Obesity is one of the most significant health problems worldwide, and the prevalence has been increasing over the past decade. Despite improvement in the performance of bariatric surgery, complications are not uncommon. These complications vary according to baseline patient characteristics, the duration of time since the operation, and the type of bariatric surgery performed. Endoscopy is the cornerstone in the diagnosis of postoperative complications after bariatric surgery, and may even be performed in the early postoperative course. With an increasing number of patients being referred for endoscopic evaluation following bariatric surgery, it is essential to develop an understanding of the anatomic changes for optimal assessment and appropriate treatment of these patients.
Early and late dumping syndrome occurs not uncommonly in patients who have undergone gastric bypass surgery when large quantities of simple carbohydrates are ingested. Early dumping typically occurs within 15 minutes of ingestion and has been attributed to rapid fluid shifts from the plasma into the bowel from hyperosmolality of the food. Late dumping occurs hours after eating and results from hyperglycemia and the subsequent insulin response leading to hypoglycemia. When hypoglycemia is severe, treatment with a low carbohydrate diet and an alphaglucosidase inhibitor may be effective. Furthermore, restoration of gastric restriction using an endoscopic approach to reduce the aperture of the GJA has also demonstrated to be effective in management of this condition.
The initial management of dumping syndrome is dietary modifications. Recommendations include consuming smaller meals by dividing daily calorie intake into six meals and delaying liquids at least 30 min after meals Rapidly absorbable simple carbohydrates should also be avoided. Adjuncts to diet modification include pectin and guar gum, which slow down gastric emptying by increasing food viscosity. Acarbose, which interferes with carbohydrate absorption in the small intestines, has also proven to relieve symptoms in small studies. After dietary modifications, medications such as somatostatin analogs (e.g., octreotide) alleviate symptoms by delaying gastric emptying and small bowel transit time, as well as inhibiting gastric hormones and insulin secretion. Multiple studies have evaluated both short- and longterm somatostatin therapies, with results showing sustained symptom control in patients refractory to dietary modifications. In severe cases refractory to medical management, surgical interventions, such as narrowing of the anastomosis, conversion of the prior bariatric surgery, and using jejunostomy parenteral feeding, may help. Follow-up with gastrointestinal specialists and the patient’s bariatric surgeon is strongly recommended if dumping syndrome is suspected.
An important metabolic complication which is attracting increasing interest is postprandial hyperinsulinemic hypoglycemia (PHH), characterized by hypoglycemic symptoms developing 1–3 h after a meal accompanied by a low blood glucose level. This condition should be distinguished from early dumping syndrome where symptoms develop within minutes to 1 h after a meal of caloric dense food, caused by the rapid and unregulated emptying of food into the jejunum, which induces rapid fluid entry into the small bowel. Early dumping often occurs early in the postoperative period, most commonly after Roux-en-Y gastric bypass, whereas PHH may develop months to years after surgery.
Symptoms related to post-PHH usually develop late after surgery in contrast to early dumping. Symptoms are wide ranging, but are usually related to Whipple’s triad: symptomatic hypoglycemia, a low plasma glucose level, and resolution of symptoms after the administration of glucose. Symptoms of hypoglycemia may include anxiety, sweating, tremors, palpitations, confusion, weakness, lightheadedness, dizziness, blurred vision, disorientation, and possibly loss of consciousness.
Because of variability in degree of symptoms and the absence of a clear pathophysiology, management of this condition can be challenging. Fortunately, a significant percentage of patients with milder forms of the condition can be managed with dietary modifications consisting of frequent small meals with a low glycemic index. This requires supervision by a dietitian and long-term patient compliance. Additional benefit has been obtained by the addition of acarbose, an α glucosidase inhibitor in doses 100–300 mg. Successful management has been also reported in case reports or small case series with diazoxide, calcium channel blockers, and somatostatin analogues. The role of GLP-1 in the pathogenesis of this condition is supported by the observation that infusions of GLP-1 antagonists corrected hypoglycemia in these patients. These agents are investigational at present, but provide opportunity for additional future treatment approaches. For patients with persistent symptoms despite medical treatment, reversal of the bariatric procedure should be considered. Partial pancreatectomy, although used in the past, is now not recommended because of the significant morbidity and poor long-term symptom control. Postprandial hyperinsulinemic hypoglycemia is an important, potentially dangerous late complication of metabolic surgery. Successful diagnosis and management of this condition requires multidisciplinary specialty resources and essential long-term follow-up capabilities.
Bariatric procedures differ in their ability to ameliorate T2DM, with intestinal bypass procedures generally associated with greater glycemic control and remission rates than purely restrictive procedures. There has been until now a paucity of data from RCTs comparing the efficacy of various bariatric procedures to treat diabetes. The recently published RCT by Schauer et al. also indicates superior efficacy of RYGB over sleeve gastrectomy in the treatment of diabetes in obese individuals. On the other hand, BPD produced greater remission of diabetes in morbidly obese patients compared to RYGB (95 % versus 75 %) in the RCT reported by Mingrone et al.
Sleeve Gastrectomy as Metabolic Surgery
Karamanakos et al. showed that LSG suppressed fasting and postprandial ghrelin levels and attributed this decrease in ghrelin to improved postoperative satiety and greater weight loss at 1year compared to LRYGB. The LRYGB group in this study had an initial decrease in ghrelin levels after surgery, but these levels returned to normal levels within 3 months. Lee et al. studied the treatment of patients with a low body mass index and type 2 diabetes mellitus between the two groups. LRYGB is reportedly more effective than LSG; they conclude that both procedures have strong hindgut effects after surgery, but LRYGB has a significant duodenal exclusion effect on cholecystokinin. The LSG group had lower acylated ghrelin and des-acylated ghrelin levels but greater concentrations of resistin than the LRYGB group. In addition to evaluations of ghrelin, there are now several small studies demonstrating that gastric emptying is increased after sleeve gastrectomy. The loss of a large reservoir in the gastric fundus and body and preservation of the antral pump provide a reasonable explanation for this finding. A secondary effect of earlier distal bowel stimulation with nutrients after meals due to increased gastric emptying time may be similar to the effects seen after gastric bypass.
Several mechanistic studies have demonstrated early and exaggerated postprandial peak levels of Peptide YY3–36 and GLP-1 after LSG. GLP-1 is an incretin that stimulates insulin production and releases from pancreatic islet cells, and the increased PYY3–36 results in satiety and reduced food intake. Karamanakos et al. have independently shown that the sleeve gastrectomy does have the effect of increasing the transit time of chyme despite an intact pylorus as measured by increased postprandial PYY levels.
Peterli et al. performed a randomized prospective trial with 13 LRYGB and 14 LSG patients to investigate the potential mechanism of LSG focusing on foregut and hindgut mechanisms. They found marked improvement in glucose homeostasis 1 week after surgery in both groups. This improvement was associated with early, exaggerated increases in GLP-1 secretion at 1 week, 3 months, and 1 year postoperatively in both groups. In addition to changes in GLP-1, PYY3–36 increased significantly and ghrelin was suppressed in both groups. It is unclear whether PYY3–36 has a direct effect on glucose homeostasis or if its effects are exhibited via appetite reduction and concomitant weight loss. Preoperatively, some patients had a blunted PYY3–36 and GLP-1 response suggesting some “resistance” to these gut hormones in obese patients. These findings suggest that the LSG should not be viewed merely as a restrictive procedure but also as a procedure that has neurohormonal and incretin effects.
Gastric Bypass versus Laparoscopic Sleeve Gastrectomy
Ramon et al. compared the effects of LRYGB and LSG on glucose metabolism and levels of gastrointestinal hormones such as ghrelin, leptin, GLP-1, peptide YY (PYY), and pancreatic polypeptide (PP) in morbid obese patients. This prospective, randomized study confirmed that the postprandial response of ghrelin, GLP-1, and PYY was maintained in patients undergoing LSG for 12 months after surgery and was similar to the LRYGB group results. A prospective, randomized study by Woelnerhanssen et al. compared the 1-year results of LRYGB and LSG for weight loss, metabolic control, and fasting adipokine levels. The authors confirmed a close association of specific adipokines with obesity and with the changes observed with weight loss after two different bariatric surgical procedures. The concentrations of circulating leptin levels decreased by almost 50 % as early as 1 week postoperatively and continued to decrease until 12 months postoperatively and adiponectin increased progressively. No differences were found between the LRYGB and LSG groups regarding adipokine changes.
How to choice a procedure?
The choice of procedure is an important determinant of outcome with a decreasing gradient of efficacy predicted from BPD, RYGB to SG and then
LAGB. Other factors that have been positively correlated with diabetes remission are percentage of excess weight loss (% EWL), younger age, lower preop HbA1c, and shorter duration of diabetes (less than 5 years). Severity of diabetes, as judged by preop treatment modality, has also been noted to be a significant factor.
Schauer et al. have reported in their series of 191 obese diabetic patients (the majority of whom were on oral agents or insulin) a diabetes remission rate of 97 % in diet-controlled, 87 % in oral agent treated, and 62 % in insulin-treated subjects. This was also confirmed by a recent retrospective analysis of 505 morbidly obese diabetic patients who underwent RYGB. In this study, a more recent diagnosis of T2DM and the absence of preoperative insulin therapy were significant predictors of remission, independent of the percentage of EWL.
Dixon et al. have recently identified diabetes duration < 4 years, BMI > 35 kg/m2, and fasting c-peptide concentration > 2.9 ng/ mL as three clinically useful cutoffs and independent preoperative predictors of remission after analyzing the outcomes of 154 ethnic Chinese subjects after gastric bypass. C-peptide > 3 ng/mL has also previously been shown to be an important predictor of diabetes resolution after sleeve gastrectomy in non-morbidly obese diabetic subjects by Lee et al.