Diabetes mellitus is a group of metabolic disorders, characterized by hyperglycemia, resulting from relative insulin deficiency, often on a background of insulin resistance (type 2 diabetes), or (near) absolute insulin deficiency related to autoimmune pathophysiology (type 1 diabetes). The most common form is type 2 diabetes, and its global prevalence is rising as a result of changes in lifestyle and lengthening of life expectancy.1,2 An estimated 463 million people aged 20 to 79 yr have diabetes,3 which corresponds to 9.3% of the adult population and represents a four-fold increase in diabetes prevalence since 1980.4 Patients with type 2 diabetes often have comorbidities such as arterial hypertension, obesity, ischemic heart disease, renal failure, and atherosclerosis. Ischemic heart disease in particular may affect younger patients compared to nondiabetic populations and may have silent ischemia. Hence, assessing the preoperative risk of diabetic patients is challenging, and can be underestimated. As shown outside the perioperative setting, these patients have equivalent or higher risk for cardiovascular events and mortality as patients with typical angina.5,6 However, current guidelines recommend against systematic stress testing in individuals without symptoms.7,8
The number of type 2 diabetes patients undergoing surgical procedures is rising worldwide. An American review from 2004 estimated that 15 to 20% of surgical patients are diabetic.4 Twenty-five percent of type 2 diabetes patients will require a surgical procedure during their lifetime9 related to chronic complications that affect the cardiovascular, ophthalmologic, renal or orthopedic systems. The risk of postoperative complications (i.e., gastroparesis, cardiovascular events, and postoperative infection) in type 2 diabetes patients is higher than in nondiabetic patients, as reported by large-scale studies.4,5,9–12 In order to adjust for potential confounders, the adverse outcomes occurring after major surgery have been compared with a propensity-matched cohort of non-diabetic patients in a cohort of more than 33,000 patients in Taiwan.13 In this study, Lin et al. found a higher risk of postoperative sepsis (odds ratio, 1.33; 95% CI, 1.01 to 1.74) and in-hospital mortality (odds ratio, 1.51; 95% CI, 1.07 to 2.13) when compared to nondiabetic patients. Karamanos et al.14 demonstrated higher mortality, cardiovascular and renal complications after urgent cholecystectomy in insulin-requiring diabetic patients than in orally treated diabetic patients, possibly pointing to worse outcomes in people with more advanced diabetes and/or more severe comorbidities.
Surgical stress induces insulin resistance and increased endogenous glucose production, resulting in stress hyperglycemia characterized by blood glucose level greater than 180 mg/dl (10 mM).15–18 When this persists, elevated glucose levels become deleterious, promoting immune dysfunction and susceptibility to infections, endothelial dysfunction and thrombosis that can culminate in stroke and acute myocardial infarction, and oxidative stress caused by enhanced reactive oxygen species production.19
In the intensive care setting, stress hyperglycemia, spontaneous or drug-related hypoglycemia (glucose less than 70 mg/dl [3.89 mM]) and high glycemic variability (coefficient of variation greater than 20%) constitute the three features of dysglycemia syndrome. All are associated with worse outcomes in acutely injured patients.20–22 Badawi et al.22 analyzed a large cohort of critically ill patients from 344 American intensive care units to evaluate the association between intensive care unit–acquired dysglycemia and in-hospital mortality. The results showed that hypoglycemia as well as hyperglycemia and glucose variability are deleterious to critically ill patients. The longer the duration of this variability, the greater is the impact on in-hospital mortality. The adjusted relative risk (95% CI) of mortality for hyperglycemia (glucose, 180 to 240 mg/dl [10 to 13.3 mM]) was 1.63 (1.47 to 1.81). The relative risk for glycemic variability was 1.61 (1.47 to 1.78), and for hypoglycemia (glucose, 40 to 60 mg/dl [2.2 to 3.3 mM]) it was 1.53 (1.37 to 1.70). The optimal glycemic control in diabetic patients should also consider the prior level of glucose control, indicated by the level of glycated hemoglobin. The “stress hyperglycemia ratio” and the “glycemic ratio” were introduced to calculate the difference between actual and average glycemia.23,24 The estimated average glucose value derived from glycated hemoglobin could be the optimal target range, as suggested by reports of lower mortality or reduced need for intensive care when glucose was closer to the average value.23,24 In surgical intensive care patients, and in particular for scheduled surgeries, perioperative care should focus on the prevention of dysglycemia in order to improve outcomes. The proper management of oral glucose-lowering treatment in type 2 diabetic patients is therefore crucial.
New Oral Glucose-lowering Drugs
Type 2 diabetes treatment has undergone major changes with the development of new drug classes, such as sodium-glucose cotransporter-2 inhibitors, and demonstration of their safety and efficacy in large-scale clinical trials. In recent trials primary outcomes were cardiovascular, whereas the primary outcome of previous trials was glycemic control.25 This change in paradigm was mandated by the United States Food and Drug Administration, which, since 2008, has required data on cardiovascular safety for new glucose-lowering drugs, after the finding increased rates of myocardial infarction and cardiovascular mortality in patients randomized to rosiglitazone in a meta-analysis of 42 studies.26 These changes show the complexity of oral glucose-lowering drug use in type 2 diabetic patients.
In practice, the current recommendations for the management of type 2 diabetes are simple for the initiation of pharmacologic treatment: metformin is the first agent to be started, unless contraindicated (such as a reduction in glomerular filtration rate less than 30 ml/min). Therapeutic efficacy is monitored by the level of glycated hemoglobin. Once metformin is no longer sufficient despite adequate dosing, a second oral medication (e.g., sulfonylurea, thiazolidinedione, a sodium-glucose cotransporter-2 inhibitor, a dipeptidyl peptidase-4 inhibitor) or a glucagon-like peptide 1 analog should be selected, taking into consideration the individual patient’s risk of atherosclerotic cardiovascular disease, heart failure, or chronic kidney disease.27 If combination therapy does not suffice, combination therapy with a third oral glucose-lowering drug or insulin must be considered. It is not uncommon therefore that type 2 diabetic patients are polymedicated. Knowledge of the mechanisms of action is important for the proper management of these drugs.
Types of Oral Glucose-lowering Drugs
Oral glucose-lowering drugs can be divided according to their mode of action: those lowering blood glucose by increasing insulin release (sulfonylurea, meglitinides, dipeptidyl peptidase-4 inhibitor), those lowering glycemia by increasing insulin action (biguanides, thiazolidinediones), those reducing glucose absorption (alpha-glucosidase inhibitors), and those increasing urinary glucose elimination (sodium glucose cotransporter-2 inhibitors). Different sites and modes of action of these medications are schematically displayed in figure 1.
Increasing Insulin Release
Sulfonylurea and Meglitinides.
Historically, the first oral treatment available for type 2 diabetes was sulfonylurea, long-acting insulin secretagogues. Meglitinides constitute another class of insulin secretagogues that are taken with meals and are more rapid and short-acting. The currently available second generation sulfonylurea include gliclazide, glimepiride and glyburide or glibenclamide. The meglitinides include repaglinide and nateglinide. These drugs bind the sulfonylurea receptor, and thereby close potassium adenosine triphosphate channels in pancreatic β cells, resulting in membrane depolarization, calcium influx and insulin release. The American Diabetes Association (Arlington, Virginia) recommends sulfonylurea or meglitinides as add-on oral treatment when metformin does not suffice, in the absence of atherosclerotic cardiovascular disease or chronic kidney disease.
Clinical studies assessed the efficacy of sulfonylurea on glycemic control.28 A meta-analysis of 31 trials with a median 16-week duration of showed that sulfonylurea in mono- or combination therapy with another drug lowered glycated hemoglobin by 1.5% (17 mmol/mol; 95% CI, 1.3 to 1.8) and 1.6% (18 mmol/mol; 95% CI, 1.0 to 2.2), respectively. Sulfonylurea are cheaper than other oral glucose-lowering drugs recommended as second step. Meglitinides were mostly tested as an adjunctive therapy to metformin, and combination therapy safely improved glycemic control (meta-analysis of 22 trials).29
As the effects of sulfonylurea are independent of circulating glucose levels, there is an increased risk of hypoglycemic events. The rate of hypoglycemia was increased by 2.4-fold in patients treated with sulfonylurea as compared with other glucose-lowering agents.28 This risk can further increase in the hospital, as a result of differences between hospital diets and usual dietary habits. Hypoglycemia appears to be less of a risk for meglitinides30 but still is relatively high. Besides the risk of hypoglycemia, the incidence of cardiovascular events and mortality has been reported to be increased in patients treated with sulfonylurea in some,31 but not all meta-analyses. Increased cardiovascular morbidity was first reported in 1970 with tolbutamide (a first generation sulfonylurea),32 and was related to weight gain and fluid retention. A recently suggested mechanism of cardiovascular toxicity of sulfonylurea relates to the prevention of preconditioning induced by ischemia or volatile anesthesia.33 Clinically, this could translate to an enlarged necrotic myocardial area.
Regarding the perioperative period (table 1), withholding of the drug depends on the time of surgery and fasting duration, according to medication half-life and risk of hypoglycemia. In case of minor or major surgery with fasting, patients should omit the dose of the day.34–39 In the case of ambulatory surgery with short-term fasting (fewer than 6 h for solid food, or fewer than 2 h for fluids) or with maltodextrin use before surgery, guidelines suggest that it is possible to continue sulfonylurea, but glycemia should be carefully monitored, with special attention to the risk of hypoglycemia.
Dipeptidyl Peptidase-4 Inhibitors.
Dipeptidyl peptidase-4 inhibitors (linagliptine, saxagliptine, sitagliptine, alogliptine) prevent the degradation of the incretin hormone glucagon-like peptide 1 by the serine protease dipeptidyl peptidase-4. By inhibiting this enzyme, dipeptidyl peptidase-4 inhibitors potentiate the incretin effect of endogenous glucagon-like peptide 1 and enhance glucose-dependent insulin secretion by pancreatic β cells. The risk of hypoglycemia associated with the use of dipeptidyl peptidase-4 inhibitors is very low. In patients undergoing noncardiac surgery, dipeptidyl peptidase-4 inhibitors lowered blood glucose to the same extent as insulin therapy with fewer hypoglycemic events40,41 ; in combination with basal insulin it was noninferior to basal–prandial insulin treatment.42 Ogawa et al showed benefits of dipeptidyl peptidase-4 inhibitors compared to other glucose-lowering drugs on long-term cardiac and cerebrovascular complications in patients undergoing cardiac surgery.43 Dipeptidyl peptidase-4 inhibitors should not be stopped before surgeries of any kind, independently of fasting duration.2,35,36
Increasing Insulin Action
Thiazolidinediones enhance insulin sensitivity in peripheral tissues as a result of their peroxisome proliferator-activated receptor agonistic effect. Peroxisome proliferator-activated receptor-γ is found predominantly in the central nervous system, macrophages, vascular endothelium, adipose tissue and pancreatic β cells, while peroxisome proliferator-activated receptor-α is found predominantly in the liver, skeletal muscle, heart, and vascular wall. Rosiglitazone and pioglitazone, the two available thiazolidinediones, bind either to peroxisome proliferator-activated receptor-γ only (rosiglitazone) or to both peroxisome proliferator-activated receptor-γ and -α (pioglitazone).
Due to potential risks of myocardial infarction, cardiovascular death and fluid retention,44 rosiglitazone has been withdrawn from several markets. Pioglitazone was found safe in a recent meta-analysis of 16 studies.45 The risk of hypoglycemia is low.30 Recent publications demonstrated that thiazolidinediones were associated with lower incidence of atrial fibrillation compared with other glucose-lowering drugs, although in cardiothoracic surgery patients this was not confirmed.46
In case of elective surgery, thiazolidinediones can be taken on the day of surgery, independently of the time of surgery and fasting duration.36
Metformin, the only marketed biguanide, exerts its glucose-lowering effect via inhibition of endogenous hepatic glucose production, as a result of blockade of mitochondrial respiratory chain and activation of the 5’ adenosine monophosphate–activated protein kinase. This drug has been used to treat type 2 diabetes for more than 60 yr and is still recommended as first-line therapy38 ; it is the most widely used glucose-lowering agent47 and one of the top ten most prescribed drugs in the United States with a well-established safety and efficacy profile. The risk of metformin-associated lactic acidosis remains of some concern, mainly in patients with chronic kidney disease (glomerular filtration rate less than 30 ml/min), heart failure, or chronic liver disease, because those conditions result in plasma metformin levels well above the therapeutic range. The mechanisms of accumulation are reduced renal metformin clearance, impaired hepatic metabolism with reduced lactate clearance, and increased lactate production, as can occur in major surgeries.48 In randomized controlled trials, metformin use in the perioperative period did not significantly affect lactate levels in coronary artery bypass or noncardiac surgery patients with type 2 diabetes and it improved glycemic control.49,50 Overall, the risk of metformin-associated lactic acidosis is extremely small (fewer than 10 cases per 100,000 patient-years) and mortality rates have improved.48,51–53
Other mechanisms of action possibly unrelated to the glucose-lowering effects of metformin could explain intriguing and partially unexplained findings of a higher survival rate in patients with diabetes and sepsis who used metformin than in those who did not (meta-analysis of five studies).54 In surgical settings, there was no difference on mortality and postoperative complications between metformin users and nonusers. The effects of metformin on mitochondrial respiratory chain and 5’ adenosine monophosphate–activated protein kinase cause increased nitric oxide and adenosine diphosphate levels, resulting in endothelial protection in the setting of ischemia/reperfusion injury.55 In the United Kingdom Prospective Diabetes Study, fewer metformin-allocated patients had diabetes-related endpoints (i.e., sudden death, death from hyperglycemia or hypoglycemia, fatal or nonfatal myocardial infarction, angina, heart failure, stroke, renal failure, amputation [of at least one digit], vitreous hemorrhage, retinopathy requiring photocoagulation, blindness in one eye, or cataract extraction [risk reduction, 19% (95% CI, 2 to 33)]), and the risk reduction persisted for 10 yr (also for myocardial infarction, 33% risk reduction56,57 ). Meta-regression analyses suggest that metformin is more beneficial in longer trials that enroll younger patients.32 Hence, it is important to avoid unnecessary withdrawal to maintain the benefits of this drug.55
The recommendations for perioperative use depend on the type of surgery, renal function and the potential risk for metformin-associated lactic acidosis. For 1-day minor surgery, it is recommended to continue metformin, except in patients with renal dysfunction or in interventions requiring a contrast medium administration or use of nonsteroidal antiinflammatory drugs, angiotensin-converting enzyme inhibitors or angiotensin II receptor antagonists.2,34–36 For major surgeries, it is recommended that metformin be stopped the day before. For both scenarios, metformin should be resumed once oral intake has been re-established and renal function is stabilized.34,36,58 Some reports of safe postoperative use of metformin in combination with insulin have recently been published.49,59
Reducing Glucose Absorption
Alpha Glucosidase Inhibitors.
This class of agents (acarbose, miglitol, voglibose) prevents intestinal glucose absorption and primarily reduces the rise in postprandial glucose. Hence, glucose-lowering efficacy depends on carbohydrate intake.60,61 The risk of hypoglycemia is very low.
Increasing Urinary Elimination of Glucose
Sodium Glucose Cotransporter-2 Inhibitors.
Sodium glucose cotransporter-2 inhibitors (canaglifozin, dapagliflozin, empagliflozin, ertugliflozin) are the newest oral type 2 diabetes drugs that are increasingly being prescribed. Their mechanism of action is to block the sodium glucose cotransporter-2 in the kidneys and promote glucosuria by inhibiting renal glucose reabsorption. This confers glycemia-dependent control, not leading to hypoglycemia. Beneficial effects on the cardiovascular system were observed for all sodium glucose cotransporter-2 inhibitors tested so far: data collected in more than 70,000 patients suggest a net protection against cardiovascular outcomes and death.62 A series of case reports describes euglycemic ketoacidosis in sodium glucose cotransporter-2 inhibitor–treated patients. Ketoacidosis probably develops due to insulinopenia, increased levels of counter-regulating hormones (glucagon, cortisol, epinephrine), and hypovolemia,63,64 and the risk increases with fasting.
Patients using sodium glucose cotransporter-2 inhibitors should therefore be considered at risk of ketoacidosis in the perioperative period. Cases have been described and the main challenge is that in some patients the event occurred at euglycemia. It is essential to have a high degree of suspicion.65,66 It is recommended to not take this drug on the day of surgery because of the risk of dehydration; some have recommended cessation several days before surgery.67 When food intake is resumed, the sodium glucose cotransporter-2 inhibitor can be restarted.2,36
Summary of the Recommendations for the Preoperative Period
During the perioperative period, table 1 summarizes recommendations for each drug class, based on guidelines from the Association of Anesthetists of Great Britain and Ireland (London, United Kingdom), a joint statement of the French Society of Anesthesia and Intensive Care Medicine and the French Society for the Study of Diabetes (Paris, Fance), a joint recommendation of the German Society for Anesthesiology and Intensive Care Medicine (Nuremberg, Germany), German Society for Internal Medicine (Wiesbaden, Germany), German Society for Surgery (Berlin, Germany), the American Diabetes Association, and the European Medicines Agency (Amsterdam, The Netherlands), and recent reviews.2,35–37,68–70 These recommendations are often based on expert opinion, as relatively few randomized controlled trials have assessed the management of oral type 2 diabetes treatment in the perioperative period. Specific considerations will help to decide whether a medication should be continued or withheld.
Guidelines suggest that insulin should be started when blood glucose is greater than 180 mg/dl (10 mM).72 Individualized insulin therapy targeting average glycemia should be assessed in clinical trials.
Patients with type 2 diabetes are commonly referred for elective or emergency surgery. In case of scheduled surgical procedures, previous guidelines recommended to withhold oral glucose-lowering drugs.35,72 Based on recent literature, this tendency has shifted toward treatment continuation.
In most cases, the continuation of glucose-lowering drugs seems safe, although it may be recommended to temporarily interrupt treatment, e.g., on the day of surgery, when there is fasting, reduced food intake, or risk of renal dysfunction. Inadvertent discontinuation may worsen glycemic control and increase complication rates, whereas an inappropriate continuation of sulfonylurea or sodium glucose cotransporter-2 inhibitors may induce hypoglycemia or ketoacidosis.
Support was provided solely from institutional and/or departmental sources.
The authors declare no competing interests.