© FD-Communications Inc. Obesity Surgery, 14, 2004 165

Obesity Surgery, 14, 165-169

Background: Controversy exists regarding the bestsurgical treatment for superobesity (BMI >50 kg/m2),and a comparison of the 2 most commonly performedprocedures in Europe, namely biliopancreatic diver-sion (BPD) and laparoscopic adjustable gastric band-ing (LAGB), has not yet been reported.

Methods: BPD has been performed in 134 morbidlyobese patients since 1996, and as the primarybariatric procedure in 23 superobese patients. 23 sex-matched patients who most closely resembled theage and BMI of the 23 BPD patients were chosen from1,319 patients who had undergone LAGB since 1996.These groups were compared using appropriate sta-tistical tests.

Results: BPD was performed laparoscopically in 12patients. Median excess weight loss at 24 monthswas 64.4% following BPD and 48.4% following LAGB.Hospital stay and complication rate were significantlygreater with BPD, although the majority of complica-tions were related to the laparotomy wound inpatients undergoing open BPD. Rate of resolution ofobstructive sleep apnea, hypertension and diabetesmellitus following LAGB was similar to BPD.

Conclusion: BPD results in significantly greaterweight loss than LAGB in superobese patients, but isassociated with a longer hospital stay and a highercomplication rate in patients undergoing open BPD.

Key words: Bariatric surgery, laparoscopy, gastric band-ing, biliopancreatic diversion, morbid obesity, superobe-sity

Introduction

Obesity is a global epidemic,1 and the only provenlong-term effective treatment is surgery.2 Morbid

obesity is defined as a body mass index (BMI) >40kg/m2; >50 kg/m2 is classified as superobesity.3

Controversy exists regarding the best surgical treat-ment for morbid obesity, with Roux-en-Y gastricbypass (RYGBP) favored in North America4 andlaparoscopic adjustable gastric banding (LAGB)more commonly performed in Europe andAustralia.5 Biliopancreatic diversion (BPD) hasbeen extensively performed in Italy, but there areconcerns regarding long-term nutritional seque-lae.6,7 For superobese patients the picture is evenless clear. RYGBP with lengthening of the alimen-tary and biliopancreatic limbs results in greater mal-absorption and weight loss than standard RYGBP insuperobesity,8,9 but the authors do not recommendthis modified RYGBP procedure because of thehigh complication rate.9 Successful treatment ofsuperobese individuals has been reported as a partof a series of morbidly obese patients undergoingboth BPD6 and LAGB.10

To our knowledge, a direct comparison of BPDwith LAGB in the treatment of superobesity has notyet been performed.

Materials and Methods

LAGB is our primary bariatric procedure, with BPDreserved for those patients in whom LAGB fails orwho have had previous surgery in the vicinity of theproximal stomach or if the patient requests BPDinstead of LAGB.

BPD has been performed in 134 morbidly obesepatients since 1996, and as the primary bariatricprocedure in 23 superobese patients. One of these

A Comparison of Laparoscopic Adjustable GastricBanding and Biliopancreatic Diversion inSuperobesity

Kevin Dolan, FRCS; Michael Hatzifotis, MBChB; Leyanne Newbury, BSc;George Fielding, FRACS

Department of Surgery, Royal Brisbane Hospital, Queensland, Australia

Reprint requests to: Kevin Dolan, Suite 109, Joondalup HealthCampus, Shenton Avenue, Joondalup, WA 6027, Australia.Fax: ++ 618 9400 9739; e-mail: medkd@iprimus.com.au

patients had a previous fundoplication and 22patients requested BPD. BPD was performed with a50-cm common channel as described byScopinaro.11

LAGB has been performed in 1,319 patients since1996, and the technique using a 9.75-cm LapBand®

(Inamed, Santa Barbara, CA, USA) has beendescribed previously.10 Twenty-three sex-matchedpatients who most closely resembled the BMI andage of the 23 BPD patients were chosen from the1,319 LAGB patients.

Patients are seen in the outpatient clinic at 3monthly intervals, and those with inadequate feel-ings of satiety or inadequate weight loss at anyclinic visit have their band expanded with 1 ml ofsaline and return to clinic 6 weeks later. This proce-dure can be repeated twice if required until 3 mlhave been injected; then any further additions to theband are made in 0.5 ml increments to a maximumof 4 ml.

Weight loss was measured as a reduction in BMIand as the percentage of excess body weight that hasbeen lost (%EWL). Medication was discontinuedwithout adverse effect with resolution of diabetesand hypertension, and continuous positive airwaypressure was stopped as obstructive sleep apnearesolved.

Comparison of age, complications, weight loss,resolution of obesity-related co-morbidities andlength of hospital stay was performed with MannWhitney U, chi-squared and Fisher exact tests.Initial analysis included 23 patients undergoingBPD and 23 matched LAGB patients. A furthercomparison excluding patients who had open BPDwas performed.

Results

Seventeen patients (69.6%) in each group werefemale, and there were no significant differences inthe ages of BPD and LAGB patients (MannWhitney U = 218, P = 0.31) (Table 1).

The first 11 BPD (47.8%) were performed vialaparotomy, and the remaining 12 were completedlaparoscopically. All 23 LAGB were performedlaparoscopically.

Weight loss following BPD was significantly

greater than that for LAGB at 3, 6, 12 and 24months following surgery (Table 1). BPD reducedBMI by 17.6 kg/m2 at 12 months and by 22.3 kg/m2

at 24 months, whereas BMI fell 12.3 kg/m2 and 17.0kg/m2 at 12 and 24 months. Median %EWL was64.4% (36.4 to 96.5%) and 48.4% (26.7 to 80.2%)24 months following BPD and LAGB respectively

Dolan et al

166 Obesity Surgery, 14, 2004

Table 1. A comparison of biliopancreatic diversion(BPD) with laparoscopic adjustable gastric banding(LAGB)

BPD LAGB Statistic P-value

Number ofpatients 23 23

Females 16 (69.6%)16 (69.6%)Age 41 39 Mann

(years) (23-67) (26-58) Whitney U 0.31Complications 13 2 Chi-squared 0.001

(56.5%) (8.7%)Re-operations 7 2 Chi-squared 0.06

(30.4%) (8.7%) MannHospital stay 8 1 Whitney U <0.001

(days) (4-146) (1-2)Follow-up 30 34 Mann

(months) (21-49) (12-59) Whitney U 0.18BMI 56.9 55.9 Mann

preoperative (50.8-83.7)(50.7-90.6)Whitney U 0.89BMI 3 mos 49.3 49.7 Mann

(41.2-57.8)(43.9-87.4) Whitney U 0.34BMI 6 mos 42.7 46.8 Mann

(32.0-55.0)(38.4-87.5) Whitney U 0.03BMI 12 mos 39.1 43.6 Mann

(31.3-50.9)(33.9-85.2) Whitney U 0.03BMI 24 mos 34.6 38.9 Mann

(26.4-48.1)(30.2-49.5) Whitney U 0.04%EWL 23.2 17.3 Mann

3 mos (10.5-48.3) (4.7-35.2) Whitney U 0.01%EWL 39.8 29.5 Mann

6 mos (23.3-72.7) (4.3-48.4) Whitney U 0.001%EWL 57.5 37.0 Mann

12 mos (36.4-82.1) (7.8-66.0) Whitney U 0.001%EWL 64.4 48.4 Mann

24 mos (36.4-96.5)(26.7-80.2) Whitney U 0.02Resolution

of OSA 4 of 5 2 of 3 Fishers exact 0.64Resolution

of HTN 4 of 6 4 of 6 Fishers exact 0.60Resolution

of DM 2 of 2 2 of 3 Fishers exact 0.65

BMI = body mass index (kg/m2); %EWL = excessweight loss expressed as a percentage; OSA = obstruc-tive sleep apnea; HTN = hypertension; DM = diabetesmellitus.

(Mann Whitney U=110, P=0.02) (Table 1). Not sur-prisingly, weight loss was significantly greater fol-lowing laparoscopic BPD then LAGB (Table 2).

The rate of resolution of obstructive sleep apnea,hypertension and diabetes following LAGB wassimilar to BPD (Table 1).

Median hospital stay after LAGB was 1 day, con-siderably shorter than 9 days for open BPD and 8days for laparoscopic BPD (Mann WhitneyU<0.001, P<0.001) (Table 2).

Complications were significantly more commonfollowing BPD than LAGB (Chi-squared=12.0,P=0.001) (Table 1), although the majority werewound-related in patients who underwent open BPD(Table 3). Complications occurred in 13 patients(56.5%) following BPD, 9 of these following openBPD. A wound infection developed in 5 patients(21.7%), 3 following open BPD; a further 3 patientswho underwent open BPD suffered a superficialwound dehiscence. Two patients (8.7%) had ananastomotic leak requiring a laparotomy, one afteropen BPD and one following laparoscopic BPD. Asingle patient (4.3%) required re-laparoscopy tounder-run a bleeding staple-line. Incisional herniashave been detected in 2 patients (8.7%) to date dur-ing a median follow-up of 30 (21 to 49) months.Two patients required shortening of the commonchannel following open BPD.

There were no in-hospital complications follow-ing LAGB. However, 10 months following LAGB asingle patient (4.3%) presented with a slipped bandwhich was repositioned laparoscopically, andanother patient required replacement of a leakingport 13 months following LAGB. Follow-up afterLAGB was 34 (12 to 59) months, which is not sig-nificantly different to that following BPD (MannWhitney U=204, P=0.18).

Five patients required further surgery followingopen BPD, one for an anastomotic leak, two inci-sional hernias and two for shortening of the com-mon channel. A return to the operating-room was

Gastric Banding and BPD for Superobesity

Obesity Surgery, 14, 2004 167

Table 2. A comparison of laparoscopic biliopancreaticdiversion (LapBPD) with laparoscopic adjustable gas-tric banding (LAGB)

LapBPD LAGB Statistic P-value

Number ofpatients 12 12

Females 9 (75.0%) 9 (75.0%)Age 35 35 Mann

(years) (23-67) (27-58) Whitney U 0.50Complications 4 1 Fisher

(33.3%) (6.3%) exact 0.32Re-operations 2 1 Fisher

(16.7%) (8.3%) exact 0.50Hospital stay 8 1 Mann

(days) (4-146) (1-2) Whitney U <0.001Follow-up 29 32 Mann

(months) (21-37) (12-54) Whitney U 0.35BMI 58.4 57.5 Mannpreoperative (50.8-66.9)(50.7-64.5)Whitney U 0.89BMI 50.7 51.0 Mann

3 mos (42.2-57.8)(45.3-60.0)Whitney U 0.31BMI 42.7 48.3 Mann

6 mos (32.0-52.7)(41.1-56.2)Whitney U 0.03BMI 38.6 45.1 Mann

12 mos (31.5-48.9)(37.1-51.6)Whitney U 0.02BMI 34.0 39.2 Mann

24 mos (29.4-45.5)(38.2-49.5)Whitney U 0.004%EWL 22.6 16.3 Mann

3 mos (10.5-48.3) (9.3-23.4) Whitney U 0.05%EWL 39.8 27.6 Mann

6 mos (23.3-72.7)(19.4-38.3)Whitney U 0.002%EWL 59.1 37.0 Mann

12 mos (36.7-82.1)(27.4-58.6)Whitney U 0.001%EWL 68.1 46.7 Mann

24 mos (45.5-96.5)(29.8-65.3)Whitney U 0.002Resolution

of OSA 2 of 2 1 of 1Resolution

of HTN 1 of 2 2 of 3Resolution

of DM 1 of 2 1 of 1

BMI = body mass index (kg/m2); %EWL = excessweight loss expressed as a percentage; OSA = obstruc-tive sleep apnea; HTN = hypertension; DM = diabetesmellitus.

Table 3. Complications following open BPD (BPD),laparoscopic biliopancreatic diversion (LapBPD) andlaparoscopic adjustable gastric banding (LAGB)

BPD LapBPD LAGB

Number of patients 11 12 23Wound infection 3 2 0Wound dehiscence 3 0 0Anastomotic leak 1 1 0Postoperative bleeding 0 1 0Incisional hernia 2 0 0Slippage 0 0 1Port-site leak 0 0 1

necessary for 2 patients who had undergone laparo-scopic BPD; a laparotomy for an anastomotic leakand a laparoscopic under-running of a bleeding sta-ple-line were performed.

Discussion

BPD has been performed for over 25 years,11 and a70% excess weight loss is maintained for at least 10years following surgery.6 A similar weight loss wasexperienced by our superobese patients followingBPD. Larger series of LAGB have reported a 52%and 58% excess weight loss 24 months after sur-gery, again not dissimilar to the weight loss experi-enced by our superobese patients. Although thenumbers in our study are small, it suggests thatweight loss following BPD and LAGB in morbidlyobese patients can be achieved in superobesepatients.

Perhaps it is not surprising that BPD with a partialgastrectomy and a 50-cm common channel resultsin greater weight loss than LAGB. BPD reducedpatients excess weight by 16% more and their BMIby 5 kg/m2 more than LAGB. The benefits of thisextra weight loss in superobese individuals is diffi-cult to measure. There were no obvious differencesin the resolution of obesity-related co-morbiditiessuch as obstructive sleep apnea, hypertension anddiabetes, although the small size of our study popu-lation may have hidden any difference.

There is a price to pay for the extra weight lossinduced by BPD, with a longer hospital stayrequired, even if performed laparoscopically.Although there were no deaths in our study, thelargest series of BPD in morbidly obese patientsreports a mortality of 0.5%,6 whereas only 1 of 5large series of LAGB in the morbidly obese10,12,14-16

has reported mortality, and the authors state thatnone of the deaths were attributable to surgicalcauses.14 More than half of the patients suffered acomplication while in hospital following BPD,including 3 emergency returns to the operating-room. Superobese patients undergoing laparoscopicBPD were part of the first 34 cases performedlaparoscopically, and there may be a learning curve.However, the most common complications follow-ing BPD were related to the laparotomy wound in

open BPD. To date, follow-up surgery after openBPD (incisional hernias and shortening of the com-mon channel) is twice as common as that followingLAGB (slippage and port leakage). These rates ofre-operation on follow-up are similar to thosereported in larger series.6,10,12,14-16 Recorded compli-cations following BPD include incisional hernia(approximately 10%), intestinal obstruction (<1%)and nutritional sequelae such as anemia (approxi-mately 5%), hypoalbuminaemia (approximately3%), night blindness due to vitamin A deficiency(<1%) and metabolic bone disease due to calciumand vitamin D deficiency (approximately 6%).6,7,13

About 5% of gastric bands will slip in the first 3years following LAGB and another 3% will have aport leak or infection.10,12,14-16

BPD results in significantly greater weight lossthan LAGB in superobese patients, but is associatedwith a longer hospital stay and a higher complica-tion rate in patients undergoing open BPD. BothBPD and LAGB resolve obesity-related co-morbidi-ties in the superobese. Our current practice ofLAGB for superobesity continues.

References

1. Deitel M. The International Obesity Task Force and“globesity”. Obes Surg 2002; 12: 613-4.

2. Klein S. Medical management of obesity. Surg Clin NAm 2001; 5: 1025-38.

3. Committee on standards for reporting results.American Society for Bariatric Surgery guidelines forreporting results in bariatric surgery. Obes Surg 1997;7; 521-2.

4. Schauer PR, Ikramuddin S. Laparoscopic surgery formorbid obesity. Surg Clin North Am 2001; 5: 1145-79.

5. Gustavsson S, Westling A. Laparoscopic adjustablegastric banding: complications and side effectsresponsible for the poor long-term outcome. SeminLaparosc Surg 2002; 9: 115-24.

6. Scopinaro N, Adami GF, Marinari GM et al.Biliopancreatic diversion. World J Surg 1998; 22:936-46.

7. Hatzifotis M, Dolan K, Fielding G. Symptomatic vita-min A deficiency following biliopancreatic diversion.Obes Surg 2003; 12: 655-7.

Dolan et al

168 Obesity Surgery, 14, 2004

8. Brolin RE, Gorman RC, Milgrim LM et al.Multivitamin prophylaxis in prevention of post-gas-tric bypass vitamin and mineral deficiencies. Int JObes 1991; 15: 661-68.

9. Sugerman HJ, Kellum JM, DeMaria EJ. Conversionof proximal to distal gastric bypass for failed gastricbypass for superobesity. J Gastrointest Surg 1997; 1:517-25.

10.Fielding G, Rhodes M, Nathanson LK. Laparoscopicgastric banding for morbid obesity. Surgical outcomein 335 cases. Surg Endosc 1999; 13: 550-4.

11.Scopinaro N, Gianetta E, Pandolfo N et al.Biliopancreatic bypass. Proposal and preliminaryexperimental study of a new type of operation for thefunctional surgical treatment of obesity. Minerva-Chir1976; 31: 560-6.

12.O’Brien PE, Brown WA, Smith A et al. Prospective

study of a laparoscopically placed adjustable gastricband in the treatment of morbid obesity. Br J Surg1999; 85: 113-8.

13.Mason EE. Bone disease from duodenal exclusion.Obes Surg 2000; 10: 585-6.

14.Angrisani L, Alkilani M, Basso N et al. LaparoscopicItalian experience with the Lap-Band. Obes Surg2001; 11: 307-10.

15.Favretti F, Cadiere GB, Segato G et al. Laparoscopicadjustable silicone gastric banding (LapBand): Howto avoid complications. Obes Surg 1997; 7: 352-8.

16.Szold A, Abu-Abeid S. Laparoscopic adjustable sili-cone gastric banding for morbid obesity. Results andcomplications in 715 patients. Surg Endosc 2002; 16:230-3.

(Received May 5, 2003; accepted July 18, 2003)

Gastric Banding and BPD for Superobesity

Obesity Surgery, 14, 2004 169