- Email: [email protected]
Should metabolic surgery be offered in morbidly obese patients with type I diabetes?
Surgery for Obesity and Related Diseases ] (2015) 00–00
Original article
Should metabolic surgery be offered in morbidly obese patients with type I diabetes? Maud Robert, M.D., Ph.D.a,b,*, Pascale Belanger, M.Sc.c, Frédéric Simon Hould, M.D.a, Simon Marceau, M.D.a, André Tchernof, Ph.D.c,d, Laurent Biertho, M.D.a a
Department of Surgery, Division of Bariatric and General Surgery, Institut Universitaire de Cardiologie et de Pneumologie de Québec, Québec, Canada b Hospices civils de Lyon, Université Lyon I, France. c Department of Nutrition, Laval University, Quebec, Canada d Institut Universitaire de Cardiologie et de Pneumologie de Québec, Quebec, Canada Received September 17, 2014; accepted December 17, 2014
Abstract
Background: Bariatric surgery has recently been endorsed as an effective treatment of type 2 diabetes (T2D) in severely obese patients. Little is known about its metabolic effects in type 1 diabetes (T1D). The objectives were to assess bariatric surgery outcomes in T1D obese patients who underwent either a biliopancreatic diversion (BPD) or a sleeve gastrectomy (SG) and who were matched with T2D obese patients requiring insulin therapy. The study setting was at a universityaffiliated tertiary care center. Methods: Through retrospective analysis of prospectively collected data, ten patients with T1D (7 BPD, 3 SG) were matched with 20 patients with T2D (14 BPD, 6 SG) according to age, gender, type of surgery, initial Body Mass Index and insulin requirements (1:2 matching). Weight loss, diabetes control, and remission of co-morbidities were compared. Results: Mean follow-up was 55.1 months. Mean Excess BMI Loss% tended to be greater in T1D patients compared to T2D (77.1% versus 68.3%, P ¼ .14). The remission and improvement rates of T2D were 55% and 45% versus 0% and 90% for T1D. The remission rate of T2D was significantly greater after BPD (71.43%) compared to SG (16.67%, P ¼ .04). Insulin requirements were significantly reduced in both groups after surgery (T1D: .44 ⫾ .24 versus 1.09 ⫾ .7 units/kg/d, P ¼ 0.03, T2D: .03 ⫾ .12 versus .89 ⫾ .77 units/kg/d, P ¼ .0001). Remission rates of hypertension and dyslipidemia were similar for T1D and T2D (66.7% versus 62.5%, P ¼ .63 and 88.9% versus 75%, P ¼ .23). Conclusions: Even if metabolic surgery has limited effect on glycemic control in T1D, it improves insulin sensitivity and other co-morbidities. It should be considered as a therapeutic option in selected obese patients with metabolic syndrome and high cardio-vascular risk. (Surg Obes Relat Dis 2015;]:00–00.) r 2015 American Society for Metabolic and Bariatric Surgery. All rights reserved.
Keywords:
Type 1 diabetes; Obesity; Metabolic surgery; Glycemic control; Insulin requirements; Weight loss; Biliopancreatic diversion; Sleeve gastrectomy
* Correspondence: Maud Robert, M.D., Ph.D., Department of Surgery, Division of Bariatric and General Surgery, Institut Universitaire de Cardiologie et de Pneumologie de Québec, 2725 Chemin Quatre-Bourgeois, G1 V 4 G5, Québec, Canada. Tel: þ418 656 8711; fax: þ418 656 4825. E-mail: [email protected]
Bariatric surgery has recently been endorsed as an effective treatment for type 2 diabetes (T2D) in severely obese patients, first by the American Diabetes Association [1] and then by the International Diabetes Federation [2]. Indeed, literature data reported a T2D remission rate of
http://dx.doi.org/10.1016/j.soard.2014.12.016 1550-7289/r 2015 American Society for Metabolic and Bariatric Surgery. All rights reserved.
M. Robert et al. / Surgery for Obesity and Related Diseases ] (2015) 00–00
2
74.6%, reaching 95.9% after biliopancreatic diversion [3]. In 2013, 2 randomized clinical trials demonstrated that bariatric surgery was more effective than medical therapy alone to achieve T2D remission [4,5], data that were recently confirmed at 3 years of follow-up [6]. Several mechanisms have been involved to explain diabetes remission including weight loss, gut hormone changes, modifications in gut microbiota, and bile acids metabolism [7–9]. The prevalence of obesity is now growing in type 1 diabetes population, making it more difficult to achieve good glycemic control [10]. The increase of insulin requirement results in further weight gain and the occurrence of comorbidities, leading to a higher cardiovascular risk. Therefore, the management of these patients not only requires a lowering of glycemia but also metabolic syndrome control, leading to change in the therapeutic strategies. The option of bariatric surgery has rarely been considered and data are remarkably scarce. The aim of this study was to assess bariatric surgery outcomes in T1D obese patients. From a series of ten T1D patients who underwent either a biliopancreatic diversion with duodenal switch (BPD/DS) or a sleeve gastrectomy (SG), and matched with 20 T2D obese patients requiring insulin, we sought to analyze the impact of the type of diabetes and the type of surgical procedure on weight loss results, diabetes control and remission of co-morbidities. Methods Population and study design From 2004 to May 2013, 3365 obese patients underwent a bariatric procedure in the department of bariatric surgery of the Quebec Heart and Lung Institute, Canada: 2662 BPD, 703 SG. We selected T1D patients (n ¼ 10) from our prospective electronic database dedicated to bariatric surgery. The diagnosis of T1D was made by endocrinologists, either by the presence of autoantibodies against glutamic
acid decarboxylase (GAD) (n ¼ 2) or, when diabetes was diagnosed in adolescence, by an array of clinical symptoms suggestive of T1D (ketoacidosis as initial presentation, n ¼ 2), a family history of T1D and abnormal glucose level that required immediate treatment with insulin. The median age at diagnosis was 13.8 years (SD ¼ 11.4). It is worth noting that among the patients with autoantibodies, one woman had a late diagnosis of diabetes, at 39 years old. She was on high dose of insulin from the beginning of the disease (390 units of insulin per day) and was diagnosed as latent autoimmune diabetes in adults. American Diabetes Criteria were used to define type 2 diabetes: HbA1 c Z6.5%, fasting plasma glucose Z 126 mg/dL, 2 hours plasma glucose 4200 mg/dL during an OGTT, classic symptoms of hyperglycemia or hyperglycemic crisis, random plasma glucose Z200 mg/dL [11]. These patients had relative insulin deficiency and were on both oral antidiabetic drugs (mean of 1.6 antidiabetic drugs, range 1–3) and high dose of insulin (mean of 124.9 units per day ⫾ 106.1). Mean duration of the disease was 7.3 ⫾ 3.6 years. Each of the 10 immune-mediated diabetes patients were matched with 2 obese T2D patients on insulin (n ¼ 20), according to age, gender, type of surgery, initial body mass index (BMI) and insulin requirements (1:2 matching). The aim of our study was to compare bariatric surgery outcomes in both types of diabetes, not only regarding diabetes control but also in terms of weight loss outcomes and evolution of other co-morbidities (obstructive sleep apnea syndrome, arterial hypertension, dyslipidemia). Baseline characteristics of the study groups are shown in Table 1. Surgical Techniques Laparoscopic Sleeve Gastrectomy. A 5-port technique was used. A long and narrow gastric tube calibrated with a 34French bougie was performed using a linear stapler and began at 5 cm from the pylorus up to the angle of His.
Table 1 Baseline characteristics of the study groups (1:2 matching)
*
BPD ratio Mean Age at surgery (years)† Female ratio* Mean initial BMI (kg/m2)† Mean Preoperative Insulin units‡ Preoperative HbA1 c (%)‡ Preoperative fasting plasma glucose (mmol/l)† Mean age at diagnosis (years)† Mean duration of diabetes (years)‡ Abbreviation: BMI ¼ Body mass index. Data are expressed as mean ⫾ SD where applicable. * Chi-square analysis. † Student's t test. ‡ Wilcoxon test.
Type 1 Diabetes n ¼ 10
Type 2 Diabetes n ¼ 20
7/10 39.2 ⫾ 5.3 8/10 46.9 ⫾ 6.3 141.4 ⫾ 94.2 7.5 ⫾ 1.9 10.5 ⫾ 3.4 13.8 ⫾ 11.4 23.1 ⫾ 11.8
14/20 40.9 ⫾ 4.2 16/20 49.4 ⫾ 5.5 124.9 ⫾ 106.1 8.1 ⫾ 0.8 8.7 ⫾ 3.1 33.6 ⫾ 6.6 7.3 ⫾ 3.6
P 1 .34 1 .28 .38 .73 .17 o.0001 .0009
Metabolic Surgery in Type 1 Diabetes Patients / Surgery for Obesity and Related Diseases ] (2015) 00–00
3
Laparoscopic Biliopancreatic Diversion with Duodenal Switch. A 6-port technique was used. A gastric sleeve was performed as described previously. The duodenum was divided 3 cm beyond the pylorus. Starting at the ileocaecal junction (ICJ), a common limb of 1 meter and an alimentary limb of 1.5 meters were measured. The bowel was transected 2.5 meters from the ICJ. An ileo-ileostomy was made using a linear stapler at the 1 meter mark. An end to side manual duodeno-ileostomy was made using running sutures. Mesenteric defects were closed with nonabsorbable sutures.
When considering the type of surgical procedure, we found a better weight loss with BPD compared to SG, which was nearly statistically significant in terms of aWL (49.5 ⫾ 9.6 kg versus 39.1 ⫾ 19.5 kg respectively, P ¼ .06) (cf. Table 2). Evolution of weight over time in each diabetes group, as a function of the type of surgical procedure is presented in Fig. 1.
Data analysis
Impact of the type of diabetes
Data were recorded preoperatively, at 3, 6, 12, 18, 24 months, and every year after surgery. Weight loss was expressed in excess weight loss % (EWL%), excess BMI loss % (EBL%), absolute weight loss in kg (aWL) and BMI loss in kg/m². Diabetes control was assessed using HbA1 c %, fasting glycemia (in mmol/l), insulin requirements and number of oral antidiabetic drugs. Type 2 diabetes remission was defined as cessation of antidiabetic treatments and HbA1 c o6% and/or fasting blood glucose levels o100 mg/dL, according to the American Diabetes Association (ADA) guidelines. Evolution of arterial hypertension (HTN), dyslipidemia, and obstructive sleep apnea syndrome were also analyzed. Improvement of these co-morbidities was considered established when medication was reduced, and considered in remission when medication was stopped. Sleep apnea was considered cured when apnea-hypopnea index (AHI) was below 15 (number of apnea-hypopnea episodes per hour) and improved when postoperative AHI decreased.
In T1D, we did not observe any statistical difference between preoperative HbA1 c (7.5 ⫾ 0.8%) and last HbA1 c (7.1 ⫾ 0.9% at the end of follow-up) (repeated measures ANOVA, P ¼ .52). However, in T2D, last HbA1 c (5.5 ⫾ 1%) was significantly lower than preoperative HbA1 c (8.1 ⫾ 1.9%, P ¼ 0.0001). Moreover, last HbA1 c was significantly lower in T2D group (5.5%) versus T1D (7.1%, P ¼ .0005) showing a stronger effect of surgery in T2D on glycemic control. When analyzing the effect of bariatric surgery on HbA1 c variation in each group, we found a significant difference (repeated measures ANOVA, P ¼ 0.01) with a stronger effect of surgery in T2D patients. In a similar manner, whereas fasting glycemia at the last follow-up was not significantly different from preoperative values in T1D patients (7.9 ⫾ 2.2 versus 10.3 ⫾ 3.4 mmol/ l, P ¼ .17 respectively), we observed a significant difference between post- and preoperative values in the T2D subgroup (6 ⫾ 1.7 versus 9.3 ⫾ 3.1 mmol/l, P ¼ .014 respectively). Insulin requirements were also significantly reduced in both groups after surgery: for T1D, .44 ⫾ .24 units/kg/d at the last follow-up versus 1.09 ⫾ 0.7 units/kg/d in the preoperative period (P ¼ 0.03), and for T2D, .03 ⫾ 0.12
Statistical analysis JMP statistical software 4.0 (SAS institute, Carry, NC) was used for all analyses. Quantitative data were compared between groups by a Student t test. When the distribution of variables was nonnormal, the nonparametric test of Wilcoxon was used. Ordinal data were compared between groups by a Χ2 test. To assess the effect of surgery with time, data were analyzed using repeated measures analysis of variance. Logistic regression analyses were performed with several dependent variables to assess the odds of diabetes remission at the last follow-up visit. P o .05 was considered significant.
2- Evolution of Diabetes control
Table 2 Weight loss outcomes according to the type of diabetes (type 1 diabetes versus type 2 diabetes) and according to the type of surgical procedure (biliopancreatic diversion versus sleeve gastrectomy) Type of diabetes
T1D n ¼ 10
T2D n ¼ 20
P (t test)
mean mean mean mean
70.17 (18.63) 16.56 (4.99) 46.14 (15.54) 77.12 (4.83)
62.78 (10.47) 16.53 (4.39) 46.47 (13.33) 68.31 (3.42)
.17 .99 .95 .15
Type of surgery
BPD n ¼ 21
SG n ¼ 9
1- Weight loss outcomes
mean mean mean mean
66.86 (13.54) 17.37 (3.51) 49.45 (9.59) 72.72 (15.37)
61.48 (14.67) 14.61 (6.07) 39.14 (19.51) 67.79 (16.51)
Weight loss outcomes were not statistically different between T1D and T2D even if we observed a trend for a greater EBL% in the T1D group (77.1 % versus 68.3%, P ¼ .14)
Abbreviations: T1D ¼ type 1 diabetes; T2D ¼ type 2 diabetes; BPD ¼ biliopancreatic diversion; SG ¼ sleeve gastrectomy; EWL ¼ excess weight loss percentage; aWL ¼ absolute weight loss; EBL% ¼ excess BMI loss percentage; ( ) ¼ SD.
Results Mean follow-up was 55.1 months (range 14-115, median 53.6). None of the patients were lost to follow-up.
EWL% BMI loss (kg/m2) aWL (kg) EBL%
EWL% BMI loss (kg/m2) aWL (kg) EBL%
.34 .13 .06 .44
4
M. Robert et al. / Surgery for Obesity and Related Diseases ] (2015) 00–00
Fig. 1. Evolution of weight over time in each diabetes group, function of the type of surgical procedure. EWL% ¼ Excess Weight Loss %; T1D ¼ type 1 diabetes (n ¼ 10); T2D ¼ type 2 diabetes (n ¼ 20). Data are expressed as mean ⫾ SD.
units/kg/d at the last follow-up versus .89 ⫾ .77 units before surgery (P ¼ .0001). As expected, we found that the postoperative dose of insulin was significantly lower in T2D group (.03 Units/kg/d) compared to T1D (.44 Units/ kg/d, P ¼ .0001).
The evolution of HbA1 c and the evolution of insulin requirements (in units/kg/d) after each procedure are shown in Fig. 2 and Fig. 3.
Impact of EWL% on diabetes control Impact of the surgical technique The overall remission rate of diabetes was 47.6% in the BPD group versus 11.1% in the SG group (ChiSquare test, P ¼ .05). Considering only the T2D group, we found that the remission rate was significantly higher in the BPD group (71%) compared to the SG group (17%, P ¼ .05). Insulin requirements as well as HbA1 c % at the last follow-up were significantly lower in the BPD group (0 unit; 5.2%) compared to the SG group (10 units, P ¼ .032; 6.3%, P ¼ .017). Remission and improvement rates of diabetes and the evolution of insulin requirements for each surgical technique are shown in Table 3.
Considering both type of diabetes, we did not find any difference in terms of mean EWL% in the group in remission (n ¼ 11) versus the improved group (n ¼ 18) (66.3% ⫾ 9.6 versus 65.9% ⫾ 15.2). In the T2D group, mean EWL% was 66.3% ⫾ 9.6 for patients in remission versus 58.4% ⫾ 10.3 for patients with improvement and it was not statistically significant (P ¼ .09). In the T1D group, mean EWL% was 73.5% ⫾ 16.2 in patients with improvement of their diabetes (n ¼ 9). One patient had no improvement after surgery, and as expected, none was in remission. Interestingly, in T1D patients, we found a negative correlation between insulin requirements (in units/kg/d) and final EWL% after surgery illustrated in Fig. 4 (Pearson correlation: r ¼ .6, P ¼ .06).
Table 3 Remission rates of diabetes and evolution of insulin requirements as a function of the type of diabetes and the type of surgical procedure Type 1 Diabetes
Type 2 Diabetes Diabetes outcomes
Remission
Improved
Biliopancreatic Diversion n ¼ 21 Sleeve Gastrectomy n ¼ 9
0/7 (0%) 0/3 (0%)
6/7 (86%) 3/3 (100%)
Biliopancreatic Diversion n ¼ 21 Sleeve Gastrectomy n ¼ 9
Before surgery 1.15 ⫾ 0.81 0.96 ⫾ 0.48
Unchanged
Remission
1/7 (14%) 10/14 (71%) 0/3 (0%) 1/6 (17%) Mean Insulin requirements ⫾ SD (units/kg/d) After surgery Before surgery 0.41 ⫾ 0.28 0.81 ⫾ 0.81 0.52 ⫾ 0.14 1.1 ⫾ 0.52
Improved
Unchanged
4/14 (29%) 5/6 (83%)
0/14 (0%) 0/6 (0%)
After surgery 0 0.13 ⫾ 0.2
Metabolic Surgery in Type 1 Diabetes Patients / Surgery for Obesity and Related Diseases ] (2015) 00–00
5
Fig. 2. Evolution of HbA1 c after biliopancreatic diversion and sleeve gastrectomy. T1D ¼ type 1 diabetes (n ¼ 10); T2D ¼ type 2 diabetes (n ¼ 20). Data are expressed as mean ⫾ SD.
Predictors of diabetes remission Using logistic regression analyses, and analyzing factors associated with remission rather than improvement of diabetes (Χ2 test) we did not find any influence of age (P ¼ .6), gender (P ¼ .48), initial BMI (P ¼ .09), EWL% (0.77), EBL% (P ¼ .91), and preoperative insulin requirements (P ¼ .91) on diabetes remission. Nevertheless, we found that diabetes duration as well as the duration of preoperative insulin treatment tended to be negatively correlated to diabetes remission (P ¼ .05 and P ¼ .09
respectively). We also found that the type of surgical procedure (BPD/DS) was a predictive factor of diabetes remission (P ¼ .05). 3- Evolution of co-morbidities. The remission rates of arterial hypertension (66.7% versus 62.5%, P ¼ .63) and dyslipidemia (88.9% versus 75%, P ¼ .22) were similar in T1D and T2D. When considering the type of surgical procedure, although we did not observe any difference in terms of
Fig. 3. Evolution of insulin requirements after biliopancreatic diversion and sleeve gastrectomy. T1D ¼ type 1 diabetes (n ¼ 10); T2D ¼ type 2 diabetes (n ¼ 20). Data are expressed as mean ⫾ SD.
M. Robert et al. / Surgery for Obesity and Related Diseases ] (2015) 00–00
6
Fig. 4. Correlation between insulin requirements (in units/kg/d) and final EWL% after bariatric surgery in type 1 diabetics (Pearson correlation: r ¼ 0.6, P ¼ .06).
arterial hypertension remission between BPD and SG (72.2% versus 42.9%, P ¼ .21), we found that the remission rate of dyslipidemia was significantly greater in the BPD group (88.2% versus 62.5%, P ¼ .01). Evolution of co-morbidities according to the type of diabetes and to the type of surgical procedure is presented in Table 4. Discussion Data assessing the impact of metabolic surgery in patients with T1D are clearly lacking: first of all, because T1D phenotype was not associated until recently with obesity, secondarily because the rapid rise in postprandial glycemia especially observed after gastric bypass was believed to increase the risk of hyperinsulinemic hypoglycemia in T1D patients, a feature that would theoretically create difficulties in optimizing insulin therapy in the postprandial period [12]. Nevertheless, we are now witnessing changes in the patterns of the disease with an increasing prevalence of obesity also in the T1D population [13]. About 50% of type 1 diabetic patients are now overweight or obese, and 8 to 40% of them meet the metabolic syndrome criteria, resulting in a high risk of cardiovascular disease, which is now the leading cause of death in this population [10]. Until now, we only found a few publications in the literature
reporting the results of bariatric surgery in T1D patients and most of them are case series or case reports [14–21]. These are retrospective studies with heterogeneous data and shortterm follow-up. Raab et al. in 2013 reported a series of 6 patients who underwent either BPD (n = 3), Roux-en-Y gastric bypass (RYGBP; n = 2), or SG (n = 1) [16]. Consistent with our results, the authors found a decrease in insulin requirements. They also found a decrease in HbA1 c levels after one year of follow-up. Nevertheless, in this study, diagnostic modalities of the type of diabetes were not specified, especially the presence of autoantibodies or C-peptide levels. Considering that most of these patients had late-onset diabetes, the possibility that they were in fact T2D patients requiring insulin cannot be formally excluded. A recent series reported 9 cases of obese women with type 1 diabetes who underwent a RYGB [14]. In this short-term study (7 weeks after surgery), the authors also reported a decrease in HbA1 c levels and basal insulin requirements (-38%) whereas none of the patients was able to discontinue insulin, consistent with the absolute insulin deficiency signing type 1 diabetes. This year, Brethauer et al. reported in a letter to the editor 10 cases of T1D patients who underwent either a RYGB (n ¼ 7), a gastric banding (n ¼ 2), or a sleeve gastrectomy (n ¼ 1) [20]. Consistent with our data, they found a remarkable weight loss and significant mean reduction in daily insulin requirements at a mean follow-up of 36.8 ⫾ 32.3 months. They also reported favorable changes in lipid profile and a 71% resolution rate of arterial hypertension. On the other hand, the authors found a significant reduction in HbA1 c after surgery what we did not observe in our series. We did not also observe any significant reduction in fasting glycemia in T1D patients suggesting that glycemic control is still difficult to achieve after bariatric surgery, even after BPD/DS which is considered as the most effective metabolic procedure. Indeed, the glycemic effect of surgery in T1D remains controversial and other authors have already reported no significant improvement [16,18]. More recently, Lannoo et al. reported 22 cases of bariatric surgery in T1D obese patients (16 RYGB and 6 SG): consistent with our data, the authors found a significant decrease in insulin dose at a
Table 4 Evolution of co-morbidities after bariatric surgery Dyslipidemia
Biliopancreatic Diversion
Sleeve Gastrectomy
Arterial Hypertension
Obstructive Sleep Apnea
preop
postop
preop
postop
preop
postop
T1D n ¼ 7
6 (86%)
6 (86%)
4 R (100%)
11 (78%)
10 (71%)
10 R (100%)
T1D n ¼ 3
3 (100%)
3 (100%)
3 R (100%)
T2D n ¼ 6
5 (83%)
2 I (40%)
4 (67%)
4 2 9 2 2 1 1 3
4 (57%)
T2D n ¼ 14
5 R (83%) 1 I (17%) 10 R (91%) 1 I (9%) 3 R (100%)
3 (50%)
3 R (100%)
12 (86%) 3 (100%)
Abbreviations: T1D ¼ type 1 diabetes; T2D ¼ type 2 diabetes; R ¼ remission; I ¼ improvement.
R (67%) I (33%) R (75%) I (17%) R (67%) I (33%) R (25%) I (75%)
Metabolic Surgery in Type 1 Diabetes Patients / Surgery for Obesity and Related Diseases ] (2015) 00–00
mean follow-up of 37.8 ⫾ 29.7 months after surgery but no substantial improvement of glycemic control as assessed by HbA1 c [21]. In our series, as expected, we clearly demonstrated a stronger effect of metabolic surgery on T2D compared to T1D on glycemic control: HbA1 c levels as well as glycemia decreased in both groups after surgery but these results were only statistically significant in T2D group. Nevertheless, insulin requirements were significantly reduced in both groups (.44 ⫾ .24 versus 1.09 ⫾ .7 units/ kg/d for T1D (P ¼ .03), and .03 ⫾ .12 versus .89 ⫾ .77 units/kg/d for T2D (P ¼ .0001), confirming that metabolic surgery increases insulin sensitivity whatever the type of diabetes. Several hypotheses have been put forward to explain the reduction of insulin requirements after bariatric surgery even in T1D patients who do not benefit from an increase of endogenous secretion: weight loss resulting in a decrease in insulin resistance, reduced caloric intake due to the restrictive component of the surgery, improvement of hepatic insulin sensitivity [14,22]. In our study, even if we did not find any correlation between EWL% and diabetes remission, we observed in the T1D subgroup a negative correlation between EWL% and insulin requirements, suggesting that weight loss might play a role in the improvement of insulin sensitivity after surgery. Dirksen et al. also reported an increase in GLP-1, PYY, and GIP after a mixed meal test in one obese woman with T1D who underwent a RYGB [15]. The authors concluded that an exaggerated secretion of anorexigenic gut hormones may contribute to the substantial weight loss and that the increase in GLP1 could improve glucose metabolism through inhibition of glucagon secretion and slowing of gastric emptying. It is interesting to note that, in our series, the woman who was diagnosed with latent autoimmune diabetes in adults and who had positive autoantibodies against GAD as well as 390 units of insulin per day was weaned off insulin 2 years after the BPD. She is currently on 2 oral antidiabetic drugs. We speculate that metabolic surgery prevented progression to total insulin deficiency and helped to preserve Beta cell mass and function in this case. Regarding weight loss outcomes, we found a trend for a greater EBL% in the T1D group although it was not statistically significant (77.1% versus 68.3% for T2D, P ¼ .15). We could hypothesize that this slight difference may be due to a better compliance with dietary recommendations in T1D patients. Not surprisingly, we also confirmed that patients with BPD lose more weight than those with SG, which was almost significant in terms of aWL (49.5 kg versus 39.1 kg, P ¼ .06): it is worth noting that the follow-up is shorter for the SG group and we can expect that these results will become significant as the follow-up increases. Considering the type of surgical procedure, we found that T2D remission rate was significantly higher in the BPD group compared to the SG group (71.4 % versus 16.7 %, P ¼ .05), which is consistent with literature data
7
[3]. We also found that BPD was significantly more effective in resolving dyslipidemia than SG whatever the type of diabetes. As the remission rate of co-morbidities was similar in both groups of diabetes, we could suggest that BPD is an interesting option in type 1 diabetes to control metabolic syndrome and to improve cardiovascular risk. This is all the more interesting in that BPD, contrary to RYGB, does not expose patients to dumping syndrome and could lower the risk of hyperinsulinemic hypoglycemia. It is also interesting to note that even if we do not have quality of life questionnaires for all our patients, they all anecdotally reported an improvement in their quality of life after bariatric surgery, which was related both to weight loss and improvement of diabetes control. We presented here the largest series of BPD/DS in type 1 diabetes patients, which is known as the most efficient procedure to achieve weight loss and resolve comorbidities. It also has the advantage of preserving the pylorus, which avoids the risk of dumping syndrome and large variations in postprandial glycemia, which could be difficult to manage in type 1 diabetes patients. This study has also the longest follow-up reported (mean ¼ 55.1 months). Another interest of this work is the study design including 1:2 matching with T2D patients. The rigorous selection of the patients allowed us to compare bariatric surgery outcomes in both types of diabetes, in terms of glycemic control but also in terms of weight loss and resolution of other co-morbidities. It is interesting to note the same drop in insulin requirements after surgery in both type of diabetes suggesting an improvement of insulin sensitivity in both groups. HbA1 c was also initially decreased but it did not last overtime in type 1 diabetes patients compared to type 2 diabetes group, confirming that it is still difficult to obtain glycemic control in type 1 diabetes after surgery. Another originality of our study was to compare the impact of the surgical procedure, restrictive versus metabolic (sleeve gastrectomy versus BPD), in both groups of diabetes, confirming that BPD has a stronger effect on weight loss, insulin requirements and comorbidities. Nevertheless, this study has some limitations: it is a retrospective study, with a limited number of patients, and although the diagnosis of type 1 diabetes was made in adolescence by the endocrinologist, the C-peptide assessment was not available for all the patients. In conclusion, our findings are consistent with previous studies and showed a significant decrease in insulin requirements in T1D obese patients after bariatric surgery, with a nonsignificant improvement of HbA1 c levels and glycemia. We also found that weight loss and the improvement rate of co-morbidities were similar in both type of diabetes. Even if metabolic surgery has limited effect on glycemic control in T1D, it improves insulin sensitivity through weight loss and seems therefore interesting as a therapeutic option in selected T1D obese patients. Metabolic surgery appears as
8
M. Robert et al. / Surgery for Obesity and Related Diseases ] (2015) 00–00
a complementary treatment helping to optimize medical therapy and the management strategy of T1D, and may decrease cardiovascular complications. Prospective studies are required to confirm these data and it is our hope this paper stimulates interest toward this goal. Disclosures The authors confirm having no commercial associations that might be a conflict of interest in relation to this article. References [1] American Diabetes A. Executive summary: standards of medical care in diabetes–2011. Diabetes Care 2011;34(Suppl 1):S4–10. [2] Schauer PR, Rubino F. International Diabetes Federation position statement on bariatric surgery for type 2 diabetes: implications for patients, physicians, and surgeons. Surg Obes Relat Dis 2011;7:448–51. [3] 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. [4] Schauer PR, Kashyap SR, Wolski K, et al. Bariatric surgery versus intensive medical therapy in obese patients with diabetes. N Engl J Med. 2012;366:1567–76. [5] Mingrone G, Panunzi S, De Gaetano A, et al. Bariatric surgery versus conventional medical therapy for type 2 diabetes. N Engl J Med 2012;366:1577–85. [6] Schauer PR, Bhatt DL, Kirwan JP, et al. Bariatric surgery versus intensive medical therapy for diabetes–3-year outcomes. N Engl J Med 2014;370:2002–13. [7] Bose M, Olivan B, Teixeira J, Pi-Sunyer FX, Laferrere B. Do Incretins play a role in the remission of type 2 diabetes after gastric bypass surgery: What are the evidence? Obes Surg 2009;19:217–29. [8] Furet JP, Kong LC, Tap J, et al. Differential adaptation of human gut microbiota to bariatric surgery-induced weight loss: links with metabolic and low-grade inflammation markers. Diabetes 2010;59:3049–57. [9] Ahmad NN, Pfalzer A, Kaplan LM. Roux-en-Y gastric bypass normalizes the blunted postprandial bile acid excursion associated with obesity. Int J Obes (Lond) 2013;37:1553–9.
[10] Chillaron JJ, Flores Le-Roux JA, Benaiges D, Pedro-Botet J. Type 1 diabetes, metabolic syndrome and cardiovascular risk. Metabolism 2014;63:181–7. [11] American Diabetes A. Diagnosis and classification of diabetes mellitus. Diabetes Care 2010;33(Suppl 1):S62–9: [12] Goldfine AB, Mun EC, Devine E, et al. Patients with neuroglycopenia after gastric bypass surgery have exaggerated incretin and insulin secretory responses to a mixed meal. J Clin Endocrinol Metab 2007;92:4678–85. [13] Conway B, Miller RG, Costacou T, et al. Temporal patterns in overweight and obesity in Type 1 diabetes. Diabet Med 2010;27: 398–404. [14] Middelbeek RJ, James-Todd T, Patti ME, Brown FM. Short-term insulin requirements following gastric bypass surgery in severely obese women with type 1 diabetes. Obes Surg. 2014;24:1442–6. [15] Dirksen C, Jacobsen SH, Bojsen-Moller KN, et al. Reduction in cardiovascular risk factors and insulin dose, but no beta-cell regeneration 1 year after Roux-en-Y gastric bypass in an obese patient with type 1 diabetes: a case report. Obes Res Clin Pract 2013;7: e269–74. [16] Raab H, Weiner RA, Frenken M, Rett K, Weiner S. Obesity and metabolic surgery in type 1 diabetes mellitus. Nutr Hosp 2013;28 (Suppl 2):31–4. [17] Chuang J, Zeller MH, Inge T, Crimmins N. Bariatric surgery for severe obesity in two adolescents with type 1 diabetes. Pediatrics 2013;132:e1031–4. [18] Mendez CE, Tanenberg RJ, Pories W. Outcomes of Roux-en-Y gastric bypass surgery for severely obese patients with type 1 diabetes: a case series report. Diabetes Metab Syndr Obes 2010;3: 281–3. [19] Czupryniak L, Wiszniewski M, Szymanski D, et al. Long-term results of gastric bypass surgery in morbidly obese type 1 diabetes patients. Obes Surg 2010;20:506–8. [20] Brethauer SA, Aminian A, Rosenthal RJ, et al. Bariatric surgery improves the metabolic profile of morbidly obese patients with type 1 diabetes. Diabetes Care 2014;37. (e51–2). [21] Lannoo M, Dillemans B, Van Nieuwenhove Y, et al. Bariatric surgery induces weight loss but does not improve glycemic control in patients with type 1 diabetes. Diabetes Care 2014;37:e173–4. [22] Lim EL, Hollingsworth KG, Aribisala BS, et al. Reversal of type 2 diabetes: normalisation of beta cell function in association with decreased pancreas and liver triacylglycerol. Diabetologia 2011;54: 2506–14.
Original article
Should metabolic surgery be offered in morbidly obese patients with type I diabetes? Maud Robert, M.D., Ph.D.a,b,*, Pascale Belanger, M.Sc.c, Frédéric Simon Hould, M.D.a, Simon Marceau, M.D.a, André Tchernof, Ph.D.c,d, Laurent Biertho, M.D.a a
Department of Surgery, Division of Bariatric and General Surgery, Institut Universitaire de Cardiologie et de Pneumologie de Québec, Québec, Canada b Hospices civils de Lyon, Université Lyon I, France. c Department of Nutrition, Laval University, Quebec, Canada d Institut Universitaire de Cardiologie et de Pneumologie de Québec, Quebec, Canada Received September 17, 2014; accepted December 17, 2014
Abstract
Background: Bariatric surgery has recently been endorsed as an effective treatment of type 2 diabetes (T2D) in severely obese patients. Little is known about its metabolic effects in type 1 diabetes (T1D). The objectives were to assess bariatric surgery outcomes in T1D obese patients who underwent either a biliopancreatic diversion (BPD) or a sleeve gastrectomy (SG) and who were matched with T2D obese patients requiring insulin therapy. The study setting was at a universityaffiliated tertiary care center. Methods: Through retrospective analysis of prospectively collected data, ten patients with T1D (7 BPD, 3 SG) were matched with 20 patients with T2D (14 BPD, 6 SG) according to age, gender, type of surgery, initial Body Mass Index and insulin requirements (1:2 matching). Weight loss, diabetes control, and remission of co-morbidities were compared. Results: Mean follow-up was 55.1 months. Mean Excess BMI Loss% tended to be greater in T1D patients compared to T2D (77.1% versus 68.3%, P ¼ .14). The remission and improvement rates of T2D were 55% and 45% versus 0% and 90% for T1D. The remission rate of T2D was significantly greater after BPD (71.43%) compared to SG (16.67%, P ¼ .04). Insulin requirements were significantly reduced in both groups after surgery (T1D: .44 ⫾ .24 versus 1.09 ⫾ .7 units/kg/d, P ¼ 0.03, T2D: .03 ⫾ .12 versus .89 ⫾ .77 units/kg/d, P ¼ .0001). Remission rates of hypertension and dyslipidemia were similar for T1D and T2D (66.7% versus 62.5%, P ¼ .63 and 88.9% versus 75%, P ¼ .23). Conclusions: Even if metabolic surgery has limited effect on glycemic control in T1D, it improves insulin sensitivity and other co-morbidities. It should be considered as a therapeutic option in selected obese patients with metabolic syndrome and high cardio-vascular risk. (Surg Obes Relat Dis 2015;]:00–00.) r 2015 American Society for Metabolic and Bariatric Surgery. All rights reserved.
Keywords:
Type 1 diabetes; Obesity; Metabolic surgery; Glycemic control; Insulin requirements; Weight loss; Biliopancreatic diversion; Sleeve gastrectomy
* Correspondence: Maud Robert, M.D., Ph.D., Department of Surgery, Division of Bariatric and General Surgery, Institut Universitaire de Cardiologie et de Pneumologie de Québec, 2725 Chemin Quatre-Bourgeois, G1 V 4 G5, Québec, Canada. Tel: þ418 656 8711; fax: þ418 656 4825. E-mail: [email protected]
Bariatric surgery has recently been endorsed as an effective treatment for type 2 diabetes (T2D) in severely obese patients, first by the American Diabetes Association [1] and then by the International Diabetes Federation [2]. Indeed, literature data reported a T2D remission rate of
http://dx.doi.org/10.1016/j.soard.2014.12.016 1550-7289/r 2015 American Society for Metabolic and Bariatric Surgery. All rights reserved.
M. Robert et al. / Surgery for Obesity and Related Diseases ] (2015) 00–00
2
74.6%, reaching 95.9% after biliopancreatic diversion [3]. In 2013, 2 randomized clinical trials demonstrated that bariatric surgery was more effective than medical therapy alone to achieve T2D remission [4,5], data that were recently confirmed at 3 years of follow-up [6]. Several mechanisms have been involved to explain diabetes remission including weight loss, gut hormone changes, modifications in gut microbiota, and bile acids metabolism [7–9]. The prevalence of obesity is now growing in type 1 diabetes population, making it more difficult to achieve good glycemic control [10]. The increase of insulin requirement results in further weight gain and the occurrence of comorbidities, leading to a higher cardiovascular risk. Therefore, the management of these patients not only requires a lowering of glycemia but also metabolic syndrome control, leading to change in the therapeutic strategies. The option of bariatric surgery has rarely been considered and data are remarkably scarce. The aim of this study was to assess bariatric surgery outcomes in T1D obese patients. From a series of ten T1D patients who underwent either a biliopancreatic diversion with duodenal switch (BPD/DS) or a sleeve gastrectomy (SG), and matched with 20 T2D obese patients requiring insulin, we sought to analyze the impact of the type of diabetes and the type of surgical procedure on weight loss results, diabetes control and remission of co-morbidities. Methods Population and study design From 2004 to May 2013, 3365 obese patients underwent a bariatric procedure in the department of bariatric surgery of the Quebec Heart and Lung Institute, Canada: 2662 BPD, 703 SG. We selected T1D patients (n ¼ 10) from our prospective electronic database dedicated to bariatric surgery. The diagnosis of T1D was made by endocrinologists, either by the presence of autoantibodies against glutamic
acid decarboxylase (GAD) (n ¼ 2) or, when diabetes was diagnosed in adolescence, by an array of clinical symptoms suggestive of T1D (ketoacidosis as initial presentation, n ¼ 2), a family history of T1D and abnormal glucose level that required immediate treatment with insulin. The median age at diagnosis was 13.8 years (SD ¼ 11.4). It is worth noting that among the patients with autoantibodies, one woman had a late diagnosis of diabetes, at 39 years old. She was on high dose of insulin from the beginning of the disease (390 units of insulin per day) and was diagnosed as latent autoimmune diabetes in adults. American Diabetes Criteria were used to define type 2 diabetes: HbA1 c Z6.5%, fasting plasma glucose Z 126 mg/dL, 2 hours plasma glucose 4200 mg/dL during an OGTT, classic symptoms of hyperglycemia or hyperglycemic crisis, random plasma glucose Z200 mg/dL [11]. These patients had relative insulin deficiency and were on both oral antidiabetic drugs (mean of 1.6 antidiabetic drugs, range 1–3) and high dose of insulin (mean of 124.9 units per day ⫾ 106.1). Mean duration of the disease was 7.3 ⫾ 3.6 years. Each of the 10 immune-mediated diabetes patients were matched with 2 obese T2D patients on insulin (n ¼ 20), according to age, gender, type of surgery, initial body mass index (BMI) and insulin requirements (1:2 matching). The aim of our study was to compare bariatric surgery outcomes in both types of diabetes, not only regarding diabetes control but also in terms of weight loss outcomes and evolution of other co-morbidities (obstructive sleep apnea syndrome, arterial hypertension, dyslipidemia). Baseline characteristics of the study groups are shown in Table 1. Surgical Techniques Laparoscopic Sleeve Gastrectomy. A 5-port technique was used. A long and narrow gastric tube calibrated with a 34French bougie was performed using a linear stapler and began at 5 cm from the pylorus up to the angle of His.
Table 1 Baseline characteristics of the study groups (1:2 matching)
*
BPD ratio Mean Age at surgery (years)† Female ratio* Mean initial BMI (kg/m2)† Mean Preoperative Insulin units‡ Preoperative HbA1 c (%)‡ Preoperative fasting plasma glucose (mmol/l)† Mean age at diagnosis (years)† Mean duration of diabetes (years)‡ Abbreviation: BMI ¼ Body mass index. Data are expressed as mean ⫾ SD where applicable. * Chi-square analysis. † Student's t test. ‡ Wilcoxon test.
Type 1 Diabetes n ¼ 10
Type 2 Diabetes n ¼ 20
7/10 39.2 ⫾ 5.3 8/10 46.9 ⫾ 6.3 141.4 ⫾ 94.2 7.5 ⫾ 1.9 10.5 ⫾ 3.4 13.8 ⫾ 11.4 23.1 ⫾ 11.8
14/20 40.9 ⫾ 4.2 16/20 49.4 ⫾ 5.5 124.9 ⫾ 106.1 8.1 ⫾ 0.8 8.7 ⫾ 3.1 33.6 ⫾ 6.6 7.3 ⫾ 3.6
P 1 .34 1 .28 .38 .73 .17 o.0001 .0009
Metabolic Surgery in Type 1 Diabetes Patients / Surgery for Obesity and Related Diseases ] (2015) 00–00
3
Laparoscopic Biliopancreatic Diversion with Duodenal Switch. A 6-port technique was used. A gastric sleeve was performed as described previously. The duodenum was divided 3 cm beyond the pylorus. Starting at the ileocaecal junction (ICJ), a common limb of 1 meter and an alimentary limb of 1.5 meters were measured. The bowel was transected 2.5 meters from the ICJ. An ileo-ileostomy was made using a linear stapler at the 1 meter mark. An end to side manual duodeno-ileostomy was made using running sutures. Mesenteric defects were closed with nonabsorbable sutures.
When considering the type of surgical procedure, we found a better weight loss with BPD compared to SG, which was nearly statistically significant in terms of aWL (49.5 ⫾ 9.6 kg versus 39.1 ⫾ 19.5 kg respectively, P ¼ .06) (cf. Table 2). Evolution of weight over time in each diabetes group, as a function of the type of surgical procedure is presented in Fig. 1.
Data analysis
Impact of the type of diabetes
Data were recorded preoperatively, at 3, 6, 12, 18, 24 months, and every year after surgery. Weight loss was expressed in excess weight loss % (EWL%), excess BMI loss % (EBL%), absolute weight loss in kg (aWL) and BMI loss in kg/m². Diabetes control was assessed using HbA1 c %, fasting glycemia (in mmol/l), insulin requirements and number of oral antidiabetic drugs. Type 2 diabetes remission was defined as cessation of antidiabetic treatments and HbA1 c o6% and/or fasting blood glucose levels o100 mg/dL, according to the American Diabetes Association (ADA) guidelines. Evolution of arterial hypertension (HTN), dyslipidemia, and obstructive sleep apnea syndrome were also analyzed. Improvement of these co-morbidities was considered established when medication was reduced, and considered in remission when medication was stopped. Sleep apnea was considered cured when apnea-hypopnea index (AHI) was below 15 (number of apnea-hypopnea episodes per hour) and improved when postoperative AHI decreased.
In T1D, we did not observe any statistical difference between preoperative HbA1 c (7.5 ⫾ 0.8%) and last HbA1 c (7.1 ⫾ 0.9% at the end of follow-up) (repeated measures ANOVA, P ¼ .52). However, in T2D, last HbA1 c (5.5 ⫾ 1%) was significantly lower than preoperative HbA1 c (8.1 ⫾ 1.9%, P ¼ 0.0001). Moreover, last HbA1 c was significantly lower in T2D group (5.5%) versus T1D (7.1%, P ¼ .0005) showing a stronger effect of surgery in T2D on glycemic control. When analyzing the effect of bariatric surgery on HbA1 c variation in each group, we found a significant difference (repeated measures ANOVA, P ¼ 0.01) with a stronger effect of surgery in T2D patients. In a similar manner, whereas fasting glycemia at the last follow-up was not significantly different from preoperative values in T1D patients (7.9 ⫾ 2.2 versus 10.3 ⫾ 3.4 mmol/ l, P ¼ .17 respectively), we observed a significant difference between post- and preoperative values in the T2D subgroup (6 ⫾ 1.7 versus 9.3 ⫾ 3.1 mmol/l, P ¼ .014 respectively). Insulin requirements were also significantly reduced in both groups after surgery: for T1D, .44 ⫾ .24 units/kg/d at the last follow-up versus 1.09 ⫾ 0.7 units/kg/d in the preoperative period (P ¼ 0.03), and for T2D, .03 ⫾ 0.12
Statistical analysis JMP statistical software 4.0 (SAS institute, Carry, NC) was used for all analyses. Quantitative data were compared between groups by a Student t test. When the distribution of variables was nonnormal, the nonparametric test of Wilcoxon was used. Ordinal data were compared between groups by a Χ2 test. To assess the effect of surgery with time, data were analyzed using repeated measures analysis of variance. Logistic regression analyses were performed with several dependent variables to assess the odds of diabetes remission at the last follow-up visit. P o .05 was considered significant.
2- Evolution of Diabetes control
Table 2 Weight loss outcomes according to the type of diabetes (type 1 diabetes versus type 2 diabetes) and according to the type of surgical procedure (biliopancreatic diversion versus sleeve gastrectomy) Type of diabetes
T1D n ¼ 10
T2D n ¼ 20
P (t test)
mean mean mean mean
70.17 (18.63) 16.56 (4.99) 46.14 (15.54) 77.12 (4.83)
62.78 (10.47) 16.53 (4.39) 46.47 (13.33) 68.31 (3.42)
.17 .99 .95 .15
Type of surgery
BPD n ¼ 21
SG n ¼ 9
1- Weight loss outcomes
mean mean mean mean
66.86 (13.54) 17.37 (3.51) 49.45 (9.59) 72.72 (15.37)
61.48 (14.67) 14.61 (6.07) 39.14 (19.51) 67.79 (16.51)
Weight loss outcomes were not statistically different between T1D and T2D even if we observed a trend for a greater EBL% in the T1D group (77.1 % versus 68.3%, P ¼ .14)
Abbreviations: T1D ¼ type 1 diabetes; T2D ¼ type 2 diabetes; BPD ¼ biliopancreatic diversion; SG ¼ sleeve gastrectomy; EWL ¼ excess weight loss percentage; aWL ¼ absolute weight loss; EBL% ¼ excess BMI loss percentage; ( ) ¼ SD.
Results Mean follow-up was 55.1 months (range 14-115, median 53.6). None of the patients were lost to follow-up.
EWL% BMI loss (kg/m2) aWL (kg) EBL%
EWL% BMI loss (kg/m2) aWL (kg) EBL%
.34 .13 .06 .44
4
M. Robert et al. / Surgery for Obesity and Related Diseases ] (2015) 00–00
Fig. 1. Evolution of weight over time in each diabetes group, function of the type of surgical procedure. EWL% ¼ Excess Weight Loss %; T1D ¼ type 1 diabetes (n ¼ 10); T2D ¼ type 2 diabetes (n ¼ 20). Data are expressed as mean ⫾ SD.
units/kg/d at the last follow-up versus .89 ⫾ .77 units before surgery (P ¼ .0001). As expected, we found that the postoperative dose of insulin was significantly lower in T2D group (.03 Units/kg/d) compared to T1D (.44 Units/ kg/d, P ¼ .0001).
The evolution of HbA1 c and the evolution of insulin requirements (in units/kg/d) after each procedure are shown in Fig. 2 and Fig. 3.
Impact of EWL% on diabetes control Impact of the surgical technique The overall remission rate of diabetes was 47.6% in the BPD group versus 11.1% in the SG group (ChiSquare test, P ¼ .05). Considering only the T2D group, we found that the remission rate was significantly higher in the BPD group (71%) compared to the SG group (17%, P ¼ .05). Insulin requirements as well as HbA1 c % at the last follow-up were significantly lower in the BPD group (0 unit; 5.2%) compared to the SG group (10 units, P ¼ .032; 6.3%, P ¼ .017). Remission and improvement rates of diabetes and the evolution of insulin requirements for each surgical technique are shown in Table 3.
Considering both type of diabetes, we did not find any difference in terms of mean EWL% in the group in remission (n ¼ 11) versus the improved group (n ¼ 18) (66.3% ⫾ 9.6 versus 65.9% ⫾ 15.2). In the T2D group, mean EWL% was 66.3% ⫾ 9.6 for patients in remission versus 58.4% ⫾ 10.3 for patients with improvement and it was not statistically significant (P ¼ .09). In the T1D group, mean EWL% was 73.5% ⫾ 16.2 in patients with improvement of their diabetes (n ¼ 9). One patient had no improvement after surgery, and as expected, none was in remission. Interestingly, in T1D patients, we found a negative correlation between insulin requirements (in units/kg/d) and final EWL% after surgery illustrated in Fig. 4 (Pearson correlation: r ¼ .6, P ¼ .06).
Table 3 Remission rates of diabetes and evolution of insulin requirements as a function of the type of diabetes and the type of surgical procedure Type 1 Diabetes
Type 2 Diabetes Diabetes outcomes
Remission
Improved
Biliopancreatic Diversion n ¼ 21 Sleeve Gastrectomy n ¼ 9
0/7 (0%) 0/3 (0%)
6/7 (86%) 3/3 (100%)
Biliopancreatic Diversion n ¼ 21 Sleeve Gastrectomy n ¼ 9
Before surgery 1.15 ⫾ 0.81 0.96 ⫾ 0.48
Unchanged
Remission
1/7 (14%) 10/14 (71%) 0/3 (0%) 1/6 (17%) Mean Insulin requirements ⫾ SD (units/kg/d) After surgery Before surgery 0.41 ⫾ 0.28 0.81 ⫾ 0.81 0.52 ⫾ 0.14 1.1 ⫾ 0.52
Improved
Unchanged
4/14 (29%) 5/6 (83%)
0/14 (0%) 0/6 (0%)
After surgery 0 0.13 ⫾ 0.2
Metabolic Surgery in Type 1 Diabetes Patients / Surgery for Obesity and Related Diseases ] (2015) 00–00
5
Fig. 2. Evolution of HbA1 c after biliopancreatic diversion and sleeve gastrectomy. T1D ¼ type 1 diabetes (n ¼ 10); T2D ¼ type 2 diabetes (n ¼ 20). Data are expressed as mean ⫾ SD.
Predictors of diabetes remission Using logistic regression analyses, and analyzing factors associated with remission rather than improvement of diabetes (Χ2 test) we did not find any influence of age (P ¼ .6), gender (P ¼ .48), initial BMI (P ¼ .09), EWL% (0.77), EBL% (P ¼ .91), and preoperative insulin requirements (P ¼ .91) on diabetes remission. Nevertheless, we found that diabetes duration as well as the duration of preoperative insulin treatment tended to be negatively correlated to diabetes remission (P ¼ .05 and P ¼ .09
respectively). We also found that the type of surgical procedure (BPD/DS) was a predictive factor of diabetes remission (P ¼ .05). 3- Evolution of co-morbidities. The remission rates of arterial hypertension (66.7% versus 62.5%, P ¼ .63) and dyslipidemia (88.9% versus 75%, P ¼ .22) were similar in T1D and T2D. When considering the type of surgical procedure, although we did not observe any difference in terms of
Fig. 3. Evolution of insulin requirements after biliopancreatic diversion and sleeve gastrectomy. T1D ¼ type 1 diabetes (n ¼ 10); T2D ¼ type 2 diabetes (n ¼ 20). Data are expressed as mean ⫾ SD.
M. Robert et al. / Surgery for Obesity and Related Diseases ] (2015) 00–00
6
Fig. 4. Correlation between insulin requirements (in units/kg/d) and final EWL% after bariatric surgery in type 1 diabetics (Pearson correlation: r ¼ 0.6, P ¼ .06).
arterial hypertension remission between BPD and SG (72.2% versus 42.9%, P ¼ .21), we found that the remission rate of dyslipidemia was significantly greater in the BPD group (88.2% versus 62.5%, P ¼ .01). Evolution of co-morbidities according to the type of diabetes and to the type of surgical procedure is presented in Table 4. Discussion Data assessing the impact of metabolic surgery in patients with T1D are clearly lacking: first of all, because T1D phenotype was not associated until recently with obesity, secondarily because the rapid rise in postprandial glycemia especially observed after gastric bypass was believed to increase the risk of hyperinsulinemic hypoglycemia in T1D patients, a feature that would theoretically create difficulties in optimizing insulin therapy in the postprandial period [12]. Nevertheless, we are now witnessing changes in the patterns of the disease with an increasing prevalence of obesity also in the T1D population [13]. About 50% of type 1 diabetic patients are now overweight or obese, and 8 to 40% of them meet the metabolic syndrome criteria, resulting in a high risk of cardiovascular disease, which is now the leading cause of death in this population [10]. Until now, we only found a few publications in the literature
reporting the results of bariatric surgery in T1D patients and most of them are case series or case reports [14–21]. These are retrospective studies with heterogeneous data and shortterm follow-up. Raab et al. in 2013 reported a series of 6 patients who underwent either BPD (n = 3), Roux-en-Y gastric bypass (RYGBP; n = 2), or SG (n = 1) [16]. Consistent with our results, the authors found a decrease in insulin requirements. They also found a decrease in HbA1 c levels after one year of follow-up. Nevertheless, in this study, diagnostic modalities of the type of diabetes were not specified, especially the presence of autoantibodies or C-peptide levels. Considering that most of these patients had late-onset diabetes, the possibility that they were in fact T2D patients requiring insulin cannot be formally excluded. A recent series reported 9 cases of obese women with type 1 diabetes who underwent a RYGB [14]. In this short-term study (7 weeks after surgery), the authors also reported a decrease in HbA1 c levels and basal insulin requirements (-38%) whereas none of the patients was able to discontinue insulin, consistent with the absolute insulin deficiency signing type 1 diabetes. This year, Brethauer et al. reported in a letter to the editor 10 cases of T1D patients who underwent either a RYGB (n ¼ 7), a gastric banding (n ¼ 2), or a sleeve gastrectomy (n ¼ 1) [20]. Consistent with our data, they found a remarkable weight loss and significant mean reduction in daily insulin requirements at a mean follow-up of 36.8 ⫾ 32.3 months. They also reported favorable changes in lipid profile and a 71% resolution rate of arterial hypertension. On the other hand, the authors found a significant reduction in HbA1 c after surgery what we did not observe in our series. We did not also observe any significant reduction in fasting glycemia in T1D patients suggesting that glycemic control is still difficult to achieve after bariatric surgery, even after BPD/DS which is considered as the most effective metabolic procedure. Indeed, the glycemic effect of surgery in T1D remains controversial and other authors have already reported no significant improvement [16,18]. More recently, Lannoo et al. reported 22 cases of bariatric surgery in T1D obese patients (16 RYGB and 6 SG): consistent with our data, the authors found a significant decrease in insulin dose at a
Table 4 Evolution of co-morbidities after bariatric surgery Dyslipidemia
Biliopancreatic Diversion
Sleeve Gastrectomy
Arterial Hypertension
Obstructive Sleep Apnea
preop
postop
preop
postop
preop
postop
T1D n ¼ 7
6 (86%)
6 (86%)
4 R (100%)
11 (78%)
10 (71%)
10 R (100%)
T1D n ¼ 3
3 (100%)
3 (100%)
3 R (100%)
T2D n ¼ 6
5 (83%)
2 I (40%)
4 (67%)
4 2 9 2 2 1 1 3
4 (57%)
T2D n ¼ 14
5 R (83%) 1 I (17%) 10 R (91%) 1 I (9%) 3 R (100%)
3 (50%)
3 R (100%)
12 (86%) 3 (100%)
Abbreviations: T1D ¼ type 1 diabetes; T2D ¼ type 2 diabetes; R ¼ remission; I ¼ improvement.
R (67%) I (33%) R (75%) I (17%) R (67%) I (33%) R (25%) I (75%)
Metabolic Surgery in Type 1 Diabetes Patients / Surgery for Obesity and Related Diseases ] (2015) 00–00
mean follow-up of 37.8 ⫾ 29.7 months after surgery but no substantial improvement of glycemic control as assessed by HbA1 c [21]. In our series, as expected, we clearly demonstrated a stronger effect of metabolic surgery on T2D compared to T1D on glycemic control: HbA1 c levels as well as glycemia decreased in both groups after surgery but these results were only statistically significant in T2D group. Nevertheless, insulin requirements were significantly reduced in both groups (.44 ⫾ .24 versus 1.09 ⫾ .7 units/ kg/d for T1D (P ¼ .03), and .03 ⫾ .12 versus .89 ⫾ .77 units/kg/d for T2D (P ¼ .0001), confirming that metabolic surgery increases insulin sensitivity whatever the type of diabetes. Several hypotheses have been put forward to explain the reduction of insulin requirements after bariatric surgery even in T1D patients who do not benefit from an increase of endogenous secretion: weight loss resulting in a decrease in insulin resistance, reduced caloric intake due to the restrictive component of the surgery, improvement of hepatic insulin sensitivity [14,22]. In our study, even if we did not find any correlation between EWL% and diabetes remission, we observed in the T1D subgroup a negative correlation between EWL% and insulin requirements, suggesting that weight loss might play a role in the improvement of insulin sensitivity after surgery. Dirksen et al. also reported an increase in GLP-1, PYY, and GIP after a mixed meal test in one obese woman with T1D who underwent a RYGB [15]. The authors concluded that an exaggerated secretion of anorexigenic gut hormones may contribute to the substantial weight loss and that the increase in GLP1 could improve glucose metabolism through inhibition of glucagon secretion and slowing of gastric emptying. It is interesting to note that, in our series, the woman who was diagnosed with latent autoimmune diabetes in adults and who had positive autoantibodies against GAD as well as 390 units of insulin per day was weaned off insulin 2 years after the BPD. She is currently on 2 oral antidiabetic drugs. We speculate that metabolic surgery prevented progression to total insulin deficiency and helped to preserve Beta cell mass and function in this case. Regarding weight loss outcomes, we found a trend for a greater EBL% in the T1D group although it was not statistically significant (77.1% versus 68.3% for T2D, P ¼ .15). We could hypothesize that this slight difference may be due to a better compliance with dietary recommendations in T1D patients. Not surprisingly, we also confirmed that patients with BPD lose more weight than those with SG, which was almost significant in terms of aWL (49.5 kg versus 39.1 kg, P ¼ .06): it is worth noting that the follow-up is shorter for the SG group and we can expect that these results will become significant as the follow-up increases. Considering the type of surgical procedure, we found that T2D remission rate was significantly higher in the BPD group compared to the SG group (71.4 % versus 16.7 %, P ¼ .05), which is consistent with literature data
7
[3]. We also found that BPD was significantly more effective in resolving dyslipidemia than SG whatever the type of diabetes. As the remission rate of co-morbidities was similar in both groups of diabetes, we could suggest that BPD is an interesting option in type 1 diabetes to control metabolic syndrome and to improve cardiovascular risk. This is all the more interesting in that BPD, contrary to RYGB, does not expose patients to dumping syndrome and could lower the risk of hyperinsulinemic hypoglycemia. It is also interesting to note that even if we do not have quality of life questionnaires for all our patients, they all anecdotally reported an improvement in their quality of life after bariatric surgery, which was related both to weight loss and improvement of diabetes control. We presented here the largest series of BPD/DS in type 1 diabetes patients, which is known as the most efficient procedure to achieve weight loss and resolve comorbidities. It also has the advantage of preserving the pylorus, which avoids the risk of dumping syndrome and large variations in postprandial glycemia, which could be difficult to manage in type 1 diabetes patients. This study has also the longest follow-up reported (mean ¼ 55.1 months). Another interest of this work is the study design including 1:2 matching with T2D patients. The rigorous selection of the patients allowed us to compare bariatric surgery outcomes in both types of diabetes, in terms of glycemic control but also in terms of weight loss and resolution of other co-morbidities. It is interesting to note the same drop in insulin requirements after surgery in both type of diabetes suggesting an improvement of insulin sensitivity in both groups. HbA1 c was also initially decreased but it did not last overtime in type 1 diabetes patients compared to type 2 diabetes group, confirming that it is still difficult to obtain glycemic control in type 1 diabetes after surgery. Another originality of our study was to compare the impact of the surgical procedure, restrictive versus metabolic (sleeve gastrectomy versus BPD), in both groups of diabetes, confirming that BPD has a stronger effect on weight loss, insulin requirements and comorbidities. Nevertheless, this study has some limitations: it is a retrospective study, with a limited number of patients, and although the diagnosis of type 1 diabetes was made in adolescence by the endocrinologist, the C-peptide assessment was not available for all the patients. In conclusion, our findings are consistent with previous studies and showed a significant decrease in insulin requirements in T1D obese patients after bariatric surgery, with a nonsignificant improvement of HbA1 c levels and glycemia. We also found that weight loss and the improvement rate of co-morbidities were similar in both type of diabetes. Even if metabolic surgery has limited effect on glycemic control in T1D, it improves insulin sensitivity through weight loss and seems therefore interesting as a therapeutic option in selected T1D obese patients. Metabolic surgery appears as
8
M. Robert et al. / Surgery for Obesity and Related Diseases ] (2015) 00–00
a complementary treatment helping to optimize medical therapy and the management strategy of T1D, and may decrease cardiovascular complications. Prospective studies are required to confirm these data and it is our hope this paper stimulates interest toward this goal. Disclosures The authors confirm having no commercial associations that might be a conflict of interest in relation to this article. References [1] American Diabetes A. Executive summary: standards of medical care in diabetes–2011. Diabetes Care 2011;34(Suppl 1):S4–10. [2] Schauer PR, Rubino F. International Diabetes Federation position statement on bariatric surgery for type 2 diabetes: implications for patients, physicians, and surgeons. Surg Obes Relat Dis 2011;7:448–51. [3] 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. [4] Schauer PR, Kashyap SR, Wolski K, et al. Bariatric surgery versus intensive medical therapy in obese patients with diabetes. N Engl J Med. 2012;366:1567–76. [5] Mingrone G, Panunzi S, De Gaetano A, et al. Bariatric surgery versus conventional medical therapy for type 2 diabetes. N Engl J Med 2012;366:1577–85. [6] Schauer PR, Bhatt DL, Kirwan JP, et al. Bariatric surgery versus intensive medical therapy for diabetes–3-year outcomes. N Engl J Med 2014;370:2002–13. [7] Bose M, Olivan B, Teixeira J, Pi-Sunyer FX, Laferrere B. Do Incretins play a role in the remission of type 2 diabetes after gastric bypass surgery: What are the evidence? Obes Surg 2009;19:217–29. [8] Furet JP, Kong LC, Tap J, et al. Differential adaptation of human gut microbiota to bariatric surgery-induced weight loss: links with metabolic and low-grade inflammation markers. Diabetes 2010;59:3049–57. [9] Ahmad NN, Pfalzer A, Kaplan LM. Roux-en-Y gastric bypass normalizes the blunted postprandial bile acid excursion associated with obesity. Int J Obes (Lond) 2013;37:1553–9.
[10] Chillaron JJ, Flores Le-Roux JA, Benaiges D, Pedro-Botet J. Type 1 diabetes, metabolic syndrome and cardiovascular risk. Metabolism 2014;63:181–7. [11] American Diabetes A. Diagnosis and classification of diabetes mellitus. Diabetes Care 2010;33(Suppl 1):S62–9: [12] Goldfine AB, Mun EC, Devine E, et al. Patients with neuroglycopenia after gastric bypass surgery have exaggerated incretin and insulin secretory responses to a mixed meal. J Clin Endocrinol Metab 2007;92:4678–85. [13] Conway B, Miller RG, Costacou T, et al. Temporal patterns in overweight and obesity in Type 1 diabetes. Diabet Med 2010;27: 398–404. [14] Middelbeek RJ, James-Todd T, Patti ME, Brown FM. Short-term insulin requirements following gastric bypass surgery in severely obese women with type 1 diabetes. Obes Surg. 2014;24:1442–6. [15] Dirksen C, Jacobsen SH, Bojsen-Moller KN, et al. Reduction in cardiovascular risk factors and insulin dose, but no beta-cell regeneration 1 year after Roux-en-Y gastric bypass in an obese patient with type 1 diabetes: a case report. Obes Res Clin Pract 2013;7: e269–74. [16] Raab H, Weiner RA, Frenken M, Rett K, Weiner S. Obesity and metabolic surgery in type 1 diabetes mellitus. Nutr Hosp 2013;28 (Suppl 2):31–4. [17] Chuang J, Zeller MH, Inge T, Crimmins N. Bariatric surgery for severe obesity in two adolescents with type 1 diabetes. Pediatrics 2013;132:e1031–4. [18] Mendez CE, Tanenberg RJ, Pories W. Outcomes of Roux-en-Y gastric bypass surgery for severely obese patients with type 1 diabetes: a case series report. Diabetes Metab Syndr Obes 2010;3: 281–3. [19] Czupryniak L, Wiszniewski M, Szymanski D, et al. Long-term results of gastric bypass surgery in morbidly obese type 1 diabetes patients. Obes Surg 2010;20:506–8. [20] Brethauer SA, Aminian A, Rosenthal RJ, et al. Bariatric surgery improves the metabolic profile of morbidly obese patients with type 1 diabetes. Diabetes Care 2014;37. (e51–2). [21] Lannoo M, Dillemans B, Van Nieuwenhove Y, et al. Bariatric surgery induces weight loss but does not improve glycemic control in patients with type 1 diabetes. Diabetes Care 2014;37:e173–4. [22] Lim EL, Hollingsworth KG, Aribisala BS, et al. Reversal of type 2 diabetes: normalisation of beta cell function in association with decreased pancreas and liver triacylglycerol. Diabetologia 2011;54: 2506–14.