Perioperative Glucose Management An Evolving Standard of Care Objectives • Harmful effects of hyperglycemia • Beneficial effects of insulin + glycemic control • Observational studies • Interventional studies • Protocol implementation Surgery • Stress + fasting associated with surgery leads to relative insulin deficiency – ↑ insulin resistance – ↓ insulin secretion • Insulin deficiency leads to hyperglycemia + excess circulating free fatty acids Gandhi GY et al. Mayo Clin Proc 2008;83:418-30 Surgical stress • Degree of hyperglycemia depends on severity + extent of tissue trauma; anesthesia factors • Elective intraperitoneal surgery +nondiabetics: ↑ 126-180 mg/dL (7-10 mmol/L) • Cardiac surgery: impressive disturbance of glucose homeostasis. – > 270 mg/dL (15 mmol/L) in nondiabetics – > 360 mg/dL (20 mmol/L) in diabetics Schricker + Carvalho. J CTVA 2005;19:684-8 Surgical stress + hyperglycemia • • • • • • Typical metabolic + endocrine alterations ↑ glucose production ↓ glucose utilization ↑ renal absorption of filtered glucose ↓ insulin activity Cardiac surgery: pancreatic hypoperfusion, excess glucose in prime + cardioplegia, hypothermia Schricker + Carvalho J CTVA 2005. Smith et al: JCTVA 2005;19:201-8 Immune system + hyperglycemia • • • • • • • • Impaired microvascular response Adhesion + transmigration of leucocytes Complement cascade Cytokine network Chemokine formation Chemotaxis, phagocytosis Generation of reactive O2 species Neutrophil apoptosis Turina M. CCM 2005;33:1624 Beneficial effects of glycemic control • • • • • • • Protects hepatocytic mitochondria Improved PMN neutrophil function Better bactericidal + opsonic activity Partial correction of abnormal serum lipids Counteracts catabolism of critical illness ↑ endothelial function + myocardial protection ↓ inflammation + apoptosis Brindley et al. CJA 2006;53:947-9 Beneficial effects of insulin • • • • • Stimulates glucose uptake + lipogenesis Inhibits lipolysis, proteolysis +glycogenolysis Multiple nonmetablic effects: ↓ levels circulating adhesion molecule + E-selectin ↓ circulating NO levels by suppressing inducible nitric oxide synthetase gene expression • Protects vascular endothelium from injury • Prevents organ system dysfunction Schricker J CTVA 2005. Langouche et al. J Clin Invest 2005. Insulin infusion + glycemic control • Avoids sustained cellular glucose overload + toxicity in many cell types • ↓ likelihood for vital organ dysfunction • Prevents damage to mitochondrion • Blunts stress response to CPB Van den Berghe Ann Int Med 2007;146;307-8 Albacker et al. Ann Thorac Surg 2008;86:20-7 Observational studies • Relationship of hyperglycemia + adverse outcomes well investigated – – – – – – Neurologic illness, trauma Myocardial infarction Burns Cardiac surgery Critical care Kidney transplant donors Gandhi et al Mayo Clin Proc 2008;83:418-30. Blasi-Ibanez et al. Anesthesiology 2009;110:333 Interventional studies • DCCT + EDIC 1993+ 2003; UKPDS 1998 + 2008 ↓ complications + mortality in diabetics • DIGAMI 2: ↓ mortality in diabetics w AMI [Eur Heart J 2005] • Leuven trial 2001: 34% ↓ mortality + 40-50% ↓ in important co-morbidities [NEJM 2001, postop surgical pts] • Krinsley: 29% ↓ mortality + ↓ renal insufficiency [Mayo Clin Proc 2004, postop ICU] Cardiac Surgery + Intraop Insulin • Furnary et al: ↓ mortality + deep sternal wound infections in diabetics [Ann Thor Surg 1999 + J TCVS 2003] • Lazar et al: ↓ infection, Afib, ischemia, inotropes, LOS, mortality in diabetics [Circulation 2004] Portland Diabetic Project • • • • 23,619 cardiac surgery pts, 1987-2005 5510 had diabetes, 40% on insulin 1987-92: subcut regular insulin, target 150-200 1992-05: continuous regular insulin, target ↓ over time, currently 70-110 in ICU • 3-BG= average value of all glucose measures on each of 3 days: POD 0, 1, 2 [24-72 measures] • Current 3- BG = 121 mg/dl Furnary AP. Endocrine Practice 2006;12[Supp 3]:22-6. http://www.providence.org/protocoldownload Portland Diabetic Project, results • ↓ risk of death by 60%, mainly pump failure + afib (4.4% vs 1.1%) • Annual CABG mortality < nondiabetics • ↓ risk of deep sternal wound infection by 77% (2% to 0.6%) • Other effects: ↓ transfusion, afib, inotropes, infections, LOS Furnary AP, Wu YX. Endocrine Practice 2006;12 (Supp 3) :22-6 Boston GIK Study • 149 diabetic pts, CABG • Randomized to GIK vs std therapy • GIK started before anesthesia induction + continued until 12 hr postop. • GIK: insulin 4.8 u/hr+ dextrose 1.5 g/h + KCL 2.4 mEq//h. Target: 125-200. • Std therapy: sliding scale subcut insulin if > 250 Lazar et al. Circulation 2004;109:1497-1502 Boston GIK Study, results • • • • • • ↓ glucose: ↓ afib: ↓ LOS ↓ infections: Other: postop: Other: 2yrs: 138 vs 260 [12 hr postop] 17 vs 42% 6.5 vs 9.2 days 0 vs 13 pts [pneumonia, wound] ↑ CI, ↓ pacing, ↓ inotropes improved outcome, ↓ ischemia, ↓ wound infections Lazar et al. Circulation 2004;109:1497-1502 Mayo Intraop Cardiac Surgery Study • 400 pts: 20% diabetics, Hg A1C 7% • Randomized to intraop intensive vs std therapy • Intensive: – Insulin started > if 100 mg/dl. Target 80-100 • Std therapy: – sliding scale IV insulin 200-250, or IV insulin if > 250 • Postop: – Both groups had intensive IV insulin, target 80-100 Gandhi GY et al. Ann Intern Med 2007;146:233-43 Mayo Intraop Cardiac Surgery Study • ↓ glucose after CPB: 123 vs 148 [19 u insulin] • ↓ glucose ICU arrival: 114 vs 157 • All pts normoglycemic afterwards by protocol design (103-104 mg/dl, 72-73 u insulin/ 24 h) • No difference in 1o or 2o outcomes: – death, sternal wound infection, cardiac arrest, arrhythmias, ARF, > 24h intubation, LOS • More deaths + strokes in treatment gp (4 vs 0, P =0.06; 8 vs 1, P = 0.02) Gandhi GY et al. Ann Intern Med 2007;146:233-43 Neurosurgical Patients • RCT 483 adult pts admitted to neuro ICU postop [20% head trauma]: IIT, target gluc 80-110, vs control (gluc < 200). Excluded pts w diabetes • LOS shorter + infection rate lower with IIT • More hypoglycemic (gluc < 50) episodes in IIT • 94% of IIT gp had hypoglycemic episodes • No difference in mortality + Glasgow outcome scale between gps Bilotta et al. Anesthesiology 2009;110:611 Rethinking Glucose Control • Tight glucose control did not prevent CV deaths + macrovascular complications in type 2 diabetes – ADVANCE – VADT – ACCORD • Tight glucose control leads to hypoglycemia – VISEP – Glucontrol ADA 2008: 68th Annual Scientific Sessions Meta-analysis of IIT • 29 RCTs, 8432 pts • No mortality difference – Tightness of control – Surgical, medical, med/surg • ↓ sepsis (10.9 vs 13.4%) • ↑ hypoglycemia (< 40 mg/dL, 13.7 vs 2.5%) Wiener et al. JAMA. 2008;300(8):933-944 Consensus recommendations Pt Population: [In-patients] General medical + surgical pts* Target glucose Rationale 1 mmol/L = 18 mg/dL ↓ mortality, LOS, infection Cardiac surgery* Fasting 90-126 Random < 200 < 150 Critically ill** < 150 ↓ mortality, morbidity, LOS Acute Neuro disorders*** 80-140 Lack of data ↓ mortality + sternal wound inf *ADA. **SCCM. ***AHA. Loh-Trivedi + Rothenberg, 2008 Protocol Implementation • Periop glycemic control depends on: – – – – – – nature of surgery severity of illness, age sensitivity to insulin modality used to achieve glycemic control body temp, caloric intake, infection Preop diabetes + treatment Protocol Implementation, cont’ • Interaction between glucose metabolism + surgical trauma + CPB is complex • Optimal intraop glycemic control cannot be achieved by occasional measurements of glucose + reactive adjustments of insulin infusion Schricker + Carvalho. J CTVA 2005 Cardiac surgery + intraop glycemic control • Hyperinsulinemic, normoglycemic clamp • Infuse insulin at constant rate • Infuse dextrose to clamp blood glucose at a specific level Carvalho et al. Anesth Analg 2004 Smith et al. J CTVA 2005 Van Wezel at al. J Clin Endocrinol Metab 2006 Cardiac surgery + intraop glycemic control • IV insulin • IV GIK • Sliding scale insulin • Adjust rate + dose based on glucose levels Furnary et al. Ann Thorac Surg 1999 + J TCVS 2003 Lazar et al. Circulation 2004 Gandhi et al. Ann Int Med 2007 Iatrogenic hypoglycemia • Brain cannot synthesize glucose or store more than a few min supply as glycogen • Brain is critically dependent on continuous supply of glucose from circulation • Signs + symptoms masked by anesthesia: – anxiety, palpitations, tremor, sweating, hunger, paresthesias, cognitive dysfct, seizures, coma, brain damage Cryer et al. Diabetes Care 2003;26:1902-12 Glycemic thresholds • 72-108 mg/dL (4.0-6.0 mmol/L) : – normal range • ~ 65-70 mg/dL (3.6-3.9 mmol/L): – neuroendocrine response (↑ glucagon + epinephrine) • ~ 50-55 mg/dL (2.8-3.0 mmol/L): – neurologic symptoms, cognitive impairment • Thresholds may be shifted in poorly controlled diabetics Cryer et al. Diabetes Care 2003;26:1902-12 My Perspective • Current treatment does not provide plasma glucose-regulated insulin replacement or secretion • Pharmacokinetics of insulin are imperfect – Time course of short acting analogues measured in hrs – Time course of endogenous insulin in nondiabetics measured in min • Iatrogenic hypoglycemia can be minimized but not eliminated if goal of treatment is near- euglycemia Glycemic control: future directions • Glucose – regulated insulin replacement via pancreatic islet transplant (diabetics) • Bio-engineered artificial β-cell (diabetics) • Closed loop insulin replacement systems: – – – – – reliable glucose sensor necessary CGMS gold: measures subcut glucose q 10 s CGMS guardian: measures interstitial glucose q 5min CSII: continuous subcut insulin infusion IV-intraperitoneal systems Cryer et al. Diabetes Care 2003. Ferrari. Curr Opin Anaesthesiol 2008 Summary • Glycemic control of benefit in pts with diabetes, + in nondiabetic critically ill patients • Optimal periop glycemic control will carry risk of iatrogenic hypoglycemia • Periop control can be achieved in most pts through coordinated protocols, but ideal glucose level not known