Anesthesiology Clin N Am 22 (2004) 93 – 123 Assessment and therapy of selected endocrine disorders Lisa E. Connery, MDa,*, Douglas B. Coursin, MDb a Departments of Surgery, Internal Medicine, and Anesthesiology, Long Island Jewish Medical Center, 270 – 05 76th Avenue, New Hyde Park, NY 11040, USA b Department of Anesthesiology, University of Wisconsin Medical School, B6/319 UW CSC, Madison, WI 53792, USA Anesthesiologists care for the entire spectrum of patients ranging from complex newborns to the most compromised geriatric patient. Clinicians must adapt techniques and anticipate special problems and complications that may arise when dealing with specific patient populations. Because patients with various endocrine disorders frequently are at risk for adverse perioperative events caused by their primary endocrinopathy or secondary complications, preemptive management should be directed when available by evidence-based guidelines. Preanesthesia evaluation entails review of medical records, the patient interview, a focused physical examination, and judicious use of preoperative testing. This assessment then guides further consultation or specific specialized tests and facilitates development of the anesthetic plan. Preoperative assessment is complicated by a number of issues. With the prevalence of same day surgeries, the patient is often first seen by the anesthesiologist on the day of surgery. This presupposes that an appropriate preoperative evaluation has been performed. There is often pressure on the anesthesiologist to proceed with even elective surgeries if information is missing or incomplete. A sea change has occurred in the area of preoperative assessment from performing a battery of tests on patients regardless of their medical history or the type of the procedure being performed, to the current situation of directed and individualized testing. This has come about from the findings that routine batteries of tests performed preoperatively on asymptomatic patients are often unlikely to be of significant benefit to the patient, and may, occasionally, harm the patient when incidental findings are unnecessarily and aggressively pursued. In general it is best not to order a test unless a result of the test would change the management plan. Preoperative assessment is not yet a science and further research is needed. The American Society of Anesthesiolo* Corresponding author. E-mail address: [email protected] (L.E. Connery). 0889-8537/04/$ – see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/S0889-8537(03)00111-1 94 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 gists task force on preanesthesia assessment concluded that based on the available scientific literature, decision-making parameters for ordering specific preoperative tests, and the timing of such tests, could not be unequivocally determined . In the evaluation of patients with various endocrine disorders, much of the ‘‘classic’’ preoperative assessment remains empiric and, unfortunately, is not prospectively established. Endocrine disorders range from the ubiquitous, diabetes, to the rare pheochromocytoma. All endocrinopathies may contribute to serious adverse events or a labile perioperative course if they are advanced, underappreciated, or inadequately managed. Although present guidelines and recommendations are often empiric, this review focuses on current concepts of preoperative assessment and perioperative management of diabetes, and also discusses some specific thyroid and adrenal pathologies. Diabetes mellitus Diabetes mellitus is the most commonly encountered endocrinopathy in western society, developing in roughly 15 to 20 million Americans or 7% to 8% of the population of the United States. Ninety percent of the diabetics in the United States have type 2 diabetes, and the remainder type 1 [2,3]. There is strong longitudinal data that long-term glycemic control limits development of some complications of diabetes. Recently a growing body of evidence supports the benefits of euglycemia in selected patients, particularly critically ill cardiac surgical patients and patients with CNS or myocardial ischemia or infarction. The clinical care of diabetes therefore should focus on appropriate diagnosis, risk stratification, and therapeutic intervention. The days of laisse faire attitude about glucose control are behind us. Type 1 diabetes results from the destruction of insulin-producing pancreatic beta cells. This destruction is mediated by autoimmune and other mechanisms. Onset of disease peaks in the teenage years . Type 2 diabetes is characterized by defective insulin secretion or use that may occur with excessive hepatic gluconeogenesis. Both genetic and environmental factors play a role in the development of this disease, which until recent years, was generally regarded as a disease of adult onset. At its current rate of increase, type 2 diabetes is projected to develop in a quarter to one third of the American population within the next 25 years. The increasing incidence of obesity and a sedentary lifestyle in our society has been blamed for causing the prevalence of type 2 diabetes to surge, and for the disease to have its onset at a younger age. Although patients with type 2 diabetes may or may not require insulin to optimize their care, the systemic complications of the disease remain the same as with type 1 diabetes . Hepatic and peripheral resistance to insulin result in increased hepatic glucose production, and decreased muscle uptake of glucose . Diabetics undergo interventional procedures and surgery more commonly than their non-diabetic counterparts. End-organ diseases resulting from diabetes often L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 95 result in the need for major surgical intervention, such as vascular and coronary bypass procedures, amputations, and renal and pancreas transplants. Coronary artery disease, peripheral vascular disease, renal insufficiency, gastroparesis, and autonomic neuropathy are common sequelae of longstanding diabetes and should be identified during preoperative evaluation. The multisystem destructive effects of diabetes places these patients at greater perioperative risk for complications such as stroke, myocardial infarction (MI), and worsening renal insufficiency. Wound complications and infection related to poor healing are also more likely than in nondiabetics. Recent data suggests control of postoperative hyperglycemia may be beneficial in cardiac surgical and critically ill patients. Coronary artery disease is prevalent in longstanding diabetics. Type 1diabetes may manifest symptoms of coronary artery disease at a young age. Diabetics with coronary disease frequently have atypical symptoms of ischemic heart disease, and often have silent ischemia. Perioperative MI has an associated mortality rate of 40% to 70% and is greater for diabetics than their non-diabetic counterparts. Diabetics who sustain a perioperative infarct are also less responsive to therapeutic intervention than their non-diabetic counterparts. Significant cardiac disease can be identified through the clinical history, including an assessment of functional capacity, physical examination, 12-lead electrocardiogram and individualized diagnostic evaluation based on initial findings and risk stratification. Patients unable to perform a 4 metabolic equivalent (MET) workload equivalent are at increased risk for perioperative cardiac events. Major, intermediate, and minor clinical predictors have been identified and are reviewed in the article by Dr. Fleisher in this issue. The presence of diabetes mellitus is classified as an intermediate clinical predictor for increased perioperative risk for adverse cardiovascular events (MI, congestive heart failure, or death). The presence of additional risk factors, such as hyperlipidemia, smoking, hypertension, and a family history of heart disease compound those odds. Preoperative electrocardiograms may reveal evidence of ischemia, prior MI, rhythm disturbance, or conduction delay. An abnormal electrocardiogram is more informative when an earlier study is available for comparison, and is more useful in patients who are classified as intermediate and high risk. The practice of evidence-based medicine forces us to critically examine the efficacy of treatment plans in a scientific manner. For example, class I evidence, based on accepted prospective randomized trials, exists for obtaining an EKG for all patients who have an intermediate risk factor including diabetes, scheduled for an intermediate or high-risk procedure . The American College of Cardiology/American Heart Association (ACC/ AHA) updated its guidelines in 2002 for the perioperative cardiovascular evaluation of patients undergoing noncardiac surgery . The committee emphasized that few patients will require revascularization procedures to minimize their risk before undergoing noncardiac surgery, unless the revascularization would have been warranted irrespective of the surgery being planned. The need for further preoperative cardiac evaluation is tailored to the circumstances and is 96 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 dependent on a number of variables, including the presence or absence of clinical predictors of disease in concert with the degree of risk of the surgical procedure. If a revascularization procedure is not deemed necessary, patients should be considered for perioperative medical therapy. Although often underused in the diabetic because of concerns about limiting the response and recognition of hypoglycemia, class I evidence exists that patients who have been on beta-blockers in the recent past for the treatment of angina, hypertension, or arrhythmias, should receive them perioperatively. In addition, diabetics have been shown in a large retrospective database review, to have improved long-term survival when maintained on chronic beta-blockade after sustaining a MI . Beta-blockade is also beneficial for high-risk patients, including those with diabetes before undergoing major vascular surgery. A class II recommendation, based on less rigorous studies than those qualifying as Class I (Class IIa data is still collected prospectively or is clearly reliable data from retrospective analyses), advises the use of beta-blockers in the perioperative period for patients with untreated hypertension, major risks for, or known coronary artery disease, in the absence of any contraindication to beta-blockade . In an emergency, one generally should proceed to the operating room and moderate risk as able. In an elective situation, if a patient has had a recent MI and post infarct risk stratification does not indicate that there is significant myocardium at risk, the ACC/AHA committee advises that it may be best to wait for 4 to 6 weeks before proceeding with the procedure. However, there are not many clinical trials that support a firm guideline on what timeframe would be best . Many diabetics are treated with angiotensin converting enzyme (ACE) inhibitors as part of an antihypertensive regimen or to limit or potentially reverse albuminuria, because ACE inhibitors and tight glycemic control have been found to reduce the rate of progression of diabetic nephropathy . Patients chronically taking ACE inhibitors have been reported to develop hypotension refractory to ephedrine administration while undergoing general anesthesia . Preoperative instructions ordinarily advise the patient to continue their antihypertensive medications on the morning of surgery. Although this is still an area of controversy, consideration should be given to advising the patient to withhold the ACE inhibitor on the day of surgery. This might be especially pertinent in the longstanding diabetic, who is more likely to have diffuse vasculopathy and hence may be more at risk for ischemic complications should problematic hypotension develop during anesthesia. The presence of significant renal disease may make the patient prone to volume overload intraoperatively. Crystalloid solutions containing potassium should be avoided in the diabetic with advanced renal disease and hyperkalemia. The effects of renally excreted medications such as aminosteroid neuromuscular blocking agents may be prolonged and should be judiciously dosed, monitored, and reversed. The half-life of insulin is also prolonged, which may make the patient more prone to hypoglycemia . Finally, patients with baseline renal insufficiency are more prone to develop acute renal failure in the perioperative period . L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 97 Blood glucose and outcome The traditional approach to diabetic treatment in the hospital focused on maintaining modest control of hyperglycemia with the philosophy that aggressive treatment and the risk of hypoglycemia and neuroglycopenic insult were potentially more detrimental than moderate hyperglycemia. Blood glucoses in the 200-mg/dL range were considered reasonable. However, it was subsequently found that blood sugars greater than 200 to 250 mg/dL are more likely to be associated with adverse outcomes. Leukocyte function and the ability of immunoglobulins to fix complement have been shown to be impaired with blood glucoses in this range . The diffuse vascular disease that is often present in diabetics also impairs the ability of the circulation to deliver adequate oxygen and substrates to the injured area. Interestingly, short-term hyperglycemia has been shown to impair the ability of the endothelium to vasodilate. Studies have shown that hyperglycemia increases endothelial release of endothelin-1, a potent vasoconstrictor, and may reduce availability of the vasodilator nitric oxide . Hyperglycemia inhibits collateral coronary blood flow and inhibits the development of new collateral vessels . Recent studies have indicated that patients treated with intensive regimens (‘‘tight control’’) that result in euglycemia versus standard regimens, have lower rates of morbidity and mortality [11,12]. Mortality rates have been found to be directly correlated with the degree of elevation of blood glucose or HbA1c levels in both critically ill patients and in those admitted with acute MI . The diabetes and insulin glucose infusion in acute myocardial infarction (DIGAMI) trial found mortality rates of 35%, 40%, and 55% in patients presenting with blood glucose concentrations of less than 235, 235 to 298, and greater than 298 mg/dL respectively, at the time of admission with an MI. This compares quite unfavorably to the 9% mortality rate found previously for patients presenting with normal blood glucose values at the time of admission for acute MI . Surprisingly, increased mortality rates have also been noted in the presence of blood glucose ranges that were previously considered to be only mildly elevated. Fasting blood glucose levels of even 110 mg/dL are associated with an increased relative risk of cardiovascular events, such as cerebrovascular accident, MI, and sudden death . More importantly, tight postoperative glycemic control maintaining blood glucose in the range of 80 to 120 mg/dL has been shown to significantly decrease mortality in these patient populations with no noted adverse consequences, compared with traditional, less aggressive diabetic management (eg, maintaining blood glucose in the range of 180 to 200 mg/dl) [11,13]. Although it is easy to extrapolate this data to the intraoperative management of the diabetic, clinical trials substantiating a benefit from tighter operative glycemic control are lacking at this time. Airway considerations in the diabetic One third of patients with longstanding type 1diabetes may be difficult to intubate . Diffuse glycosylation of proteins in patients with longstanding 98 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 diabetes may result in ‘‘stiff joint syndrome.’’ This is manifested by the appearance of tight, waxy skin, short stature, and joint rigidity. Inability of the patient to approximate the palms and fingers together is called the ‘‘prayer sign.’’ This is believed to be a manifestation of stiff joint syndrome and an indication that the patient may be difficult to intubate. Stiff joints in the diabetic result in decrease of atlantoaxial mobility. This may result in difficult laryngoscopy . Gastroparesis is also more common in patients with autonomic neuropathy. Awake fiberoptic intubation may be a safer way to proceed with these patients, who thus have a combination of a potentially difficult airway and are at greater risk for aspiration. Sodium citrate, 30 mL orally just before proceeding to the operating room will help to decrease gastric acidity, and metoclopramide 10 mg intravenously may be administered a half-hour before induction to improve gastric emptying. H2 blockers such as ranitidine or famotidine may also be administered, but require more than 45 minutes to reduce gastric acidity. Patients with diabetes on occasion will develop a neuropathy of the vagus and the recurrent laryngeal nerves, resulting in bilateral vocal fold immobility. Symptoms of this disorder include the insidious onset of dysphonia and stridor . Autonomic dysfunction Autonomic neuropathy has been reported to be present in 20% to 40% of diabetics . Identifying the presence of autonomic neuropathy preoperatively may influence the anesthetic plan. The presence of the prayer sign, peripheral neuropathy, orthostatic changes in blood pressure, loss of normal respiratory variations of the heart rate, and resting tachycardia are suggestive findings on physical examination that are simple and expeditious to perform . Patients with autonomic neuropathy are more likely to have problems with intraoperative blood pressure lability . Invasive arterial monitoring may be helpful to more closely monitor such blood pressure variability. A study by Kitamura and colleagues  found that diabetics with autonomic neuropathy developed significantly lower core temperatures under general anesthesia for abdominal surgery than control nondiabetic patients and diabetics without autonomic neuropathy. Patients with autonomic neuropathy were found to have impaired vasoconstriction, which was felt to be responsible for the predisposition to hypothermia. Perioperative hypothermia has been found to be associated with complications such as poor wound healing. More aggressive intraoperative measures to maintain normothermia may be warranted. Preoperative instructions Perioperative control of diabetes is complicated by the induction of counterregulatory hormones by surgical stress, infection, and concurrent corticosteroid administration. This is compounded by a situation where the patient is either NPO, unable to tolerate their normal diet, or has sporadic caloric intake. Ideally, diabetic patients should have their surgeries scheduled for early in the morning to L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 99 reduce the time period of fasting and to minimize the disturbances in the patient’s regimen. Type 1 diabetics Patients with type 1 diabetes will always require at least basal amounts of insulin, even when fasting. This prevents the onset of ketosis. Regular insulin is usually held on the morning of surgery, unless the blood glucose is greater than 200 mg/dL. Sliding scale insulin should not be used as a sole substitute for an insulin infusion or a partial dose of intermediate- acting insulin. Doing so risks precipitating ketoacidosis . Table 1 [3,17] reviews various insulin preparations whereas Table 2 [9,17] provides guidance on insulin dosing pre and intraoperatively. External continuous subcutaneous insulin pumps administer insulin preparations through a subcutaneous needle. A reservoir is filled with short-acting insulins, either lispro, aspart, regular insulin, or buffered regular insulin (Velosin SR). The pump may be programmed to have a variable output throughout the day and night, and can administer boluses . These short acting insulins can be discontinued, and the patient converted to a regular insulin infusion perioperatively when judged appropriate (Box 1) . Type 2 diabetes Type 2 diabetes is frequently associated with obesity and is characterized by the presence of insulin resistance. Patients may be treated with diet, oral medications, insulin or combinations of the three. The goals of medical therapy of type 2 diabetes are to stimulate insulin production by the pancreas, decrease peripheral and hepatic resistance to insulin, and modulate hepatic gluconeogenesis . Five classes of oral hypoglycemic agents are currently available and highlighted in Box 2 [3,9,11,17 –19]. Several classes are associated with increased risk of perioperative hypoglycemia, including sulfonylureas, glitazone compounds, and the biguanide, metformin. With the exception of acarbose, the alpha-glucosidase inhibitor, patients should be instructed to routinely hold their oral agent on the morning of surgery. Type 2 diabetics with inadequate glycemic control with diet and oral agents require the addition of insulin to their regimen. Type 2 diabetics with insulin resistance may require much higher doses of insulin than expected to achieve euglycemia. Diet controlled type 2 diabetics may be able to maintain normoglycemia perioperatively simply by avoiding dextrose in intravenous fluids . However, many patients with type 2 diabetes who did not require insulin at home may require insulin perioperatively to maintain glycemic control . Subcutaneous insulin is generally not advised for intraoperative glucose control because of its potentially erratic absorption secondary to altered regional blood flow, tissue edema, or fluid shifts during surgery . If an insulin infusion (see Box 1) is chosen as the method to control blood glucose intraoperatively in 100 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 Table 1 Insulin preparations and guidelines Rapid-acting Insulin lispro (Humalog) Insulin aspart (Novolog)  The altered amino acid sequence of these analogs favors rapid subq absorption and rapid onset. Usually administered immediately prior to a meal. May be used subq or via an insulin pump, but not recommended for continuous infusion Short acting Regular insulin Administered around 30 – 60 min before meals. Main insulin used in continuous iv infusions. Intermediate-acting NPH (neutral Made by adding protamine protamine to insulin Hagedorn) Lente Made by adding zinc to an acetate buffered solution of insulin Long-acting Ultralente Zinc suspension of insulin. Its peak levels are less prominent than those of NPH. Glargine Another insulin analog, It most closely provides a constant, basal level of insulin. It is administered at 10 PM. It cannot be mixed with other types of insulin in the same syringe. It is less likely to cause hypoglycemia.  Although long-acting insulins are often held or halved in dose before surgery, glargine may be administered as usual to provide basal insulin levels during surgery.  Ultralente and glargine are used as basal insulin regimens 70/30: 70% NPH/30% regular Premixed/combination Premixed insulin formulas help 50/50: 50% NPH/ 50% regular to reduce the 75/25: 75% insulin neutral protamine lispro (NPL, similar likelihood of mixing errors. to NPH) and 25% insulin lispro. This mixture is intermediate in action and is usually administered twice a day, and is also known as Humalog mix. Onset 5 – 15 min Peak effects: 60 – 120 minutes Duration: 4 – 5 hours.  Onset 30 – 60 mins Peak: 2 – 4 h Duration: 6 – 8 h Onset 1 – 3 h Peak 4 – 6 h Duration 12 – 14 h Onset 1 – 3 h Peak 4 – 8 h Duration 12 – 20 h Onset 2 – 4 hours. Peak 14 – 18 hours. Lasts 18 – 24 hours. Onset of 1 – 2 h Peakless Duration: 20 – 24 h 70/30 usually given before breakfast 50/50 before dinner L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 101 Table 2 Insulin protocols on day of surgery Insulin infusion Hold all insulins the morning of surgery. Obtain a blood glucose and start an insulin infusion at 1 – 2 units per h, along with D5W or D5 0.45 NS at 75 – 100 cc/h, approximately 5 grams of glucose per hour. Administration of glucose during surgery helps to prevent ketosis, hypoglycemia, and catabolism.  The dextrose solution is not intended for volume replacement. Any additional fluid necessary for volume resuscitation should not contain dextrose. Blood glucose should be checked every 1 – 2 h, and the insulin infusion adjusted to achieve a blood glucose between 100 – 150 mg/dl  Intermediate-acting Give half to two thirds of intermediate to long acting insulin on the morning insulin use of surgery. Dose with regular insulin intravenously, from 1 – 4 units/h, with a goal blood glucose of 100 – 150 mg/dL  Potassium Glucose- insulin-potassium (GIK) regimen: Patients with normal renal function supplementation and normal potassium levels may receive dextrose containing fluids with additional potassium (10 – 20 mEq/L) in addition to the insulin infusion. Insulin pump Options are to turn the pump off and use a continuous insulin infusion or continue pump at a basal rate supplemented with dextrose and potassium as needed with rate adjustment based on serial blood glucose measurements. the type 2 diabetic, the patient should receive their last normal dose of insulin the evening before surgery, and the insulin infusion should generally be initiated around 2 hours before surgery to allow for equilibration (See options in Table 2). In the immediate postoperative period, blood glucose should be checked every 1 to 2 hours for several hours in type 1 diabetics and every 4 hours for type 2 diabetics . Dextrose should be infused and insulin administered as directed by laboratory data. The sympathetic response provoked by surgical trauma results in neuroendocrine changes, including increased serum levels of catecholamines, adrenocorticotrophic hormone (ACTH) and cortisol. The stress reaction thus produces an environment that favors catabolism and gluconeogenesis. The relative insulin resistance and insulin deficiency invoked by surgical stress predisposes to hyperglycemia and also enhances lipolysis, which may lead to ketosis and acidosis in type I diabetics . The renal threshold for glucose resorption in normal kidneys is between 10 mmol/L and 11.1 mmol/L (180 mg/dL) . Higher glucose levels may result in osmotic diuresis and compromised wound healing and therefore require appropriate insulin therapy. Diabetic ketoacidosis and surgery Acutely ill type 1 diabetics who present for emergency surgery may present in diabetic ketoacidosis. Diabetic ketoacidosis may occur even in the absence of significant hyperglycemia, provoked by inadequate availability of insulin at a 102 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 Box 1. Insulin infusion protocol Standard insulin infusion orders 1. Stop all previous insulin orders. 2. Draw stat serum glucose and compare to glucose meter measurement within 15 minutes. 3. Prepare solution of regular insulin 250 units/250 mL (1 unit/1 mL) in: (Check one) __ Dextrose 5% or __sodium chloride 0.9% (Use if diabetic ketoacidosis is present; may switch to Dextrose 5% when glucose is <250 mg/dL.) a. Run 50 mL of the solution through the tubing and waste. Should also be done with each tubing change. b. If admitted to a medical/surgical unit use the micro drip infusion pump. c. Target range for blood glucose __– __ mg/dL (suggest 100– 160). d. If blood glucose is not decreasing by at least 50 mg/dL/h, notify physician. e. If blood glucose is less than 200 mg/dL, initiate drip at __ units/h (suggest 1– 2 units/h). f. If blood glucose is greater than or equal to 201 mg/dL, administer IV bolus of __ units regular insulin (suggest 2 – 4 units). g. Check blood glucose after 1 hour. Maintenance titration 1. Do not stop the infusion! 2. Perform glucose meter measurement every hour, adjust infusion rate as follows: <70 mg/dL Decrease rate by __units/hr (suggest 1– 2 unit/h) and administer 12.5 G (1/2 vial) of D50 IV (25 mL). 70 – 100 mg/dL Decrease rate by __ units/h (suggest 0.5-1.0 unit / hour). 100 – 130 mg/dL No change. 131 – 160 mg/dL Increase rate by __ units/h (suggest 0.5 unit/hour). 161 – 200 mg/dL Increase rate by __ units/h (suggest 1.0 unit/hour). 201 – 250 mg/dL Increase rate by __ units/h (suggest 1.5 unit/hour). L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 103 >251 mg/dL Increase rate by __ units/h, give __ unit IV bolus (suggest 2 units/h & 2 units bolus). Notify MD 3. When feasible, all other drips should be in normal saline. 4. When glucose remains low (<70 mg/dL) for two consecutive tests please page MD. 5. If the glucose is consistently within the goal and the infusion rate is stable, glucose meter measurements may be changed to every 2 –4 hours. Discontinuation 1. For conversion to subcutaneous insulin, refer to the new insulin orders. 2. Infusion may only be stopped 30 minutes after the administration of the scheduled subcutaneous insulin dose. time when there is increased demand . Type 2 diabetics are far less ketosis prone, but elderly patients with poor oral intake may develop a nonketotic, hyperglycemic, hyperosmolar state . Urine or serum ketones should be measured if the blood glucose is greater than 240 mg/dL in a type 1 diabetic patient . Therapy of evolving diabetic ketoacidosis centers on identifying and eliminating the initiating event, providing adequate insulin therapy, fluid resuscitation and correction of electrolyte abnormalities . Perioperative glycemic control Several recent studies resulted in various clinicians advocating euglycemia as the goal of perioperative glucose control. The single center prospective controlled study by Van Den Berghe and colleagues in 1548 patients compared euglycemic or ‘‘tight control’’ of blood glucose between 80 and 110 mg/dL with conventional therapy. Investigators only initiated insulin therapy in the control group if the blood glucose was greater than 210 mg/dL, aiming for a treated level of 150 to 180 mg/dL. Most of those reported were surgical patients with cardiac surgical patients representing roughly 60% of enrollees. The tightly controlled group required insulin infusions quite often, but had significant decrease in mortality (4.6% versus 8% for controls). The effect was most notable in patients who required prolonged intensive care, longer than 5 days. For this group, patients had fewer blood stream infections, a lower incidence of renal failure requiring dialysis, required less red cell transfusion therapy, and had a lower incidence of critical illness polyneuropathy. Various mechanisms have been proposed to 104 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 Box 2. Oral hypoglycemic agents Sulfonylureas: First generation long acting sulfonylureas: chlorpropamide (Diabinese), tolbutamide (Orinase), tolazamide (Tolinase), acetohexamide (Dymelor). Second-generation sulfonylureas include glyburide (Diabeta, Micronase), glipizide (Glucotrol), and glimepride. Sulfonylureas work by closing the adenosine triphosphate regulated potassium (Katp) channel in pancreatic islet cells. Closure of this channel results in insulin release. This mechanism is clinically relevant because ischemic preconditioning (whereby myocardium that has been subjected to brief episodes of ischemia is more resistant to infarction) is dependent on activation and resultant opening of Katp channels. This physiology is a potential explanation for the increased mortality rate noted in diabetic patients being treated with sulfonylureas after angioplasty for acute myocardial infarction, and their worsened appearance of myocardial ischemia as determined by dipyridamole stress endocardiography . Sulfonylureas become less effective with progression of the disease . Long-acting sulfonylureas are best discontinued 48 –72 hours before surgery . Shorter acting sulfonylureas may be held the night before or the morning of surgery, to minimize the risk of perioperative hypoglycemia . Meglitinides are nonsulfonylurea insulin secretagogues such as repaglinide (Prandin), and nateglinide (Starlix), and have a short duration of action geared towards treating postprandial hyperglycemia . Insulin sensitizers act to decrease hepatic gluconeogenesis and reduce peripheral insulin resistance and increase insulin uptake into muscle. Thiazolidinediones and biguanides are the two classes of insulin sensitizer currently available. Thiazolidinediones: rosiglitazone (Avandia) piaglitazone (Actose) These drugs improve insulin uptake and decrease hepatic gluconeogenesis by activating genes that stimulate transcription for proteins that enhance cellular insulin action. The onset of action of these drugs is about 6 weeks. They may result in preserved pancreatic beta cell function and improve endothelial function. L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 105 Another drug in the thiazolidinedione class, trogliazone, was associated with cases of hepatic failure and was removed from the market. Sporadic cases have been associated with rosiglitazone also. It is recommended that liver function be assessed every 2 months in patients taking this class of drugs , thus it may be wise to obtain liver function studies in the perioperative period. These may be given on the morning of surgery, although missing a dose will not have much of an effect because of the long duration of action of this class of drugs. Thiazolinediones may also be continued postoperatively as long as hepatic function and cardiac function are normal. The liver metabolizes them. Biguanides The biguanide, phenformin, was banned in the United States in the 1970s because of the potential development of lactic acidosis. Metformin (Glucophage), the only biguanide currently available in the US, is commonly administered as part of a treatment regimen for type 2 diabetes. Although it’s association with lactic acidosis is much less common than that of phenformin, lactic acidosis may occur in the perioperative period with patients receiving this drug . Risk factors for developing lactic acidosis related to metformin use include the presence of renal or hepatic insufficiency, a fasting state, surgery, and age >60 years . Metformin should be discontinued at least 24 h prior to surgery or before performing any studies involving administration of radiographic contrast . Metformin should be held for 48 h after major surgery until it is clear that the patient has not suffered a decline in renal function postoperatively . Alpha-glucosidase inhibitors (acarbose [Precose], miglitol [Glyset]) These agents act by blocking the breakdown and absorption of complex carbohydrates, and are effective only when administered with food. This helps to reduce postprandial hyperglycemia, but has no effect on fasting blood glucose [3,9,18]. It should be noted that oral sucrose, maltose, and starch will not be effective for treatment of hypoglycemia in patients taking this drug . 106 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 explain the benefits of normal blood glucose. These include maintenance of normal white blood cell and macrophage function, positive trophic and anabolic effects of insulin, improved erythropoiesis and decreased hemolysis, reduced cholestasis, and less axonal dysfunction and damage associated with hyperglycemia and insulin deficiency. A recent retrospective single center study by Krinsley corroborated Van Den Berghe’s work in a larger, but more diverse medical and surgical evaluation of over 1800 patients . He showed that even a modest increase in the mean blood glucose level above normal in critically ill medical, surgical, and coronary care patients substantially increased hospital mortality. Other evidence supporting the deleterious effects of hyperglycemia in postoperative patients include reports of decreased sternal infections in cardiac surgical patients who had tighter intraoperative and postoperative glucose control, improved survival in acute MI patients who had their blood glucose controlled, and worsened outcome after stroke in patients who were concurrently hyperglycemic [24 – 26]. More data is likely to be forthcoming regarding the potential benefits of euglycemia in the perioperative period and a better understanding of the exact mechanisms that generate improved outcome and limited morbidity. Thyroid disease Hypothyrdoidism Hypothyroidism is a common condition in the United States, affecting approximately 1% of all patients and 5% of the population over age 50 . Hypothyroidism develops 10 times more often commonly in females than males. The most likely cause of hypothyroidism is iatrogenic secondary to either surgical resection or radioactive ablation of the gland as part of treatment for hyperthyroidism. Hashimoto’s thyroiditis is the most common noniatrogenic cause of hypothyroidism. If Hashimoto’s is present, one must seek other autoimmune diseases such as systemic lupus erythematosus, rheumatoid arthritis, primary adrenal insufficiency (AI), pernicious anemia, diabetes mellitus, and Sjogren’s syndrome. Clinical findings The diagnosis and treatment of hypothyroidism are facilitated by timely history, physical examination and application of modern thyroid hormone assays. Fatigue, memory loss, headaches, thinning hair, lethargy, constipation, cold intolerance, and anorexia are the most common symptoms whereas weight gain may be the most noticeable of the multitude of presenting physical signs. Hypothyroidism should be considered in inactive elderly patients who have become increasingly withdrawn and selected patients who live sedentary lives. Cardiovascular effects of hypothyroidism include a decrease in cardiac output, as a result of a decreased stroke volume and heart rate, caused by the loss of the inotropic and chronotropic effects of thyroid hormone. The decrease in circula- L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 107 tion to the skin results in its cool and pale appearance. Pericardial effusion may be present with severe hypothyroidism, but will rarely result in tamponade. Patients with hypothyroidism may also have obstructive sleep apnea and depressed respiratory drive. Pleural effusions may be observed. Renal water excretion is impaired, resulting in hyponatremia and an increase in total body water, although plasma volume tends to be reduced . Treatment of hypothyroidism Correction of the hypothyroid state entails thyroid hormone replacement, most commonly with tetraiodothyronine (T4) (levothyroxine). Triiodothyronin, T3, the most active thyroid moiety, is subsequently formed in the body by intracellular conversion of T4 to T3. Levothyroxine, a common replacement agent, has a halflife of 6 to 7 days; thus missing a morning dose on the day of surgery has little impact on the patient . Levothyroxine prescriptions seem to be prone to transcription errors. Prescribers occasionally will misplace the decimal point when ordering levothyroxine in milligrams, resulting in a 10-fold error in dosage. Errors have also been reported when the dose has been converted to micrograms from milligrams. It has been suggested that levothyroxine be ordered in micrograms, rather than milligrams, to avoid decimal point errors and conversion errors. The anesthesiologist needs to be alert to the possibility of thyroid medication errors. Intravenous forms of levothyroxine are available, if a patient is unable to take oral medication for a prolonged period of time, say greater than 5 to 7 days. The intravenous dose is 100% bioavailable, whereas oral doses are only 50% bioavailable. Thus, when converting from oral to intravenous doses of levothyroxine, the dose should be halved. The hypothyroid patient with coronary artery disease Timing of initiation of thyroid hormone replacement in patients with coexisting coronary artery disease and hypothyroidism remains an area of controversy and clinical judgment. The competing concerns involve the risk for precipitating unstable angina or an MI by increasing the patient’s metabolic rate and cardiac work with levothyroxine, versus precipitating myxedema coma, heart failure, or neurologic complications of severe hypothyroidism should the patient proceed for cardiac intervention before hormone replacement is initiated. It is generally believed that most patients may proceed with cardiac surgery before replacing thyroid hormone . The surgical patient with hypothyroidism There is not convincing evidence that mild to moderate hypothyroidism necessitates postponing elective surgery. Elective surgery should be delayed for patients with more advanced, severe hypothyroidism. If a patient with severe hypothyroidism (manifestations being myxedema coma, pericardial effusion, heart failure or extremely low levels of thyroid hormone) requires urgent surgery, 108 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 they should receive intravenous T4 or T3 perioperatively, in addition to glucocorticoids as a certain proportion of patients with hypothyroidism may also have occult AI. If urgent treatment is deemed necessary, 200 to 500 micrograms (mcg) of intravenous levothyroxine (T4) may be administered slowly. Subsequent daily replacement doses are 50 to 100 (mcg) intravenously (IV) per day . Treatment of the hypothyroidism without providing steroid replacement may precipitate adrenal crisis [20,30]. Myxedema coma Patients may develop myxedema coma postoperatively . Myxedema coma is uncommon but carries a significant mortality rate of up to 60%. It can be precipitated in a patient with hypothyroidism by several medications, events or environmental factors, including hypothermia, trauma, infections, cerebrovascular accidents, anesthetics, sedatives and analgesics, amiodarone, and lithium carbonate . Myxedema coma is manifested by depressed mental status, delirium or coma, hypothermia, bradycardia, and hypopnea. CO2 narcosis caused by hypoventilation may also be contributory to the changes seen in mental status. Pericardial and pleural effusions may be present. Myocardial contractility is decreased, resulting in a diminished stroke volume and low cardiac output. Cardiac tamponade may also occur in the presence of a significant pericardial effusion. Echocardiography should be considered in hemodynamically unstable patients. Despite an increase in total body water, intravascular volume is decreased, which in combination with the depressed cardiac output makes the patient susceptible to hypotension should they become vasodilated. External warming is thus inadvisable because the resulting vasodilation may precipitate cardiovascular collapse. Laboratory findings include an elevated TSH and depressed levels of free T4 and T3, with the exception of patients who have hypothyroidism caused by pituitary disease whereupon TSH levels will be normal or low. Hyponatremia is common. This can also exacerbate mental status changes and occasionally promote seizure activity. Hypoglycemia can be present and may be a manifestation of associated AI, because of autoimmune causes or secondary AI caused by pituitary disease. Initiation of levothyroxine replacement can be complicated by the onset of arrhythmias. T4 is converted to T3 in the periphery by deiodinases, but the activity of these enzymes is reduced in the presence of hypothyroidism. Since T3 is the active form of thyroid hormone, initiating thyroid hormone replacement with T4 rather than T3 will result in a more gradual onset of effect. However, using T3 alone for acute replacement may precipitate arrhythmias because of its more rapid onset. Different protocols have thus been advised, with varying combinations of T3, T4, or both, depending on the age of the patient and the suspicion for the presence of coronary artery disease. Initial loading doses of intravenous T4 are generally in the range of 200 to 500 mcg. Typical (low) doses of intravenous T3 are 10 mcg every 8 hours, until the patient regains consciousness. In general, the use of intravenous T3 alone for replacement is not recommended . L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 109 Hyperthyroidism Grave’s disease is an autoimmune disorder that results in excessive thyroid hormone production because of abnormal stimulation of the thyroid gland by thyroid stimulating hormone (TSH) receptor antibodies. Hyperthyroidism may be clinically overt, manifested by signs and symptoms of hyperthyroidism (tachycardia, tremor, weight loss), goiter, and opthalmopathy. Gastrointestinal symptoms, such as nausea, vomiting, and diarrhea may also be present. At times, patients may only have subclinical or biochemical evidence of hyperthyroidism. Subclinical hyperthyroidism is manifested by increased nocturnal pulse rates, frequent atrial premature beats, or in the elderly, the onset of atrial fibrillation. Total and free T4 and T3 levels are normal or elevated, but TSH is suppressed . Postpartum patients may also develop relapses of Grave’s disease. The anesthesiologist needs to remain alert to the possibility of an undiagnosed case of thyrotoxicosis in patients with unexplained tachycardia, ectopic beats, fever, tremor or other suggestive signs and symptoms. Laboratory evidence of thyrotoxicosis includes an elevated free T4 level and suppressed TSH levels. Some patients will have normal free T4 levels, but will have suppressed TSH levels and an elevated T3 level . Treatment of hyperthyroidism Surgical treatment of Grave’s disease is not as common as it was in the past, with options for medical treatment and radioactive iodine being available, safe, and effective. Radioactive iodine is usually the treatment of choice, and generally achieves a euthyroid state within 6 to 18 weeks. The incidence of hypothyroidism after radioiodine treatment is at least 50% by 10 years after treatment . Pregnant women presenting with hyperthyroidism are generally placed on medical regimens because of the risks of surgery and the concerns of using radioactive iodine for the fetus. Beta-blockers, antithyroid medications, and iodides are used in various combinations for the patient being medically treated for hyperthyroidism. Goals of medical treatment are to reduce symptomatology caused by beta-adrenergic excess with beta-blockers, and to reduce thyroid hormone production with the antithyroid agents and iodides. Pregnant women presenting with hyperthyroidism are generally placed on medical regimens because of the risks of surgery and the concerns of using radioactive iodine for the fetus. Beta-blockers inhibit peripheral conversion of T4 to T3, the most active thyroid hormone, and help to limit the adrenergic effects of hyperthyroidism. The goal of beta-blockade should be to achieve control of the heart rate to less than 90 . Propranolol or more beta-1 selective agents such as atenolol or metoprolol may be used . Antithyroid medications such as propylthiouracil (PTU) and methimazole are actively transported into the thyroid gland and decrease thyroid hormone synthesis. PTU also inhibits peripheral conversion of T4 to T3, although this effect may not be clinically significant. Starting doses of PTU are 100 to 300 mg by mouth daily, and are generally adjusted downwards as treatment progresses. 110 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 Methimazole achieves a euthyroid state more rapidly than PTU, and is less likely to be associated with agranulocytosis, hepatitis, or vasculitis. Initial doses of methimazole are 10 to 20 mg by mouth daily [20,28,30]. Thionamides only block formation of new thyroid hormone. Iodines block the release of thyroid hormone from the colloid space . Thyroidectomy is considered when patients have a large goiter, have failed or are intolerant of antithyroid medications, or refuse radioactive iodine. Surgery is also performed in patients with Grave’s disease that have cold nodules and have a suspicion for thyroid cancer. A euthyroid state should be achieved before surgery to avoid precipitating thyroid storm . The hyperthyroid patient presenting for surgery Patients with medically treated hyperthyroidism should take their antithyroid medications on the morning of surgery . In the event of the need for emergent surgery in a patient who has thyrotoxicosis, emergency preparation entails administering an antithyroid medication, such as PTU or methimazole (Tapazole), followed by iodide. Iodides should not be administered first, as supplemental iodide provides more substrate for thyroid hormone synthesis. This can result in an increased amount of thyroid hormone released from the thyroid gland and potentially precipitate thyroid storm. One should wait for 2 to 3 hours after administering the thionamide before initiating iodines . PTU may be administered as a loading dose of 1 g orally or by nasogastric (NG) tube and later followed by a dose of 200 mg orally or NG every 6 hours. Iopanoic acid, an oral contrast agent, (Telepaque) is considered the iodine of choice. This is administered as 1 gram every 8 hours for the first 24 hours, then 500 mg twice daily. Lugol’s iodine or saturated solution of potassium iodide (SSKI) are alternatives. These are administered as oral drops in a dose of 4 to 8 drops every 6 to 8 hours . Patients with thyrotoxicosis are also at risk for AI and should receive stress doses of corticosteroids [20,30]. Glucocorticoids also have the effect of decreasing peripheral conversion of T4 to T3. Anesthetic agents that are vagolytic or sympathomimetic should be avoided in patients with thyrotoxicosis. Thyroid storm Thyroid storm may occur in the perioperative period in patients who have undiagnosed or undertreated hyperthyroidism. It is also more commonly seen in poor or underserved populations who have limited access to medical care . Malignant hyperthermia and pheochromocytoma crisis are in the differential diagnosis. Thyroid storm is recognized by the onset of fever, tachycardia, and delirium, which may progress to cardiovascular collapse and death. It may be necessary to treat the patient based on clinical suspicion alone while waiting for confirmatory thyroid function tests. The laboratory value itself does not distinguish between hyperthyroidism and thyroid storm per se. Thyroid storm remains a clinical diagnosis. Thyroid storm carries a mortality rate of 10% to 75% and mandates treatment in the intensive care unit (ICU) . L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 111 Treatment encompasses thionamides, beta-blockers, antipyretics, nutritional support with dextrose and vitamins, and treatment of cardiac complications such as atrial fibrillation and high output heart failure. Acetylsalicylic acid should not be used as an antipyretic in patients with thyroid storm, as it interferes with the protein binding of T4 and T3 and can thus increase free thyroid hormone concentrations . Methods to enhance thyroid hormone clearance may also be used, including cholestyramine to bind the hormone and clear it through the gastrointestinal tract. Rarely, charcoal hemoperfusion, hemodialysis, or plasmapheresis may be needed to increase thyroid hormone clearance (Table 3) [27,31]. Adrenal insufficiency Hypothalamic – pituitary-adrenal axis The hypothalamic – pituitary-adrenal axis (HPA) regulates adrenal output of glucocorticoids, which are intricately involved with the metabolism and production of nutritional substrates and maintenance and regulation of immune and circulatory function. Hypothalamic release of corticotrophin releasing hormone stimulates the pituitary to produce ACTH. ACTH output from the anterior pituitary subsequently stimulates the adrenal cortex to produce cortisol, which completes the cycle by providing negative feedback for CRH and ACTH release . Normal cortisol output of the adrenal gland in nonstressed conditions is between 15 mg and 30 mg per day [32,33]. The intact HPA axis is responsive to stressors such as surgery, trauma, burns, exercise, and psychologic trauma. Cortisol output amplifies proportionate to the degree of stress, and may increase to 60 to 100 mg/m2 per day [32,33]. Glucocorticoids affect the circulation by facilitating the effects of catecholamines such as norepinephrine and epinephrine on vascular tone and by their positive inotropic effects . Glucocorticoids have an inhibitory effect on endothelial production of prostacyclin (PGI2) . Relative glucocorticoid deficiency thus allows enhanced PGI2 production, which results in a vasodilated state . The ability of the adrenal axis to increase its output in the presence of stressful circumstances and enhance substrate availability and cardiac output to accommodate the increase in energy demand is essential for the survival of the organism. Etiology of adrenal insufficiency Endogenous causes of AI are uncommon disorders. Primary AI results from the destruction or malfunction of more than 90% of the adrenal cortex. Tuberculosis used to be the most common etiology of primary AI. With the wane of tuberculosis, autoimmune adrenalitis became the most likely primary cause in the United States. However, 30% of patients with advanced HIV infections develop primary AI . Aldosterone production is deficient in 112 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 Table 3 Management of thyroid storma Treatment Beta-blockade Esmolol Propranolol Atenolol or metoprolol Thionamides Propylthiouracil (PTU) Methimazole Iodinated contrast agents (administer only after PTU or methimazole given) Iopanoic acid or ipodate Iodine Lugol’s solution or Saturated solution of potassium iodide (SSKI) or Sodium iodide Glucocorticoids Hydrocortisone or Dexamethasone Dose and medication route Action 250 mg – 1,000 microg/kg iv bolus followed by 50 – 300 mg/kg/min 1 mg iv as needed, can convert to 60 – 80 mg mouth (po) or nasogastric(NG) tube every 4 hours Adjust for equivalent doses to propranolol Antagonizes effect of increased adrenergic tone and inhibits conversion of T4 to T3 800 – 1000 mg orally immediately, then 200 mg every 4 hours by mouth or NG 30 mg by mouth immediately, then 30 mg every 6 h by mouth or NG Blocks new thyroid hormone synthesis, blocks T4 to T3 (PTU only) 0.5 – 1.0 g/day by mouth or NG Blocks T4 to T3 conversion, block thyroid hormone release (via iodine release) 10 drops three times per d by mouth or NG 5 drops every 6 h by mouth or NG Blocks thyroid hormone release 0.5 – 1.0 g every 12 h 50 mg every 6 h IV 2 mg every 6 h IV Blocks conversion of T4 to T3 and provide stress doses of glucocorticoids a Supportive therapy in an intensive care unit to maintain blood pressure, heart rate, and temperature, while aggressive identification of a precipitating etiology is undertaken. Medications as outlined above are used to control thyroid hormone release and peripheral activity. Modified from Ginsberg J. Diagnosis and management of Graves disease. CMAJ 2003;168(5):584; with permission of the publisher, 2003 Canadian Medical Association. patients with primary AI also, and may be apparent because of the presence of hyponatremia and hyperkalemia. Secondary AI results from pituitary disease that has resulted in reduced ACTH production. Tertiary AI is caused by hypothalamic disease, or is iatrogenic in nature. Iatrogenic AI because of exogenous administration of corticosteroids is the most common cause overall of AI in the general population. Because aldosterone production is primarily regulated by the renin-angiotensin system, hypoaldosteronemia is not expected in AI of secondary and tertiary causes, and the electrolyte disturbances are generally not seen [20,33,34]. L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 113 Screening of the surgical patient for adrenal insufficiency Preoperative screening should help identify individuals who are at risk for HPA axis suppression. Categories of patients include patients receiving chronic corticosteroids for conditions other than primary AI (eg, transplant patients), patients who may be receiving corticosteroids intermittently or who had a course of steroids within the past year, those being treated for diagnosed AI, and patients with undiagnosed/suspected AI. One may suspect the possibility that a patient may have received a course of corticosteroids in the last year by noting certain diagnoses, such as asthma, inflammatory bowel disease, collagen vascular disease, rheumatoid arthritis, and in those who have had central nervous system tumors, or neurosurgery. Patients may also develop adrenal suppression from the use of topical corticosteroids for dermatologic disorders, especially when applied to large skin surface areas, and with the use of occlusive dressings and glucocorticoids preparations of higher potency. Adrenal suppression may also occasionally occur with the use of potent inhaled glucocorticoids . Chronic corticosteroid administration, with the exception of low doses and alternate day regimens, will suppress the HPA axis. The duration of treatment with corticosteroids that results in HPA axis suppression, and the daily threshold dose necessary for this to occur has not been precisely defined. Generally, however, patients who have received doses in excess of 20 mg per day for more than 5 days may be considered to be at risk for adrenal suppression . Patients who have been receiving doses of prednisone or prednisone equivalent in the physiologic range are not considered to have HPA axis suppression. This corresponds to doses of prednisone 5 mg per day, hydrocortisone 25 mg per day, and dexamethasone 0.75 mg per day. Patients who have been receiving steroid doses in excess of the physiologic range, but less than 20 mg of prednisone or equivalent per day, may be expected to have HPA axis suppression after 4 weeks of treatment . Alternate day regimens, especially when dosing the corticosteroid in the morning rather than the evening, are less likely to cause HPA axis suppression. Recovery from HPA axis suppression after removal from endogenous (Cushing’s syndrome) or exogenous corticosteroids has been shown to occur over a time course of approximately 9 months. Pituitary function normalizes first, with ACTH secretion resuming its diurnal pattern. Return of adrenocortical function is more gradual. The response to provocative testing generally normalizes 9 months or more after withdrawal from glucocorticoid therapy . Patients with chronic primary AI present with fatigue, weight loss, nausea, vomiting, and diarrhea. Hyperpigmentation may be seen. Electrolyte abnormalities, eosinophilia, and hypoglycemia may be noted on laboratory examination. Patients with primary AI generally receive routine replacement doses of hydrocortisone in the range of 20 to 30 mg daily in divided doses, with most of it administered in the morning. Such patients may still develop symptomatic AI when subjected to the stresses of surgery and the perioperative period because 114 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 they remain unable to compensate for the increased cortisol levels necessitated by these circumstances. Diagnostic provocative testing of the HPA axis can be performed if one suspects undiagnosed primary AI, but these circumstances are rare. Cosyntropin, a synthetic ACTH, stimulation testing assesses the ability of the adrenal gland to accelerate production of cortisol in response to a surge in ACTH levels. Baseline cortisol levels are drawn, and then a high dose, 0.25 mg of cosyntropin, is administered. Cortisol levels are repeated at 1/2 hour and again at 1 hour after the dose. Baseline cortisol levels in normal, non-stressed individuals are in the range of 6 to 18 mg/dL. Cortisol levels are expected to increase to more than 18 to 20 mg/dL after the stimulation test in non-stressed individuals. Critically ill, stressed patients are expected to have higher than normal baseline cortisol levels. Acute adrenal insufficiency Acute AI is manifested by nausea, vomiting, and hypotension and may progress to cardiovascular collapse. Hyponatremia and hyperkalemia may be present. Shock caused by acute AI is typically distributive, associated with a low systemic vascular resistance. The shock state is frequently unresponsive to catecholamines. Acute AI can mimic septic shock. Some degree of hypovolemia may be present also. A high index of suspicion for acute AI needs to be maintained in this scenario in patients who have not been previously diagnosed with AI. To circumvent this outcome, supplemental corticosteroids should be administered to patients identified to be at risk. Once one has identified the patient at risk for HPA suppression, the dose of corticosteroid supplementation with hydrocortisone or equivalent (Tables 4,5)  should be determined on an individualized manner, based on an estimate of the degree of stress anticipated to occur in association with an illness or planned surgical procedure. Excessive supplementation is unnecessary and may predispose the patient to untoward side effects such as hyperglycemia, poor wound healing, catabolism, and corticosteroid psychosis. Under normal circumstances, cortisol production increases for approximately 2 days after surgery and loses its normal diurnal variation. Afterwards, cortisol levels generally decline to the normal range and resume a diurnal pattern . Two other specialized groups have an increased risk for AI with evidence of inadequate cosyntropin responsiveness and improvement with corticosteroid supplementation. These include a subgroup of septic shock patients and elderly (>55 years old) patients who are vasopressor dependent post operatively after general surgical procedures despite adequate volume resuscitation [35,36]. Annane and colleagues  performed a French multicenter, randomized, prospective, double-blind study in 299 septic shock patients septic shock patients who were vasopressor and fluid dependent for less than 8 hours. A cosyntropin stimulation test was performed to identify responders (>9 mg/dL increase after cosyntropin) or non-responders (<9 mg/dL after stimulation). All patients were then randomized to receive intermittent intravenous (IV) boluses of hydrocorti- L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 115 Table 4 Guidelines for adrenal supplementation therapy Medical or surgical stress Corticosteroid dosage Minor Inguinal hernia repair Colonoscopy Mild febrile illness Mild-moderate nausea/vomiting Gastroenteritis Moderate Open cholecystectomy Hemicolectomy Significant febrile illness Pneumonia Severe gastroenteritis Severe Major cardiothoracic surgery Whipple procedure Liver resection Pancreatitis 25 mg of hydrocortisone or 5 mg of methylprednisolone intravenous (IV) on day of procedure only 50 – 75 mg of hydrocortisone or 10 – 15 mg of methylprednisolone IV on day of procedure Taper quickly over 1 – 2 days to usual dose 100 – 150 mg of hydrocortisone or 25 – 30 mg of methylprednisolone IV on day of procedure Rapid taper to usual dose over next 1 – 2 days From Coursin D, Wood K. Corticosteroid supplementation in adrenal insufficiency. JAMA 2002;287: 236 – 40; with permission. sone 50 mg every 6 hours and daily 50 mcg enterally of fludrocortisone or placebo. Patients who were non-responders and were treated with both glucocorticoid and mineralocorticoid had faster recovery and improved survival. It is important to emphasize that responders treated with glucocorticoids had no benefit and may have been compromised by steroid therapy . Rivers et al  described a ‘‘transient’’ AI developing in surgical patients older than 55 years of age who had sepsis and hypotension. In a prospective, Table 5 Comparative steroid potency (mg basis)a Steroid preparation Hydrocortisone (equivalent to cortisol) Prednisone Methylprednisolone Dexamethasone Fludrocortisone Glucocorticoid effect 1 4 5 30 0 Mineralocorticoid effect 1 0.1 – 0.2 0.1 – 0.2 x <0.1 20 Biologic 1/2 life (h) 6–8 18 – 36 18 – 36 36 – 54 18 – 36 Available formulations PO, IV, IM PO IV PO, IV PO a Supplementation via the intravenous route is preferred for those who are NPO, have unpredictable or poor absorption of medications, or have major stresses or critical illness. Prednisone and cortisone are not recommended in patients who are unable to methylate these preparations into an active form. 116 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 observational study, Rivers and colleagues  evaluated 104 patients with a mean age of 65.2 +/ 16.9 years who underwent surgery and were admitted to the ICU with hypotension requiring vasopressors despite adequate fluid resuscitation. After high dose synthetic cosyntropin stimulation testing, AI was defined as a baseline serum cortisol less than 20 mg/dL with an increase in cortisol of less than 9 mg/dL after stimulation, or functional (relative) hypoadrenalism, was identified by a serum cortisol less than 30 mg/dL with change in cortisol of less than 9 mg/dL. The latter was present in almost a third of patients (32.7%), a far higher incidence than predicted for the general surgical population. The patients with relative AI benefited from corticosteroid therapy. Pheochromocytoma Pheochromocytomas are chromaffin cell tumors that secrete excessive amounts of catecholamines. Patients develop secondary hypertension, which is often paroxysmal, and the classic triad of headaches, sweating, and palpitations. Anxiety and panic attacks may also be seen . Most pheochromocytomas are located in the adrenal medulla. Classical teaching is that 10% of patients will have bilateral adrenal tumors, 10% will be extraadrenal, and less than 10% are malignant [38,39]. Most pheochromocytomas are sporadic in nature. Pheochromocytomas may also be familial, inherited in an autosomal dominant manner, or may be seen with VonHippel Lindau syndrome. Approximately 16% will be associated with other endocrine disorders, such as multiple endocrine syndrome type 2 (MEN II), which is comprised of medullary thyroid carcinoma, pheochromocytoma, and parathyroid hyperplasia. Pheochromocytoma has also been associated with neurofibromatosis type I [37,40]. Diagnosis of pheochromocytoma Lenders and colleagues  compared the sensitivity and specificity of plasma free metanephrines, plasma catecholamines, urinary catecholamines, urinary total and fractionated metanephrines, and urinary vanillylmandelic acid (VMA) for the diagnosis of pheochromocytoma. They found that plasma free metanephrines and urinary fractionated metanephrines had the greatest sensitivity for the diagnosis, with 99% and 97% sensitivities, respectively. Urinary vanillylmandelic acid had the lowest sensitivity at 65%. The greatest specificity was found with urinary VMA (95%) and urinary total metanephrines (93%). The lowest specificity was seen with urinary fractionated metanephrines at 69%. The authors concluded that plasma free metanephrines should be the first test used when evaluating the patient for a possible pheochromocytoma, and that a negative test would virtually exclude this diagnosis. False positive test results may be seen in the presence of caffeic acid (present in coffee), tricyclic antidepressants, and phenoxybenzamine . Patients with positive biochemical screening tests for pheochromocytoma should then proceed L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 117 to an imaging study to attempt to localize the tumor . Imaging tests include MRI or CT scans, nuclear imaging studies such as metaiodobenzyl guanine (MBIG), and 6-(18 F) fluorodopamine positron emission tomography. CT scans are superior for detecting adrenal tumors, whereas MRI scans are superior in the detection of extraadrenal tumors. Neither of these two techniques is very specific for pheochromocytoma. Nuclear scans for pheochromocytoma are more specific but are limited in availability. At times biochemical screening tests are positive, but conventional scans are unable to localize the tumor. Nuclear testing may be helpful in such situations  (see reference  for algorithms to diagnose pheochromocytoma). The clonidine suppression test and glucagon stimulation test may be useful in some circumstances, but should be performed with caution because of their potential to result in severe hypotension or hypertension, respectively. Cardiac manifestations of pheochromocytoma Excessively high levels of catecholamines have been demonstrated to have toxic effects on the myocardium. Not surprisingly, patients with pheochromocytoma may be noted to have significant baseline ECG changes. Some may also present with chest pain and ECG changes suspicious for ischemia. Despite striking repolarization changes, many patients who proceed to coronary angiography preoperatively are found to have normal coronary arteries. Patients presenting with chest pain, ECG changes, and known pheochromocytoma may be better served by proceeding to angiography rather than thrombolysis . Some believe the ECG changes are, in fact, a manifestation of a toxic myocarditis. Such changes have often been noted to resolve rapidly in the postoperative period. In addition to ECG changes that seem suspicious for MI, many patients with pheochromocytoma are noted to have a long QTc interval, which may predispose the patient to ventricular arrhythmias. Liao and colleagues  noted a 16% incidence of significant QTc prolongation in patients presenting for pheochromocytoma resection, and an 80% incidence in those who were evaluated with coronary angiography before surgery. The QTc intervals also normalized after surgery. Pheochromocytoma multisystem crisis Occasionally, patients may present with pheochromocytoma multisystem crisis. This syndrome is manifested by wide swings in blood pressure, fever, metabolic acidosis, (which may be lactic acidosis or hyperglycemic and hypersosmolar nonketotic in nature), renal insufficiency, myocarditis, respiratory failure, and pulmonary edema. Mental status changes may also be present. Lactic acidosis may be caused by tissue hypoperfusion, or possibly by an increase in hepatic lactate production secondary to glycogenolysis. Tumor necrosis may also play a role in this presentation, which is often associated with large, right-sided adrenal masses . 118 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 Surgical resection of pheochromocytoma Although most of these tumors are not malignant, surgery is necessary to eliminate the humoral disturbances and limit growth of the mass. Surgical resection is curative in up to 90% of cases [37,44]. Pheochromocytoma resections have mortality rates reported between 0% and 6.5% in recent years (a considerable improvement from the 20% mortality rate noted in 1951), and morbidity rates of 3% to 36%. This may be more reflective of risk associated with open, rather than laparoscopic resections [44,45]. Historically, two open surgical approaches have been commonly performed for pheochromocytoma resection. The midline abdominal approach was advantageous for thorough exploration of the abdominal cavity and contralateral adrenal gland or bilateral resection. This approach also facilitates proximal vascular control and enables early ligation of the adrenal vein. The flank incision has the advantage of decreasing the likelihood of adhesions or organ injury (eg, splenic injury), but greatly limits exploration . The first laparoscopic adrenalectomy for pheochromocytoma was performed in 1992 [40,46]. Bentrem and colleagues  reviewed the outcomes of patients who underwent open, laparoscopic, or laparoscopic-assisted resections of adrenal or extraadrenal pheochromocytomas by the same surgeon at Northwestern Memorial Hospital between 1997 and 2001. Both laparoscopic and laparoscopic-assisted procedures were performed in the lateral decubitus position. The surgeon planned laparoscopic-assisted procedures when it was unclear if the procedure could be completed by way of a laparoscopic technique alone. After initial laparoscopic dissection, a subcostal incision was made while still in the lateral decubitus position. Seventeen patients, who had similar weight, age, and preoperative preparation were studied. Patients who had laparoscopic resections had a smaller mean tumor size (4.2 cm) versus the open (6.7 cm) and laparoscopic-assisted (6.3 cm) groups, because of the selection criteria used. All extraadrenal tumors were removed with an open technique. They found that laparoscopic resections were longer than open procedures (average operative time of 218 versus 202 minutes) but were associated with less blood loss (187 versus 562 mL), and the patients had a shorter hospital stay. Exposure is more difficult with the laparoscopic technique, but the patient benefits postoperatively with significantly less incisional pain. Laparoscopic assisted procedures, on the other hand, were even longer than laparoscopic cases (at 260 minutes) and were associated with more blood loss than the open technique (average of 925 mL). These patients were also not discharged earlier than those in the open group. The investigators concluded that laparoscopic adrenalectomies were preferable to open resections, although they still advised open resections for tumors greater than 6 cm in size and those extraadrenal in location. They also concluded that one should proceed directly to open rather than laparoscopicassisted adrenalectomies if it was unclear preoperatively whether the tumor would be able to be completely resected laparoscopically. Open procedures also remain preferred for patients in whom malignancy is suspected based on CT, MRI, or MBIG nuclear scan findings of periaortic L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 119 adenopathy. Patient selection for open versus laparoscopic procedures should be individualized. Preparation Preparing the patient with pheochromocytoma for surgery entails the institution of alpha and beta-adrenergic blockade. Preoperative preparation with alpha antagonists has been credited with much of the decrease in perioperative mortality that has been demonstrated over the past few decades. Phenoxybenzamine, a longacting noncompetitive alpha-adrenergic antagonist, is initiated at least 1 week and generally for a period of 2 to 4 weeks, before surgery. Beta-blockade is initiated in patients more than 3 days before surgery in patients who have persistent tachycardia or reflex tachycardia related to initiation of alpha blockade, or who are having arrhythmias. It is important to initiate alpha adrenergic blockade before beta adrenergic blockade to avoid a situation of ‘‘unopposed alpha’’ agonism whereby the patient suffers from intense vasoconstriction from the alpha adrenergic excess and is at risk for extreme hypertension and increases in myocardial workload. The pharmacologic adrenergic blockade helps to blunt the intense surges in blood pressure that occur with surgery and tumor manipulation. Preoperative preparation with alpha and beta-blockade also allows the heart to recuperate before surgery from catecholamine-induced stress and cardiomyopathy. Intravascular volume is also decreased in patients with pheochromocytoma. This is manifested by hemoconcentration and orthostatic changes in blood pressure. Alpha adrenergic-mediated vasoconstriction and, possibly altered capillary permeability is felt to be responsible for these findings . Treatment with metyrosine (alpha-methyl-para-tyrosine) preoperatively results in depletion of tumor catecholamine stores caused by competitive inhibition of tyrosine hydroxylase, and gives rise to decreased blood pressure lability, and decreased blood loss intraoperatively . Alpha adrenergic blockade enables the patient to have repleted intravascular volume. If this has been successful, one expects the hemoconcentration to resolve or improve before surgery. The presence or absence of orthostasis, and changes in the hematocrit should be assessed at the time of the preoperative visit. Certain medications should be avoided in patients with pheochromocytoma, including tricyclic antidepressants, metoclopramide, droperidol, and naloxone . Intraoperative management Various anesthetic techniques have been used for pheochromocytoma resections, a full discussion of which is beyond the scope of this article. Short acting antihypertensive infusions are preferred for blood pressure control, because of anticipated hemodynamic lability before tumor resection and the frequency of hypotension afterwards. Nitroprusside is commonly used. Others have used fenoldopam for this purpose, which has the advantage of not having toxic metabolites. Fenoldopam acts by stimulation of dopamine-1 receptors that cause 120 L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 peripheral vasodilation while simultaneously increasing renal blood flow. Dose ranges are 0.2 to 0.8 mg/kg/min. Fenoldopam also results in a diuresis and natriuresis, however, which may not be desirable in these patients who are often volume contracted . Intraoperative use of magnesium sulfate with administration of a 40 to 60 mg/kg bolus before intubation, followed by a 2 g/h infusion has also been advocated, because of inhibition of catecholamine release from the adrenal medulla, and antiarrhythmic and vasodilatory properties . Pheochromocytoma and surgical outcomes Despite careful preoperative preparation for surgery, many patients will still have a labile perioperative course. In a retrospective study by Plouin , patients with higher systolic blood pressures, recurrent tumor resections, and higher urine metanephrine concentrations were more likely to have complications or die in the perioperative period. Plasma catecholamine levels did not correlate with risk. Kinney and colleagues  did a retrospective review of all patients presenting for an initial pheochromocytoma or paraganglionoma resection at the Mayo Clinic. Of the 149 patients in the review period, there were no perioperative deaths. Almost 30% of patients had sustained episodes of hypertension or hypotension intraoperatively, despite most having been premedicated with both alpha and beta adrenergic blockers. A correlation was noted between large tumor size, higher levels of urinary metanephrines and catecholamines, and a prolonged surgical time and perioperative complications. However, the incidence of significant perioperative complications even in the presence of labile intraoperative blood pressure was surprisingly small in this review. None of the patients suffered MI or stroke. There were also no reports of ventricular dysrhythmias intraoperatively. Postoperative complications occurred in 6.3% of patients, mostly made up of a 4.2% incidence of prolonged intubation, and a 1.4% incidence of renal dysfunction. Summary Diabetes remains the most commonly encountered endocrinopathy with the incidence of type 2 doubling in the past decade. The prevalence of diabetes is projected to continue to increase dramatically over the next several decades unless major public health initiatives are successful in stemming this growth. Both type 1 and 2 diabetics more frequently require surgical and critical care than their non-diabetic counterparts. Type 1 and 2 diabetics also sustain greater perioperative morbidity and mortality. Careful preoperative assessment and appropriate perioperative intervention may limit this. There is increasing evidence that maintenance of normal blood glucose in the perioperative period and during critical illness is beneficial for diabetic and nondiabetic patients. More data will hopefully be forthcoming to substantiate recent reports and identify the mechanisms of improved outcome. L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 121 Thyroid disease remains a commonly encountered pathology that is more readily identified and controlled in the modern era of radioimmune assays of thyroid hormone and successful medical and surgical therapies. Severe hypothyroidism and thyroid storm are associated with significant increases in perioperative morbidity and mortality. Recognition of these entities or those at risk for developing them post operatively is crucial in initiating timely and effective therapy. Primary AI is uncommon, but results in glucocorticoid and mineralocorticoid deficiency. Tertiary AI is far more common, most often secondary to iatrogenic therapy with exogenous glucocorticoids for the management of chronic diseases such as connective tissue disorders, anti-rejection regimes, and severe asthma. Glucocorticoid replacement or supplementation is needed on a case-by-case basis and should be individualized based on chronic steroid dose, duration, and stress of the surgical procedure. Perioperative steroid dosing regimes now recommend lower doses for shorter periods than previously suggested. More recently AI has been recognized in two populations, elderly patients undergoing major surgery and a subgroup of patients with septic shock. Timely diagnosis using synthetic ACTH stimulation testing and stress glucocorticoid, and possibly mineralocorticoid therapy, seems to reverse these processes and improve recovery. Although uncommon, patients with pheochromocytoma who undergo open or laparoscopic resections remain diagnostic and therapeutic challenges. Perioperative outcome seems to have improved, in part, related to newer therapies and less invasive surgeries when indicated. The appropriate preoperative assessment and management of patients with various endocrinopathies is important to optimize outcome and limit avoidable complications. Hopefully additional evidence based guidelines will be forthcoming particularly in caring for the ever increasingly encountered perioperative diabetic. References  American Society of Anesthesiologists Task Force on Preanesthesia Evaluation. Practice advisory for preanesthesia evaluation: a report by the American Society of Anesthesiologists Task Force on preanesthesia evaluation. Anesthesiology 2002;96(2):485 – 96.  Glister B, Vigersky R. Perioperative management of type 1 diabetes mellitus. Endocrinol Metab Clinics 2003;32(2):411 – 36.  Weiland D, White R. Diabetes mellitus. Clini Fam Pract 2002;4(3):703 [Serial on-line. Available from MD consult. Accessed August 8, 2003].  Nathan D. Clinical review 146: the impact of clinical trials on the treatment of diabetes mellitus. J Clin Endocrinol Metab 2002;87(5):1929 – 37.  Eagle K, Berger P, Calkins H, Chaitman BR, Ewy GA, Fleischmann KE, et al. ACC/AHA guideline update for perioperative cardiovascular evaluation for noncardiac surgery- executive summary. A report of the American College of Cardiology/American Heart Association Task Force on practice guidelines (committee to update the 1996 guidelines on perioperative cardiovascular evaluation for noncardiac surgery). Anesth Analg 2002;94:1052 – 64.  Gottleib SS, McCarter RJ, Vogel RA. Effect of beta-blockade on mortality among high-risk and low-risk patients after MI. N Engl J Med 1998;339(8):489 – 97.  Boccara G, Outtara A, Godet G, Dufresne E, Bertrand M, Riou B, et al. Terlipressin versus 122                       L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 norepinephrine to correct refractory arterial hypotension after general anesthesia in patients chronically treated with renin-angiotensin system inhibitors. Anesthesiology 2003;98(6):1338 – 44. Ryckwaert F, Boccara G, Frappier JM, Colson PH. Incidence, risk factors, and prognosis of a moderate increase in plasma creatinine early after cardiac surgery. Crit Care Med 2002;30(7): 1495 – 8. Metchick L, Petit W, Inzucchi S. Inpatient management of diabetes mellitus. Am J Med 2002; 113(4):317 – 23. Bagg W, Ferri C, Desideri G, Gamble G, Ockelford P, Braatvedt GD. The influences of obesity and glycemic control on endothelial activation in patients with type 2 diabetes. J Clin Endocrin and Metab 2001;86(11):491 – 7. Gu W, Pagel P, Warltier DC, Kersten JR. Modifying cardiovascular risk in diabetes mellitus. Anesthesiology 2003;98(3):774 – 9. Van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M, et al. Intensive insulin therapy in critically ill patients. N Engl J Med 2001;345(19):1359 – 67. Malmberg K. Prospective randomised study of intensive insulin treatment on long term survival after acute myocardial infarction in patients with diabetes mellitus. DIGAMI (diabetes mellitus, insulin glucose infusion in myocardial infarction) study group. BMJ 1997;314:1512 – 5. Doyle DJ, Arellano R. Upper airway diseases and airway management: a synopsis. Anesth Clin North Am 2002;20(4):767 – 87. Gardner GM. Posterior glottic stenosis and bilateral vocal fold immobility: diagnosis and treatment. Otolaryngologic Clin N Am 2000;33(4):855 – 78. Kitamura A, Hoshino T, Kon T, Ogawa R. Patients with diabetic neuropathy are at risk of a greater intraoperative reduction in core temperature. Anesthesiology 2000;92(5):1311 – 8. Marks J. Perioperative management of diabetes. Am Fam Phys 2003;67(1):93 – 100. Levetan CS. Hospital management of diabetes. Endocrinol Metab Clin N Am 2000;29(4): 745 – 70. Luft FC. Lactic acidosis update for critical care clinicians. J Am Soc Nephrol 2001;12(Suppl 17): S15 – 9. Mercado DL. Perioperative medication management. Med Clin N Am 2003;87(1):41 – 57. Hemmerling T, Schmid MC, Schmidt J, Kern S, Jacobi KE. Comparison of a continuous glucose-insulin-potassium infusion versus intermittent bolus application of insulin on perioperative glucose control and hormone status in insulin-treated type 2 diabetics. J Clin Anesth 2001; 13(4):293 – 300. Angelini G, Ketzler JT, Coursin DB. Perioperative care of the diabetic. ASA Refresher Courses in Anesthesiology 2001;29:1 – 9. Krinsley JS. Association between hyperglycemia and increased hospital mortality in a heterogeneous population of critically ill patients. Mayo Clin Proceed 2003;78:1471 – 8. Furnary AP, Zerr KJ, Grunkemeier GL, Starr A. Continuous intravenous infusion reduces the incidence of deep sternal wound infection in diabetic patients after cardiac surgical procedures. Ann Thorac Surg 1999;67:352 – 62. Weir CJ, Murray GD, Dyker AG, Lees KR. Is hyperglycemia an independent predictor of poor outcome after acute stroke? Results of a long-term follow up study. BMJ 1997;314: 1303 – 6. Sala J, Maisa Gonzalez de Molina FJ, Fernandez-Real JM, Gil M, Bosch D, Ricart W, et al. Short-term mortality of MI patients with diabetes or hyperglycaemia during admission. J Epidemiol Comm Health 2002;56(9):707 – 12. Ringel MD. Management of hypothyroidism and hyperthyroidism in the intensive care unit. Crit Care Clin 2001;17(1):59 – 74. Larsen RP, Davies TF. Hypothyroidism and thyroiditis. In: Larsen PR, Kronenberg HM, Shlomo M, Polonsky KS, editors. Williams textbook of endocrinology. 10th edition. Philadelphia: WB Saunders; 2003. p. 423 – 8. Spell III NO. Stopping and starting medications in the perioperative period. Med Clin N Am 2001;85(5):1127 – 8. L.E. Connery, D.B. Coursin / Anesthesiology Clin N Am 22 (2004) 93–123 123  Schiff R, Welsh G. Perioperative evaluation and management of the patient with endocrine dysfunction. Med Clin N Am 2003;87(1):175 – 92.  Ginsberg J. Diagnosis and management of Grave’s disease. CMAJ 2003;168(5):575 – 85.  Jabbour SA. Steroids and the surgical patient. Med Clin N Am 2001;85(5):1140 – 7.  Coursin D, Wood K. Corticosteroid supplementation in adrenal insufficiency. JAMA 2002; 287(2):236 – 40.  Axelrod L. Perioperative management of patients treated with glucocorticoids. Endocrinol Metab Clin N Am 2003;32(2):367 – 83.  Annane D. Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock. JAMA 2002;288(7):862 – 71.  Rivers EP, Gaspari M, Saad GA. Adrenal insufficiency in high-risk surgical ICU patients. Chest 2001;119(3):889 – 96.  Pacak K, Linehan WM, Eisenhofer G, Walther MM, Goldstein DS. Recent advances in genetics, localization, and treatment of pheochromocytoma. Ann Intern Med 2001;134(4):315 – 29.  Caputo C, Fishbane S, Shapiro L, Courgi RG, Kostadinov S, Donovan V, et al. Pheochromocytoma multisystem crisis in a patient with multiple endocrine neoplasia type II B and pyelonephritis. Am J Kidney Dis 2002;39(6):E23.  Bogolioubov A, Keefe D, Groeger J. Circulatory shock. Crit Care Clin 2001;17(3):697 – 719.  Bentrem D, Pappas SG, Ahuja Y, Murayama KM, Angelos P. Contemporary surgical management of pheochromocytoma. Am J Surg 2002;184(6):621 – 5.  Lenders J, Pacak K, Walther MM, Linehan WM, Mannelli M, Friberg P, et al. Biochemical diagnosis of pheochromocytoma: which test is best? JAMA 2002;287(11):1427 – 34.  McHenry C. What’s new in general surgery: endocrine surgery. J Am Coll Surg 2002;195(3): 364 – 71.  Liao WB, Liu CF, Chiang CW, Kung CT, Lee CW. Cardiovascular manifestations of pheochromocytoma. Am J Emerg Med 2000;18(5):622 – 5.  Plouin PF. Factors associated with perioperative morbidity and mortality in patients with pheochromocytoma: analysis of 165 operations at a single center. J Clin Endocrin Metab 2001;86(4): 1480 – 6.  Kinney M, Warner ME, van Heerden JA, Horlocker TT, Young Jr WF, Schroeder DR, et al. Perianesthetic risks and outcomes of pheochromocytoma and paraganglionoma resection. Anesth Analg 2000;91(5):1118 – 23.  Gagner M, Lacroix A, Bolte E. Laparoscopic adrenalectomy in Cushing’s syndrome and pheochromocytoma. N Engl J Med 1992;327(14):1033.  Cooper Z, Mihm F. Blood pressure control with fenoldopam during excision of a pheochromocytoma. Anesthesiology 1999;91(2):558 – 60.  Pivalizza E. Magnesium sulfate and epidural anesthesia in pheochromocytoma and severe coronary artery disease. Anesth Analg 1995;81(2):414 – 6.
© Copyright 2018