OVERVIEW: What every practitioner needs to know

Drug-induced hyperglycemia is a more common complication of various treatment regimens than many physicians may realize.

– Glucocorticoid (GC) or steroid-induced hyperglycemia has been well described in rheumatology, oncology and transplant literature for several decades. GC-induced hyperglycemia is the most common form of drug-induced hyperglycemia. GC treatment routes include oral, IV, intranasal, inhaled, topical, and localized injections. While all forms of GC administration carry the risk of diabetes, the oral and IV routes are by far the most commonly associated with hyperglycemia and will thus be the only treatment routes described in this chapter. GCs may be the most diabetogenic of all drugs known to induce hyperglycemia.

– L-Asparaginase induced hyperglycemia has also been well known to complicate acute lymphocytic leukemia (ALL) induction therapy, particularly in children.

Continue Reading

– Tacrolimusand cyclosporine (calcineurin inhibitors – CNI) are also diabetogenic. In transplant patients, tacrolimus is more likely to result in hyperglycemia, especially in pediatric transplant recipients and African Americans, and is estimated to be 5 times as diabetogenic as cyclosporine.

– In both transplant and oncology patients, GCs, in addition to either L-asparaginase or tacrolimus, has resulted in significantly higher rates of hyperglycemia than seen with any of the agents alone. Because diabetes in transplant patients is unlikely to be due to one factor, the condition has been given a new name, NODAT (New Onset Diabetes After Transplant).

– More recently, hyperglycemia has been shown to be associated with
a typical or second generation antipsychotic (SGA) use, particularly in conjunction with weight gain. The risk of hyperglycemia varies among the different SGAs available on the market. Olanzapine (Zyprexa®) has the highest risk, and aripiprazole (Abilify
®) has the lowest risk. Clozapine (Clozaril®) also carries a high risk, and ziprasidone (Geodon®) carries a low risk. Risperidone (Risperdal
®) and quietapine (Seroquel®) appear to have an intermediate risk, although the incidence of diabetes varies in the literature. Discussion of each specific SGA is beyond the scope of this chapter and thus, unless specifically stated, SGAs will be discussed as a group.

This chapter will focus primarily on GC-induced hyperglycemia. L-Asparaginase-, CNI- and SGA-induced hyperglycemia will also be discussed.

Are you sure your patient has drug-induced hyperglycemia? What are the typical findings for this disease?

The classic symptoms are the same as those in any other type of diabetes.

1. Polyuria, Polydipsia and Nocturia.

2. Weight loss or failure to gain weight.

3. Polyphagia.

4. Dehydration.

High random blood glucose of >200 mg/dL or detection of glycosuria only.

– In the inpatient setting, random hyperglycemia is often discovered when glucose levels are measured in conjunction with routine labs and the classic symptoms above are not noted.

– Random hyperglycemia may be the only presentation in the outpatient setting as well.

– Screening protocols have become a part of the treatment protocols in the inpatient setting. In many pediatric oncology protocols, especially for ALL, urine is routinely screened for glucose and ketones. Transplant patients are also screened, although the exact protocol varies.

Rarely, drug-induced hyperglycemia can present with either DKA or HHS (decompensated diabetes).

In patients with ALL, increased risk of infections is seen with hyperglycemia, and infection can be an initial presentation of hyperglycemia.

In patients who are post-transplant, especially kidney transplant recipients, the success of the graft can be compromised if hyperglycemia is present; thus, graft failure can be an initial presenting sign.

In both cases, screening enables detection of hyperglycemia before serious complications occur, such as graft failure or infection.

What other disease/condition shares some of these symptoms?

Type 1 diabetes

– Patients may be at risk for autoimmune disease, such as type 1 diabetes, which can make the distinction between type 1 diabetes and drug-induced hyperglycemia difficult. Examples of such patients include those receiving GC for rheumatologic diseases and patients with Down syndrome diagnosed with ALL.

Type 2 diabetes

– Obese patients taking either GC or atypical antipsychotics may present with an insulin resistance picture that mimics type 2 diabetes.

Stress hyperglycemia

– Patients who have elevated blood glucose levels detected during times of illness but do not have diabetes. The likely mechanism is stress hormones (such as cortisol, catecholamines, and growth hormone) driving increased glucose production and insulin resistance.


– Patients with CF who are post transplant or receiving GCs can mistakenly be diagnosed with CFRD.

What caused this disease to develop at this time?

  • Dose and duration of drug exposure. These factors clearly affect the development of steroid-induced hyperglycemia and likely atypical antipsychotic-induced hyperglycemia. While many cases of L-asparaginase-induced hyperglycemia occur early (even after the first dose), most episodes occur during the induction protocol, when the dose of L-asparaginase is highest.

  • Multiple drugs causing hyperglycemia used together.The addition of GCs to other drugs known to be diabetogenic significantly increases the likelihood of diabetes.

  • Underlying disease and intercurrent illness. In some cases, particularly in oncology or transplant patients, the exact cause of hyperglycemia may not be known. Some authors have suggested that leukemia alone increases the likelihood of diabetes. Critical illness or sepsis also increase the risk. In both leukemia and sepsis, cytokines may play a role in causing hyperglycemia. L-asparaginase, tacrolimus, and GCs each can cause pancreatitis, which can cause secondary hyperglycemia. Uremia in renal failure can cause insulin resistance, and cirrhosis in hepatic failure can result in an abnormal OGGT in pre-transplant evaluation. Resolution is expected with a successful transplant, but may not if hyperglycemia causes graft dysfunction.

  • Combination of risk factors. The risk of diabetes increases sharply when more than one risk factor is present, and appears to be proportionate to the number of risks present.

What laboratory studies should you request to help confirm the diagnosis? How should you interpret the results?

  • A serum glucose is mandatory. A high glucose obtained by glucometer or D-stick should be confirmed on a venous sample. A random glucose over 200 mg/dL associated with classic symptoms, glucose > 200 mg/dL 2 hours after a glucose load as part of a formal OGTT, and a fasting glucose over 125 mg/dL are confirmatory for diabetes per the ADA (American Diabetes Association) and World Health Organization (WHO) guidelines. A confirmatory laboratory test based on measurements of venous blood glucose must be done on another day in the absence of unequivocal hyperglycemia accompanied by acute metabolic decompensation. A random glucose value of >200 mg/dL (confirmed on repeat testing) without symptoms in the context of drug treatment can be considered hyperglycemia.

  • Urine for ketones. In patients at risk for beta cell failure, urine should be tested for ketones. The presence of ketones implies insulin deficiency and increases the risk of diabetic ketoacidosis (DKA).

  • Serum electrolytes with bicarbonate and/or pH levels should be drawn if the urine is strongly positive for ketones and DKA is suspected. DKA is diagnosed when ketones are present, the bicarbonate level is low (under 20 mEq/L in oncology literature or under 15 mEq/L in endocrine literature) with hyperglycemia (serum glucose > 200 mg/dL).

  • Serum osmolarity. In patients suspected of hyperglycemic hyperosmolar syndrome (HHS), a serum osmolarity should be measured directly or calculated. Many hyperglycemic patients have some degree of hyperosmolarity, but patients with HHS have significantly elevated levels (> 320 mOsm/L) with clinical evidence of significant dehydration and possible mental status changes. The diagnosis of HHS should also prompt a urine test for myoglobulin and creatine kinase (CK) levels for evaluation of rhabdomyolysis. The calculated total body water deficit can be as high as 20%-25%.

  • A hemoglobin A1c (HbA1c) of > 6.5% is now also diagnostic of diabetes per recent ADA guidelines. A normal HbA1c, however, is not reassuring as the duration of hyperglycemia may not have been sufficient to raise the level to the diabetic threshold.

  • A lipid panel may be helpful in patients receiving either GC or atypical antipsychotics, and may reveal dyslipidemia as a complication of insulin resistance (high triglyceride level and/or low HDL level).

  • A diabetes auto-immune panel may be drawn in cases where type 1 diabetes is also suspected. Antibodies against beta cell antigens are present as markers of disease. Anti-glutamic acid decarboxylase (GAD) 65 and anti-islet cell antibodies are most specific and sensitive for the diagnosis of type 1 diabetes. Anti-insulin antibodies are less specific but are also included in the panel.

Would imaging studies be helpful? If so, which ones?

  • No imaging studies are necessary.

Confirming the diagnosis

There are no formal decision algorithms cited in the literature specifically addressing the evaluation and confirmation of drug-induced hyperglycemia in children. The following suggestions are based on literature and the author’s best clinical reasoning:

Confirm hyperglycemia (see above).

Evaluate the possibility of DKA or HHS.

Determine likely offending agent. A review of current AND past medications is necessary.

Exclude the possibility of other causes of hyperglycemia, particularly in those patients at risk for type 1 and type 2 diabetes and other less common types of diabetes.

Ideally, patients and families should be counseled about the risk of hyperglycemia before treatment has been initiated.

In 2003, international consensus guidelines on NODAT were published. The panel adopted the same criteria for NODAT as the ADA and WHO use for the diagnosis of diabetes. No specific recommendations were made as to the diagnosis of NODAT in children. Transplantation 2003;75(10):SS3–SS24

In 2004, a consensus development conference was held to address the issue of diabetes and obesity in patients treated with SGAs. No specific definition of diabetes was proposed; however, recommendations were made about screening for diabetes in patients treated with an SGA. Baseline BMI, family history, fasting glucose, BP and lipid profiles are suggested, and continued frequent monitoring of weight is recommended. A fasting glucose level should be measured every 3 months and then annually. In addition, they suggested that health care professionals, patients, families, and caregivers should be informed about the signs and symptoms of diabetes, especially those suggestive of metabolic decompensation such as DKA. There were no recommendations specific to the use of an SGA in children. Diabetes Care 2004;27(2):596-601

If you are able to confirm that the patient has drug-induced hyperglycemia, what treatment should be initiated?

Significant metabolic derangements (DKA or HHS):

  • Treatment should be initiated urgently and is identical to the treatment of DKA or HHS in children with either type 1 or type 2 diabetes.

Inpatient insulin treatment:

  • Inpatients with persistent hyperglycemia, even without DKA or HHS, may require insulin treatment.

  • The type and dose of insulin should be determined based on the patient’s clinical status, blood sugars, and age. The various types of insulin available are the same as those used in patients with type 1 diabetes.

  • Attention should be paid to the timing of the peak glucose with an aim to choose an insulin best suited to minimize glucose excursions while avoiding hypoglycemia and minimizing the number of daily injections.

  • GC given IV or PO often results in hyperglycemia 4-8 hours after dosage administration, which can be treated with an insulin whose peak of action also occurs 4-8 hours after administration, such as regular or neutral protamine Hagedorn (NPH) insulin.

  • Postprandial glucose excursions can be best managed with an ultra-rapid insulin such as lispro insulin (Humalog®) or insulin aspart (Novolog®).

  • Basal insulin may not be required, but if needed, can be administered as a once daily injection of a long-acting insulin such as glargine (Lantus®), a twice daily injection of an intermediate-acting insulin such as NPH or detemir (Levemir®), or Regular insulin can be added to an IV bag of fluids or TPN.

  • In most patients, an initial starting dose of 0.25-0.5 units per kg per day is reasonable. Subsequent doses of insulin are determined by blood sugars.

  • The frequency of monitoring of blood glucose levels depends on the type of insulin therapy chosen.

  • The glycemic goals will differ based on the underlying disease (see below).

  • Urine should be regularly monitored for ketonuria if blood glucose remains significantly elevated (above 250-300 mg/dL), with evidence of infection or if there is an overall clinical deterioration.

Management of hyperglycemia in oncology patients:

The goal of therapy is focused on avoiding extreme hyperglycemia to prevent development of DKA or HHA, and to stop the continued loss of fluids and calories in the urine.

Target blood glucose levels should be under 180 mg/dL, the renal threshold of glycosuria, but above the hypoglycemic cut-off of 80 mg/dL.

Strict glucose control can be dangerous, because insulin requirements vary greatly depending on the timing of steroid and L-asparaginase administration.

Hyperglycemia resolves spontaneously in most cases. Insulin doses should thus be decreased rapidly in response to lowered glucose levels.

In some cases, GC or L-asparaginase doses can be decreased, but this decision should be made in consultation with the treating oncology team.

Dietary modification may provide additional therapeutic benefits, but precautions should be taken with patients receiving chemotherapy because adequate protein and calorie intake are necessary to prevent the lean body mass loss associated with chemotherapy.

Dehydration may frequently accompany persistent hyperglycemia and should be managed with IVF.

For patients with persistent hyperglycemia managed as outpatients, exercise and dietary modifications may be sufficient. Insulin therapy is added if these measures fail to adequately control hyperglycemia. The choice and dose of insulin, and frequency of blood glucose monitoring are based on the same considerations as in the inpatient setting.

HbA1c levels are not usually necessary because the goals of therapy are different than for patients with other types of diabetes. If levels are measured, consideration of recent blood transfusions is necessary during interpretation of results.

Patients receiving insulin as outpatients must be counseled about the signs, symptoms and treatment of hypoglycemia. A prescription and instruction on the emergency use of glucagon for severe hypoglycemia should be included.

Patients and families should also be taught to check urine for ketonuria under the conditions described above.

Management of NODAT (New Onset Diabetes After Transplantation):

  • The period after reperfusion post transplant is critical for graft survival. A normal metabolic milieu promotes graft performance and decreases the incidence of peri-operative infections.

  • Tight glycemic control may be necessary.

  • The dose of GC should be reduced as early as possible post transplant. A reduction in calcineurin inhibitor (CNI) dosage or a switch from tacrolimus to cyclosporine should also be considered. An individualized titration of agents and close blood monitoring to ensure optimum immunosuppression is essential. Patients should also be monitored carefully for clinical evidence of rejection.

  • Acute hyperglycemia (> 250 mg/dL) should be managed with an insulin infusion to restore normoglycemia as quickly as possible.

  • In adults, a treat-to-target/step-wise approach is recommended for persistent diabetes. A similar approach is reasonable in children although the glycemic targets may differ because of the increased risks associated with hypoglycemia in children. (See section on type 1 diabetes treatment goals).

  • In adults, non-pharmacologic interventions such as lifestyle modification are the first step and can be done safely in children.

  • The age of the patient may determine the choice of pharmacologic agents used in subsequent steps. The precise intervention to achieve optimal glycemic control will vary with each individual patient.

  • Insulin is the only treatment for diabetes that is FDA approved for all ages.

  • The choice of insulin, dose of insulin, and frequency of blood glucose monitoring are based on the same considerations as in the inpatient setting.

  • With the initiation of insulin therapy, patients and families should be educated about hypoglycemia and ketonuria as described above.

  • Self-blood glucose monitoring should performed with age-based glycemic targets similar to that used in children with type 1 diabetes. If treatment is non-pharmacologic or is limited to metformin monotherapy, a measurement 2 hours after the largest carbohydrate-containing meal should be sufficient. Additional fasting measures may be indicated depending on dosage schedule of GC or CNI. Once insulin treatment is initiated, the monitoring schedule is determined by the specific type of insulin used.

  • HbA1c levels should be monitored but should not be drawn for at least 3 months post transplant because of the high likelihood of transfusion in the peri-operative period. HbA1c goals should be age-based as per ADA recommendations of treatment of diabetes in children (link to section on type 1 diabetes treatment goals).

GC-induced diabetes:

  • Decrease the dose of GC. Even small decrements can produce significant reduction in blood glucose levels enabling decreased doses of insulin or even cessation of treatment.

  • If mild hyperglycemia in the outpatient setting persists despite GC dose adjustment and the patient is older than 10 years of age, one could consider treatment with metformin using the same doses as used in children with type 2 diabetes. A slow upward titration of dose will minimize gastrointestinal side effects and increase compliance. More traditional therapy includes low-dose insulin titrated carefully to manage blood sugars.

  • Insulin therapy is necessary if hyperglycemia persists in patients under 10 years despite GC dose adjustments or in older children, if significant hyperglycemia exists despite maximal treatment with metformin or if metformin is not tolerated. GC-induced diabetes may require initially higher insulin doses to overcome the GC-induced insulin resistance and glucose toxicity. Otherwise, the choice of insulin and frequency of blood glucose monitoring are based on the same considerations as in the inpatient setting.

  • With initiation of insulin therapy, patients and families should be educated about hypoglycemia and ketonuria as described above.

SGA-induced hyperglycemia:

  • A switch to a less diabetogenic SGA should first be considered.

  • Outpatient treatment of patients with either AP-induced hyperglycemia should focus on treating the insulin resistance and the hyperglycemia.

  • Lifestyle and behavioral modification should be initiated with the diagnosis of diabetes.

  • The use of an insulin sensitizer such as metformin has improved IR and decreased BG levels. The initial dose of metformin is the same as in children with type 2 diabetes. A slow upward titration of the metformin dose will minimize gastrointestinal side effects and increase compliance.

  • Insulin therapy may be needed if hyperglycemia persists. SGA-induced hyperglycemia may also require higher doses of insulin to overcome insulin resistance. Otherwise, the choice of insulin and frequency of blood glucose monitoring are based on the same considerations as in the inpatient setting.

  • With initiation of insulin therapy, patients and families should be educated about hypoglycemia and ketonuria as described above.

What are the adverse effects associated with each treatment option?


– Hypoglycemia.


– Diarrhea and abdominal pain, which are minimized by titrating the dose slowly.

– Lactic acidosis very rarely in patients with renal or hepatic failure.

Lifestyle and behavioral modifications have no known adverse effects, but high rates of non-adherence.

What are the possible outcomes of drug-induced hyperglycemia?

***The natural history of drug-induced hyperglycemia in children is largely unknown. In many cases, the hyperglycemia resolves with discontinuation or even a dose decrease of the offending medication. In many studies, the likelihood of persistent rather than transient hyperglycemia was seen to increase if the patient had more than one risk factor, such as for the development of drug-induced hyperglycemia.

– In children receiving induction therapy for ALL, approximately 25% of patients experienced a spontaneous resolution of hyperglycemia without pharmacologic intervention. Insulin is required in 2/3 of patients, but the average duration of therapy is only about 10 days.

– In adults with NODAT after solid organ transplantation, the likelihood of diabetes increases with time, although many patients do not require pharmacologic intervention. The specific rates of transient versus long-term diabetes are not available in children with NODAT. While NODAT is multifactorial, GC use is responsible for 74% of the variation in rates. Decreased GC doses or transition to steroid-free immunosuppressive regimens significantly and positively impact the natural history of NODAT, although the rates of resolution of diabetes related to decreased steroid use are not available.

– Hyperglycemia from SGA use is a more recently recognized phenomenon; thus, information about the natural history is limited.

– In hyperglycemic adults treated with GC for indications that were not related to cancer or transplantation, up to half required pharmacological intervention. In most cases, however, the diabetes eventually resolved but took months to years to do so in some patients. The resolution rates after discontinuation of GC therapy are not available. Very little information is available on children in this diagnostic criteria as GC-induced hyperglycemia not associated with cancer or transplantation is rarely studied.

***The risks and benefits of the above-described treatment options vary, based on the nature of the underlying disease, exact drug that caused the hyperglycemia, and the specific treatment regimen.

– In children receiving insulin for drug-induced hyperglycemia, hypoglycemia has been reported in up to 25% of children. However, insulin can both treat and prevent metabolic decompensation and is the cornerstone of inpatient therapy. Insulin regimens can be individualized to the patient to minimize the risk of hypoglycemia and trauma related to insulin injections or finger-stick measurements of blood glucose.

– In transplant patients, hyperglycemia poses threats to both the patient and the graft, and should thus be treated aggressively. Changes in the immunosuppressive schedule, doses, or agents should be balanced against increasing the risk of rejection. The treatment plan should thus be individualized and developed in conjunction with the treating team.

– Similarly, changes in the treatment protocols for ALL and other childhood cancers may decrease the likelihood of remission, but mitigate the detrimental effects of hyperglycemia on survival.

– Metformin is largely well tolerated but can have significant gastrointestinal side effects, particularly if the dose is not titrated up slowly. Caution should also be applied when using metformin for the treatment of NODAT, particularly after kidney transplantation, as renal insufficiency increases the likelihood of lactic acidosis. Lactic acidosis due to metformin, however, remains very rare.

– Discontinuation of an SGA can acutely exacerbate the underlying psychiatric illness, but switching from a more diabetogenic to a less diabetogenic SGA has not had a negative impact.

What causes this disease and how frequent is it?

Incidence of drug-induced hyperglycemia:

  • The incidence of GC-induced hyperglycemia in the pediatric outpatient setting is unknown, but frank diabetes is thought to occur in less than 1%. GC-induced diabetes has been described in 15% of adult patients receiving > 20 mg of prednisone equivalent a day. In the adult rheumatological literature, rates as high as 54% have been reported, but the definition of hyperglycemia has varied.

  • Treatment with either a GC or a CNI alone causes diabetes in 1% of patients. Synergism between the two agents results in diabetes occurring in 10%-35% of children during ALL induction treatment. Rates as high as 50+% have been reported in the literature, but combine those children who had glucose levels in the diabetic range (“overt hyperglycemia”) with those in the impaired fasting glucose range (“mild hyperglycemia”).

  • The use of dexamethasone in the induction, rather than prednisone, doubled the likelihood of diabetes. The relative potency of dexamethasone compared with prednisone is not exactly known, but is thought to be 10-15 times greater. The differences in rates of diabetes between the two GC formulations likely reflects a difference in potency rather than factors specific to the formulation.

  • The risk of DKA in ALL treatment was greater if the patients were older than 10 years at diagnosis of ALL, were earlier in the treatment protocol (remission induction versus maintenance therapy), and had previous drug-induced hyperglycemia.

  • The rates of NODAT are reported, but no distinction is made as to the relative contributions of each drug and of non-pharmacologic factors. The specific immunosuppressive regimen explains 74% of the variability in rates of NODAT; thus, drug-induced hyperglycemia is the most specific factor in the development of NODAT.

  • NODAT is diagnosed in 20% of pediatric kidney transplant patients.

  • In adult kidney transplants, diabetes was seen in 24% of patients over 10-year follow-up. The rates of NODAT in adults after any solid organ transplant are 9% at 3 months, 16% at 6 months and 24% at 36 months (from Medicare beneficiary data).

  • Blood glucose values in the impaired or frankly diabetic range were found in 90% of pediatric heart transplant patients upon entry to the ICU, and resolved without treatment in most within 72 hours. A quarter of patients had blood glucose levels > 180 mg/dL, and 10%-30% of these patients received peri-operative steroids versus 4%-18% of non-hyperglycemic patients.

Mechanism of hyperglycemia:

  • Insulin resistance (IR). IR Is defined as the impaired ability of plasma insulin at usual concentrations to adequately promote peripheral glucose disposal, suppress hepatic glucose, and metabolize triglycerides. The exact pathophysiology of IR is not known, but obesity, particularly visceral and myocellular adiposity (versus subcutaneous), is critical to the development of IR. Secretion of resistin and other adipokines, particularly by visceral adipocytes, plays a role, as does inhibition of adiponectin secretion from subcutaneous adipoctyes.

  • Pancreatic beta cell failure.

Other risk factors for the development of drug-induced diabetes:

  • Obesity is a risk factor in drug-induced hyperglycemia and likely exerts its effects through insulin resistance, although other mechanisms are possible.

  • Family history of diabetes and a personal history of previous drug-induced hyperglycemia are consistent risk factors regardless of underlying disease, the dose or duration of drug, and the presence of other risk factors.

  • Patients with Down syndrome and ALL are at increased risk of hyperglycemia during induction therapy.

How do these pathogens/genes/exposures cause the disease?

  • Insulin resistance (IR):

  • Glucocorticoids cause hyperglycemia by increasing gluconeogenesis as a result of decreased hepatic insulin sensitivity and by inhibiting peripheral utilization of glucose.

  • Atypical antipsychotics may cause diabetes because of weight gain, change in body fat distribution, or by a direct effect on insulin-sensitive target tissues.

  • Pancreatic beta cell failure:

  • L-asparaginase decreases insulin synthesis by depletion of asparagine and decreasing insulin secretion from beta cells. Insulin deficiency results in hyperglucagonemia, and the drug may also impair insulin receptor function.

  • CNIs also exert their diabetogenic effects largely through pancreatic beta cell toxicity, decreased insulin synthesis, and decreased insulin secretion. Histological reports show cytoplasmic swelling, vacuolization, apoptosis, and abnormal immunostaining of islet cells in patients treated with CNI therapy. Islet cell damage is more profound in tacrolimus-treated patients versus cyclosporine-treated patients.

  • Combination of mechanisms:

  • In post-transplant and oncology patients, the effects of drugs causing beta cell failure are magnified by concomitant use of GCs. GCs result in increased insulin requirements, for which the L-asparaginase or CNI-damaged beta cell is unable to compensate.

Other clinical manifestations that might help with diagnosis and management

GC and CNI inclusion in post-transplant immunosuppressive regimens largely explains the development of NODAT. However, sirolimus, a mammalian target of rapamycin (mTOR) inhibitor used in immunosuppressive protocols, may also be diabetogenic.

What complications might you expect from the disease or treatment of the disease?

Children who were overtly hyperglycemic during induction treatment for ALL had poorer risk-free survival and overall survival at 5 years. In addition, these children had a significantly increased risk of death independent of risk group and type of steroid. Overt hyperglycemia may be an independent predictor of survival in children with ALL.

Hyperglycemia may be particularly detrimental to transplant patients. In the adult solid organ transplantation literature and in pediatric heart and kidney transplant experiences, hyperglycemia increases morbidity and mortality. In kidney and other solid organs, hyperglycemia predisposes to graft dysfunction, intercurrent infection, and death. In pediatric heart transplant patients, hyperglycemia increased the likelihood of longer intensive care/hospital stays, lactic acidosis, cardiac arrest, and longer ventilation time.

Are additional laboratory studies available; even some that are not widely available?

Monitoring of complications and cardiac risk factors. In patients with long-term diabetes, monitoring for both microvascular and macrovascular complications and an evaluation of cardiac risk factors are necessary. Blood pressure should be measured at least quarterly, and lipid profiles should be measured within the first year of diagnosis. A routing screening schedule similar to that recommended by the ADA for children with type 1 diabetes is useful (link to section on type 1 diabetes). Caution should be applied when interpreting microalbumin results in post-renal transplant patients.

How can steroid/drug-induced hyperglycemia be prevented?

Glucocorticoids: Promotion of a healthy lifestyle with regular exercise may moderate the diabetogenic risk of GCs. Efforts should be made to use the lowest possible dose of GC and shortest duration of therapy.

Chemotherapy: There may not be any prevention of L-asparaginase-induced beta cell failure, but careful monitoring, especially if GCs are also part of the treatment protocol, may prevent the progression to DKA or HHA. Minimizing dextrose loads in IVF may moderate glucose excursions and can result in lowered blood sugar levels and lowered insulin requirements. In ALL, some studies have shown that changing the schedule of GC and L-asparaginase so that they are not given concurrently may decrease the risk of diabetes while still maintaining excellent cure rates.

Transplantation: As in L-asparaginase, careful monitoring may prevent acute complications such as DKA or HHS. Selection of steroid-free immunosuppression protocols and those minimizing the frequency and dose of diabetogenic drugs may greatly reduce the likelihood of NODAT.

Atypical antipsychotics: Lifestyle modifications, particularly in conjunction with behavioral therapy, reduce the risk of weight gain, insulin resistance, and frank diabetes. Each atypical antipsychotic drug confers different risk for developing diabetes. Switching to a different medication within the same class with a lower diabetogenic profile also can both prevent and treat diabetes.

– In all cases, maintenance of a healthy weight reduces the risk of diabetes. Care should be taken with patients receiving chemotherapy as described above.

What is the evidence?

The treatment of NODAT is based on the 2003 guidelines, although strength of evidence for each recommendation was not clarified. There is no data from randomized controlled trials comparing various oral agents for the treatment of non-decompensated NODAT, although one adult trial is ongoing.

No randomized controlled studies have been done to determine the ideal treatment regimen or optimum glycemic target in hyperglycemic pediatric oncology patients. Treatment protocols have been suggested based on experience and cohort studies.

Cohort studies and randomized controlled studies have supported the use of behavioral therapy and metformin for the treatment of obesity and insulin resistance in patients, including children, treated with an SGA but similar data in SGA-induced diabetes is lacking. Switching from a high-risk to a low-risk SGA has proven to treat weight gain, but the effect of this intervention on diabetes is less clear. Because diabetes from SGA use is thought in part to be mediated by obesity, switching was recommended for the treatment of diabetes in the 2004 consensus guidelines.

Ongoing controversies regarding etiology, diagnosis, treatment

-Blood glucose targets in critically ill patients.Target blood glucose levels have been debated even in the pediatric literature after associations between mortality and hyperglycemia were noted in adult critical care (surgical) patients in 2001. The results of the NICE-SUGAR study in the New England Journal of Medicine in 2009 showed more definitively that the practice of intensive insulin treatment with tight glycemic control did not confer any benefits in mortality and was associated with significant risk. Several guidelines have since been published but do not specifically address the pediatric patient population nor oncology or transplant patients.

-Route of GC administration and risk of diabetes. Glucocorticoids given by other routes of administration such as inhalation, transdermally, intra-articular or by nasal spray, have been posited to cause diabetes as well, but the evidence is contradictory and unconvincing.

-Metformin for the prevention of SGA-induced diabetes. The pathophysiology behind the development of SGA-associated diabetes is unclear, but because insulin resistance appears to play a role, the use of metformin has been advocated by some authors in the psychiatry literature as prophylaxis. Some authors have recommended that metformin be prescribed in conjunction with the SGA before clinical or biochemical evidence of IR exists. Well-designed randomized controlled studies in pediatrics are limited and have not consistently shown that this practice is beneficial.