Feline-diabetes Mellitus-evaluation Of Treatment

Vol. 22, No. 5 May 2000

CE

Refereed Peer Review

FOCAL POINT ★Assessment of diabetic patients is a multifactorial, continuous process that requires evaluation of all available data to ensure maintenance of adequate glycemic control.

Feline Diabetes Mellitus: Evaluation of Treatment * Auburn University

Ellen N. Behrend, VMD, MS Colorado State University

Deborah S. Greco, DVM, PhD

KEY FACTS ■ Serial blood glucose monitoring is the gold standard for assessing glycemic control but can be difficult to perform in cats. ■ When adjusting insulin dose based on a blood glucose curve, both the lowest measured concentration of glucose and the duration of action must be considered. ■ Hypoglycemia should be treated aggressively; intravenous glucose infusion may be required for up to 4 days. ■ If insulin resistance is suspected, problems with owner administration of insulin must be ruled out first. ■ If insulin resistance is truly present, possible causes should be assessed in a logical order.

ABSTRACT: Accurate assessment of diabetic patients requires cooperation between owners and veterinarians as well as evaluation of all data, including resolution of clinical signs, serial blood glucose levels, and possibly serum glycosylated protein concentrations. In evaluating serial blood glucose curves, both the duration of action and nadir of glucose concentration must be assessed and changes in insulin dose or frequency made accordingly. Glycosylated protein concentrations may help determine the adequacy of ongoing control. Development of hypoglycemia or hyperglycemia in a previously well-controlled patient suggests a change in either endogenous or exogenous factors, and a cause should be determined. Hypoglycemia should be avoided and treated aggressively if it does occur. If insulin resistance is present, the cause should be delineated and removed if possible.

T

reating and monitoring diabetic cats require a partnership between the owners and attending veterinarian. The importance of home monitoring cannot be overemphasized—owner observations (e.g., resolution of such clinical signs as polyuria/polydipsia, urine glucose determinations) are a necessary adjunct to the assessment of diabetic cats. However, more sophisticated monitoring techniques may be necessary to determine the cause of poor glucose regulation in a previously well-controlled diabetic cat or a cat undergoing initial insulin therapy. Monitoring techniques can be divided into two broad categories: those that determine the dose and duration of insulin for initial diabetic regulation (i.e., glucose curves) and those that assess adequacy of long-term diabetic regulation for the prevention of such complications as diabetic polyneuropathy (i.e., measurement of glycosylated proteins).

MONITORING THERAPY Home Monitoring Clients should be instructed to monitor the effects of insulin and gross regula*Copyright 1998 by the Western Veterinary Conference from the booklet “Clinical Pharmacology, Principles and Practice.” Modified with permission.

Compendium May 2000

Small Animal/Exotics

tion of hyperglycemia by keeping daily stress of hospitalization thus avoided. The Blood Glucose records of changes in appetite, attitude, results can then be reviewed by a veteriCurve Protocol body condition, polydipsia, polyuria, narian and appropriate changes made. and urine glucose and ketone levels.1 If Water consumption can be quantitated to ■ Collect an initial, water intake, urine output, and appetite give a more accurate assessment of subtle pre–insulin injection are normal and body weight is constant, changes in diabetic control. Food intake blood sample. a diabetic patient is usually well concan be similarly measured. The size and ■ Feed the same amount trolled.2 This information is extremely number of urine clumps in the litterbox and type of food as important when assessing blood glucose can be recorded to assess polyuria. concentrations. For example, stress durusual. Serial Blood Glucose Curve ing blood collection can falsely elevate ■ Have the owner The serial blood glucose curve is the serum glucose concentration, implying administer insulin after ideal and time-honored method of assessinadequate control; thus knowledge of a the patient has eaten, ing glycemic control in diabetic patients. cat’s home behavior must be used to and assess the owner’s However, serial blood glucose curves may evaluate laboratory data. technique. be difficult to interpret in some cats beOlder articles and texts recommend cause of the effects of stress on the blood that owners evaluate urine glucose levels ■ Take blood samples at glucose concentration. In addition, apand provide charts and tables for owners 2-hour intervals for 12 petite may be inconsistent in hospitalized to use to adjust insulin doses according to 24 hours. feline patients. to urine concentration. This approach is Ideally, the serial blood glucose curve no longer recommended 2,3 because urine glucose represents only an average should be used to evaluate feline diabetic value over the previous hours; it is not an accurate picpatients (see Blood Glucose Curve Protocol). In this ture of what happened from hour to hour. For example, procedure, blood glucose is measured every 2 hours for a patient could have a very low blood glucose concenone interval between insulin administrations (i.e., for tration followed by very high values, but urine collected 12 hours if insulin is administered twice daily; for 24 at the end of this period would merely show a large hours if insulin is given once daily). Normal insulin amount of urinary glucose. and feeding schedules must be maintained as much as If urine can be collected, however, measurement of possible. A diabetic cat should be fed its standard diet urine glucose concentration at home can play a role. and given its insulin at the usual time. Urine glucose levels can be determined as needed to aid Obtaining a fasting blood sample for measuring in the assessment of glycemic control, especially when serum glucose concentration before insulin is adminisother data are conflicting. Consistently negative readtered can aid in the appraisal of glycemic control; this ings on urine glucose may indicate that insulin doses may not be possible, however, if a patient’s normal are adequate or excessive. A serial glucose curve can diffeeding time occurs outside clinic hours or if the cat has ferentiate between adequate insulin therapy and the use a poor appetite on the morning of the assessment. Furof excessive doses that could result in hypoglycemic thermore, feeding an animal at home may ensure that it shock. Uniformly high urine glucose readings coupled eats. If a patient is fed at home, the owner should adwith unresolved clinical signs indicate a potentially inminister the insulin as normal, either at home or in the adequate insulin dose.1 Negative urine glucose concenhospital under supervision of a technician or veterinaritrations in the afternoon followed by high levels (4+) an2; the latter scenario is particularly advisable if owner technique is questionable or needs to be assessed. Cothe next morning may indicate Somogyi phenomenon operation between clients and veterinarians is needed to (i.e., hypoglycemia-induced hyperglycemia); however, maximize information obtained while minimizing disdocumentation requires serial blood glucose monitorturbance to routine. If given the choice between obing. In addition, urine glucose concentrations can be taining a fasting blood glucose sample or assessing owndetermined regularly (at least weekly) to help assess oner injection technique, the latter is preferable. going control. Changes in urine glucose levels may alert Blood glucose concentrations can be determined in a owners and clinicians to the loss of glycemic control laboratory, on an in-house chemistry machine, or by and need for reevaluation. use of glucose reagent strips with or without a glucomeOther methods of home monitoring can also be ter. Glucometers are easy and relatively inexpensive to used. With proper instruction, some owners may be use, but accuracy needs to be considered. Blood glucose able to perform blood glucose curves at home. Blood concentrations as determined by a glucometer differ can be obtained from ear veins relatively easily and the CLIENT RECORDS ■ URINE GLUCOSE LEVELS ■ HOME MONITORING ■ FASTING BLOOD SAMPLE

Small Animal/Exotics

Compendium May 2000

Blood Glucose (mg/dl)

from values obtained in a labois to be fed at home, a decision ratory using bench-top methmust be made about which is odologies. A recent study4 evalmore important. 300 uated the accuracy of several The three-syringe sample glucometers (iSTAT analyzer, collection technique miniGlucose 200 Heska Sensor Devices, Waumizes blood loss and maxidifferential Nadir sheka, WI; Glucometer Enmizes accuracy. Using 3-ml 100 core, Bayer, Elkhart, IN; Exsyringes, the first syringe acTech R.S.G., MediSense, should contain 0.5 ml hepWaltham, MA; Glucometer arinized saline and is used to 2 4 6 8 10 12 Elite, Bayer; Accu-Chek Easy, collect 2.5 ml of the patient’s Time (hour) Boehringer Manheim, Indiblood; saline and blood anapolis, IN; and Glucometer should be mixed gently. This Figure 1—Ideal blood glucose curve for twice-daily insulin 3, Bayer) in assessing blood administration at 0 and 12 hours. blood is not used for glucose glucose levels in dogs; iSTAT, measurement because part or AccuCheck Easy, and Gluall of it may have been sitting cometer Encore were most reliable. At low blood glucose in the catheter and thus may not accurately reflect concentrations (below 60 mg/dl), glucometers tend to unserum glucose concentrations; instead, it is returned to derestimate the measured concentration. This added safety the patient after the second syringe is used to collect feature alerts human diabetic patients to the possibility of sufficient blood to measure blood glucose. To minimize hypoglycemic shock well before the onset of clinical signs. blood loss, the minimal amount needed for testing Other factors about glucometers and serial blood glushould be collected. The contents of the first syringe cose curves also need to be considered. High altitude, are then injected into the catheter, and the catheter is blood oxygen content, hematocrit, shock, dehydration, flushed with 2 ml of heparinized saline. and severe infections can affect glucometer accuracy. The third syringe remains attached to the catheter to Out-of-date or improperly stored test strips may give maintain a closed system; alternatively, the third syringe erroneous results. Certain home glucose monitoring can be replaced with an injection cap. Three syringes devices have different reagent strips for venous versus are used each time blood is collected. Nonsterile gloves capillary blood, and the correct reagent strip for the should always be worn to protect patients from nosocoglucometer should be used. If a glucose strip or glumial infections.3 If blood is to be submitted to a laboratory, it can be collected into a red-topped (plain) tube cometer is used, at least one of the blood samples coland allowed to sit for 20 minutes (to clot); serum is lected during a glucose curve should always be meathen separated and placed in another tube. Delayed sured by both the in-house technique and a laboratory separation can lower the blood glucose concentration as to determine accuracy.3 If the accuracy of a glucometer or glucose strip varies with different glucose concentraerythrocytes metabolize glucose in the sample. Samples tions, two samples—one containing a low glucose concan also be collected in sodium fluoride tubes and centration and one with a high concentration—should stored unseparated under refrigeration.3 be evaluated by a laboratory. A disadvantage of sending ADJUSTMENTS TO INSULIN DOSES blood to a laboratory is that higher volumes are needed, In an ideal glucose curve, the nadir blood glucose and thus the risk of anemia is increased. Because techconcentration should be 80 to 150 mg/dl. Glucose difnique can also affect results obtained with glucose strips ferential refers to the difference in blood glucose conor glucometers, the fewest people possible should percentrations between the nadir and immediately before form all blood glucose measurements.3 To minimize patient stress and damage to veins during the next insulin dose (Figure 1) and can measure insample collection, a small (25-gauge) needle may be used sulin effectiveness.5 If the curve is relatively flat (e.g., a glucose differential of 50 to 100 mg/dl), insulin may to collect a single drop of blood from the cephalic or menot be having the desired effect. Blood glucose concendial saphenous vein. This method prevents a patient from trations must also be considered: If blood glucose conbecoming anemic from multiple blood sampling. Altercentration is always below 200 mg/dl, insulin is very efnatively, a jugular catheter can be used for sample collecfective; blood glucose concentrations of 350 to 400 tion. The catheter should be placed the night before serial mg/dl indicate that the prescribed insulin dose is inefblood glucose measurements are to be plotted so that the fective. In assessing any glucose curve, two basic quesstress of placement does not affect the results. However, tions need to be asked: Has the insulin succeeded in because a catheter cannot be placed in advance if the cat GLUCOMETERS ■ ACCURACY ■ THREE-SYRINGE TECHNIQUE

Compendium May 2000

Small Animal/Exotics

TABLE I Initial Insulin Regimens for Diabetic Catsa Insulin Type Lente® (Eli Lilly, Indianapolis, IN) Neutral protamine Hagedorn Protamine zinc insulin ®

Ultralente (Eli Lilly)

Initial Dose (U)

Frequency

Feeding Schedule

Insulin Adjustment (U/dose)

2–3

Twice daily

Twice dailyb or free choice

0.5–1

1

Twice daily

Twice dailyb or free choice

0.5–1

1–3

Once or twice dailyc

Twice dailyb or free choice

0.5–1

1–3

c

Once or twice daily

b

Twice daily or free choice

0.5–1

aModified from Greco DS, Broussard JD, Peterson ME: Insulin therapy. Vet Clin North Am Small Anim Pract 25:684, 1995; with permission. bPreferred. cTherapy can be instituted once daily; twice-daily therapy may ultimately be required.

lowering blood glucose, and how long has the insulin lasted? By answering these questions, logical changes in the dosing regimen can be made if necessary. The first goal in regulating a diabetic patient is to achieve an acceptable blood glucose nadir. In general, if an acceptable nadir is not achieved, the insulin dose should be adjusted based on the animal’s size and degree of hyperglycemia. For cats, increases of 0.5 to 1 U per dose of insulin are appropriate (Table I).1,3 An acceptable glucose nadir may not be obtainable in some animals, however, if short-acting insulin is used.2 In these patients, the blood glucose concentration is typically high in the morning because control has been inadequate for most of the previous day. Even if an injection of short-acting insulin can lower blood glucose, its effective period is insufficient to maintain glucose within an acceptable range. In this situation, a blood glucose curve would show a noticeable but brief decrease in serum glucose concentration after the insulin injection. Hypoglycemia should always be avoided. Insulin doses should be reduced if the blood glucose nadir is below 80 mg/dl3—regardless of other blood glucose concentrations detected during the day. A decrease of 10% to 25% is usually adequate in cats if the insulin dose is at or below 6 to 8 U. If the insulin dose is above 2.2 U/kg, the initial dose should be reinstituted and the process of regulation restarted.2 Once an acceptable nadir is achieved, a blood glucose curve can be used to determine duration of action2 (roughly defined as the time from the insulin injection, through the nadir, until blood glucose concentration exceeds 200 to 250 mg/dl). If the insulin dose is inadequate and the target glucose nadir has not been achieved, the dose must be increased until the nadir is acceptable before insulin’s duration of action can be de-

termined. Duration of action and nadir cannot be assessed at the same time if either is insufficient. If duration of action is 22 to 24 hours, once-daily insulin therapy is adequate. If duration of action is 15 to 20 hours and Ultralente® (Eli Lilly, Indianapolis, IN) insulin is being used, Lente® (Eli Lilly) insulin can be given twice daily at a dose 10% to 25% smaller than that of Ultralente®. Alternatively, regular insulin can be given in the evening, preferably 16 to 18 hours after the morning dose; and the morning Ultralente® dose can remain the same or be decreased by 10%. Because regular insulin is more potent than is Ultralente®, the dose of the shorter-acting (i.e., regular) insulin is usually lower than that of the longer-acting (i.e., Ultralente®) insulin.2 If the duration of action is 10 to 14 hours, twice daily insulin or a longer-acting insulin, if available, is required. The latter is better for owners because the number of daily injections is minimized. Ultralente® or protamine zinc insulin (PZI) may be needed only once daily but is more often required twice daily. When changing from Lente® or neutral protamine Hagedorn (NPH) insulin to Ultralente® or PZI and maintaining once-daily injections, the insulin dose should not be changed initially if the glucose nadir is 100 to 180 mg/dl.2 Alternatively, the same type and dose of insulin can be administered twice daily.2,3 In assessing a patient with an apparently short duration of insulin action (8 hours or less), the blood glucose nadir must always be above 80 mg/dl. If the blood glucose drops to less than 60 mg/dl at any time, the Somogyi phenomenon can occur. The body tries to correct hypoglycemia by releasing counterregulatory hormones (e.g., epinephrine, glucagon), and blood glucose quickly increases to high levels, falsely decreasing the

BLOOD GLUCOSE NADIR ■ DURATION OF ACTION ■ SOMOGYI PHENOMENON

Small Animal/Exotics

Compendium May 2000

apparent interval of effect.2 The appropriate response is to decrease the insulin dose so that the nadir is above 80 mg/dl; when counterregulatory hormones no longer interfere with the action of exogenous insulin, the true duration of effect becomes apparent. If the duration of action is truly less than 8 hours, the insulin used must be administered more frequently than twice daily, which is an impractical situation. Accordingly, the type of insulin should be changed to a longer-acting formulation (if available) or another type or species of origin of insulin should be used. In cats, human insulin can be substituted for beef or pork insulin. Switching to a different type of intermediate-acting insulin can also be beneficial. For example, a cat may metabolize NPH insulin quickly, resulting in a duration of action that is too brief; but Lente® may have a longer duration of action. After control has been achieved, blood glucose curves should be plotted to assess adequacy of glycemic control every 3 to 6 months (more frequently if clinical signs suggest that control has been lost). The more precarious the control is, the more frequently rechecks should be done. Again, if the nadir is unacceptable, the insulin dose must be lowered or raised accordingly. If duration of action appears to have changed, modifications can be made as discussed.

GLYCOSYLATED PROTEINS Serial blood glucose curves can be affected by the stress of hospitalization, especially in cats, and by deviation from normal routine. For example, a patient’s refusal to eat while hospitalized can invalidate a glucose curve. Therefore, although such clinical parameters as owners’ reports of control of clinical signs or maintenance of body weight suggest good control, a blood glucose curve may demonstrate consistent hyperglycemia throughout the day. In these patients, the glucose curve is suspect. In a recent study investigating glipizide,6 clinical signs resolved in 6 of 50 cats but hyperglycemia persisted on serial blood glucose tests. Glycosylated hemoglobin (GHb; in the human literature, Hb A1c) steadily decreased, however, suggesting good control.6 The discrepancy between the glucose curve and the owner’s report of apparent resolution of clinical signs was likely due to the stress of hospitalization.6,7 Measurement of glycosylated proteins may be especially helpful in evaluating glycemic control in these scenarios. Glycosylated hemoglobin is formed by nonenzymatic, irreversible binding of glucose to hemoglobin.8 Fructosamine refers to glycosylated serum proteins, mainly albumin.3 Both GHb8 and fructosamine9 form at a rate proportional to the average blood glucose concentration present: The higher the mean blood glucose con-

centration is over time, the greater their concentrations should be. Levels of glycosylated proteins are also affected by the half-life of the native protein: The shorter the half-life is, the more quickly the concentration of glycosylated protein falls after correction of hyperglycemia. Albumin has a short half-life9 compared with erythrocytes.10 Thus GHb reflects glycemic control over the previous 2 to 3 months, whereas fructosamine reflects control over the previous 2 to 3 weeks. As with any laboratory test, care must be taken to use correct sample-handling techniques, and methodology for measurement and limitations of the test must be recognized. GHb is measured in whole blood collected in EDTA tubes; samples can be refrigerated for up to 1 week.2 Affinity chromatography may be preferred to inhouse automated analyzers.2 Normal GHb concentrations may be lower in cats than in dogs,6 and normal ranges vary among laboratories; thus it is necessary to ensure that the laboratory has established its own normal ranges for dogs and cats and uses those in interpreting test results. Anemia may lower GHb concentrations.11 Fructosamine samples can be stored at 4˚C for 8 days12 or at 25˚C for 5 days13 without a significant change in serum concentration. Because the effect of freezing is unclear,12,13 storage at temperatures below 0˚C should be avoided. As with GHb, ranges for serum fructosamine concentration vary among laboratories. Within the normal range, serum protein does not affect fructosamine; when serum protein concentration is below or above normal, however, fructosamine levels may decrease or increase, respectively.14 Results of one study15 suggested that fructosamine may be more sensitive than is GHb for monitoring control, but measurement of either can be helpful. Both parameters are correlated to blood glucose concentrations and typically not affected by stress.12–14,16–18 However, the value obtained from the laboratory must be interpreted in conjunction with all other data. Normal animals or well-controlled diabetic patients can have elevated concentrations of either GHb or fructosamine; conversely, uncontrolled diabetic animals can have normal levels of either.6,14,17 Fructosamine may be elevated in nondiabetic cats that are sick and hyperglycemic.14,17 Therefore, the best use for GHb or fructosamine concentrations may be as one more piece of information used to differentiate stress hyperglycemia from diabetes mellitus (DM) or, if measured at each recheck, to evaluate trends in glycemic control.

HYPOGLYCEMIA AND HYPERGLYCEMIA Hypoglycemia Hypoglycemia should be avoided at all costs because it is a life-threatening complication. Numerous causes

GLIPIZIDE ■ STRESS OF HOSPITALIZATION ■ FRUCTOSAMINE

Compendium May 2000

Causes of Hypoglycemia in Diabetic Cats ■ Use of the incorrect syringe type (e.g., use of a U40 syringe with U100 insulin) ■ Insulin overdose ■ Transient diabetes mellitus ■ Overlap of insulin action ■ Anorexia (e.g., caused by oral disease, ketosis, concurrent illness) ■ Vomiting ■ Maldigestion (e.g., exocrine pancreatic insufficiency, bacterial overgrowth) ■ Malabsorption (e.g., inflammatory bowel disease, neoplasia, lymphangiectasia) ■ Concentrated insulin (e.g., insulin improperly mixed by owner)

exist (see Causes of Hypoglycemia in Diabetic Cats). The development of hypoglycemia in a previously wellcontrolled animal suggests that a change has occurred; exogenous factors include the use of an incorrect syringe and increased insulin concentration caused by improper mixing; endogenous causes include vomiting, anorexia, maldigestion, and malabsorption. Transient DM has been documented in cats6,19–21 and may be partly related to resolution of glucose toxicity (see the companion article entitled “Treatment of Feline Diabetes Mellitus: Overview and Therapy” in this issue of Compendium). Insulin overdose can occur at any time during therapy; in a recent study of 20 overdosed cats,20 the overdose occurred 10 days to 6 years (median, 8 months) after initiation of insulin therapy.20 Overlap of insulin action may occur when insulin with a duration of action exceeding 14 hours is given in equal doses twice daily and the second injection is administered while insulin from the first injection is still present and active. Overlap of insulin action can usually be avoided by making the evening dose smaller than the morning dose. Gastrointestinal signs (e.g., vomiting, diarrhea) usually accompany maldigestion and malabsorption. Clinical signs of hypoglycemia in diabetic patients are initially consistent with the epinephrine release that occurs to counter hypoglycemia. Nervousness, anxiety, vocalization, muscle tremors, ataxia, and pupillary dilation should alert owners to the possibility of hypoglycemia. At this point, the animal should be offered food and the owner should seek veterinary advice. Late in the course of hypoglycemic shock, animals may become recumbent or comatose or have seizures. Humans INSULIN OVERDOSE ■ EPINEPHRINE RELEASE

Small Animal/Exotics

can experience hypoglycemic unawareness, a condition associated with poorly controlled DM in which mild hypoglycemia is not sensed and signs do not develop until hypoglycemia is severe. Although not documented, this condition may also occur in cats. If venous access is not readily available or an owner is administering hypoglycemia therapy in a patient that is unable to eat, 50% dextrose (corn syrup, pancake syrup) may be applied to the oral mucous membranes using a large syringe. Owners should be advised to drip the syrup on the gums (maintaining a reasonable distance from the animal’s teeth to prevent accidental injury from biting) and then to transport the animal to a veterinarian as soon as possible. At the veterinary office, treatment should consist of administration of a slow intravenous bolus of 50% dextrose (0.5 g/kg diluted 1:4). Thereafter, continuous infusion of 5% dextrose should be administered until the animal can eat. A single dose of dextrose may not adequately correct hypoglycemia if insulin is still present and active. Cats may need to be on a dextrose drip for 3 to 4 days after hypoglycemia becomes apparent.20 Endogenous glucose stores may have been depleted by the insulin overdose, and several days may be needed for hyperglycemia to re-

Interested in writing for COMPENDIUM? For small animal articles, please contact Dr. Douglass Macintire (email macindk@ vetmed.auburn.edu; phone 334-844-6032). For exotics articles, please contact Dr. Branson Ritchie (phone 706-542-6316; email [email protected]).

Small Animal/Exotics

cur. Insulin therapy should be discontinued until hyperglycemia is documented. Many animals receiving an overdose of insulin develop cerebral edema and temporary blindness or behavior changes; these signs are temporary and resolve after several weeks or months.

Compendium May 2000

waning, the timing of meals should be adjusted.22 Alternatively, addition of an oral hypoglycemic agent (e.g., acarbose) can be considered. If no response ■ Insufficient insulin dose to one type of insulin occurs, an alter■ Insufficient duration of nate should be tried. Because insulin insulin action absorption can vary among subcuta■ Out-dated, inactive neous injection sites, 22 another site Hyperglycemia (e.g., the lateral thorax, abdomen) insulin A variety of problems can cause hypershould be used. Finally, a glucose curve ■ Owner administration glycemia despite administration of adeshould be plotted to eliminate other problems quate doses of insulin (see Causes of Hypossible mimics of insulin resistance ■ Somogyi phenomenon perglycemia in Diabetic Cats). Insulin (e.g., the Somogyi phenomenon, inad■ Stress resistance should be suspected in any cat equate duration of insulin action). In in which marked hyperglycemia persists the presence of the Somogyi phe■ Insulin resistance throughout the day despite insulin doses nomenon, a blood glucose nadir below of 1.5 U/kg per injection or higher or 80 mg/dl followed by a swift rebound when large doses of insulin (i.e., above 2.2 U/kg per inof hyperglycemia is seen. jection) are necessary to maintain adequate glycemic Diagnostic Differentials for Insulin Resistance control.22 However, use of these doses does not necesInsulin Antibodies sarily mean that insulin resistance is present. The probInsulin antibodies are a commonly discussed cause of lem could be associated with the owner’s insulin adinsulin resistance. Although antibodies may form, associministration technique, patient management (e.g., ated clinical insulin resistance is apparently rare.25 Conexercise, diet), or choice of insulin. Lack of response to fusion over the importance of antibodies may derive in high doses of one type of insulin does not mean that all part from the lack of a commercially available assay for insulins will be ineffective (e.g., 20% of cats do not redetecting anti-insulin antibodies. A radioimmunoassay spond to high doses of Ultralente® but can be effectively managed by twice-daily Lente®).23,24 In addition, longerfor insulin may help investigate this potential diagnosis acting insulins (PZI, Ultralente®) are more slowly abfor insulin resistance.26 Circulating insulin antibodies can sorbed and less bioavailable compared with shorter-actinterfere with radioimmunoassay measurement of serum ing insulins; cats may therefore require more than 2.2 insulin by causing spuriously high insulin values, and U/kg of long-acting insulin, particularly if it is being this artifact can suggest antibody presence. Typically, 24 given only once daily. hours after an insulin injection in a diabetic cat, serum Before a thorough and costly workup for insulin reinsulin concentration is below 50 µU/ml. In comparisistance is initiated, the following factors that mimic inson, apparent serum insulin concentration is above 400 sulin resistance should be ruled out. The owner’s inµU/ml if antibodies are present.26 In some species, antibodies appear to be more likely if sulin administration and handling techniques should using dissimilar insulin (e.g., beef insulin in dogs).27 always be evaluated first. Possible causes for an unsatisOne study25 did not uncover similar findings in cats, but factory response to insulin treatment include inadethis could have been related to the small sample size. If quate mixing before filling the syringe, use of the incorantibodies are believed to be causing insulin resistance, rect syringe type (e.g., using a U100 syringe with U40 the insulin source should be changed. If human insulin insulin), misreading the syringe, improper injection is currently being used, switching to PZI, which is 90% technique, insulin inactivation caused by improper beef insulin, is the best option for cats in the United handling, and improper dilution.2,22 The current belief is that a bottle of insulin should be discarded after 2 States. According to one author, if insulin-binding antimonths because of potential reduced activity. bodies are causing resistance, glycemic control usually If any of these factors are suspected, a glucose curve improves within 2 weeks after changing to a different should be plotted after the owner administers insulin source of insulin.26 using a new, undiluted bottle and while being obInfection served.22 The owner should also be questioned to enInfections can stimulate secretion of counterregulasure consistent and appropriate diet and exercise. If hytory hormones (e.g., cortisol, glucagon, possibly perglycemia is believed to be a result of postprandial epinephrine) that antagonize insulin and thus can lead surge related to feeding a meal when insulin’s effects are

Causes of Hyperglycemia in Diabetic Cats

ORAL HYPOGLYCEMIC AGENTS ■ RADIOIMMUNOASSAY FOR INSULIN ■ COUNTERREGULATORY HORMONES

Compendium May 2000

to insulin resistance.22 These infections may not be clinically apparent except by virtue of causing inadequate insulin response. The urinary tract and oral cavities are common sites22; urinalysis, urine culture (regardless of urinalysis findings28), and complete oral examination should always be performed.

Ketoacidosis or Concurrent Illness Ketoacidosis and concurrent illness can cause insulin antagonism via the same mechanism as infection.22 Renal disease, hepatic insufficiency, cardiac insufficiency, pancreatitis, and starvation should be considered as possible causes,2,3,22 and appropriate diagnostic testing performed as indicated. Malnutrition or Obesity Malnutrition can lead to insulin resistance and diminished insulin secretion.29,30 Obesity has been linked to glucose intolerance in cats,29,31 but its role in creating insulin resistance is unclear because obese diabetic cats generally remain responsive to insulin.32 Weight loss is recommended to improve insulin sensitivity and perhaps decrease the required dose.22 Weight loss should be gradual because severe restriction in caloric intake can lead to deterioration in glucose tolerance (see the companion article entitled “Treatment of Feline Diabetes Mellitus: Overview and Therapy” in this issue of Compendium).30 Endocrine Disorders Acromegaly, hyperthyroidism, and hyperadrenocorticism can cause insulin resistance through diverse mechanisms.2,22,33,34 Acromegaly in cats is predominantly caused by a growth hormone–secreting pituitary tumor and is most commonly seen in middle-aged to older males.35 Although progestogens do not stimulate secretion of the potent insulin-antagonist growth hormone in cats like they do in dogs, they do lead to insulin resistance through other mechanisms. 34 Hyperthyroidism can cause insulin resistance in cats,2,22 but concomitant hyperthyroidism and severe clinical insulin resistance are uncommon.36,37 Hyperadrenocorticism is likely to cause a pot-bellied appearance, hair loss, and thin or easily torn skin—signs not typical of DM alone.38 Miscellaneous Certain drugs, most notably progestogens and glucocorticoids, can cause insulin resistance. Cats may readily develop glucocorticoid-associated glucose intolerance. Hyperglycemia was experimentally induced in cats after administration of prednisolone at 2 mg/kg for 8 days.39 Use of these medications should be slowly discontinued if possible. Otherwise, patients may need to be treated for insulin resistance. Neoplasia that does

Small Animal/Exotics

not cause hyperadrenocorticism or acromegaly has been associated with insulin resistance in 5% to 10% of diabetic cats and dogs.2 In cats, pheochromocytoma, glucagonoma, lymphoma, mast cell tumor, and exocrine pancreatic adenocarcinoma are of particular concern.2 Hyperlipidemia should also be considered as a possible cause of insulin resistance.2,22 When trying to diagnose the cause of insulin resistance, the ones that are most likely and easiest to rule out should be eliminated first. The following order is generally recommended in cats: concurrent drugs, obesity, concurrent disease (including infection and ketoacidosis), hyperthyroidism, hyperadrenocorticism, acromegaly, and insulin antibodies.22 This recommended procession is not absolute; if strong evidence exists for a differential diagnosis lower in the order, that possibility should be ruled out first. Caution must be exercised in making a diagnosis of hyperadrenocorticism in animals with DM. In chronically ill cats, the corticotropin stimulation test may show an exaggerated (i.e., false-positive) response.40 In comparison, the dexamethasone suppression test (0.1 mg/ kg intravenous dexamethasone) is unlikely to do so.40 The diagnosis of concurrent DM and hyperadrenocorticism becomes more difficult because of the similarity in clinical signs (e.g., polyuria, polydipsia, polyphagia, hepatomegaly). Thus the diagnosis of hyperadrenocorticism must be based on consideration of the animal’s history, clinical signs, and routine laboratory findings and not only on serum cortisol concentrations. In general, hyperadrenocorticism is an uncommon cause of insulin resistance in cats and should be considered later in the diagnostic workup.41

Management Managing insulin resistance focuses on correcting the underlying cause, if possible. For such causes as obesity or concurrent administration of diabetogenic medications, eliminating the underlying problem is relatively easy. If insulin antibodies are suspected, a less antigenic form of insulin (e.g., PZI) can be used. Other problems, such as acromegaly in cats, may be difficult to correct. If the cause cannot be determined or eliminated, the following guidelines are recommended22: ■ Once daily insulin administration should be avoided; instead, at least two daily injections are preferred. ■ Long-acting insulin should be avoided unless regular insulin is added; intermediate-acting insulins are more effective in overcoming insulin resistance and lowering blood glucose. ■ A mixture of short- and long-acting insulins should be considered:

ACROMEGALY ■ GLUCOCORTICOID-ASSOCIATED GLUCOSE INTOLERANCE ■ NEOPLASIA

Small Animal/Exotics

Compendium May 2000

—Regular insulin and NPH insulin (mixed 1:2 in same syringe; i.e., one third regular insulin; two thirds NPH insulin) or a commercially available combination of 70% NPH insulin–30% regular insulin —Lente® and regular insulin (ideally administered separately) —Ultralente® and regular insulin (ideally administered separately) —Lente® and Ultralente® (can be mixed in one syringe) ■ Insulin should be administered shortly before or at feeding to help control postprandial hyperglycemia. Large doses of insulin may be required, but the actual dose must be determined using serial blood glucose curves, as is true for any diabetic patient. When interpreting the results, the dose of short-acting insulin should be adjusted to control blood glucose in the hours immediately following the insulin injection and the dose of long-acting insulin to regulate blood sugar during the hours before the next injection. Regardless of the cause of resistance, twice-daily injections are believed to be more effective in maintaining glycemic control.22

REFERENCES 1. Greco DS, Broussard JD, Peterson ME: Insulin therapy. Vet Clin North Am Small Anim Pract 25:677–689, 1995. 2. Feldman EC, Nelson RW: Diabetes mellitus, in Canine and Feline Endocrinology and Reproduction. Philadelphia, WB Saunders Co, 1996, pp 339–391. 3. Miller E: Long-term monitoring of the diabetic dog and cat: Clinical signs, serial blood glucose determinations, urine glucose and glycated blood proteins. Vet Clin North Am Small Anim Pract 25:571–584, 1995. 4. Cohn LA, McCaw DL, Tate D, et al: Assessment of five portable blood glucose meters, a point-of-care analyzer, and color test strips for measuring blood glucose concentration in dogs. JAVMA 216:198–202, 2000. 5. Nelson RW: Disorders of the endocrine pancreas, in Nelson RW, Couto CG (eds): Essentials of Small Animal Internal Medicine, ed 2. Philadelphia, Mosby, 1998, pp 734–774. 6. Feldman EC, Nelson RW, Feldman MS: Intensive 50-week evaluation of glipizide administration in 50 cats with previously untreated diabetes mellitus. JAVMA 210:772–777, 1997. 7. Elliott DA, Nelson RW, Feldman EC: Glycosylated hemoglobin concentration for assessment of glycemic control in diabetic cats. J Vet Intern Med 11:161–165, 1997. 8. Dennis JS: Glycosylated hemoglobins in dogs. Compend Contin Educ Pract Vet 11(6):717–720, 1989. 9. Kawamoto M, Kaneko JJ, Heusner AA: Relation of fructosamine to serum protein, albumin, and glucose concentrations in healthy and diabetic dogs. Am J Vet Res 53:851– 854, 1992. 10. Jain NC: Erythrocyte physiology and changes in disease, in Essentials of Veterinary Hematology. Philadelphia, Lea & Febiger, 1993, pp 133–158. 11. Elliott DA, Nelson RW, Feldman EC, et al: Glycosylated

12.

13. 14. 15.

16. 17.

18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

29. 30. 31.

hemoglobin concentrations in the blood of healthy dogs and dogs with naturally developing diabetes mellitus, pancreatic β-cell neoplasia, hyperadrenocorticism, and anemia. JAVMA 211:723–727, 1997. Kaneko JJ, Kawamoto M, Heusner AA, et al: Evaluation of serum fructosamine concentration as an index of blood glucose control in cats with diabetes mellitus. Am J Vet Res 53: 1797–1801, 1992. Jensen AL: Serum fructosamine in canine diabetes mellitus: An initial study. Vet Res Commun 16:1–9, 1992. Lutz TA, Rand JS, Ryan E: Fructosamine concentrations in hyperglycemic cats. Can Vet J 36:155–159, 1995. Reusch CE, Dlough U, Heusner AA: Evaluation of statistically significant changes in the level of glycosylated hemoglobin and serum fructosamine (Abstr). J Vet Intern Med 9:186, 1995. Reusch CE, Liehs MR, Hoyer M, et al: Fructosamine: A new parameter for diagnosis and metabolic control in diabetic dogs and cats. J Vet Intern Med 7:177–182, 1993. Crenshaw KL, Peterson ME, Heeb LA: Serum fructosamine concentration as an index of glycemia in cats with diabetes mellitus and stress hyperglycemia. J Vet Intern Med 10:360– 364, 1996. Thoresen SI, Bredal WP: Clinical usefulness of fructosamine measurements in diagnosing and monitoring feline diabetes mellitus. J Small Anim Pract 37:64–68, 1996. Nelson RW, Griffey SM, Feldman EC, et al: Transient clinical diabetes mellitus in cats: 10 cases (1989–1991). J Vet Intern Med 13:28–35, 1999. Whitley NT, Drobatz KJ, Panciera DL: Insulin overdose in dogs and cats: 28 cases (1986–1993). JAVMA 211:326–330, 1997. Goossens M, Nelson RW, Feldman EC, et al: Response to insulin treatment and survival in 104 cats with diabetes mellitus (1985–1995). J Vet Intern Med 12:1–6, 1998. Peterson ME: Diagnosis and management of insulin resistance in dogs and cats with diabetes mellitus. Vet Clin North Am Small Anim Pract 25:691–713, 1995. Nelson RW, Feldman EC, DeVries S: Use of Ultralente insulin in cats with diabetes mellitus. JAVMA 200:1828–1829, 1992. Bertoy EH, Nelson RW, Feldman EC: Effect of Lente insulin for treatment of diabetes mellitus in 12 cats. JAVMA 206: 1729–1731, 1995. Harb-Hauser M, Nelson RW, Gershwin L, et al: Prevalence of insulin antibodies in diabetic cats (Abstr). J Vet Intern Med 12:245, 1998. Nelson RW: Diabetes mellitus, in Ettinger SJ, Feldman EC (eds): Textbook of Veterinary Internal Medicine, ed 4. Philadelphia, WB Saunders Co, 1995, pp 1510–1537. Harb-Hauser M, Nelson RW, Gershwin L, et al: Prevalence of insulin antibodies in diabetic dogs (Abstr). J Vet Intern Med 12:213, 1998. Forrester SD, Troy GC, Dalton MN, et al: Retrospective evaluation of urinary tract infection in 42 dogs with hyperadrenocorticism or diabetes mellitus or both. J Vet Intern Med 13:557–560, 1999. Nelson RW, Himsel CA, Feldman EC, et al: Glucose tolerance and insulin response in normal-weight and obese cats. Am J Vet Res 51:1357–1362, 1990. Biourge V, Nelson RW, Feldman EC: Effect of weight gain and subsequent weight loss on glucose tolerance and insulin response in healthy cats. J Vet Intern Med 11:86–91, 1997. Kirk CA, Feldman EC, Nelson RW: Diagnosis of naturally acquired type-I and type-II diabetes mellitus in cats. Am J Vet Res 54:463–467, 1993.

Compendium May 2000

32. Ihle SL, Nelson RW: Insulin resistance and diabetes mellitus. Compend Contin Educ Pract Vet 13(2):197–202, 1991. 33. Middleton DJ, Watson ADJ: Glucose intolerance in cats given short-term therapies of prednisolone and megestrol acetate. Am J Vet Res 46:2623–2625, 1985. 34. Peterson ME: Effects of megestrol acetate on glucose tolerance and growth hormone secretion in the cat. Res Vet Sci 42:354–357, 1987. 35. Peterson ME, Taylor RS, Greco DS: Acromegaly in 14 cats. J Vet Intern Med 4:192–201, 1990. 36. Peterson ME, Kintzer PP, Cavanagh PG: Feline hyperthyroidism: Pretreatment clinical and laboratory evaluation of 131 cases. JAVMA 183:103–110, 1983. 37. Broussard JD, Peterson ME, Fox PR: Changes in clinical and laboratory findings in cats with hyperthyroidism from 1983 to 1993. JAVMA 206:302–305, 1995. 38. Duesberg CA, Peterson ME: Adrenal disorders in cats. Vet Clin North Am Small Anim Pract 27:321–347, 1997. 39. Middleton DJ, Watson ADJ, Howe CJ, et al: Suppression of cortisol responses to exogenous adrenocorticotrophic hor-

Small Animal/Exotics

mone, and the occurrence of side effects attributable to glucocorticoid excess, in cats during therapy with megestrol acetate and prednisolone. Can J Vet Res 51:60–65, 1987. 40. Zerbe CA, Refsal KR, Peterson ME, et al: Effect of nonadrenal illness on adrenal function in the cat. Am J Vet Res 48:421–454, 1987. 41. Peterson ME, Nesbitt GH, Schaer M: Diagnosis and management of concurrent diabetes mellitus and hyperadrenocorticism in thirty dogs. JAVMA 178:66–69, 1981.

About the Authors Dr. Behrend is affiliated with the Department of Small Animal Surgery and Medicine, Auburn University, Alabama. Dr. Greco is affiliated with the Department of Clinical Sciences, Colorado State University, Fort Collins, Colorado. Drs. Behrend and Greco are Diplomates of the American College of Veterinary Internal Medicine.