen. The band was re m oved endoscopically in 14 patients (Figure 12). Open surgical intervention was performed in 26 patients, and 12 of these had open surgery for removal only. Eight had band removal with revision of the gastroenterostomy, of which seven had band replacement. Six had band removal and revision to a distal Roux-en-Y gastric bypass (DRYGBP) without band replacement. Three patients with band replacement and surgical revision developed a leak that healed spontaneously. Two of these developed band erosion again, and the band was removed endoscopically and these patients had revision to a DRYGBP. Two Table 1. Presenting symptoms of band erosion Symptoms No. of (%) Patients Weight Regain Outlet Obstruction Pain Bleeding of the eight patients who we re revised to a DRYGBP were re-revised to lengthen the common limb because of intractable protein malnutrition. Forty-three of the 48 patients are thus without a band, and five have a replaced band. Six of the 43 without a band have a DRYGBP instead of a band. Twenty-one of the patients have regained an average of 14 of the percent excess weight lost (%EWL). The five patients who had the band replaced have an average of 6 %EWL gained. The six patients who were revised to a DRYGBP and the 16 who had the band removed in the first 9 months after the initial surgery have not had any weight gain but have recorded weight loss. Discussion Total Band Erosion 48 (100.0%) Expectant Endoscopic Surgical Treatment Removal Treatment 8 (16.67%) 14 (29.17%) 26 (54.17%) Band Band Removal & Band removal & Removal revision to DRYGBP revision of 12 without band replacement gastroenerostomy (25.0%) 6 (12.50%) 8 (16.67%) DRYGBP Revised to With band Without 4 (14.58%) RYGBP replacement replacement Revision to Band eroded Band replaced DRYGBP Figure 12. Treatment of band erosion. Stomal dilatation has been long recognized as one of the causes of weight regain after short limb gastric bypass. 8,16-21 Linner attempted to prevent this by reinforcing the gastroenterostomy anastomotic site with a silastic ring prosthesis. 8 The resultant high incidence of band erosion forced him to abandon the use of the prosthetic band and use a fascial band obtained from the linea alba (Figure 7). This concept of reinforcing the anastomotic site has been reintroduced by Kini and Gagner, 22 using a biodegradable porcine graft around the anasto- 4 Obesity Surgery, 11, 2001

5 motic site. Linner s concept of preventing weight regain by reinforcing the stoma against dilating was reintroduced in 1989 by Fobi et al 11 by placing the band around the gastric pouch, as used in the vertical banded gastroplasty and silastic ring vertical gastroplasty, rather than around the gastroenterostomy anastomosis. The anastomosis in the TBVGBP is made 1.5 to 2.5 cm distal to the band. No band erosion was seen in the first 200 cases that were evaluated. This resulted in the silastic ring vertical gastric bypass (Figure 9). 11 This banded gastric bypass evolved to what is now TBVGBP (Figure 10). This method of banding the pouch to control stoma dilatation has been substantiated by Crampton, 23 Zorilla 24 and Capella. 25,26 This modified gastric bypass appears to provide more weight loss in more patients that is maintained over a longer period of time. 27 However, one of the complications of this modification is band erosion (BE). Incidence of Band Erosion The overall incidence of BE in this series was 1.63% (48/2,949). The incidence was 0.92% (22/2,386) in primary operations, 5.53% (21/380) in secondary operations and 28.57% (2/7) in p atients with band replacement after ero s i o n (Figure 11). Causes of Band Erosion Possible causes of BE include: 1) constricting bands, 2) suturing the band to the stomach, 3) imbricating the band with stomach and 4) infection. A band that is placed tightly around the stomach may result in necrosis and erosion. A review of the immediate post-operative x-rays of patients who subsequently had BE showed narrower outlets than in the other patients, indicating that the bands might have been placed tighter than thought (Figure 13A and B). This was confirmed in three patients where we found that the Prolene inside the tubular band had a shorter length than the tubing. Incorporating the stomach wall in placing the band or placing the sutures through the stomach wall to hold the band in place appears to increase the incidence of BE. Anecdotal review of the frequency of BE in our experience with the vertical Erosion of Silastic Ring in Transected Gastric Bypass Figure 13.A. X-ray of the stoma after TBVGBP with subsequent band erosion. B. X-ray of the stoma, after TBVGBP without band erosion. banded gastroplasty and silastic ring vertical gastroplasty and reports in the literature, indicate that there have been more cases of silastic ring erosion by passing the suture that held the ring through the stomach wall as described by Laws (Figure 3) 3 than with the Marlex mesh placed through a wind ow with no sutures through the stomach as described by Mason (Figure 2). 2 We changed the method of band placement after the first 1,100 cases and noticed a significant decrease in the incidence of BE (Figures 14A and B) (27/1,107 [2.43%] vs 21/1,842 [1.14%]). The diameter of the band is usually larger than that of the pouch (Figure 15). Covering the band with the s t o m a ch has been rep o rted to increase BE. 1 2 A A Figure 14. A. Band placement up to B. Band placement after B B Obesity Surgery, 11,

6 Fobi et al Figure 15. Cross-section of silastic ring band and pouch. ERODED SILASTIC RING A Figure 16. A. Partial band erosion. 6 Obesity Surgery, 11, 2001 B Figure 16. B. Complete band erosion.

7 Infection can contribute. Infection around a foreign body is a nidus for long-term smoldering infection, which is rarely controlled with antibiotics and ultimately results in rejection and extrusion of the foreign body. Five patients with BE had had postoperative leaks that were treated without band removal. Another four of the patients with BE had documented post-operat ive subclinical leaks treated with antibiotics. Five other patients had persistent tachycardia for 2-3 days that was consistent with leaks or infection around the pouch, but since they responded to fluids, pulmonary toilet and antibiotic therapy, surgical intervention was not necessary; three of these patients not only had BE but also developed a gastro-gastric fistula, suggesting a leak and infection between the pouch and the bypassed stomach in the immediate post-operative period. The role of infection in BE has also been reported with the Lap-Band. 13,14 There is also a higher incidence of BE in patients who had a previous bariatric operation [21/380 (5.53%)] and in patients who had a revision operation with band replacement [2/7 (28.57%)] (Figure 11). Mechanism of Band Erosion BE usually starts with an inflammatory reaction between the stomach wall and the band. This ultimately results in band extrusion into the pouch lumen, as is documented in stitch abscess formation and extrusion of the stitch. Operative findings in patients with BE confirm the dense fibrous reaction around the banded pouch. Symptoms of Band Erosion The inflammatory reaction around an eroding band results in swelling, causing narrowing of the outlet, resulting in symptoms of outlet stenosis or obstruction or progressive weight loss. Seventeen patients in this series presented with symptoms of stenosis or obstruction. Upper GI x-rays showed narrowing of the outlet. Endoscopic eva l u ation showe d inflammatory changes and ulceration, with partial erosion of the band (Figure 16A and B). In seven cases, the ulceration caused bleeding that presented as hematemesis, melena or anemia. In 18 cases the patients presented with complaints of weight gain, the ability to eat more and an increased frequency Erosion of Silastic Ring in Transected Gastric Bypass of feeling hungry. The finding at endoscopy in these patients showed almost complete extrusion of the band with a dilated stoma (Figure 16B). A careful history from some of these patients confirmed that they had had transient episodes of dyspepsia, nausea, symptoms of stenosis with solid food intolerance and weight loss, but gradually the symptoms improved to a point where they could eat more, with resultant weight gain. Nine patients presented with epigastric pain that radiated to the back, worsened by eating, resulting in progressive weight loss. These patients also had symptoms of obstruction, and BE was confirmed on endoscopic evaluation. Three also had gastro-gastric fistulas on contrast study. Diagnostic Evaluation The best diagnostic test for BE is endoscopic evalu ation. Occasionally plain x-rays may show absence of the band at the pouch level, with or without an abnormal location of the band. Three patients were diagnosed on plain films but had endoscopic evaluation for confirmation. Management BE can be managed expectantly, endoscopically or by open surgical intervention. Expectant treatment is indicated for patients who are diagnosed with BE but the endoscopist did not remove the band at the time of diagnosis and the patient is asymptomatic and either cannot afford or does not want another endoscopic procedure. Eight patients were treated expectantly with spontaneous extrusion of the band in this series. They were also treated with H2- blockers because of the findings of pouch ulcers. Endoscopic band removal is the management of choice for BE. Fourteen patients in this series had 16 successful endoscopic band removal. These patients are placed on H2-blockers for 30 to 60 days because of ulceration at the site where the band extrudes into the lumen. They are usually hospitalized for 24 hours after the band removal, and are monitored without oral intake for symptoms of leak or bleeding. An upper GI series is done after 24 hours to rule out a leak, and the patient is started on liquids and advanced to a regular diet. The last four endoscopic band removals Obesity Surgery, 11,

8 Fobi et al have been done as outpatient procedures; they had no evidence of leaks on contrast studies, and feeding was resumed on the same day. There were five failed attempts at endoscopic band removal. In two patients, it was due to the limitations of the endoscopist and in three patients it was due to lack of appropriate equipment. Four of the patients had the band removed at a second trial by another endoscopist or when the correct instruments were available. One patient required open surgical intervention. Five had repeat endoscopic evaluation within 6 weeks, with findings of complete healing. Open surgical treatment was done in 26 patients. Twelve involved surgery for band removal only. This was in the early stages of our experience, when the ga s t ro e n t e ro l ogists we re reluctant to attempt endoscopic band removal. Patients subjected to open surgery for band removal had perigastric drains placed and temporary gastrostomy tubes for interim feeding until x-rays confirmed no evidence of a leak on the fourth or fifth post-operative day. Eight patients had band removal with revision of the gastric bypass to a DRYGBP. These patients were the ones who had not lost sufficient weight with the TBVGBP before the BE. We also used open band removal in patients who needed another surgical operation such as cholecystectomy, hernia repair, and/or revision of the gastroenterostomy. Outcome after Band Erosion Of the eight patients who were diagnosed and treated at other facilities, four were treated with only endoscopic removal after a phone consultation with us, three with only surgical band removal, and one with band removal and revision of the gastroenterostomy. All reported uncomplicated outcome. Five patients developed leaks after operative band removal: three after band removal and revision of the gastroenterostomy and two after band removal and revision to DRYGBP. Drains were used in all five patients, and the leaks healed spontaneously while the patients were fed via the gastrostomy tube that is inserted routinely. None of the 14 patients that had the band removed endoscopically bled or leaked. Intra-operative band removal is difficult and more complicated in patients who have BE. There is usually a dense fibrous inflammatory reaction associated with a BE that makes o p e rat ive tre atment difficult. Endoscopic band removal is our treatment of choice for BE. Even in patients who need another operation, we prefer to remove the band endoscopically before the other operation. The effect of band removal on weight loss and maintenance is still to be determined. Appreciable weight gain occurred in 21 patients who had BE and re m ova l, with an ave rage of 14% EWL regained (75% to 61% EWL). The average followup in this group of 21 since the original operation is 6 years. The EWL regained in the patients with 6 years follow-up is 8% (75% to 67% EWL), 28 which is less than that in those with the band removed, 14% (75% to 61% EWL). Patients who had the band removed during the first 6 months after the initial surgery have lost further weight. The patients who had a replaced band have only regained 6% of the EWL. The patients who had revision to the DRYGBP after band removal lost more weight. The net effect of BE in TBVGBP is the approximation of a short-limb gastric bypass. Conclusion Band erosion is an uncommon complication of the TBVGBP (< 1% in primary TBVGBP). The technique of placing the band, whether it is a primary, secondary or revision operation, and the presence of any infection appear to be contributing factors to BE. BE can be asymptomatic or it can present with symptoms of outlet stenosis or obstruction, weight rega i n, dy s p ep s i a, ep i ga s t ric pain, a n e m i a, hematemesis or melena. The best diagnostic test is endoscopic eva l u ation. In some cases, e ro d e d bands extrude spontaneously into the GI tract and are passed out with the stool. Endoscopic band removal is the management option of choice. Band replacement after erosion will like ly result in another erosion. Open surgical treatment of BE should be the last resort because of the inherent risks. The findings in this series further confirm the role of the band in weight loss maintenance. 8 Obesity Surgery, 11, 2001

9 Reference 1. Wilkinson LH. Reduction of gastric reservoir capacity. Am J Clin Nutr 1980; 33: Mason EE. Vertical banded gastroplasty for morbid obesity. Arch Surg 1982; 117: Laws HL. Standardized gastroplasty orifice. Am J Surg 1981; 141: Molina M. Gastric banding: an experience with more than 500 cases. Presented at Symposium on Surgical Treatment of Obesity, Los Angeles, CA, Kuzmak LI. A review of seven years experience with silicone gastric banding. Obes Surg 1991; 1: Kuzmak LI. Stoma adjustable silicone gastric banding. Surg Rounds Jan 1991: Kolle K, Bo O. Gastric banding. Proceedings of the 7th Congre s s, O rga n i z ation Mondiale de Gastroenterologie, Stockholm, Sweden, 1982: Linner JH. Gastric operations: In: Linner JH, ed. Surgery for Morbid Obesity. New York: Springer 1984: Forsell P, Hallberg D, Hellers G. Gastric banding for morbid obesity: initial ex p e rience with a new adjustable band. Obes Surg 1993; 3: Fobi MAL. The surgical technique of the Fobi-pouch operation for obesity. Obes Surg 1998; 8: Fobi MAL, Lee H, Flemming AW. The surgical technique of the banded gastric bypass. J Obes and Weight Regulation 1989; 8: Fobi MAL, Lee H, Holness R. Gastric bypass operation for obesity. World J Surg 1998; 22: Fried M. Open and laparoscopic non-adjustable gastric banding. In: Deitel M, ed. Update: Surgery for the Morbidly Obese Patient. To ro n t o : F D - Communications 2000: Hell E, Miller KA. Comparison of vertical banded gastroplasty and adjustable silicone gastric banding. In: Deitel M, ed. Update Surgery for the Morbidly Obese Patient. Toronto: FD-Communications 2000: Eckout GV, Willbanks OL, Moore JT. Vertical ring gastroplasty for morbid obesity: 5 years experience with 1,463 patients. Am J Surg 1986; 152: Erosion of Silastic Ring in Transected Gastric Bypass 16. Griffen WO. Gastric bypass. In: Griffen WO, Printen KJ. Surgical Management of Morbid obesity. New York: Marcel Dekker 1987: Hemreck AS, Jewell WR. Hardin CA. Gastric bypass for morbid obesity: results and complicat i o n s. Surgery 1976; 80: Halverson JD, Koehler RE. Assessment of patients with failed gastric operations for morbid obesity. Am J Surg 1983; 145: Mason EE, Printen KJ, Lewis JW et al. Gastric bypass criteria for effectiveness. Int J Obes 1981; 5: C a rey LC, M a rtin EW Jr, M o j z i s i k C. The surgical treatment for morbid obesity. Curr Probl Surg 1984; 62: 21(10). 21.Halverson JD, Koehler RD. Gastric bypass: analysis of weight loss and factors determining success. Surgery 1981; 90: Kini S, Gagner M, de Csepel J et al. A biodegradable membrane from porcine intestinal submucosa to re i n fo rce the ga s t ro j e j u n o s t o my in lap a ro s c o p i c Roux-en-Y gastric bypass: preliminary report. Obes Surg 2001: 11: Campton NA, Isvornicov V, Stubbs RS. Silastic ring gastric bypass: results in 64 patients. Obes Surg 1997; 7: Zorilla PG, Salinas-Martinez AN. Vertical banded gastroplasty gastric bypass with and without the interposition of jejunum: preliminary report. Obes Surg 1999; 9: Capella RF, Capella JF, Mandac H et al. Vertical banded ga s t ro p l a s t y - ga s t ric by p a s s : p re l i m i n a ry report. Obes Surg 1991; 1: Capella JF, Capella RF. The weight reduction operation of choice: vertical banded gastroplasty or gastric bypass. Am J Surg 1996; 17: Fisher BC, Barber AE. Gastric bypass procedures. Eur J Gastroenterol Hepatol 1999; 11: Fobi MAL, Lee H, Igwe D et al. Prospective comparative evaluation of stapled vs transected silastic ring gastric bypass: a 6 year follow-up. Obes Surg 2001; 11: 1-7. ( R e c e ived September 6, 2001; accepted October 9, ) Obesity Surgery, 11,