Abstract
Hypoglycaemia is rare in healthy individuals owing to the numerous elegant hormonal and neuronal mechanisms that maintain glucose homeostasis. Glucose is an obligate metabolic fuel for cerebral tissue and therefore hypoglycaemia, if uncorrected, can have disastrous consequences including death. Clinical hypoglycaemia is defined as a plasma (or serum) glucose concentration low enough to cause symptoms and/or signs, including impairment of brain function. However, no single plasma (or serum) glucose concentration categorically defines hypoglycaemia. Hypoglycaemia is probably the most common endocrine and metabolic emergency in clinical practice. The overwhelming majority of occurrences of hypoglycaemia occur in patients with diabetes, either as a result of treatment-induced hypoglycaemia and/or abnormalities that affect the normal counterregulatory response to hypoglycaemia. The differential for nondiabetes-associated hypoglycaemia is broad and includes insulinoma, drugs, hormone deficiencies, and critical illness. The acute management of hypoglycaemia is discussed along with a review of the pathophysiology and aetiology of this commonly encountered clinical problem.
Introduction
Acute hypoglycaemia is one of the most feared medical emergencies owing to the associated morbidity and mortality, and significant impedance on daily living and quality of life in those patients with frequent risk of hypoglycaemia (i.e. diabetic patients). Beyond its personal costs, hypoglycaemia has a wide-reaching economic impact. The word hypoglycaemia literally means ‘under-sweet blood’, taken from ancient Greek, and the exact prevalence is difficult to quantify, as many events are self-diagnosed and managed outside of a medical facility. Despite this, hypoglycaemia is probably the most common endocrine and metabolic emergency in clinical practice.
Clinical hypoglycaemia is defined as a plasma (or serum) glucose concentration low enough to cause symptoms and/or signs, including impairment of brain function [Cryer et al. 2009]. However, no single plasma (or serum) glucose concentration categorically defines hypoglycaemia [American Diabetes Association, 2005].
Hypoglycaemia is significantly more common in diabetes patients, particularly if treated with hypoglycaemic agents (i.e. capable of lowering blood glucose to the extent of severe hypoglycaemia and neuroglycopaenia, e.g. insulin, sulfonylureas and meglitinides). Indeed, hypoglycaemia is the critical limiting factor in glycaemic management in both the short and long term [American Diabetes Association, 2011].
Hypoglycaemia results in recurrent morbidity in most persons with type 1 diabetes mellitus (T1DM). Whilst the landmark Diabetes Control and Complications Trial (DCCT) in T1DM patients reported unequivocal salutary effects of intensive diabetes control on long-term complications, severe hypoglycaemia (i.e. an episode requiring assistance to treat) was increased threefold in the intensively treated group [Diabetes Control and Complications Research Group, 1993]. In addition, it has been estimated that 2–4% of people with T1DM die from hypoglycaemia [Cryer et al. 2009]. Although prolonged, profound hypoglycaemia can cause neurological damage and brain death, the mechanism(s) of sudden death in less profound hypoglycaemia is unknown, but may be due to cardiac arrhythmia related to the prolonged QT interval caused by hypoglycaemia-induced sympathetic stimulation [Graveling and Frier, 2010]. Sudden nocturnal death (‘dead in bed’ syndrome) in young people with T1DM may be due to this cause.
In comparison, hypoglycaemia is less frequent in type 2 diabetes mellitus (T2DM) than in T1DM. However, hypoglycaemia becomes progressively more limiting to glycaemic control especially in advanced (i.e. absolute endogenous insulin-deficient) T2DM [Fisher, 2010]. As in T1DM patients, hypoglycaemia can result in increased mortality in T2DM patients. In younger T2DM patients (aged 20–49 years), between 6% and 18% of deaths have been attributed to hypoglycaemia [Graveling and Frier, 2010]. Several recent trials of intensive glucose control in long-standing T2DM have refocused attention on the possible relationship of hypoglycaemia with excess mortality. For example, an excess of deaths in the intensive treatment arm of the ACCORD (Action to Control Cardiovascular Risk in Diabetes) study led to early discontinuation of the trial [ACCORD Study Group, 2008]. Whilst no definitive cause of the increased mortality has been proven, increased hypoglycaemia occurrence (i.e. annual prevalence of severe hypoglycaemia of 3.3% in the intensive group compared with 1.1% in the controls) is likely to be relevant especially in those patients at high risk or with established cardiovascular disease [Zoungas et al. 2010]. Hypoglycaemia provokes sympatho-adrenal activation and counterregulatory hormone secretion resulting in profound physiological effects on a diseased coronary vasculature and a dysfunctional cardiac conductive system [Graveling and Frier, 2010]. Recurrent severe hypoglycaemia in T2DM patients has also been associated with increased long-term risk of developing dementia [Bailey and Day, 2010].
Tight glycaemic control in critically ill patients became a therapeutic paradigm after the 2001 publication of a single-centre intervention trial that targeted euglycaemia in mechanically ventilated patients in a surgical intensive care unit (ICU) [Van den Berghe et al. 2001]. However, subsequent studies (including one set in the medical ICU from the above group) and a meta-analysis failed to demonstrate a survival benefit for tight glycaemic control in critically ill patients [Weiner et al. 2008]. Indeed, the large multicentre NICE-SUGAR (Normoglycaemia in Intensive Care Evaluation—Survival Using Glucose Algorithm Regulation) trial, targeting intensive glucose control (4.5–6 mmol/l [81–108 mg/dl])
In comparison, spontaneous hypoglycaemia in the nondiabetic population is relatively uncommon but creates both diagnostic and therapeutic challenges to healthcare providers.
Appropriate initial evaluation and treatment of hypoglycaemia during the acute presentation are not only crucial in preventing morbidity and mortality, but also necessary for establishing the underlying aetiology.
Pathophysiology
Numerous elegant mechanisms have evolved to ensure that blood glucose levels remain within a narrow window thus allowing normal cerebral function (i.e. glucose is an obligate metabolic fuel for brain neurones), and also to provide rapid counterregulatory mechanisms, predominantly via hormones (including insulin, glucagon, adrenaline [epinephrine], cortisol and growth hormone [GH]), when the blood glucose drops below this level [Matfin, 2009]. Therefore, hypoglycaemia below a threshold will result in neuroglycopaenia with consequent cerebral dysfunction.
The liver functions as a glucose buffer system, maintaining tight control of plasma glucose concentrations. In the fasting (postabsorptive) state, hepatic glycogenolysis and gluconeogenesis maintains glucose concentrations. In the postprandial state, glucose directly (and indirectly by the secretion of incretins such as glucagon-like-peptide-1 [GLP-1]) stimulates the pancreatic β cell to secrete insulin. Insulin acts as both a regulatory and counterregulatory hormone in glucose homeostasis, through mediating increased glucose uptake in insulin-responsive tissues (i.e. liver, adipose tissue and skeletal muscle), suppressing hepatic glycogenolysis and gluconeogenesis, and conversely allowing both these counterregulatory processes if its output is inhibited.
Glucagon normally acts on the liver to increase glycogenolysis and gluconeogenesis, and is secreted from the pancreatic α cells in response to low extracellular glucose concentrations. Adrenaline (epinephrine) has a less significant role than glucagon, but severe or absolute deficiency can promote hypoglycaemia. Slower acting responses to hypoglycaemia include the release of cortisol and GH, which prevent uptake of glucose by the peripheral tissues.
Hypoglycaemic unawareness
Awareness of hypoglycaemia is mainly the result of the perception of neurogenic symptoms. In diabetes, the development of autonomic dysfunction can result in an impaired adrenergic response, with resultant loss of warning symptoms. This phenomenon is known as ‘hypoglycaemia unawareness’. Hypoglycaemic unawareness affects ∼25% of T1DM patients and is increasingly recognized in T2DM. The risk of severe hypoglycaemia is increased sixfold to sevenfold in patients with hypoglycaemia unawareness.
Hypoglycaemia unawareness can result from recurrent hypoglycaemia (termed ‘hypoglycaemia-associated autonomic failure’ [HAAF]); or in long-standing diabetes as a consequence of decreased glucagon and sympatho-adrenal responses, and autonomic neuropathy [Cryer et al. 2009].
Aetiology
Hypoglycaemia results from a reduction in available glucose in the context of reduced intake, increased utilization, or reduced glycogen storage; or in the context of a relative increased circulating concentration of either endogenous or exogenous insulin (Table 1). Rarely, deficiencies in hormones involved in the normal counterregulatory response against hypoglycaemia can be responsible (e.g. Addison’s disease and hypopituitarism)
Examples of the aetiology of hypoglycaemia.
NIPHS, non-insulinoma pancreatogenous hypoglycaemia syndrome.
Drugs are the most common cause of hypoglycaemia. In this context, one must always consider dispensing error, accident, factitious or malicious disorders. Drug-induced hypoglycaemia, as a result of hypoglycaemic agent usage (e.g. insulin, sulfonylureas and meglitinides), is the most common cause of diabetes-associated hypoglycaemia and represents the overwhelming majority of hypoglycaemic episodes encountered by most healthcare providers. Clinicians have recognized the problem of iatrogenic hypoglycaemia since the first use of insulin in 1922 [Fletcher and Campbell, 1922]. The proportion of diabetic patients affected by one or more severe hypoglycaemic episodes has been variously documented as 10–30% annually for T1DM; generally <5% annually for insulin-treated T2DM; and typically <1% annually for sulfonylurea-treated T2DM [Bailey and Day, 2010]. In contrast, antihyperglycaemic agents (i.e. lower blood glucose without predisposing to severe hypoglycaemia) such as metformin, thiazolidinediones (TZDs), α-glucosidase inhibitors, and incretin-based therapies (e.g. GLP-1 analogues and dipeptidyl peptidase [DPP]-4 inhibitors) are not associated with hypoglycaemia when used as single agents [Bailey and Day, 2010].
Alcohol is perhaps the next most common cause of drug-induced hypoglycaemia, and acts predominantly by inhibiting gluconeogenesis (Table 1). Salicylates also reduce basal glucose output if taken in large doses and can induce hypoglycaemia. Pentamidine, an antimicrobial agent used in the treatment of
Hypoglycaemia is not an infrequent finding in the critically ill patient and may be related to sepsis, hepatic or renal failure, or to general malnourishment. Hypoglycaemia is seen in approximately 1% of patients admitted to an ICU, although this figure rises to 2.1–11.5% when intensive insulin therapy is administered as per many guidelines [Prieto-Sanchez, 2011; Qaseem et al. 2011; Krinsley and Keegan, 2010; Waeschle et al. 2008; Vriesendorp et al. 2006]. Hepatic failure leads to an inability to maintain adequate fasting glucose levels despite adaptive renal gluconeogenesis. However, significant hepatic function must be lost before hypoglycaemia is seen. Renal failure may be associated with hypoglycaemia although the causes are multifactorial. Of particular note, though, is the reduction in both insulin and sulfonylurea dosages that must be made so as to avoid hypoglycaemia in the diabetic patient with progressive renal dysfunction due to decreased clearance [Shrishrimal et al. 2009].
Deficiency of cortisol and/or GH can lead to hypoglycaemia through impairment of the counterregulatory mechanisms, but hypoglycaemia is a rare finding in the context of GH deficiency. Hypoglycaemia complicates many cases of childhood adrenal insufficiency, but is an uncommon finding if Addison’s disease develops in later life [Arlt, 2009; Artavia-Loria et al. 1986]. Non-islet cell tumour hypoglycaemia (NICTH) can result from either increased glucose utilization or via the production and secretion of insulin-like factors (i.e. precursor forms of insulin-like growth factor-2 [IGF-2], termed ‘big’ IGF-2) which stimulate the insulin receptor. These tumours are typically large mesenchymal tumours (∼50% of cases of NICTH); or hepatocellular tumours (∼25% of cases of NICTH). Other rare causes of hypoglycaemia include inherited metabolic disorders (e.g. hereditary fructose intolerance).
Hypoglycaemia associated with endogenous hyperinsulinaemia is very rare. Pancreatic islet cell adenomas with autonomous insulin production, commonly termed an insulinoma, are rare gastropancreatic neuroendocrine tumours (NETs) with an estimated incidence of 1 or 2 per million [ENETS guidelines, 2007]. As with other causes of endogenous hyperinsulinaemia, hypoglycaemia commonly presents in the fasting state, with the patient complaining of symptoms on waking in the morning. The minority of insulinomas are malignant (10%) and occasionally multiple (10%), although both of these features are more common in the 10% of insulinomas associated with multiple endocrine neoplasia type 1 (MEN 1). Non-insulinoma pancreatogenous hypoglycaemia syndrome (NIPHS) is associated with nesidioblastosis (focal or diffuse islet cell hyperplasia) with resultant dysfunctional insulin secretion [Barnard et al. 2010]. NIPHS is characterized by postprandial attacks of neuroglycopaenia due to endogenous hyperinsulinaemia. The frequency of this condition, and indeed its true aetiology, are debated, although it has been associated with a history of bariatric surgery [Barnard et al. 2010; Cryer et al. 2009; Clancy et al. 2006]. Autoimmune hypoglycaemia, mediated via antibodies against the insulin molecule or the insulin receptor, has also been described [Basu et al. 2005].
Reactive (postprandial) hypoglycaemia occurs exclusively after meals. Alimentary hypoglycaemia, with hypoglycaemia occurring within 4 hours of a meal, and secondary to marked early hyperinsulinaemia and GLP-1 release, is well recognized, and is associated with previous gastric surgery. The existence of idiopathic reactive hypoglycaemia is still debated, not least because a significant proportion of the normal population (∼25%) can have asymptomatic hypoglycaemia following an oral glucose tolerance test (OGTT) [Lev-Ran and Anderson, 1981]. The diagnosis should only be considered when Whipple’s triad (see the next section) is also satisfied. It should also be noted that causes of hypoglycaemia usually presenting in the fasting state, can occasionally cause reactive hypoglycaemia (e.g. insulinoma).
Diagnostic considerations
Traditionally, pathological hypoglycaemia has been confirmed by documentation of Whipple’s triad: symptoms, signs, or both consistent with hypoglycaemia; a low plasma glucose concentration; and resolution of those symptoms or signs after restoration of normoglycaemia [Whipple, 1938]. In the nondiabetic patient, Whipple’s triad should always be present before embarking on further investigation into possible hypoglycaemia. Conversely, in the diabetic patient on insulin or an oral hypoglycaemic agent who exhibits typical symptoms or signs of hypoglycaemia, documentation of a low plasma glucose concentration during an event is recommended but not mandatory, as the likelihood of true hypoglycaemia is significant in this scenario.
The diagnosis of hypoglycaemia should not be based solely on capillary blood glucose measurements (i.e. using a glucometer), which are often inaccurate in the hypoglycaemic range. However, a capillary blood glucose measurement is useful if hypoglycaemia is suspected, and if low should be confirmed ideally with a laboratory glucose measurement. If no clinically obvious cause of hypoglycaemia is evident (i.e. in a seemingly healthy individual), a further blood sample (∼20 ml) should be centrifuged and the serum saved for later analysis. However, correction of the hypoglycaemia should not be delayed if severe neuroglycopaenic signs are present.
After stabilization, subsequent investigations will be directed at finding the cause of the hypoglycaemia. Renal and liver function tests should be measured to exclude renal impairment (e.g. diabetic nephropathy) or liver failure as the cause of the hypoglycaemia. The serum saved previously during the hypoglycaemic episode should be assayed for insulin, C-peptide, pro-insulin, β-hydroxybutyrate (blood ketones can be low in hyperinsulinaemic states; and high with counterregulatory failure, as in hypopituitarism), and cortisol and GH as indicated. If the insulin concentration is not suppressed, a concomitant low C-peptide level suggests exogenous insulin, whilst high C-peptide levels suggest endogenous insulin (e.g. insulinoma, stimulation by oral hypoglycaemic agents). Urine and/or serum sulfonylurea screen should be measured for further investigation of endogenous insulin production leading to hypoglycaemia (if detected this could have been taken as prescribed, accidently, factitiously or maliciously). Serum levels of incompletely processed IGF-2 (i.e. ‘big’ IGF-2) may be high with NICTH. These patients have suppressed insulin levels, low IGF-1 and a raised ratio of IGF-2 to IGF-1. Blood alcohol may not be useful in alcohol-induced hypoglycaemia, because of a lag of 6–24 hours between ingestion and hypoglycaemia onset. Insulin and insulin receptor antibodies can be measured as warranted.
When a spontaneous hypoglycaemic episode cannot be observed, hypoglycaemia can be provoked in controlled situations so as to obtain the above samples at a time of hypoglycaemia. The standard test is a 72 hour fast (allowing unlimited noncaloric fluids), and involves observation of the patient in a controlled environment. The patient should be active during waking hours, and facilities to reverse hypoglycaemia should be at hand. If postprandial hypoglycaemia is suspected, a mixed meal can be administered, with regular timed collection of the blood samples noted above. A full description of the diagnostic algorithm for hyperinsulinaemic hypoglycaemia, including a protocol for provocation of hypoglycaemia, has been comprehensively outlined elsewhere [Cryer et al. 2009].
Clinical signs and features
The clinical manifestations of hypoglycaemia are nonspecific and symptoms cannot be defined by plasma glucose levels. However, the signs and symptoms can be divided into the effects caused by adrenergic symptoms and signs, and those resulting from neuroglycopaenia. Sympatho-adrenal activation results in tremor, sweating, hunger, anxiety and palpitations. However, these symptoms can be found in many other acute medical states (e.g. acute coronary syndromes [ACS] and shock).
The central nervous system relies primarily on glucose as the main substrate for normal function. It takes several days for the brain to shift to using ketones as the major fuel source in prolonged starvation. Neuroglycopaenic symptoms are protean and include impaired judgement, blurred vision, diplopia, slurred speech, incoordination, ataxia, focal neurological deficit, headaches, atypical behaviour as a result of loss of normal inhibitions, aggressiveness, confusion, somnolence, seizures and coma. Prolonged episodes of severe hypoglycaemia may result in irreversible brain damage and even death, whilst the adrenergic response to hypoglycaemia may place intolerable strain upon the cardiovascular system if there is underlying cardiovascular disease [Graveling and Frier, 2010].
Neuroglycopaenic symptoms are usually preceded by adrenergic symptoms, especially in patients with diabetes. This does not always apply to insulinoma patients, who frequently present with neuroglycopaenic symptoms such as diplopia, blurred vision, abnormal behaviour, confusion or coma. Patients with diabetes can also suffer from neuroglycopaenia without preceding adrenergic warning symptoms in hypoglycaemic unawareness.
Acute hypoglycaemia is a great impersonator. For example, the presentation of slurred speech with a focal deficit may lead the clinician to diagnose a transient ischaemic attack (TIA) or stroke, whereas emotional lability and confusion may mimic alcohol intoxication, drug overdose or encephalitis. For this reason it is imperative to check the blood glucose level in any scenario involving an acutely unwell patient. The test itself is quick, simple and minimally invasive; the result, however, may be life saving.
Nocturnal hypoglycaemia
Nocturnal hypoglycaemia represents an atypical presentation of hypoglycaemia. Normal insulin requirements fall during the night, but increase again in the early hours of the morning (termed ‘dawn phenomenon’). Patients with residual background endogenous insulin production, who require exogenous insulin to optimize glycaemic control, are at particular risk of hypoglycaemia at this time. Whilst the episode of hypoglycaemia itself may be missed, patients often report waking in the morning feeling unwell (e.g. headaches and night sweats). Frequently, fasting glucose levels are high following nocturnal hypoglycaemia as a consequence of hormonal counterregulatory mechanisms (known as the ‘Somogyi effect’). The clinician should consider the possibility of occult nocturnal hypoglycaemia whenever reviewing disproportionately raised fasting glucose measurements in the diabetic patient and can be further evaluated by occasional 3 am glucose measurements.
Sudden nocturnal death (‘dead in bed’ syndrome) in young people with T1DM may be due to cardiac arrhythmia related to a prolonged QT interval caused by hypoglycaemia-induced sympathetic stimulation [Graveling and Frier, 2010].
Management of hypoglycaemia
Many cases of hypoglycaemia can be appropriately managed in the community setting. However, hypoglycaemia requiring third-party intervention (i.e. severe hypoglycaemia) should prompt assessment at a hospital and, if required, the patient should be managed in an appropriate location such as an acute medical unit (AMU), high-dependency area or ICU. As with all acute medical patients, prompt assessment and management of the ABCDEs should occur (i.e. Airway; Breathing; Circulation; Disability [i.e. conscious level]; and Examination and Evaluation).
The management of hypoglycaemia can be divided into three phases:
acute intervention – to prevent and minimize neurological damage; maintenance therapy – to prevent recurrence of hypoglycaemia; subsequent measures – to search for and treat the underlying cause.
Acute intervention
Prolonged severe hypoglycaemia can cause irreversible cerebral dysfunction and must be treated promptly. However, as indicated above, it is important if possible to obtain a blood sample for laboratory glucose measurement before glucose administration and to save serum for more sophisticated investigation if the cause of hypoglycaemia is not obvious (i.e. hypoglycaemia in a seemingly healthy, nondiabetic patient).
If the patient has a history of malnutrition or chronic alcohol abuse, intravenous (IV) thiamine at a bolus dose of 1–2 mg/kg should be given before initiation of glucose treatment, to avoid precipitating Wernicke’s encephalopathy.
A treatment algorithm for acute hypoglycaemia is outlined in Figure 1. The treatment for all hypoglycaemia events is the administration of glucose. The route and amount of administration will depend on the glucose level as well as the patient’s level of consciousness and available access. Where possible an oral carbohydrate load should be administered urgently, followed by careful blood glucose monitoring. Consider the ‘rule of 15s’ during therapy (i.e. 15 g of carbohydrate will raise the glucose level about 15 mg/dl [∼0.8 mmol/l] in about 15 minutes). If the oral route is not possible, such as in cases of impaired consciousness or an uncooperative patient, intramuscular (IM) or subcutaneous (SC) glucagon or IV glucose can be considered (Figure 1).
Oral carbohydrate is ideally provided in the form of 15–20 g (or 20–30 g if blood glucose ≤2.8 mmol/l [≤50 mg/dl]) glucose tablets as the administered load is quantifiable, thus reducing the risk of rebound hyperglycaemia [Tomky, 2005]. However, any form of carbohydrate that contains glucose can be used [American Diabetes Association, 2011]. This dose of glucose will typically maintain euglycaemia for up to 2 hours, and therefore a complex carbohydrate (e.g. snack or meal) should be administered as soon as it is safe to do so. Failure of the hypoglycaemia to correct within 15 minutes following one dose of glucose should lead to administration of a second dose, and occasionally a third, but failure thereafter should prompt the clinician to consider other interventions. The patient should be warned not to drive for at least 45 minutes after correction of hypoglycaemia.
An algorithm for the management of acute hypoglycaemia.
Oral treatments are clearly inappropriate in the unresponsive patient or those who are unable to take oral medications. In this situation, the presence of IV access allows 25–50 ml (i.e. 12.5–25 g) 50% dextrose (D50) to be administered and is adequate in most circumstances. D50 is highly irritating and should be administered through a large gauge needle into a large vein if possible and followed by a saline flush. Larger volumes of less concentrated dextrose in IV infusions (e.g. 250 ml of 10% dextrose [D10]) may be used to minimize irritation. For inpatients with hypoglycaemia, D50 mixed with equal parts of water can also be given through a feeding tube if available.
If IV access is not available, or is delayed, glucagon 1 mg IM (or SC) can be administered. Glucagon stimulates hepatic glycogenolysis and gluconeogenesis and thereby raises blood glucose in 5–10 minutes, but its action is short lived. However, glucagon may not be effective in cases where gluconeogenesis is defective, such as in cases of extreme fasting, liver failure, alcohol-induced hypoglycaemia or adrenal insufficiency because of glycogen depletion. Glucagon in these situations may paradoxically aggravate the hypoglycaemia by stimulating insulin secretion. Glucocorticoid replacement (after plasma is saved for cortisol testing) is required for patients with suspected adrenal insufficiency. Owing to these various issues, IV glucose remains the treatment of choice for severe hypoglycaemia.
Maintenance therapy
The clinical response of hypoglycaemia to IV glucose administration should be rapid and dramatic. Patients with hypoglycaemic coma are expected to regain consciousness and become coherent within 5–10 minutes. The response to IM (or SC) glucagon is slightly slower, with an average time difference of 2–3 minutes when compared with IV glucose. However, complete cognitive recovery may be delayed for 30–60 minutes after restoration of normoglycaemia. If there is no obvious improvement in symptoms or consciousness within 10–15 minutes, alternative diagnoses (e.g. stroke or drug overdose) should be reconsidered.
A common mistake is to assume that once the glucose has been corrected, it will maintain itself. Depending on the initial cause for hypoglycaemia, a concurrent source of glucose may need to be administered for some time. As the effect of IV glucose is relatively transient, patients should receive an additional form of glucose along with protein (e.g. milk, cheese and crackers) if between meals or the regular meal is more than one hour away, to replenish hepatic glycogen stores. If the hypoglycaemic episode is expected to be prolonged or recurrent (e.g. due to long-acting insulin or sulfonylurea), an IV infusion of 5–10% dextrose (D5 or D10) should be commenced and continued as necessary. For example, sulfonylurea-induced hypoglycaemia, which due to long duration of action can cause prolonged hypoglycaemic events (especially in the elderly or in patients with renal impairment). Octreotide, a synthetic somatostatin analogue, inhibits insulin release and has been used to treat hypoglycaemia in this context. Various regimens are described but 50 µg 6–8 hourly administered IV or SC are commonly used [Roberts, 2010]. The use of this agent should be considered in any situation where endogenous hyperinsulinaemia is apparent.
Subsequent measures
After initial stabilization, subsequent management should be directed at searching for the underlying aetiology of hypoglycaemia and preventing further attacks [Cryer et al. 2009]. Repeated hypoglycaemia in an otherwise stable diabetic patient should alert the healthcare provider of the onset of nephropathy, concomitant Addison’s disease, hypothyroidism, hypopituitarism or interfering medications (e.g. angiotensin-converting enzyme [ACE] inhibitors). The symptoms and their relation to meals is important, as is drug history (prescribed and over-the-counter [OTC]), and history of gastric surgery. Once the underlying cause is established, definitive therapy should be offered.
Treatment of non-diabetes-related hypoglycaemia
Nondiabetic hypoglycaemia definitive management depends on the underlying aetiology. Hypoglycaemia induced by medications improves promptly once the medication is removed (with the exception of pentamidine), whilst correction of sepsis and improvement in hepatic and renal function improves hypoglycaemia of the critical illness. Deficiencies of counterregulatory hormones can be corrected with replacement of relevant hormone(s).
Medications available for the treatment of hypoglycaemia.
IM, intramuscular; IV, intravenous; SC, subcutaneous.
Diazoxide is a potassium channel activator, first developed as an antihypertensive agent, but now more commonly used in the context of hypoglycaemia due to inhibition of insulin secretion it engenders. It is administered at a dose of 5 mg/kg/day (with higher doses in refractory cases up to 15 mg/kg/day) in two or three divided oral doses (e.g. 200–1200 mg/day) [BNF, 2011]. Diazoxide has numerous side effects including peripheral oedema, nausea, vomiting, hypotension and arrhythmias. Of concern, particularly to females, is the tendency for diazoxide to promote hypertrichosis, whilst pancytopaenia is occasionally seen in an idiosyncratic manner. Thiazide diuretics synergise the hyperglycaemic effect of diazoxide as well as reduce the fluid retention, and can be used in addition.
As noted above, somatostatin inhibits insulin production, and analogues can be used in any state of chronic hyperinsulinaemia [Vezzosi et al. 2008]. As with many well-differentiated NETs, insulinomas frequently retain a response to somatostatin and insulin output can be reduced significantly with the use of this medication [Oberg et al. 2004]. Octreotide therapy is effective in reducing hypoglycaemia in over 50% of patients with an insulinoma and can be administered as a long-acting formulation [Vezzosi et al. 2008]. However, a paradoxical fall in blood glucose levels can occur in approximately 50% of patients because of suppression of counterregulatory hormones such as glucagon. Octreotide is commenced at a dose of 50 µg three times daily by SC injection, and can be titrated to a maximum dose of 500 µg three times daily. Patients are usually stabilized with short-acting octreotide for 10–28 days before converting them to long-acting somatostatin analogues. Long-acting somatostatin analogues include: Sandostatin LAR, administered as an IM depot, given at an initial dose of 20 mg every 4 weeks and adjusted after 3 months to a maximal dose of 40 mg every 4 weeks; Lanreotide LA, administered as an IM depot 30 mg every 2 weeks and can be increased to 30 mg every 7 days; and Lanreotide Autogel, administered as a deep SC injection of 60 mg every 4 weeks and increased after 3 months to 120 mg every 4 weeks if needed.
Everolimus, an inhibitor of the mammalian target of ramamycin (mTOR), has shown promise in the treatment of NETs [Yao et al. 2011]. A frequent consequence of treatment is hyperglycaemia, purported to be secondary to mTOR regulated downregulation of insulin receptors, which has prompted some groups to use everolimus to treat symptomatic malignant insulinoma. A recent report has suggested efficacy of this approach, although further data to confirm effectiveness and to define the role of everolimus in non-malignant insulinoma are required [Kulke et al. 2009].
Many other medications have been used in the management of hyperinsulinaemic hypoglycaemia including verapamil (i.e. because calcium influx is required for insulin secretion, calcium channel blockers have been tried for treatment of hyperinsulinaemia), α-glucosidase inhibitors (e.g. acarbose or miglitol) and glucocorticoids, but evidence supporting their efficacy is based primarily on case reports only [ENETS guidelines, 2007].
Glucocorticoid and/or GH therapy have been used with some success in patients with NICTH, for tumors which cannot be resected completely.
Treatment of diabetes-related hypoglycaemia
Hypoglycaemia is the critical limiting factor in the glycaemic management of diabetes in both the short- and long-term [American Diabetes Association, 2011; Bailey and Day, 2010]. It is important to address hypoglycaemia at each patient contact [Cryer et al. 2009]. Evaluation of hypoglycaemia in a diabetic patient is focused on intensity of glycaemic control and the treatment regimen. If hypoglycaemic agents (e.g. insulin, sulfonylureas or meglitinides) are prescribed, adjustments may be required to prevent further episodes. Patient education is vital with respect to diet, exercise, timing of medications, insulin injection sites and frequent self-monitoring of blood glucose (SMBG) or continuous glucose monitoring (CGM), to try to avoid hypoglycaemia. The patient should be concerned about the possibility of developing hypoglycaemia when the SMBG (or CGM which lags SMBG) is falling rapidly or is less than 4 mmol/l (∼70 mg/dl). In the UK, it has been recommended that ‘four is the floor’ for blood glucose targets in order to limit hypoglycaemia frequency. Similarly from a healthcare providers perspective, recurrent hypoglycaemia or hypoglycaemia in a patient for whom it confers considerable risk, should be avoided by employing less tight glycaemic control (i.e. goals need to be individualized) [American Diabetes Association, 2011]. Glucagon should be prescribed for all individuals at significant risk of severe hypoglycaemia.
Several studies showing increased mortality in critically ill patients treated with intensive glycaemic control [Prieto-Sanchez, 2011; Qaseem et al. 2011; Krinsley and Keegan, 2010; Mayfroidt et al. 2010; NICE-SUGAR Study Investigators, 2009], have led to a relaxing of glycaemic targets in this setting with goals of 7.8–10 mmol/l (140–180 mg/dl) [American Diabetes Association, 2011]. For noncritically ill patients, glycaemic goals are premeal and random blood glucose levels <140 mg/dl (7.8 mmol/l) and <180 mg/dl (10 mmol/l), respectively [American Diabetes Association, 2011; Prieto-Sanchez, 2011]. More recently, the American College of Physicians published guidelines which recommends a target blood glucose range of 140–200 mg/dl [7.8–11.1 mmol/l] if intensive insulin therapy is used in ICU patients [Qaseem et al. 2011]. In addition, a hypoglycaemia management protocol should be adopted and implemented in each hospital [American Diabetes Association, 2011].
As discussed earlier, the phenomenon of hypoglycaemic unawareness is due to an attenuated adrenergic response, resulting in a deficiency of hypoglycaemic ‘warning signs’ for the patient. This can be due to HAAF or long-standing diabetes (with defective counterregulation and/or autonomic neuropathy). HAAF can be managed by complete removal of hypoglycaemia for a period of at least 2–3 weeks by raising glycaemic targets. This frequently results in a return of hypoglycaemic awareness [Cryer et al. 2009].
Emerging therapies
Improvements in insulin administration (i.e. insulin analogues, insulin pumps, pancreas and islet transplantation, and experimental artificial pancreas ‘closed-loop’ systems); blood glucose monitoring tools (i.e. SMBG and CGM); and antihyperglycaemic agents (especially incretin-based therapies, and experimental agents such as renal sodium–glucose contransporter [SGLT] inhibitors), may result in less frequent and severe hypoglycaemic attacks in some diabetes patients. Unfortunately, the ultimate goal of achieving intensive glycaemic control, whilst at the same time protecting the brain from the potentially devastating effects of neuroglycopaenia (i.e. using neuroprotection), and/or augmenting glucose counterregulation, is still a long way off [Heller, 2008].
Conclusions
Hypoglycaemia is a common occurrence, especially in the context of diabetes. Acute hypoglycaemia is a medical emergency and must be promptly treated to avoid serious morbidity and even mortality. Work up of seemingly healthy patients presenting with a hypoglycaemic episode is more challenging and numerous causes must be excluded. In all cases, patient education is vital for prevention and to ensure early recognition and intervention of hypoglycaemia.
Footnotes
Funding
This article received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest statement
None declared.