Diabetes, Obesity and Metabolism 18: 217–223, 2016. © 2015 John Wiley & Sons Ltd
Hypoglycaemia after gastric bypass: mechanisms and treatment P. Ritz, C. Vaurs, M. Barigou & H. Hanaire Pôle Cardio-vasculaire et métabolique, Centre Intégré de l’obésité, Inserm U1027, CHU de Toulouse, Hôpital Larrey, Toulouse, France
Hypoglycaemia after gastric bypass can be severe, but is uncommon, and is sometimes only revealed through monitoring glucose concentrations. The published literature is limited by the heterogeneity of the criteria used for diagnosis, arguing in favour of the Whipple triad with a glycaemia threshold of 55 mg/dl as the diagnostic reference. Women who lost most of their excess weight after gastric bypass, long after the surgery was performed, and who did not have diabetes before surgery are at the greatest risk. In this context, hypoglycaemia results from hyperinsulinism, which is either generated by pancreas anomalies (nesidioblastosis) and/or caused by an overstimulation of 𝛽 cells by incretins, mainly glucagon-like peptide-1 (GLP-1). Glucose absorption is both accelerated and increased because of the direct communication between the gastric pouch and the jejunum. This is a post-surgical exaggeration of a natural adaptation that is seen in patients who have not undergone surgery in whom glucose is infused directly into the jejunum. There is not always a correspondence between symptoms and biological traits; however, hyperinsulinism is constant if hypoglycaemia is severe and there are neuroglucopenic symptoms. The treatment relies firstly on changes in eating habits, splitting food intake into five to six daily meals, slowing gastric emptying, reducing the glycaemic load and glycaemic index of foods, using fructose and avoiding stress at meals. Pharmacological treatment with acarbose is efficient, but other drugs still need to be validated in a greater number of subjects (insulin, glucagon, calcium channel blockers, somatostatin analogues and GLP-1 analogues). Lastly, if the surgical option has to be used, the benefits (efficient symptom relief) and the risks (weight regain, diabetes) should be weighed carefully. Keywords: bariatric surgery, continuous glucose monitoring, glucose load, hyperinsulinism, hypoglycaemia Date submitted 26 June 2015; date of first decision 1 August 2015; date of final acceptance 17 October 2015
Introduction In 2005, two publications discussed severe hypoglycaemia in people who had undergone gastric bypass surgery [1,2]. Today, if we cross-reference the terms ‘hypoglycaemia’ and ‘gastric bypass’ in Pubmed, we find 91 articles in humans, including 10 that are reviews of the literature. In a recent editorial in the journal Surgery for Obesity and Related Diseases, Roslin [3] provides a very good account of the impact of this phenomenon. Severe hypoglycaemia after gastric bypass is well known to physicians and surgeons, but much less so to patients. It can have serious consequences. Low glycaemia levels observed using continuous glucose monitoring are much more common among operated patients than severe episodes, and frequently asymptomatic. Given the major benefits of this surgery, the question of the impact of this adverse side effect on patients’ health and quality of life is important. The present review had three objectives: to analyse the incidence and prevalence of hypoglycaemia after gastric bypass; to analyse the mechanisms responsible for hypoglycaemia; and to analyse the therapeutic options.
Incidence and Prevalence Cryer [4] has contributed to the clarification of the definition of hypoglycaemia. The clinical symptoms of hypoglycaemia are Correspondence to: Patrick Ritz, Pôle Cardio-vasculaire et métabolique, Centre Intégré de l’obésité, Inserm U1027, CHU de Toulouse, Hôpital Larrey, 26 Chemin de Pouvourville, 31059 Toulouse Cedex 9, France. E-mail:
[email protected]
non-specific, therefore, a positive diagnosis has to be made in a context when many symptoms can occur associated with other complications (e.g. dumping syndrome, fistula). Documentation of Whipple’s triad [(i) symptoms, signs, or both, consistent with hypoglycaemia; (ii) a low measured plasma glucose concentration; and (iii) resolution of those symptoms and signs after the plasma glucose concentration is raised] is then particularly important. Symptoms and signs of hypoglycaemia are shown Table 1. Neuroglucopenic symptoms are a direct result of brain glucose deprivation, and can have dramatic consequences, such as car accidents. In the context of diabetes, a severe neuroglucopenic episode is an event requiring the assistance of another person to actively administer carbohydrate, glucagon or other resuscitative actions. Most of the episodes after bariatric surgery with neuroglucopenic symptoms can therefore be considered severe. Neurogenic (or autonomic) symptoms are largely the result of the perception of the physiological changes caused by sympathoadrenal discharge triggered by hypoglycaemia. They are considered responsible for subjective awareness, allowing an intervention before the onset of neuroglucopenic symptoms. The special issue of unawareness is discussed below. Episodes of severe hypoglycaemia are uncommon. Their prevalence is estimated to be between 0.2 and 0.36%, based on >8000 patients in cohorts or registries [6,7]. In a US study using the BOLD registry and including >100 000 bypasses, the prevalence of the symptoms declared by the patients and compatible with hypoglycaemia was 0.02–0.1% [8]. That is twice to seven times the rate observed with other surgeries [8]. The intensity and severity of the clinical presentation of
REVIEW ARTICLE
review article
review article
DIABETES, OBESITY AND METABOLISM
Table 1. Clinical presentation of hypoglycaemia after bariatric surgery [5]. Clinical presentation
Description/definition
Whipple triad
(1) Symptoms, signs, or both, consistent with hypoglycaemia; (2) a low measured plasma glucose concentration; (3) resolution of those symptoms and signs after the plasma glucose concentration is raised A direct result of brain glucose deprivation. Symptoms include cognitive impairments, behavioural changes, psychomotor abnormalities, and, at lower glucose levels, seizure and coma. Symptoms can be feeling warm, weak, having difficulty thinking, feeling confused, being tired or drowsy, sensation of faint, dizziness, having difficulty speaking, complaining of blurred vision. Largely the result of the perception of physiological changes caused by sympathoadrenal (particularly the sympathetic neural) discharge triggered by hypoglycaemia. They include adrenergic (catecholamine-mediated) symptoms such as palpitations, tremor, and anxiety/arousal and cholinergic (acetylcholine-mediated) symptoms such as sweating, hunger and paresthesias. An event requiring assistance of another person to actively administer carbohydrate, glucagon, or other resuscitative actions. This has emerged from the diabetes field, when hypoglycaemia is likely and the recognition of the symptoms by the patients is very important.
Neuroglycopenic symptoms
Neurogenic (or autonomic) symptoms
Severe episode
severe hypoglycaemia varies. Using a clinical hypoglycaemia self-questionnaire (including the Edinburgh glycaemia scale) sent by post, Lee et al. [9] reported a prevalence of 34.2% (for 1119 patients, with a 40.2% response rate). Fifty-two patients (11.5%) had severe hypoglycaemia, which is much higher than the prevalence described above. Another presentation of hypoglycaemia is low glycaemia after a glucose load, with or without symptoms. This is often reported, but there is confusion about the threshold that defines hypoglycaemia. Some authors use 2.8 mmol/l (50 mg/dl) [10,11] and others use 3.3 mmol/l (60 mg/dl) [10,12]. Hence, the prevalence of a low glycaemia value after receiving a glucose load [test meal or 75-g oral glucose tolerance test (OGTT)] varies greatly, between 10.4 and 68%. This variability in prevalence also depends on the glycaemic load, which varies among studies from ∼30 g of glucose in a test meal to 100 g in certain OGTTs [10]. The incidence of low glycaemia has been reported by Pigeyre et al. [11], based on the “Atlas Biologique de l’Obésité Sévère” registry. By choosing patients with normal preoperative OGTT results (neither diabetes, nor values
