Revision of Roux-en-Y gastric bypass to distal bypass for failed weight loss

Surgery for Obesity and Related Diseases 7 (2011) 45– 49

Original article

Revision of Roux-en-Y gastric bypass to distal bypass for failed weight loss M. Logan Rawlins, M.D.a,b,*, Donovan Teel, II, M.D.b,c, Kim Hedgcorth, C.M.A.c, John P. Maguire, M.D.b,c a

Department of Surgery, Wright Patterson Medical Center, Wright Patterson Air Force Base, Ohio b Department of Surgery, Wright State University Boonshoft School of Medicine, Dayton, Ohio c Premier Metabolic and Bariatric Surgery, Miami Valley Hospital, Dayton, Ohio Received April 19, 2010; accepted August 28, 2010

Abstract

Background: Weight loss failure after proximal Roux-en-Y gastric bypass (RYGB) occurs in ⱕ35% of cases. No consensus has been reached on the best revisional operation for these patients. Our objective was to review our data on the conversion to distal gastric bypass at a universityaffiliated private practice in the United States. Methods: A retrospective review was completed of 29 patients from 2002 to 2009 who had undergone conversion of RYGB to distal gastric bypass because of failure to lose weight, defined as a body mass index ⬎35 kg/m2 or a percentage of excess weight loss of ⬍50%. The RYGB of each patient was converted to a 100-cm distal common limb with a total in-continuity length of 250 cm. Attention was primarily paid to the percentage of excess weight loss at each yearly follow-up visit and any metabolic complications. Results: The average excess weight loss and body mass index were significantly improved from 26.6% and 48.1 kg/m2 before revision to 60.9% and 35.3 kg/m2 at 1 year and 68.8% and 31.5 kg/m2 at 5 years. Of the 29 patients, 6 developed protein calorie malnutrition requiring parenteral nutrition, and 1 required reversal. Diabetes had completely resolved with the additional weight loss. Conclusion: Revision of RYGB to distal gastric bypass can allow patients to achieve sustainable weight loss similar to what they were seeking from their primary surgery. From the results of the present series, the operation was safe, with no perioperative mortality or anastomotic complications. The development of protein calorie malnutrition and vitamin deficiencies is real, and patients require close monitoring and follow-up. (Surg Obes Relat Dis 2011;7:45– 49.) Published by Elsevier Inc. on behalf of American Society for Metabolic and Bariatric Surgery.

Keywords:

Revision distal gastric bypass; Failed weight loss

The most commonly performed bariatric surgical procedure worldwide has been Roux-en-Y gastric bypass (RYGB) [1]. Many prospective studies with long-term follow-up have shown RYGB to provide good weight loss and resolution of co-morbid conditions [2]. Despite

ⴱCorrespondence: Logan Rawlins, M.D., Department of Surgery, Wright State University Boonshoft School of Medicine, Weber CHE, Suite 7000, One Wyoming Street, Dayton, OH 45409. E-mail: [email protected]

these promising results, a fraction of patients still either do not lose sufficient enough weight or regain weight after RYGB. Weight loss failure after proximal RYGB has been reported to occur in 15–35% of cases [3–5]. Failure to achieve a body mass index (BMI) of ⬍35 kg/m2 has been much greater in the superobese, occurring in up to 60% [3]. Although several options are available for revision, many bariatric surgeons are rather hesitant to convert RYGB to a more malabsorptive state. Also, others are reluctant to perform revisions at all, because of the greater risk of complications [6,7]. Regardless, many

1550-7289/11/$ – see front matter Published by Elsevier Inc. on behalf of American Society for Metabolic and Bariatric Surgery. doi:10.1016/j.soard.2010.08.013

46

M. L. Rawlins et al. / Surgery for Obesity and Related Diseases 7 (2011) 45– 49

patients with insufficient weight loss will seek additional surgery to attain their goal. Methods A retrospective chart review was completed of patients since 2002 who had undergone conversion of standard RYGB to distal gastric bypass (D-RYGB) for failed weight loss. Weight loss failure was defined as a BMI ⬎35 kg/m2 or a percentage of excess weight loss (%EWL) of ⬍50% [8]. All patients in the present study had met both criteria. The patients were identified from the operative scheduling logs, and the office charts were examined. A minimum of 1 year of follow-up was required. The institutional review board approved the study. Each patient had undergone contrasted upper gastrointestinal series and/or endoscopy preoperatively to rule out gastrogastric fistula as the cause of weight regain. All the operations were performed using an open approach by surgeon preference. The patients underwent conversion from a 150-cm proximal Roux limb to a 100-cm distal common limb by the same surgeon (J.P.M.). The total in-continuity limb length was 250 cm (alimentary plus common limb). The Roux limb was transected at the jejunojejunal anastomosis and reconnected to the ileum using a linear stapled technique, creating a long biliopancreatic limb. The mesenteric defect was closed, and appendectomy was performed. The gastrojejunal (GJ) anastomosis was not revised. In addition to the weight loss criteria, the patients chosen for this operation had been carefully selected by the primary surgeon. Close follow-up was stressed owing to the significant risk of malabsorption. Those who seemed unreliable for follow-up were not offered this type of revision. The presence of unresolved co-morbidities was not an absolute inclusion criterion, although they were present in most patients. The standard patient selection for primary bariatric procedures as defined by the National Institutes of Health guidelines was followed. Preoperative and ongoing counseling with a licensed dietician was mandatory, with an emphasis on protein intake. An adjunctive physical activity regimen was highly recommended. The patient consent process emphasized the additional risks of revisional surgery and the metabolic complications and need for close followup. All patients were given vitamin supplementation similar to that for those undergoing a duodenal switch procedure: 2 multivitamins, calcium citrate (2400 mg), vitamin D3 (5000 IU), vitamin A (25,000 IU), and B50 complex vitamin, plus elemental iron (151 mg) for premenopausal women. At each postoperative visit, we examined the %EWL, resolution of co-morbid conditions, common postoperative complaints, vitamin deficiencies requiring additional supplementation, and complications from surgery. Most importantly, we reviewed the number of patients who required parenteral nutrition for protein calorie malnutrition (PCMN; i.e., hypoalbuminemia), and how many patients required

reversal because of PCMN. The typical follow-up schedule included visits every 3 months for the first year, every 6 months for the second year, and at 6 –12-month intervals thereafter. Those with problems were seen more frequently. Results From 2002 to 2009, 29 patients were identified who met the selection criteria. The mean age at revision was 47.7 years (range 25– 67). Most of the patients were women (90%). The mean BMI at the primary RYGB was 57.9 kg/m2 (range 38 – 81) compared with 48.1 kg/m2 (range 35– 67) before revision. Of the 29 patients, 24 were superobese (BMI ⬎50 kg/m2) before the primary RYGB, and 13 were still superobese before revision. The average %EWL was 26.6% (range 0 – 46%) before revision and 60.9% (range 39 – 83%) and 68.8% (range 53–91%) at 1 and 5 years after revision, respectively. The follow-up rate was 100% at each follow-up interval (Table 1). D-RYGB constituted 2.5% (29 of 1175 cases) of the surgeon’s practice during the study period. The patients were queried regarding postoperative symptoms. The most common complaint was diarrhea in 79%, with patients reporting an average of 4.4 bowel movements daily (range 1–7). Tincture of opium or diphenoxylate/atropine was commonly prescribed to help alleviate this issue. Oral metronidazole therapy (250 mg/d) was liberally prescribed to most patients if diarrhea/bloating were believed to be due to bacterial overgrowth of the biliopancreatic limb. Other symptoms included fatigue (54%), abdominal pain (46%), nausea (39%), bloating (36%), alopecia (25%), and headaches (21%). Two patients who reported visual changes were found to have low vitamin A levels and were given additional supplementation. Two other patients reported paresthesia in their extremities. Both were found to have normal vitamin B1 levels, and their symptoms had spontaneously resolved by their next office visit. Co-morbidities, such as hypertension and/or diabetes mellitus, were present in 72% of patients. Only 4 patients in the present study had a BMI of 35–39.9 kg/m2, all of

Table 1 Comparison of weight loss before and after revision Variable

Total patients (n) Follow-up (n) Average %EWL Average BMI (kg/m2)

Before RYGB

Before revision

After D-RYGB 1y

3y

5y

29 — — 57.9

29 — 26.6 48.1

29 29/29 60.9* 35.3

13 13/13 69.8* 31.9

5 5/5 68.8* 31.5

RYGB ⫽ Roux-en-Y gastric bypass; D-RYGB ⫽ distal gastric bypass; %EWL ⫽ percentage of excess weight loss; BMI ⫽ body mass index. * P ⬍.01.

M. L. Rawlins et al. / Surgery for Obesity and Related Diseases 7 (2011) 45– 49

whom had both unresolved hypertension and diabetes mellitus. Of the 18 patients with hypertension and 11 patients with diabetes mellitus, 11 (61%) and 11 (100%) respectively no longer required their medication by 1 year after revision. No anastomotic leaks or perioperative deaths occurred. Of the 29 patients, 4 developed deep venous thrombosis, all related to the peripherally inserted central catheter (PICC) lines placed for intravenous access at surgery. Another 10 patients experienced a surgical site infection, only 4 of which were deep and required opening of the wound. In terms of late complications, 1 partial small bowel obstruction occurred that resolved with nonoperative treatment, and 6 patients developed a ventral incisional hernia requiring repair, 3 of whom had had a previous wound infection. The interval to hernia repair was 1–5 years after revision. Only 1 late death occurred at 3 years postoperatively because of cardiac arrest that was attributed to severe coronary artery disease. The overall combined rate of ⱖ1 complication was 59%. The most concerning sequelae of D-RYGB revision are the metabolic complications. All patients were encouraged to consume high-protein shakes daily. In 9 (31%) of the 29 patients, the serum albumin level decreased to ⬍3.0 g/dL. These 9 patients were given pancreatic enzyme replacement. However, 6 of these 9 patients had inadequate response to this treatment, and each completed a 6 –12-week course of total parenteral nutrition. Of the 6 patients, one did not respond to this therapy and had persistent hypoalbuminemia, anorexia, weakness, muscle wasting, pale skin tone, and peripheral edema. That patient underwent revision to provide an additional 100-cm length to the common channel. The vitamin and trace element levels were checked twice in the first year and yearly thereafter. If multiple or significant deficiencies existed, they were measured again every 3– 6 months until corrected. Many patients required additional treatment doses greater than the baseline supplementation they were already taking. Vitamins D (82%) and A (61%) were the most frequent subnormal value. Others included zinc (36%), iron (36%), vitamin B12 (32%), copper (25%), and selenium (14%). Eight patients (27%) developed nonclinical secondary hyperparathyroidism because of vitamin D deficiency. No pathologic fractures were observed or reported. On the follow-up questionnaires, 89% of the patients rated their satisfaction with their results from surgery as “very satisfied” or “satisfied.” Only 11% rated their satisfaction as “somewhat satisfied,” and none rated it as “dissatisfied.” Every patient reported they would undergo the surgery again, regardless of any complications and/or additional treatments they might have required. Most did not regularly attend the support group meetings.

47

Discussion Weight loss failure after standard RYGB is one of the most difficult issues facing bariatric surgeons today. Some surgeons believe the weight regain results from poor and exuberant food choices by the patient, leading to dilation of the GJ stoma or pouch. However, it is likely that most patients will experience some dilation over time, regardless of their food choices [9]. No consensus has yet been reached on the favored operation for RYGB revision. For surgeons willing to perform revisions, which are known to have a greater risk of complications, the risk/benefit ratio from the revisional surgery must carefully be considered [7,10]. Little evidence is available supporting the idea that restoring the restriction of the GJ stoma/pouch will provide the best weight loss. Endoluminal therapies such as endoscopic plication of the stoma have been associated with temporarily reducing the stoma diameter and length. However, the weight loss has been poor, and multiple attempts have been necessary because of failure of the suture plication to hold. The BMI decreased only 2.8 kg/m2 (8.8-kg average weight loss) by 5 months in one study, and the %EWL at 6 months was only 18% in another study [9,11,12]. Müller et al. [13] surgically revised the GJ to create a smaller 15-cm3 pouch. At 1 year, the mean BMI decrease was only 3.9 kg/m2. Schwartz et al. [10] also reported on patients who had undergone surgical revision of their dilated GJ pouch. The additional weight loss was “negligible,” and the immediate complication rate was quite high at 50% [10]. The leak rates from this revision have been reported at 10 –15% [7]. Adjustable gastric banding over the dilated pouch has been attempted, with good early results. The best case series to date of 22 patients was recently reported by Bessler et al. [14]. Their results showed a %EWL of 27%, 42%, and 47% at 1, 3, and 5 years after the revisional procedure, respectively [14]. That operation is attractive because no anastomosis or change in absorption is required. The last option for revision of failed RYGB is a more malabsorptive procedure. Parikh et al. [3] reported a series of 12 patients who underwent conversion to biliopancreatic diversion-duodenal switch (100-cm common channel). At 1 year, the BMI had decreased by 10 kg/m2, and the patients had lost 63% of their prerevision excess weight [3]. Given the short follow-up period of their study, more time is needed to assess the rate of PCMN. One of the concerns of biliopancreatic diversion-duodenal switch is the significant number of staple lines and anastomoses. Furthermore, the gastric pouch will be larger, similar to a nonrevised dilated pouch of RYGB. This lends credence to the idea that better weight loss is more attributable with malabsorption than with restriction. Finally, D-RYGB (100 –150-cm common channel) is an option. As a primary operation, it has had mixed results.

48

M. L. Rawlins et al. / Surgery for Obesity and Related Diseases 7 (2011) 45– 49

Table 2 Comparison of RYGB revision to distal bypass Investigator

Primary operation

Patients (n)

Common limb length (cm)

Preprimary %EWL

Prerevision %EWL

BMI decrease after revision (kg/m2)

Present study

P-RYGB

29

100

Brolin and Cody [5], 2007

P-RYGB, 44/54; 8 pure restrictive; 2 loop GB FPO (P-RYGB equivalent) P-RYGB

54

75–100

61 (1 y) 69 (5 y) NA

46 (1 y) 58 (5 y) 48

13 (1 y) 17 (5 y) 13

65

200–300

NA

NA

7

22

150

NA

5

50

61 (1 y) 69 (5 y) NA

9 (1 y) 13 (5 y) NA

Fobi et al. [16], 2001 Sugerman et al. [4], 1997 Sugerman et al. [4], 1997

P-RYGB

NA

Maximal follow-up (y)

PCMN

TPN

D-RYGB reversed/ revised

5

9 (31)

6 (21)

1 (3)

1

4 (7)

4 (7)

3 (6)

⬎1

15 (23)

6 (9)

6 (9)

5

4 (18)

3 (14)

3 (14)

2

5 (100)

5 (100)

5 (100)

%EWL ⫽ percentage of excess weight loss; BMI ⫽ body mass index; PCMN ⫽ protein calorie malnutrition; TPN ⫽ total parenteral nutrition; D-RYGB ⫽ distal Roux-en-Y gastric bypass; P-RYGB ⫽ proximal RYGB (100 –150-cm Roux limb); FPO ⫽ Fobi pouch operation; GB ⫽ gastric bypass; NA ⫽ not available.

Sugerman et al. [4] demonstrated improved weight loss, and Müller et al. [15] found no difference. However, 25% of the patients in the study by Sugerman et al. [4] required reversal because of PCMN. D-RYGB as a revisional procedure such as we have described has also been previously reported (Table 2). Sugerman et al. [4] initially performed conversion to a 50-cm common limb but increased the length to 150 cm after the first 5 patients had unacceptable results and underwent reversal. In addition, 2 patients died of hepatic failure because of the shorter length [4]. Brolin and Cody [5] revised or created a new pouch/GJ anastomosis. In contrast, Fobi et al. [16] and Sugerman et al. [4] modified the small bowel only. The weight loss results from the present study are consistent with, and have enhanced, the existing data supporting D-RYGB as an option for revisional surgery after failed RYGB. Our prerevision %EWL at 1 year was similar that reported by Brolin and Cody [5] (46% versus 48%), the preprimary %EWL at 5 years was similar to that reported by Sugerman et al. [4] (69% versus 69%), and the postrevision BMI reduction at 1 year was similar to or better than that reported by Brolin and Cody [5], Sugerman et al. [4], and Fobi et al. [16] (13 kg/m2 versus 13, 9, and 7 kg/m2, respectively). Despite the durable weight loss after D-RYGB, it is indisputable that revisional surgery has a significantly increased risk of complications. In our series, 59% of patients experienced a nonmetabolic complication. However, this percentage could be reduced dramatically with simple changes to the operational protocol. The removal of PICC lines as the standard method of vascular access would eliminate the rate of deep venous thrombosis. Switching to a laparoscopic approach instead of open would certainly eliminate the development of most wound infections and ventral hernias.

Compared with the combined data from other published reports of metabolic complications of D-RYGB (Table 2), our rate of PCMN (31% versus 16%, P ⫽ .07) and those requiring total parenteral nutrition (21% versus 9%, P ⫽ .10) was greater, although the difference was not statistically significant [4,5,16]. This difference might be explained by our longer follow-up period. In the present series, the average interval to the development of PCMN was ⬎3 years (range 1–5), much longer than the 1-year follow-up period reported by Fobi et al. [16] and Brolin and Cody [5]. Our rate of reversal was slightly less (3.4% versus 8.5%, P ⫽ .47) [4]. Given the small number of patients (5 of 29) with 5 years of follow-up, it is possible this rate could increase. The fact that Brolin and Cody moved the biliopancreatic limb distal as compared to the roux limb might also contribute to their lower rate of malnutrition. Fox et al. [17] concluded that D-RYGB was quite drastic in terms of malabsorption and was best reserved for the superobese; however, we did not find that to be true in our subgroup analysis. The superobese actually showed a trend toward a lower %EWL at 1 year (58% versus 63%, P ⫽ .20) and 3 years (69% versus 71%, P ⫽ .87) and a greater rate of PCMN (46% versus 19%, P ⫽ .23) and the need for total parenteral nutrition (31% versus 13%, P ⫽ .36), although none of the differences were statistically significant. We experienced a very high rate of vitamin D (82%) and A (61%) deficiency in our patients. Most required ⱖ50,000 IU of vitamin D3, 10 times more than the standard supplementation. Similarly, the vitamin A doses were often doubled to 50,000 IU. We now know these patients might require greater starting doses than previously suspected as many of them have pre-existing deficiencies prior to surgery. Many methods can be used to define and judge a successful revision. We believe those criteria should include

M. L. Rawlins et al. / Surgery for Obesity and Related Diseases 7 (2011) 45– 49

patient satisfaction, the incidence of complications and metabolic sequela, the resolution of co-morbidities, and the resolution of morbid obesity (BMI ⬍35 kg/m2). In our study, the number of patients in whom the morbid obesity had resolved continued to increase with time: 59% (17 of 29) at 1 year, 77% (10 of 13) at 3 years, and 80% (4 of 5) at 5 years. Determining the acceptable rate of PCMN will be the most difficult question to answer. After reviewing the results from our study and the published data, a 100-cm common limb with a total in-continuity length of 250 cm seems to be the most appropriate for weight loss. If the original alimentary limb was ⬍150 cm, we recommend lengthening that proximal segment before the distal bypass first, although this will add an additional anastomosis to the procedure. We are not suggesting D-RYGB should be used as a primary operation. We believe these patients select themselves out for this type of revision. Those who fail to lose enough weight after RYGB will have shown that a primarily restrictive operation might not be enough for them.

[2] [3]

[4]

[5] [6]

[7]

[8] [9]

[10]

Conclusion [11]

Revision of RYGB to D-RYGB will allow patients to achieve sustainable weight loss similar to what they were seeking from their primary surgery. From the results of the present series, the operation can be considered safe, with no perioperative mortality or anastomotic complications. The occurrence of PCMN and vitamin deficiencies is real and patients require close monitoring and follow-up.

[12]

[13]

[14]

Disclosures The authors have no commercial associations that might be a conflict of interest in relation to this article. References [1] Buchwald H, for the Consensus Conference Panel. Consensus conference statement: bariatric surgery for morbid obesity: health

[15]

[16]

[17]

49

implications for patients, health professionals, and third-party payors. Surg Obes Relat Dis 2005;1:371– 81. Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA 2004;292:1724 –37. Parikh M, Pomp A, Gagner M. Laparoscopic conversion of failed gastric bypass to duodenal switch: technical considerations and preliminary outcomes. Surg Obes Relat Dis 2007;3:611– 8. Sugerman HJ, Kellum JM, DeMaria EJ. Conversion of proximal to distal gastric bypass for failed gastric bypass for superobesity. J Gastrointest Surg 1997;1:517–24. Brolin RE, Cody RP. Adding malabsorption for weight loss failure after gastric bypass. Surg Endosc 2007;21:1924 – 6. Cates JSA, Drenick EJ, Adedin MZ, Doty JE, Aaungers KD, Roslyn JJ. Reoperative surgery for the morbidly obese: a university experience. Arch Surg 1990;125:1400 –3. Coakley BA, Deveney CW, Spight DH, et al. Revisional bariatric surgery for failed restrictive procedures. Surg Obes Relat Dis 2008; 4:581– 6. Reinhold RB. Critical analysis of long term weight loss following gastric bypass. Surg Gynecol Obstet 1982;155:385–94. Horgan S, Jacobsen G, Weiss GD, et al. Incisionless revision of post-Roux-en-Y bypass stomal and pouch dilation: multicenter registry results. Surg Obes Relat Dis 2010;6:290 –5. Schwartz RW, Strodel WE, Simpson WS, Griffen WO. Gastric bypass revision: lessons learned from 920 cases. Surgery 1988;104: 806 –12. Mullady DK, Lautz DB, Thompson CC. Treatment of weight regain after gastric bypass surgery when using a new endoscopic platform: initial experience and early outcomes. Gastrointest Endosc 2009;70: 440 – 4. Thompson C, Slattery J, Bundga M, Lautz D. Peroral endoscopic reduction of dilated gastrojejunal anastomosis after Roux-en-Y gastric bypass: a possible new option for patients with weight regain. Surg Endosc 2006;20:1744 – 8. Müller MK, Wildi S, Scholz T, Clavien PA, Weber M. Laparoscopic pouch resizing and redo of gastro-jejunal anastomosis for pouch dilatation following gastric bypass. Obes Surg 2005;15:1089 –95. Bessler M, Daud A, Digiorgi M, et al. Adjustable gastric banding as revisional bariatric procedure after failed gastric bypass-intermediate results. Surg Obes Relat Dis 2010;6:31–5. Müller MK, Räder S, Wildi S, Hauser R, Clavien PA, Weber M. Long term follow up of proximal versus distal laparoscopic gastric bypass for morbid obesity. Br J Surg 2008;95:1375–9. Fobi M, Lee H, Igwe D, et al. Revision of failed gastric bypass to distal Roux-en-Y gastric bypass: a review of 65 cases. Obes Surg 2001;11:190 –5. Fox SR, Fox KM, Oh KH. The gastric bypass for failed bariatric surgical procedures. Obes Surg 1996;6:145–50.