Surgery for Obesity and Related Diseases 9 (2013) 42 47 Original article Medium-term outcomes of patients with insulin-dependent diabetes after laparoscopic adjustable gastric banding Rishi Singhal, M.R.C.S.*, Missba Ahmed, M.B.Ch.B., Avril Krempic, M.B.Ch.B., Mark Kitchen, M.B.Ch.B., Paul Super, F.R.C.S. Upper Gastrointestinal Unit and Minimally Invasive Unit, Heart of England National Health Service Foundation Trust, Birmingham, United Kingdom Received May 1, 2011; accepted July 19, 2011 Abstract Keywords: Background: Bariatric surgery is effective in the control of type 2 diabetes mellitus. Laparoscopic adjustable gastric banding is a popular form of bariatric surgery, but very limited data are available on its long-term effect on type 2 diabetes mellitus. The present study examined the effect of gastric banding on a consecutive cohort of unselected patients with insulin-dependent diabetes mellitus at a teaching hospital in Birmingham, United Kingdom. Methods: From April 2003 to December 2008, 200 patients with diabetes underwent laparoscopic adjustable gastric banding at our unit. All those with insulin-dependent diabetes and 1 year of follow-up were included in the present analysis. Data collection included the body mass index, weight, blood pressure, glycosylated hemoglobin, fasting glucose, total cholesterol, triglycerides, and medication dose preoperatively and 1, 2, and 3 years postoperatively. Results: Preoperatively, 69 patients were taking insulin, with a mean daily preoperative dose of 132.3 U (range 15 500). At 1 year, 27 of these patients had discontinued using insulin (34.8%). At 2 years, 34 patients had discontinued using insulin (54.8% of the patients taking insulin preoperatively and who had also completed 2 yr of follow-up). At 3 years, 40 patients had discontinued using insulin (80% of patients who were taking insulin preoperatively and who had also completed 3 yr of follow-up). These changes were accompanied by an improvement in glycosylated hemoglobin, fasting glucose, total serum cholesterol, triglycerides, and mean arterial pressures. Conclusion: Laparoscopic gastric banding can be considered a powerful treatment option in the management of obese patients with insulin-dependent diabetes and becomes increasingly effective with time 3 years after surgery. (Surg Obes Relat Dis 2013;9:42 47.) 2013 American Society for Metabolic and Bariatric Surgery. All rights reserved. Laparoscopic adjustable gastric banding; Diabetes; Insulin dependent; Bariatric The worldwide epidemic of obesity has reached epic proportions. It is estimated that obesity is fast approaching tobacco as the top underlying preventable cause of death in the United States [1]. Obesity is also an independent risk factor in the development of diabetes [2], the severity of which increases with increasing excess weight [3]. The role of weight loss in the control of type 2 diabetes mellitus has been known for some time [4]. The role of *Correspondence: Rishi Singhal, M.R.C.S., Upper Gastrointestinal and Minimally Invasive Unit, Heart of England National Health Service Foundation Trust, Birmingham, UK. E-mail: singhal_rishi@rediffmail.com surgery in the management of diabetes mellitus was first described in 1955 [5]. In 2004, a review by Buchwald et al. [6] confirmed the role of bariatric surgery in the management of diabetes and the related metabolic syndrome. There has been a very steep increase in the number of bariatric procedures performed worldwide. Compared with just 40,000 procedures in 1998, 344,221 bariatric procedures were performed in 2008. Roux-en-Y gastric bypass seems to be the most commonly performed procedure, followed closely by adjustable gastric banding [7]. The available published data examining the effects of gastric banding on diabetes resolution has either been selective (with regard to patient weight or diabetes duration) 1550-7289/13/$ see front matter 2013 American Society for Metabolic and Bariatric Surgery. All rights reserved. doi:10.1016/j.soard.2011.07.014
Insulin-Dependent Diabetes and LAGB / Surgery for Obesity and Related Diseases 9 (2013) 42 47 43 [1] or limited in numbers [9 12]. Although studies have examined the outcomes in patients with non-insulin-dependent diabetes [13], studies examining the in-depth outcomes for those with insulin-dependent diabetes are not available. Thus, we performed the present study to evaluate the outcomes and examine the factors affecting the resolution of diabetes in an unselected consecutive population of patients with insulin-treated diabetes who underwent laparoscopic adjustable gastric banding at our unit. Methods All cases of laparoscopic banding performed from April 2003 to December 2008 at a single consultant unit were included. These cases were either performed by, or under the supervision of, a single surgeon. All patients fulfilled the criteria for surgical treatment of obesity (body mass index [BMI] 40 kg/m 2 without co-morbidities or BMI 35 kg/m 2 with co-morbidities) [14]. Data were collected from the medical records, operation notes, and computer records and from the general practitioners for all patients who were contacted for additional data. Data collection included the BMI, weight, fasting glucose level, glycosylated hemoglobin (HbA1c), total serum cholesterol, serum triglycerides, blood pressure, and medications taken for blood pressure and diabetes preoperatively and during the subsequent years of follow-up. All assays were performed in the laboratories at the Heart of England National Health Service Foundation trust with internal and external quality control. The ideal body weight was determined according to the Metropolitan Life Insurance Company s 1983 height/weight tables [15]. The percentage of excess BMI loss was calculated as recommended by Deitel et al. [16]. The metabolic outcomes were calculated for the patients with diabetes at the end of 1, 2, and 3 years of follow-up. The following criteria were used to define diabetes: fasting plasma glucose of 7.0 mmol/l, random plasma glucose of 11.1 mmol/l, and ketonuria and clinically symptomatic. Impaired fasting glucose was defined as a fasting plasma glucose of 6.1 7.0 mmol/l. Impaired glucose tolerance was defined as a 2-hour oral glucose tolerance test of 7.8 but 11.1 mmol/l. Finally, gestational diabetes was defined as for diabetes or impaired glucose tolerance but occurring during pregnancy only [17]. Hypertension was defined as a history of hypertension, taking an antihypertensive medication preoperatively, or a persistent, increased blood pressure 140/90 mm Hg, as defined by the UK National Institute for Clinical Excellence criteria [18]. All patients were admitted on the morning of the operation. The pars flaccida technique was used to place all the bands. The band was secured in place using 3 plicating sutures. The first was a gastropexy suture ( Birmingham stitch ), anchoring the fundus to the left crus. The remaining 2 were gastrogastric sutures [19]. Three different bands were used for these procedures (Allergan Vanguard, Allergan AP large, Irvine, CA, and Swedish bands, Obtech, Zug, Switzerland). All patients received low-molecular-weight heparin postoperatively and graduated compression stockings during surgery. Immediate postoperative mobilization and discharge on the same day or the day after surgery was encouraged. Postoperatively, the patients began oral fluids for 2 4 weeks, followed by a soft puree diet for an additional 2 weeks. In most cases, a normal band diet was achieved by the end of 6 weeks. Fluoroscopic-guided adjustments were performed at about 2, 5, and 8 months, and additional fluoroscopic-guided adjustments were performed only if clinically indicated. For follow-up, the patients were seen initially at 6 weeks and then every 3 months for the first 6 months and every 6 months until 2 years postoperatively. Subsequently, the patients were seen annually or at patient request. Statistical analysis was performed using the Statistical Package for Social Sciences, version 17 (SPSS, Chicago, IL). A comparison of the mean values was performed using the Mann-Whitney U test. Spearman s rho was used for nonparametric correlations. The percentage resolution was calculated using only the patients available for follow-up at a given point. Statistical significance was accepted for P.05. Values are presented as the mean and range. Results From April 2003 to December 2008, 200 consecutive patients with diabetes (175 patients with diabetes and 25 patients with impaired glucose tolerance or gestational diabetes) underwent laparoscopic adjustable gastric banding at our unit. Of these, 69 patients were taking insulin preoperatively. These 69 patients were considered for additional analysis in the present study. The data capture for weight loss and medications was 100% at the end of the first and second years. Thus, data were available for 69 of 69 patients at the end of the first year, 62 of 62 patients at the end of the second year, and 50 of 50 patients at the end of 3 years of follow-up. The data capture for HbA1c, glucose, blood pressure, and lipid profile was 90% at all the points examined. Demographics The mean age of these patients was 47 years (range 29 74). The mean preoperative weight and BMI was 140 kg (range 93 222) and 49.2 kg/m 2 (range 37 73), respectively. The ideal weight and excess weight for these patients was 64.2 kg (range 53.8 78.7) and 77.9 kg (range 39.5 156.1), respectively. Of the 69 patients, 49 were also taking metformin, 7 sulfonylurea, 3 thiazolidinedione, and 1 exenatide. The as-
44 R. Singhal et al. / Surgery for Obesity and Related Diseases 9 (2013) 42 47 Table 1 Excess percentage of body mass index loss, weight loss, and metabolic profile of patients with diabetes at examined points after laparoscopic adjustable gastric banding Reference range Pre LAGB 1 year 2 year 3 year Excess % BMI loss 36 19.8 41.3 21.7 45.2 21.8 Excess % weight loss 31.7 17.7 36.5 19.7 39.8 19.9 Glucose metabolism HbA1c (%) 5% 8.9 1.9 7.9 2 7.7 1.8 7.8 1.8 Fasting glucose (mmol/l) 5mmol/L 10.6 4.5 7.4 3.1 7.2 1.9 7.2 2.9 Diabetic medications Insulin (units) 137.8 93.5 62.3 46.5 51 34 50.8 34.9 Metformin (mg) 2108.2 828.9 2032.4 830 1802.7 755.5 1734.4 729.7 Dyslipidemia Total cholesterol 5.0mmol/L 4.5 1.1 4.6 1.1 4.6 1.2 4.3 1.1 Serum triglycerides 2.0mmol/L 2.1 1.2 1.9 1 1.7.9 1.5 0.7 Hypertension Systolic pressure (mm Hg) 140 137.9 16.8 129.8 15.4 130.6 15.3 128.5 15.8 Diastolic pressure (mm Hg) 90 81.4 11.5 76.3 8.8 77.4 8.5 76.3 8.3 Mean arterial pressure (mm Hg) 100 11.5 93.9 9.3 92.8 15.3 90.3 19.5 Values have been expressed as means standard deviation. sociated conditions included hypercholesterolemia (28.9%), hypertriglyceridemia (34.8%), and hypertension (44.9%). The percentage of excess weight loss and the percentage of excess BMI loss in these patients are listed in Table 1.An analysis of the patients co-morbidities is also given in Table 1. The fasting glucose levels are shown in Fig. 1. The mean preoperative glucose level was 10.6 mmol/l (range 4.6 26.0). The mean glucose levels at the various points are listed in Table 1. Of the 69 patients, 83.3% experienced an improvement in the fasting glucose levels at the end of the first year (mean improvement 3.4 mmol/l); 66.7% patients experienced an improvement in glucose levels at the end of 2 years (mean improvement 4.3 mmol/l). Overall, 53.3% of patients experienced an improvement in glucose levels at the end of 3 years (mean improvement 7.2 mmol/l). The percentage of excess BMI loss did not have a significant effect on the decrease in fasting glucose levels at any of the follow-up points. The HbA1c levels are shown in Fig. 1. The mean preoperative HbA1c was 8.95% (range 5.7 13.3%). The mean HbA1c levels at the various follow-up points are listed in Table 1. Of the 69 patients, 73.9% experienced an improvement in the HbA1c level at the end of the first year (mean improvement 1.01); 77.6% patients experienced an improvement in the HbA1c level at the end of 2 years (mean improvement 1.24). Overall, 75% of patients experienced an improvement in the HbA1c level at the end of 3 years (mean improvement 1.2). This percentage increased to 82.3% at the end of 4 years of follow-up. The percentage of excess BMI loss had a significant effect on the decrease in HbA1c levels at the end of the first year (r.25; P.038), second year (r.36; P.003), and third year (r.29; P.03). The insulin dose is shown in Fig. 1 and listed in Table 1. Preoperatively, the 69 patients taking insulin had a mean daily dose of 137.8 U (range 30 500). At 1 year, 27 of these patients had discontinued using insulin (34.8%), 36 had had their dose reduced, 5 patients required the same daily dose, and 1 patient had required an increase in insulin at the end of 1 year. The mean dose for patients still taking insulin at the end of 1 year was 62.3 U (range 8 230). At 2 years, 34 patients had discontinued using insulin (54.8% of the patients who were taking insulin preoperatively and who had also completed 2 yr of follow-up). The mean dose for patients still taking insulin at the end of 2 years was 47.3 U (range 6 150). At 3 years, 40 patients had discontinued using insulin (80% of patients who were taking insulin preoperatively and had completed 3 yr of follow up). The mean dose of insulin at the end of 3 years was 46.6 U (range 14 140). The percentage of excess BMI loss had a significant effect on the decrease in insulin levels at the end of the first year (r.282; P.03); however, this correlation did not achieve statistical significance for the second and third years. Preoperatively, 49 patients were taking metformin, with a mean dose of 2108.2 mg (range 500 3000; Table 1). At 1 year, 14 of these patients had discontinued taking metformin (28.6%), 15 had had their dose reduced, 16 had experienced no change, and 4 patients had required an increase in the metformin dose. The mean dose for patients still taking metformin at the end of 1 year was 2032 mg (range 500 4000). Of these 35 patients, an additional 3 patients had discontinued using metformin at the end of 2 years. The mean dose for patients still taking metformin at the end of 2 years was 1802.7 mg (range 500 3000). Of these, an additional 2 patients had discontinued using metformin at 3 years. The percentage of excess BMI loss did not have a significant effect on the decrease in metformin use at any of the follow-up intervals.
Insulin-Dependent Diabetes and LAGB / Surgery for Obesity and Related Diseases 9 (2013) 42 47 45 Fig. 1. Error bars for mean glucose levels, HbA1c levels, and mean insulin dose at examined points. Preoperatively, elevated total serum cholesterol and serum triglyceride levels were present in 28.9% and 34.8% of the patients, respectively (Table 1). Although some variations were noted in the serum cholesterol levels at the examined follow-up points, none of these changes achieved statistical significance. There was a change in the serum triglyceride level from a mean of 2.14 to 1.85 mmol/l at the end of 1 year (mean improvement.41 mmol/l); 66% of the patients experienced an improvement in the serum triglyceride levels at 1 year. The mean triglyceride level was 1.7 mmol/l and 1.5 mmol/l at the end of 2 and 3 years of follow-up. This improvement was seen in 71.8% and 80% of the patients at 2 and 3 years of follow-up, respectively, compared with their baseline levels. The percentage of excess BMI loss did not have a significant effect on the decrease in serum triglyceride levels at any of the examined points. Of the 69 patients, 44.9% had hypertension at the preoperative assessment. The mean preoperative arterial pressure was 100 mm Hg. The mean arterial pressure improved in 71% of the patients at the end of the first year of followup. In the subsequent years, an overall improvement in 60.9% and 67.2% of the patients was noted at the end of the second and third years, respectively. The percentage of excess BMI loss did not have a significant effect on the decrease in mean arterial pressures at any of the follow-up intervals. Diabetes Duration The mean duration of diabetes for these 69 patients was 84.8 78 months (range 1 420). No statistically significant correlation was found between the duration of diabetes and changes in any of the parameters examined, other than in the
46 R. Singhal et al. / Surgery for Obesity and Related Diseases 9 (2013) 42 47 reduction of the metformin dose at 12 months (r.288; P.047). Complications In the 69 insulin-treated patients, the mean length of stay was 1.1 days (range 0 2), with an overall operative complication rate of 2.9% (2 patients). These complications included 1 punctured band balloon, for which the band was replaced with a different band type and 1 partial slip that was noted at 38 months after surgery. This patient underwent immediate deflation of the band, followed by elective repositioning. Each of these reoperations was performed as outpatient surgery with same day discharge. No cardiorespiratory or anesthetic-related complications developed in this cohort. Discussion The role of bariatric surgery in the treatment of diabetes and its related co-morbidities is now well established. In their review of bariatric surgery in 2004, Buchwald et al. [6] noted that a substantial majority of patients with diabetes, hyperlipidemia, hypertension, and obstructive sleep apnea experienced complete resolution or improvement. Similar results were echoed in the Swedish Obese Subjects study a few months later [20]. There is, however, a lack of data about the factors influencing these changes, specifically with regard to gastric banding. This issue has been further clouded by selective studies that might have inadvertently suggested that the resolution of diabetes is best seen within a particular range of BMI levels and a shorter duration of diabetes [8]. In the present study, an unselected consecutive population of patients with insulin-dependent diabetes was evaluated to determine the efficacy of gastric banding. The heaviest diabetic patient in the present study had a BMI of 73 kg/m 2. In contrast to the other studies and widely held beliefs, the excess percentage of BMI loss did not influence the resolution or improvement in the metabolic syndrome, as measured in the present study. Of all the parameters examined, only the reductions in HbA1c had a consistent statistically significant correlation with the percentage of excess BMI loss. This is in contrast to the finding of Dixon and O Brien [12]. They found a strong and statistically significant association between weight loss and the change in most parameters [12]. These results compare quite favorably with the data available for Roux-en-Y gastric bypass. According to the meta-analysis by Buchwald et al. [21] in 2009, Roux-en-Y gastric bypass achieved a resolution of diabetes (discontinuation of medication) in 58.29% of the patients at 2 years of follow-up (296 patients). In our study, 34% of the patients were off all medications and 80% had discontinued using insulin at the end of 3 years of follow-up. The present study included a cohort of only patients with insulin-dependent diabetes; thus, the results for an unselected population of patients with diabetes can only be better. For weight loss outcomes, it has been generally agreed that the excess weight loss after a Roux-en-Y gastric bypass is more than that after laparoscopic adjustable gastric banding. It is important to note that only a small proportion of patients who underwent surgery at our unit were treated at a specialist diabetes unit postoperatively. Most of the management of their diabetes and their medication was performed by their individual general practitioners. This can partly explain the diverse range of medications, along with the increases in dose of some medications, despite patients losing weight in response to gastric banding surgery. What the present study has demonstrated is that the effect of gastric banding is much improved glycemic control, which was seen uniformly in all patients, irrespective of their insulin requirements. Moreover, the percentage of excess BMI loss played only a small non-uniform role in the resolution of these co-morbidities. We can therefore conclude that all obese patients with type 2 diabetes mellitus can be considered for gastric band surgery, regardless of the severity of their diabetes, severity of diabetes and pre-op BMI. Disclosures The authors have no commercial associations that might be a conflict of interest in relation to this article. References [1] Stein R. Obesity passing smoking as top avoidable cause of death. The Washington Post, March 10, 2004, A01. [2] Must A, Spadano J, Coakley EH, Field AE, Colditz G, Dietz WH. The disease burden associated with overweight and obesity. JAMA 1999;282:1523 9. [3] Chan JM, Rimm EB, Colditz GA, Stampfer MJ, Willett WC. Obesity, fat distribution, and weight gain as risk factors for clinical diabetes in men. Diabetes Care 1994;17:961 9. [4] Newburgh L. Control of hyperglycaemia of obese diabetics by weight reduction. Ann Intern Med 1942;17:935 42. [5] Friedman MN, Sancetta AJ, Magovern GJ. The amelioration of diabetes mellitus following subtotal gastrectomy. Surg Gynecol Obstet 1955;100:201 4. [6] Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA 2004;292:1724 37. [7] Buchwald H, Oien DM. Metabolic/bariatric surgery worldwide 2008. Obes Surg 2009;19:1605 11. [8] Dixon JB, O Brien PE, Playfair J, et al. Adjustable gastric banding and conventional therapy for type 2 diabetes: a randomized controlled trial. JAMA 2008;299:316 23. [9] Sultan S, Gupta D, Parikh M, et al. Five-year outcomes of patients with type 2 diabetes who underwent laparoscopic adjustable gastric banding. Surg Obes Relat Dis 2010;6:373 6. [10] Pontiroli AE, Pizzocri P, Librenti MC, et al. Laparoscopic adjustable gastric banding for the treatment of morbid (grade 3) obesity and its
Insulin-Dependent Diabetes and LAGB / Surgery for Obesity and Related Diseases 9 (2013) 42 47 47 metabolic complications: a three-year study. J Clin Endocrinol Metab 2002;87:3555 61. [11] Pontiroli AE, Folli F, Paganelli M, et al. Laparoscopic gastric banding prevents type 2 diabetes and arterial hypertension and induces their remission in morbid obesity: a 4-year case-controlled study. Diabetes Care 2005;28:2703 9. [12] Dixon JB, O Brien PE. Health outcomes of severely obese type 2 diabetic subjects 1 year after laparoscopic adjustable gastric banding. Diabetes Care 2002;25:358 63. [13] Greenway SE, Greenway FL III, Klein S. Effects of obesity surgery on non-insulin-dependent diabetes mellitus. Arch Surg 2002;137:1109 17. [14] Hubbard VS, Hall WH. Gastrointestinal surgery for severe obesity. Obes Surg 1991;1:257 65. [15] Metropolitan Height and Weight Tables. New York: Metropolitan Life Foundation. Stat Bull 1983;64:2 9. [16] Deitel M, Gawdat K, Melissas J. Reporting weight loss 2007. Obes Surg 2007;17:565 8. [17] Longmore M. Wilkinson IB, Rajagopala SR. Oxford handbook of clinical medicine. 6th ed. Oxford, UK: Oxford University Press; 2004. [18] Management of Hypertension in Adults in Primary Care; NICE, June 2006. [19] Singhal R, Kitchen M, Ndirika S, Hunt K, Bridgwater S, Super P. The Birmingham stitch avoiding slippage in laparoscopic gastric banding. Obes Surg 2008;18:359 63. [20] Sjöström L, Lindroos AK, Peltonen M, et al. Lifestyle, diabetes, and cardiovascular risk factors 10 years after bariatric surgery. N Engl J Med 2004;351:2683 93. [21] Buchwald H, Estok R, Fahrbach K, et al. Weight and type 2 diabetes after bariatric surgery: systematic review and meta-analysis. Am J Med 2009;122:248 56, e5.