Diabetes mellitus (Proceedings)

Definition

Fasting hyperglycemia

Glucosuria suggests diabetes but is not definitive for the diagnosis because, for example, renal tubular functional defects (Fanconi syndrome in Basenji and other breeds) can cause glucosuria

Glucosuria with ketonuria is almost definitive evidence of diabetes

Classification

Type I and Type II are the main two types of diabetes mellitus

Type I diabetes

Type I diabetes is characterized by lack of insulin production by the pancreatic beta cells of the islets of Langerhans; it is also known as IDDM or Insulin Dependent Diabetes Mellitus because insulin injections are required for treatment

Type I is the type most commonly diagnosed in dogs because it causes profound obligatory solute (glucose) osmotic diuresis resulting in urinary accidents which prompts dog owners to seek veterinary care

The catabolic state that follows is severe and weight loss is often profound for two reasons:

glycogenolysis (recall that stored glycogen in terrestrial mammals' liver is limited) is followed within a few hours of insulin lack by gluconeogenesis

gluconeogenesis is much less limited and occurs by way of skeletal muscle protein catabolism and amino acids (principally alanine) from this source are the basis of hepatic over production of new glucose that cannot be used by peripheral cells due to the lack of insulin

under use and over production of glucose contribute to solute diuresis as the renal tubular threshold for glucose re-absorption continues to be exceeded and protein catabolism also continues

in the quest for a utilizable fuel (energy source), in another day or two a second major catabolic process is up regulated: ketone production

the ketone bodies (acetoacetic acid, β-hydroxybutyric acid and acetone) are made by the liver from the mobilization and catabolism of fat stores;

within 7 to 10 days after marked reduction in insulin production, severe weight loss is obvious in the dog with Type I diabetes because of the catabolism of both muscle and fat

Type II diabetes

Type II diabetes is characterized more by peripheral resistance to the effects of insulin at the receptor sites than to an absolute lack of insulin; it is also known as Non-Insulin Dependent Diabetes Mellitus or NIDDM because insulin injections are not always required for treatment

More cats have Type II diabetes than is recognized in dogs and the severity of weight loss and degree of solute diuresis is often less obvious than it is in dogs

Depending on the cat population studied, it is believed that from 50% to 80% of diabetic cats have Type II diabetes at the time of diagnosis

diabetic cats, like dogs, have glucose diuresis due to hyperglycemia resulting in glomerular filtrate glucose concentration that exceeds renal tubular re-absorptive capabilities but the magnitude of these pathophysiologic events are often not as great as in dogs

Pathogenesis of type I diabetes

Type I diabetes, the type most commonly diagnosed in dogs, is characterized by marked reduction in insulin production by the pancreatic beta cells

The events leading to impaired insulin secretion by beta cells appear to be immune mediated and evidence exists to suggest that both humoral and cell-mediated immune responses may be involved

The lines of evidence are derived from observational and experimental studies of rodents and human beings as well as dogs

Genetic predisposition appears to be a factor as some strains of mice, certain HLA types of human beings and specific breeds of dogs are either at increased or reduced risk for developing Type I diabetes

Some environmental factor then appears to interact with an at risk host to initiate an immune attack on insulin producing beta cells or on some part of the insulin producing machinery

Viruses are the environmental factors most often implicated and the mechanisms involved may include direct viral destruction of insulin producing beta cells, infection of beta cells with associated inflammatory response resultant diminution in insulin producing capacity, or in epitope creation, modification or mimicry that stimulates cell-mediated and/or humoral immune attack on insulin producing cells or insulin related molecular associates

Epitope mimicry associated with the intestinal absorption of some "look alike protein" that is ingested is also proposed as a mechanism that might stimulate an immune response directed at some portion of the insulin producing apparatus

In the aggregate, the evidence supporting a general sequence leading to Type I diabetes by the above outline is reasonably strong. For example, the diabetes producing variant of the murine EMC virus replicates in the beta cells of nearly all strains of mice experimentally infected with the virus but produces diabetes in a variable but predictable percent of infected mice depending on the mouse strain. When athymic (nude) mice are infected, viral replication in beta cells occurs but diabetes does not. When thymically transplanted nude mice are infected with the virus, both viral replication and diabetes results. Diabetes can be prevented by administration of appropriate immunosuppressive agents.

Pathophysiology of type I diabetes of dogs

Although lack of insulin production is the apparent primary endocrine event that initiates Type I diabetes, continued production of glucagon by unaffected alpha cells of the Islets of Langerhans is apparently an important determinant of the severity of disease characteristic of Type I diabetes

For this reason, type I diabetes is referred to as a bi-hormonal disease as it is the lack of insulin in the presence of normal or increased amounts of glucagon that results in keto-acidosis characteristic of Type I diabetes

Although insulin lack is the principal event that favors mobilization of amino acids from muscle and fatty acids from adipose, it is glucagon that promotes hepatic over-production of glucose and especially ketones

In addition to osmotic effects, it is important to recall that two of the three ketones are organic acids and their accumulation titrates normal physiologic buffering systems until metabolic acidosis results

Metabolic acidosis that occurs in diabetic keto-acidosis (DKA) is the most severe metabolic acidosis commonly encountered in small animal practice

Pathogenesis of type II diabetes of cats

Some investigators believe that the initiating event leading to Type II diabetes in most cats is related to diet

Cats are considered to be obligate carnivores and feeding free choice relatively high carbohydrate low protein cat chow may promote feline obesity because protein is apparently responsible for cat satiety and low protein food promotes increased total caloric intake

Obesity causes peripheral insulin resistance

Moreover, the long term feeding of relatively high carbohydrate low protein free choice cat chow may be responsible for an abnormally high requirement for long-term insulin production

Although a Utrecht study provided some evidence exonerating high carbohydrate diets, the study clearly established that hypersecretion of insulin is a consequence of high carbohydrate low protein feeding but determination of the long term effects of chronic hyperinsulinism was beyond the scope of the study

Co-secreted on a molar for molar basis with cat insulin is an amyloid protein termed amylin or IAPP (Islet Amyloid Polypeptide) which is capable of forming amyloid fibrils that may accumulate in pancreatic islets and induce beta cell apoptosis

The net effect is peripheral insulin resistance caused by obesity with concurrent diminished insulin secretory capacity caused by amyalin deposition in islets

Pathophsiology of Type II Diabetes in cats

Peripheral insulin resistance caused by obesity is initially overcome by increased insulin secretion

Continued hyper secretion of insulin is followed by amyalin deposition in islets which eventually results in diminished insulin production

The type of diabetes usually affecting cats, therefore, is initially Type II or NIDDM that gradually becomes Type I or IDDM

The phenomenon of glucose toxicity also apparently plays a role in the pathophysiology of cat diabetes

Chronic high glucose concentration in and of itself (glucose toxicity) seems to contribute to diminished insulin production as well as increasing peripheral insulin resistance by causing a defect in post-transport insulin action

Of the pathophysiologic events leading to diabetes in cats, glucose toxicity in particular appears to be somewhat reversible if diabetes is intensively treated when first diagnosed

At the time that a cat is diagnosed with diabetes it may have Type II diabetes or it may be keto-acidotic and a Type I diabetic.

The nomenclature used in the classification of diabetes is borrowed from human diabetes and is not precisely applicable to diabetes of cats and dogs. Of the approximately 24 million human diabetics in the United States, about 10 % are Type I or have Insulin Dependent Diabetes Mellitus (IDDM) which is really an insulin requiring diabetes (insulin injections required for treatment). Because most human beings are young at age of onset, Type I diabetes is also known as juvenile diabetes; patients are prone to keto-acidosis. The majority of human diabetics have Type II or adult onset diabetes which is strongly associated with obesity and insulin resistance. This is the type that is currently referred to as having reached epidemic proportions. The long-term side effects of diabetes such as nephropathy, retinopathy, neuropathy and vascular disease leading to myocardial infarction and stroke all take about 5 years of diabetes before human patients are in increased risk for these complications. Good control of blood glucose concentration using intensive monitoring and intensive treatment with multiple daily insulin injections including management with insulin pumps clearly minimizes the likelihood of long term complications in Type I human diabetics. This benefit has not been clearly demonstrated for Type II human diabetics.

Most dogs and cats at the time of diagnosis of diabetes do not have a life expectancy (usually less than 5 years) that poses much risk for these long-term complications and for this reason extremely tight control of blood glucose is not a usual therapeutic goal. Dogs do develop diabetic cataracts much more frequently than human diabetics but the development of diabetic cataracts in dogs does not seem to be related to how well their glucose concentration is controlled. A small percent of diabetic cats and dogs do develop peripheral neuropathy that does appear to be related to how well their diabetes is controlled. The motor neuropathy in diabetic cats and dogs is most frequently manifested as hind limb weakness and does seem to respond to better control of their blood glucose concentrations. Neuropathy (manifested as dropped hocks) in some cats seems to respond to methylcobalamin (Xobaline®) but not to cyanocobalamin.

Comparative treatment of diabetic cats and dogs related to pathophysiology

Because diabetic dogs are almost always Type I diabetics, insulin is required for successful therapy. The prognosis with insulin therapy is very good and life expectancy with treatment is measured in years. Although diabetic cats are frequently Type II diabetics at the time of diagnosis, their response to oral hypoglycemic agents commonly used to treat human Type II diabetics is successful in less than half of diagnosed cats. The oral hypoglycemic agent most commonly used is glypizide (Glucotrol, a sulfonylurea) but no predictors of successful response have yet been identified. Furthermore, early treatment of cat diabetes with a combination of high protein low, carbohydrate diet and intensive insulin therapy may actually reverse the pathophysiologic process and "cure" the diabetes. Diabetic cats that require indefinite treatment with insulin, like diabetic dogs, typically enjoy good quality life for several years.

Insulin preparations for treating diabetic cats and dogs

Many commercially available insulin preparations familiar to veterinarians are no longer available and recombinant insulin analog products are now on the market. As of late spring 2008, Idexx Pharmaceuticals Inc. has discontinued its PZI VET (protamine zinc insulin of beef-pork origin) and the US FDA has approved Intervet Inc.'s Vetsulin (porcine insulin zinc suspension lente) for treating diabetic cats after having previously approved its use in dogs.

Making logical decisions regarding which insulin products to use is best based on an understanding of insulin and the commercially available products.

Mammalian insulins are structurally similar, consisting of a 21 amino acid A chain and a 31 amino acid B chain that are connected by two disulfide bonds. Within pancreatic beta cells, the insulin molecule is derived from the proinsulin molecule by the cleavage of a connecting peptide (C-peptide). Although mammalian insulins vary by 1 to 3 amino acid substitutions in each chain, overall they are remarkably similar not only in structure but are also nearly identical with respect to biological activity (potency) and are minimally antigenic across species lines. These facts made possible the 75 year-long-historical use of ethanol-acid extracted insulin from a harvest species (usually cattle or pigs) for use in another mammalian species (usually humans, cats or dogs) that suffered from diabetes. In the last 30 years, recombinant DNA techniques have allowed for the commercial production of synthetic insulin with an amino acid sequence identical to human insulin and these products now dominate the human therapeutic market. Fortunately, high cross species biological activity and low antigenicity allow for cross species use with virtual impunity.

In the 1920s it was quickly learned that the early insulin extractions, although life saving for diabetic patients, had the drawback of such a short duration of action that multiple daily injections were necessary. Through experimentation it was learned that the addition of zinc to the extracted insulin caused the formation of zinc-insulin crystals that resulted in a product with prolonged duration of action. It was also determined that the addition of protamine (a protein derived from fish semen) also prolonged the action of insulin. Insulin preparations with additives were commercialized and given names by the appropriate laboratory or manufacturer. Examples are Lente (slow acting), Ultralente (very slow acting), PZI (protamine zinc insulin) and NPH (neutral pH, protamine, Hagedorn) also known as isophane insulin. Regular insulin does not have additives that appreciably prolong the duration of action.

In addition to labeling that indicated the duration of action, the species of origin of the modified insulin was included in the labeling such as beef, pork or beef-pork. By the 1980s, insulin made by recombinant DNA techniques with an amino acid sequence identical to human insulin were generally assigned copyrighted names such as Humalin (Eli Lilly) and Novalin (Novo Nordisk).

Lastly, the concentration of insulin biological activity appears on insulin vials. Currently, FDA approved veterinary products are U-40 (40 units of insulin/ml) and FDA approved human products are U-100 (100 units of insulin/ml) with the exception of the rarely encountered human U-500 preparation.

Currently the only FDA approved insulin preparation for veterinary use is Intervet's Vetsulin, which is a lente (intermediate duration of action) preparation of pork origin. It has been available in Europe for several years where it is sold as Caninsulin. In this country it is a U-40 (40 units/ml) product and should be used with U-40 insulin syringes. Vetsulin has FDA approved labeling for treating diabetic dogs with once daily injections although the manufacturer concedes that twice daily administration is necessary for satisfactory control of diabetes in 80% of affected dogs. The manufacturer touts the product as superior for dog use because the amino acid sequence of porcine insulin is identical to dog insulin. The firm has recently obtained FDA approval for use in cats but, to my knowledge, has made no specific comment regarding the porcine amino acid sequence as it relates to use in cats.

The only other FDA insulin approved for veterinary use is Idexx Pharmaceutical's PZI VET, a long acting beef-pork insulin for use in cats which the firm has announced will be discontinued as of spring 2008 although substantial supply is reportedly on hand.

FDA approved products for use in human beings are numerous and are either of human insulin amino acid sequence or are insulin analogs with substitutions designed to provide for a prolonged (24 hour) duration of action, such as glargine (Lantus) or for short but almost immediate onset of action such as lispro (Humalog). In general, long acting insulin analogs are used in human beings to provide basal insulin whereas immediate, short acting insulin analogs are designed to provide an insulin pulse at meal time.

Because neither natural mammalian insulins nor synthetic insulin analogs are very immunogenic, all can generally be used across species. When human beings, cats or dogs make anti-insulin antibodies, the antibodies are usually against all insulins regardless of whether the insulin is a substituted synthetic analog or a natural mammalian amino acid sequence. This is true because the epitope is generally located at a site on the insulin molecule other than where amino acid substitutions have occurred. Regardless of the specificity or lack of specificity of anti-insulin antibodies, the clinical relevance is usually trivial in that a slight increase in dose (~10%) will usually overcome the effects of anti-insulin antibodies.

In those rare instances in human beings when anti-insulin antibodies are a major problem, hyposensitization protcols have been devised that require specialized care.

In as much as insulins and insulin analogs are biologically active across species regardless of the amino acid sequence, other factors such as duration of action and cost per unit become more important considerations. FDA approved veterinary Vetsulin™ is a porcine origin Lente insulin that is an intermediate acting insulin with duration of approximately 12 hours (± 6 hours). In most cases, Vetsulin™ must be administered twice daily. Because it is a U-40 preparation, 2.5 times greater volume is necessary to administer the same biologic activity as a U-100 preparation. The greater volume is not usually a problem as higher doses are usually needed for large breeds of dogs. Cost, however, can be a concern as cost per unit of insulin activity is substantially more (4 times greater) than human preparations with similar duration of action such as NPH. The cost of U-40 insulin syringes is also a factor as we can purchase U-100 insulin syringes for almost half the cost. Available insulin preparations suitable for lone term treatment of diabetic cats and dogs can be summarized:

Intermediate (~12 hour) or long (~24 hour) acting insulins are appropriate

One veterinary approved product for cats and dogs is currently made which is Vetsulin™ a U-40, porcine origin, Lente insulin that cost about 4 times as much as similar insulins approved for human use

NPH insulins approved for human are similar to Lente insulin in terms of duration , are U-100, cost about one-fourth less and are made by Lilly (Humalin® NPH) and Novo Nordisk (Novolin® NPH)

A long acting (24 hour) human approved insulin analog product, glargine (Lantus©, Aventis), has gained popularity, particularly for use in cats. It is a U-100 product, is sometimes administered twice daily to cats in spite of 24 hour duration and costs 4-5 times as much as human approved NPH preparations.

Other insulin products are available but have little application for long term, maintenance treatment of diabetic cats and dogs:

A competing human approved insulin analog product, detemir (Levemir©, Novo Nordisk) has been evaluated very little in either cats or dogs

The use of regular insulin (Humalin© R, Lily and Novolin© R, Novo Nordisk) is primarily restricted for veterinary use in treating cats and dogs with diabetic ketoacidosis (DKA)

Other rapid, short acting insulin analogs are used primarily at meal time to supplement long acting insulin injections in treating human diabetics. The rapid, short acting insulin analogs include insulin glulisine (Apidra®, Sanofi-Aventis), insulin lispro (Humalog®, Lilly) and insulin aspart (Novolog®, Novo Nordisk).

Drugs other than insulin and insulin analogs used to treat diabetic human beings

Several classes of drugs have evolved primarily for treating human patients with Type II diabetes, the most common type of diabetes in this species. Because many diabetic cats have Type II diabetes, some representative drugs have been used in cats with generally mixed or disappointing results. Use of these drugs in diabetic dogs has been attempted rarely and generally without success, as dogs usually have Type I diabetes.

Biquanides – metformin (Glucophage) acts in an insulin-dependent manner by enhancing sensitivity of hepatic and peripheral tissues to insulin. Use in cats has not been encouraging.

Sulfonylureas – glipzide (Glucotrol) and glyburide (DiaBeta) as well as others act by stimulating insulin release from pancreatic beta cells and, to a lesser extent, by enhancing peripheral insulin effectiveness at the receptor or post-receptor site. Sufonylureas are also called secretagogues. Have been used with limited success in Type II diabetic cats. About 30 % of cats have some response but no reliable predictor of response has yet been identified. Used primarily in instances of extreme reluctance by cat caregivers to administer insulin.

Meglitinides – repaglinide (Prandin) and nateglinide (Starlix) like sulfonylureas are secretatogues but are shorter acting and typically used to treat Type II human diabetics at mealtime. No reported use in cats or dogs.

Alpha-glucosidase inhibitors – acarbose (Precore) and miglitol (Glyset) are complex oligosaccharides that inhibit certain intestinal brush border digestive enzymes (α – glucosidases), the inhibition of which retards absorption of glucose and other simple sugars. Although quite expensive, these drugs have been evaluated in normal and dogs with Type I diabetes (IDDM) and do aid in glycemic control. Use in cats is generally not recommended as low carbohydrate, high protein foods are recommended to be fed.

Thiazolidinediones (TZDs or glitazones) – rosiglitazone (Avandia), pioglitazone (Actos) and the recently withdrawn hepatitis associated troglitazone (Rezulin) act by binding to PPARs (peroxisome proliferator-activated receptors) for treating type II human diabetes. They have not been used in cats or dogs and recently published studies (ACCORD) suggest increased mortality when used intensively to control human Type II diabetes associated cardiovascular events.

Amylin analogs- pramlintide (Symlin) is an injectable used as adjunct treatment for human Type I and type II diabetes. Its has not been used for treating diabetic cats and dogs but information leading to its discovery was reported (1987) in part by University of Minnesota veterinarians who discovered islet amyloid polypeptide (IAPP or Amylin) while studying spontaneously diabetic cats. Amylin is co-secreted on a molar for molar basis with insulin and has hormone effects. When hypersecreted in pancreatic islets by type II diabetic cats, however, the amylin forms amyloid deposits and is toxic to insulin secreting beta cells. The synthetic amylin (pramlintide or Symlin) does not form amyloid fibrils while maintaining the desired effects of inhibiting inappropriate glucagon secretion, slowing gastric emptying and promoting satiety.

Incretins are hormones originating from endocrine cells of the intestinal epithelium. Incretins are secreted into the circulation in response to oral carbohydrate (glucose) ingestion and signal the pancreatic islets to suppress glucagon secretion and stimulate additional insulin secretion. Two incretins have been identified: GIP (gastric inhibitory peptide also known as glucose-dependent insulinotropic peptide) and GLP-1 (glucagon-like peptide-1). Identification of the proinsulin incretins has been followed by the approval of two related pharmacologic agents for treating type II human diabetes. Exenatide (Byetta) is an incretin analog somewhat of veterinary interest because it is an injectable synthetic derivative of Gila monster saliva (lizard spit). The second drug, intensely advertised sitagliptin (Januvia), is administered orally and increases incretins by inhibiting the incretin inactivating enzyme, DPP-4 (dipeptidyl peptidase 4).

Surgical treatment for type II human diabetes is currently receiving substantial publicity and the effectiveness of gastric bypass surgery is apparently related to the incretins or, rather, removal of anti-incretins. Because obesity causing peripheral insulin resistance certainly plays an important role in type II human diabetes, both gastric banding and gastric bypass surgery have been used to promote weight loss. Observed effectiveness of gastric bypass surgery (but not gastric banding) in "curing" diabetes has been unanticipated and apparently unrelated to surgically induced weight loss. The dramatic effectiveness of gastric bypass bariatric surgery has resulted in the "anti-incretin theory" and the organization of the Diabetes Surgery Summit. Previously investigated surgical approaches to type I diabetes include segmental pancreatic and isolated pancreatic islet transplantation and have had limited success.

Treating non-ketoacidotic (uncomplicated) diabetes in dogs

Because most diabetic dogs have type I or insulin dependent diabetes mellitus (IDDM), insulin administration is required for satisfactory treatment. IDDM can be confirmed by reserving a serum sample for simultaneous determination of insulin and glucose. If the insulin concentration is low at a time that fasting glucose is high, IDDM or type I diabetes is confirmed. If insulin is not decreased below reference values, diabetes may be the result of peripheral insulin resistance which may be secondary to high progesterone induced acromegaly in the older non-spayed bitch, secondary to some other identifiable cause of peripheral insulin resistance or may be a rare case of NIDDM. It is important in the initial evaluation to include urine analysis, serum chemistry determinations and complete blood count as reasonable assurance that IDDM is not complicated by concurrent illness. Some deviations from normal are anticipated as a consequence of diabetes and do not warrant undue concern. Examples include moderate increases in hepatic enzyme activity, modest deviations in serum electrolyte concentrations, the presence of ketonuria, and moderate decrease in serum bicarbonate and total CO2 . Although the presence of ketones in urine and evidence of metabolic acidosis in the strictest sense may classify the patient as a diabetic ketoacidotic, if physical examination findings are normal and if the appetite remains good, the dog should be treated as an uncomplicated or non-ketoacidotic patient.

At this time, commercially available insulin preparations include one FDA approved product for canine use. Vetsulin® is a porcine origin, intermediate acting Lente insulin with a 40 unit per ml concentration (U40). The duration of action is approximately 12 hours and although Vetsulin® is approved for once daily administration, 80-90% of dogs require twice daily administration for adequate glycemic control. Because the product is less than half the concentration of products approved for use in human beings (U40 vs U100), the volume per unit of biologic activity is correspondingly greater. The cost per unit of insulin is generally greater than human approved products such as Humalin® NPH and Novalin® NPH but less expensive per unit than long acting (24 hour) insulin analog products such as glargine (Lantus®) insulin. The difference in cost per unit is seldom an issue in small dogs but may become a consideration when contemplated for use in large dogs. Aside from cost per unit of insulin, a second consideration is the number of daily insulin injections required for control of diabetes. Although the total daily dose of administered insulin is likely to be the same, Vetsulin®, Humalin®NPH and Novalin®NPH usually must be administered twice daily, whereas Lantus® can usually be administered once daily. Factors such as family obligations, work schedules and life-style are determinates of whether twice daily or once daily insulin administration is most appropriate for the individual canine patient and the caregiver.

Dose of insulin required for adequate glycemic control can only be determined by monitoring response to treatment but the usual total daily dose for dogs is about 1.0 unit per kg of body weight per day (0.5 units per pound per day). Although large dogs may require fewer insulin units per kg of body weight than do small dogs, the usual appropriate daily dose of insulin is much more linear with respect to body weight than are many drugs, such as most cancer chemotherapeutic agents, which are more appropriately dosed on the basis of body surface area. Some sources recommend a starting dose of insulin that is approximately 0.5 units per kg of body weight per day (0.25 units per pound per day) with incremental increases until the required dose is reached. Although this extremely cautious approach is correct in theory, the usual practical consequence is a protracted interval, often approaching two months, before the appropriate insulin dose is finally reached. During this time, home caregivers may become frustrated and diabetes associated ketosis may actually worsen. For these reasons, unless there is evidence to suggest that the dog is unusually insulin sensitive, the initial insulin dose is best calculated at only slightly less than 1.0 unit per kg of body weight per day.

Published monitoring recommendations of the diabetic dog during the first few days and weeks after initiating insulin administration are diverse and somewhat confusing. A rational approach can be based on simple review of monitoring methods available and the information provided in the context of reversing the pathophysiologic events of type I diabetes (IDDM) in the dog. The following is a list of monitoring methods and the relevance of the methods in the initial days of insulin administration for the type I diabetic dog:

Initial home monitoring

Clinical Signs: PUPD or increased drinking and urination caused by obligatory solute (glucose) diuresis are hallmarks of diabetes. If PUPD is not caused by coexisting disease then substantial glucosuria must still be present during the early stages of insulin administration to account for PUPD. Periodic blood glucose determinations, blood glucose curves and fructosamine determinations provide little useful information. Periodic urine glucoase estimates confirm the magnitude of glucosuria.

Urine Glucose Estimates: Urine glucose estimated with KetoDiastix® by the caregiver at home confirm glucose diuresis, provide the caregiver practice in collecting and testing urine and document the presence of ketonuria which is often present and typically lags behind reduction in glucosuria.

Blood Glucose Determinations: Periodic, timed or un-timed blood glucose determinations whether done by the caregiver at home or on an out-patient basis by the veterinarian provide little information in the initial stages of insulin administration if PUPD and glucosuria are present.

Blood Glucose Curves: In the early stages of insulin administration, no useful information is gained by performing multiple blood glucose determinations because hyperglycemia is constantly present and reversal of diabetic metabolic pathways is not accomplished sufficiently well as to provide reliable glucose curves.

Fructosamine Determinations: Because fructosamine is an estimation of the non-enzymatic glycation of serum proteins that has occurred during the preceding 3-4 weeks, this determination is of little value at the early stage of insulin administration except to provide base line data. Because of the relatively short duration of diabetes, at the early stages of insulin administration fructosamine is usually only modestly increased and may actually increase more as a reflection of duration of hyperglycemia than as a reflection of magnitude of hyperglycemia.

Foods and feeding the diabetic dog

Considerable objective information exists regarding available food choices and feeding plans suitable for the diabetic dog. Each diabetic dog is an individual, however, and plans must be adjusted to the circumstances while rigid adherence to plans predicated on the best scientific information may lead to substantial problems.

High fiber foods containing complex carbohydrates and minimal amounts of simple sugars specifically formulated for diabetic dogs are commercially available usually as a prescription diet. There is abundant evidence to support their use. The high fiber retards carbohydrate absorption and diabetic dogs (as a group) fed prescription foods require slightly less insulin on a per weight basis than do dogs (as a group) that are not fed diets specifically formulated for diabetes. Palatability, however, is sometimes and issue and if the diabetic dog refuses the ideal diet it is not in the patient's best interest to rigidly insist that the caregiver only feed the prescribed food. The consequences of feeding a non-prescription diabetes diet are usually minimal as the increased dose of required insulin is seldom greater that 2-4%. Furthermore, is not unusual to encounter middle-aged dogs, typically toy breeds, which have never been fed commercial dog food of any kind. Attempts at convincing such dogs that they should eat any commercial dog food, let alone a high fiber food, may be futile. In as much as food intake is of particular importance in the early stages of treating the newly diagnosed diabetic dog, it is important that our dietary recommendations take into account such exigencies. A home prepared relatively high protein diet will often work well in such circumstances. It is also important that some latitude in feeding times be considered. Although timed feedings that correspond to insulin administration maybe ideal, some dogs (although with less frequency than cats) refuse to adhere to timed feedings, having been intermittent nibblers all of their lives. In as much as most dogs are treated with basal insulin only (no feeding associated pulse insulin administration), dogs that refuse to eat specifically at the time of intermediate acing insulin administration can usually be satisfactorily regulated. In the overall context of initial home insulin administration, efforts should be made to keep dogs eating even if the food type and feeding schedule is not textbook optimized. This is particularly important because at this stage of diabetes control, residual ketosis is often still present and failure to eat may exacerbate ketosis, which may suppress appetite.

Usual phases of initiating treatment of an dog with uncomplicated diabetes

It is helpful to consider phases of bringing diabetes under control. Successful treatment of a diabetic dog requires the coordinated team action between the veterinarian and the home caregiver. In most instances the veterinarian must educate the home caregiver as well as assess the medical progress of the patient in an ongoing manner. Few medical disorders require the home caregivers active participation to the degree that diabetes demands. In this context it is useful to consider the process in phases.

Hospital and Assessment Phase – 1 or 2 days is usually sufficient to assess the dog's preliminary response and, based on phone conversations, to partially assess the caregivers abilities to devote efforts for home care. Titrating insulin dose can only be done in the home environment.

Discharge Phase – Although the discharge from hospital or clinic may take over one hour and may include practice injections, verbal instructions and review of a comprehensive written handout, the education process is usually only started; daily or twice daily phone consultations should be anticipated for the next few days.

Initial Home Phase- Daily phone conversations between the veterinarian and home caregiver should focus on providing moral and instructional support as it relates to twice daily insulin administration (if an intermediate acting insulin such as Vetsulin or NPH is used), feeding and assessment of the diabetic dog's appetite and monitoring with KetoDiastix for the presence of urine ketones and glucose as often as is conveniently possible. At this stage, PUPD, glucosuria and ketonuria are expected and appetite should be increasing.

Confidence Building Phase – With in 2-4 days, home caregiver confidence usually builds with the phone guidance of the veterinarian. Caregivers' education continues as the veterinarian answers daily questions. Glucosuria and attendant PUPD are usually still present as incremental increases in insulin dose are made at the approximate rate of 10% per day. Ketonuria usually wanes but remains at a trace to small (5-15 mg/dl) concentration.

Fine Tuning Phase – By the end of the first week at home, the caregivers' questions usually reflect more insight, ketonuria is expected to be abolished, glucosuria is minimized to 1+ (1000 mg/dl) to trace, nocturia is no longer present, there is noticeable decrease in PUPD and the dog is much more active. Although these events can typically take place by means of phone consultations, out-patient rechecks can also occur, particularly if the dog travels well, is not "white coat hypertensive" and the distance is not too great. At the start of the second week the veterinarian can determine whether a "blood glucose curve" is needed although often there is no need if the diabetic dog is clinically doing well. At this stage, fructosamine determinations are not yet particularly contributory to monitoring as the duration of diabetes is usually less than three weeks and diabetes control has been in a flux state during this interval. Multiple daily urine glucose estimates by the home caregiver are no longer necessary although periodic urine monitoring is advisable.

Diabetes monitoring methods

Several methods of monitoring control of diabetes can be used and all have utility in certain circumstances. Logical use is best made with an understanding of the information provided.

Clinical Signs – Increased thirst and urination (PUPD), when present or persistent in a diabetic dog that does not have a concurrent condition such as renal insufficiency or hyperadrenocorticism is likely explained by insufficient control of diabetes (obligatory solute diuresis due to insufficient dose of insulin). Nocturia, particularly if not enuresis and not associated with other lower urinary tract signs, is also likely to be explained by hypoinsulinization of diabetes. Failure to regain regain weight or weight loss in a diabetic may be due to inadequate insulin dosing if caloric intake is appropriate and other disease is not present.

Urine Glucose Testing – Estimating urine glucose concentration is a useful method for home-monitoring of canine diabetes control. Many years ago, daily urine glucose testing was advocated for daily adjustments of insulin dose. This scheme for monitoring and treating canine diabetes proved unsatisfactory, however, and resulted in a general distrust of urine monitoring in general. More recently, the proper use and interpretation of urine glucose testing has re-established the utility of this technique. Urine glucose test results are not used for daily modulation of insulin dose. Instead, the results are used to detect trends in diabetes control and are used to adjust insulin dose only after 2-3 days assessment. It is important to recall that urine glucose concentration is a reflection of the blood glucose concentration and the glomerular filtration rate (GFR) during the interval that the tested urine was formed (usually 2-6 hours). The extent, during the urine formation interval, that tubular maximum for glucose resorption was exceeded is reflected in the in the urine glucose concentration. At usual GFR, the tubular maximum is not exceeded unless blood glucose concentration is about 180 mg/dl or greater. For this reason, lack of detectable glucose in urine does not necessarily indicate that hyperglycemia was absent during the time that tested urine was formed.

When multiple daily urine glucose determinations are performed, the pattern of glucosuria may provide useful information. Persistent glucosuria greater than 1000 mg/dl throughout the day, is clear evidence that the insulin dose needs to be increased whereas variation in urine glucose concentration from negative to 1000 mg/dl demands some analysis. Urine formed during the anticipated waning period of action after insulin administration, for example, may have urine concentration of 1000 mg/dl (1%) while urine formed earlier during peak insulin action may be negative. On the other hand, if urine glucose at peak insulin action is 500 mg/dl (1/2%) and 1000 mg/dl near the end of insulin activity, increased insulin dose is needed whereas in the first example, it is more likely that the duration of insulin action is slightly less than anticipated.

Urine Ketone Estimating – When the concentration of urine ketone bodies in glomerular filtrate exceeds tubular resorptive capacity, acetoacetate can be detected by KetoDiastix urine reagent pad analysis. Urine ketone bodies are often present in urine of diabetic dogs when initially treated with insulin at home. Resolution of ketonuria is usually accomplished in 3 or 4 days. Thereafter, ketones are seldom detected in urine unless the insulin dose is chronically less than required and diabetes is poorly controlled. An exception to this generalization is the occasional diabetic dog that is prone to form ketones in spite of adequate control of diabetes.

Blood Glucose Curves – The determination of blood glucose concentration at approximately 2 hour intervals for 12 to 24 hours has been termed a blood glucose curve. The blood sampling and determinations are usually done in the veterinarian's office. The results are used to gain a more complete understanding of the effects of each insulin administration and the overall control of diabetes. It has been documented, however, that the results are not as reliable as previously thought. Lack of reliability and relevance of the procedure is thought to be related to factors such as apprehension and excitement. Although validity of glucose curves can be improved by performing them 3 days in succession, overall patient care is disrupted and compromised. For these reasons, blood glucose curves are not routinely done on every diabetic but are reserved for problem cases. In general, the shape of the curve obtained is of greater validity than are the individual determinations.

Spot Blood Glucose Determinations – Individual blood glucose determinations have also been used to assess diabetes control but limitations must be kept in mind. Blood glucose may not be representative of the concentrations an hour before or after and the blood glucose nadir cannot be predicted. A previously determined blood glucose curve may be helpful in predicting nadir because the shape of the curve does tend to be reliable although the actual individual determinations are subject to lack of reliability.

Caregiver Home Blood Glucose Monitoring – It is increasingly common for caregivers to monitor blood glucose using commercially available glucometers used by human diabetics. Selection of the site for procuring the necessary drop of blood is highly individualized and dependent on the response of the dog and the preference of the caregiver. Sites include ear tips, footpads and elbow calluses.

Glucometers – Commercially available glucometers designed for home monitoring by human diabetics are readily attainable at a reasonable price. The instruments are quite accurate when used with dog or cat blood except at the lower range of blood glucose concentration (< 80 mg/dl). In this range, human glucometers tend to underestimate glucose concentration of cat and dog blood because of species difference in erythrocyte size. In the course of home monitoring diabetes, it is not common for the home caregiver to be dealing with hypoglycemia. Regardless, the glucometers tend to under estimate rather than over estimate blood glucose. Glucometer systems marketed for dogs and cats include AlphaTRAK™ by Abbott Laboratories and GlucoPet® and GlucoVet® by Animal Diabetes. Dog and cat glucometer systems are more expensive than human products.

Continuous Glucose Monitors – Glucose monitoring systems that provide interstitial fluid glucose determinations at 5-minute intervals for 3 days are also available. Systems such as Medronics' Continuous Glucose Monitoring system have been used in cats and dogs. A small, light-weight (less than 1 ounce) subcutaneous sensor with transmitter sends glucose readings at 5-minute intervals to a remote receiver that stores the data until retrieved for analysis. Calibrating glucose determinations from an external source must be performed and entered into the system every 12 hours. Although some study time must be devoted to gain familiarity with the system, remarkable data are generated for understanding the diabetic's response to insulin.

Fructosamine-An assessment of glucose control over the preceding weeks or months is possible by determining the degree to which body proteins have become non-enzymatically combined with glucose. Fructosamine is the established term foserum or plasma proteins that are glycated or glycosylated (non-enzymatically combined with glucose). Fructosamine determinations are useful for assessing glycemic control during the preceding 1-3 weeks and are commonly used by veterinarians to assess diabetes control in dogs and cats. Physicians more commonly use glycated hemglobin (hemglobin A1c) to assess glycemic control over the preceding 3-4 months.

Treating and managing non-ketoacidotic (uncomplicated) diabetes in cats

Approaches to treating diabetic cats usually differs from the approaches used in dogs because of differences in pathogenesis in the two species. Whereas dogs are usually Type I diabetics and require insulin administration for life, cats more frequently have Type II diabetes and the common therapeutic approaches used for cats reflect this difference. There is a 15 fold documented increase in the incidence of cat diabetes in the last 30 years and many investigators postulate that feeding this obligate carnivore relatively high carbohydrate diet may be the underlying cause. Cats surviving on rodent and bird prey consume negligible carbohydrates whereas cats fed commercial cat food consume substantially more carbohydrate and relatively less protein. Carbohydrates contribute 1-35% of calories in canned cat food and 25-60% of calories in extruded dry cat chow. High carbohydrate diets place an un-natural demand on insulin secretion and, coupled with a sedentary life, promote obesity. Obesity imparts peripheral insulin resistance that demands even greater insulin secretion than does high carbohydrate diet alone. Co-secreted with insulin is amylin (islet amyloid polypeptide) that, when hypersecreted, tends to form amyloid fibrils that are deposited in pancreatic islets. Deposited amyloid is injurious to insulin producing pancreatic beta cells. Furthermore, resultant hyperglycemia is toxic (glucose toxicity) to beta cells. This outlined sequence of events results in Type II diabetes that can transition to Type I diabetes. Although not yet intensively studied, some estimation of developmental stage can be approximated by determining circulating insulin and glucose concentration at the time of diagnosis. Even if and especially when high insulin suggests Type II diabetes, the combination of intensive insulin therapy and feeding low carbohydrate high protein diet is highly recommended because the diabetic state can be reversed in approximately 80% of diabetic cats.

Intensive insulin therapy is recommended to minimize glucose toxicity by striving for euglycemia. Two insulin products are preferred: the human-use approved insulin analog, glargine (Lantus®) or the feline-use approved PZI VET® insulin. Both insulin products have a long duration of action that approaches 24 hours when administered to cats. In spite of the prolonged action, it is recommended that twice daily administration be used to minimize fluctuations in systemic insulin action and achieve a near euglycemic state. The dose administered is that necessary to control clinical signs of PUPD, stabilize body weight, increase activity and maintain fructosamine concentration approaching 400 µmol/l. The dose necessary to achieve these goals is usually approximately 0.4 units/kg of body weight twice daily (0.8 units/kg total daily dose).

At the same time that insulin therapy is used to reduce glucose toxicity, the diet is changed to reduce endogenous insulin and amylin (islet amyloid polypeptide or IAPP) secretory demands. Reduction in amylin secretion minimizes amyloid fibril deposition in pancreatic islets and the subsequent deleterious effects on pancreatic beta cells. The most important aspect of diet change appears to be reduction in dietary carbohydrate irrespective of fiber and metabolizable protein content although maintenance of caloric needs is nearly impossible with out relatively high protein. In general, diet change goals are most easily attained by feeding canned foods, although some habitual "cruncher-nibbler" cats are disinclined to readily accept canned foods. At the time that cats are diagnosed with Type II diabetes, they are usually no longer obese and body weight stabilization rather than weight reduction is the usual goal.

Using this two-pronged approach of insulin and diet, the number of diabetic cats that revert to a non-insulin dependent state approaches 70%. Remission of diabetes generally occurs over a period of 1 to 4 months and many cats that have been diabetic for more that 2 years can revert to a non-insulin dependent state. Cats that have been diabetic for over 3 years rarely revert in spite of the combination of low carbohydrate diet and intensive insulin therapy. Failure to achieve a non-insulin dependent state is presumed to reflect lack of capacity for insulin secretion characteristic of type I diabetes.

A question that surfaces is this context is what defines low carbohydrate diet? In one study of treating diabetic cats, low carbohydrate was defined as 3.5g/100kcal metabolizable energy whereas moderate carbohydrate was defined as 7.8 g/100kcal. Both low and moderate carbohydrate diets were efficacious in achieving remission of diabetes when fed in combination with intensive insulin therapy but the low carbohydrate diet was substantially more efficacious. Feline DM® diet (Purina) in both dry and canned form contains less than 3.5 g/100kcal and feline MD® diet (Hills) in both dry and canned preparations contain only slightly more.

In as much as therapeutic insulin and diet may reverse type II diabetes in cats and the use of oral hypoglycemic agents (eg glipizide) is of marginal efficacy, the latter's use is discouraged and reserved for caregivers who refuse to administer insulin.

Remission of long-standing type II diabetes in cats treated with insulin in combination with feeding low carbohydrate diets prompts the question as to whether diabetes can be prevented by feeding low carbohydrate diets. Many have concluded that the answer is obvious enough that only low carbohydrate foods are fed their cats.

Diabetic ketoacidosis: methods of management

For the purpose of "triage," diabetic ketoacidosis (DKA) is defined by clinical status as well as clinical chemistries. Although any diabetic with ketone bodies detected in urine or serum and reduced serum bicarbonate or total CO2 with or without blood pH below the reference range is technically a diabetic ketoacidotic, only those cats and dogs that are clinically ill, not eating, inactive and dehydrated are intensively treated for diabetic ketoacidosis by the methods outlined below. Furthermore, many detailed protocols have been published, many of which are quite complex. Board certified veterinary emergency and critical care specialists who routinely treat DKA patients usually have an established protocol with which they are familiar and comfortable using. Although varying in complexity, substantial commonality exists in the underlying factors on which the protocols are based. Only one protocol for intravenous infusion of insulin (CRI) will be outlined here although some comments on other methods will be included.

Until the mid 1970s DKA patients, whether cat, dog or human being, were treated with relatively high dose IV boluses of regular insulin. Since that time IV bolus insulin administration has been replaced with IV infusions of relatively low doses of regular insulin and is currently termed constant rate of infusion (CRI), a term that is a bit of a misnomer in that the rate is modulated as needed rather than being constant. Low dose IM insulin has also been used but has the disadvantage of substantially less flexibility in dose adjustment and will not be outlined. In addition to type, route, and rate of insulin administration, consideration is also given to water deficits, electrolyte imbalances. acid- base status and anticipated contingencies.

Fluids – Most DKA patients are so profoundly dehydrated that circulating volume is reduced and must be initially treated. Accordingly, the first step is to treat aggressively with IV balanced electrolyte solution as if the patient is in shock. During this fluid resuscitation period, until pulse strength returns and capillary refill time normalizes, the nest action plans can be formulated. The total fluid deficit can be estimated and usually approaches a volume that is the equivalent of 10% of body weight. This calculated volume (less the amount administered for restoration of circulating volume) is scheduled to be administered over the next 24-48 hours.

Insulin – Regular insulin is used and can be simply added to the balanced electrolyte solution at the rate of 1.0 unit per 100 ml of IV fluids. This should not be done until after the initial fluid resuscitation of circulating volume is completed. Although several protocols suggest that an insulin containing fluid be administered independently of hydrating fluids, this simple one bag system is almost always satisfactory. Some protocols suggest that a few ml of the insulin containing solution be run through the IV line and discarded. This procedure is to saturate insulin binding sites in the plastic tubing and has been found to be not necessary. The insulin containing balanced electrolyte solution is then administered at a rate consistent with re-hydration scheduled over 24-48 hours.

Electrolytes - Balanced electrolyte solutions are usually supplemented with potassium because DKA patients invariably have total body potassium deficits irrespective of serum potassium concentration. Because potassium is the chief intracellular cation, serum concentration is not necessarily an accurate reflection of body deficit. For this reason a minimum of 20mEq of potassium, usually as the chloride salt, is added to each liter of re-hydrating fluid. More often, 40mEq/l is added and may be increased to 60-80mEq/l if hypokalemia is documented. During the course of insulin administration, serum potassium concentration sometimes is observed to decrease as potassium relocates intracellularly in response to insulin administration and correction of acidosis. Serum inorganic phosphorus concentration may similarly decrease and intravascular hemolysis may occur if serum phosphorus concentration decreases to 1.0mEq/l. In this case, half of the potassium may be supplemented as the phosphate and half as the chloride but all of the potassium should not be added as the phosphate because of the threat of hyperphosphatemia. During the course of insulin infusion and re-hydration, clinical suggestion of hypokalemia may be evident as developing skeletal muscle weakness during the time that blood glucose is decreasing but hypoglycemia is not present. If this clinical situation is confirmed with biochemical determinations, increased potassium supplementation is indicated.

Acidosis and Bicarbonate – It has been amply documented that acidosis in human DKA patients reverses satisfactorily with or without the addition of bicarbonate. There is no reason to anticipate that cats and dogs are likely to differ in this regard from human patients; for this reason bicarbonate supplements are not generally recommended. Probably because of the real threat of inducing paradoxical CSF acidosis, some veterinary sources still cite formulas for calculating bicarbonate replacement at a such a markedly reduced rate that a negative number of mEq is sometimes the result. Regardless, if bicarbonate is supplemented, the balanced electrolyte solution should not be one that contains calcium (eg Ringer's or Ringer's lactate).

Monitoring – During insulin CRI for DKA it is important to monitor response with glucose determinations done at 1-2 hour intervals. Glucometers are frequently used for this purpose and although they useful, it is worth noting that glucometers designed for human use tend to underestimate blood glucose concentration in the lower range (<80mg/dl), often by as much as 20mg/dl. Particularly when using protocols that specify that insulin CRI be done using a glucose containing fluid when blood glucose is in this lower range, intervals of confusing apparent hypoglycemia followed by hyperglycemia can result. For this reason, the use of glucose containing fluids for insulin CRI is to be discouraged. Rather, the rate of insulin administration can be simply reduced or discontinued. The hypothesis that the administration of insulin in a glucose containing fluid reverses ketosis more quickly has not been substantiated.

Transitioning – Making the transition from insulin CRI to the administration of intermediate or long acting insulin can be a challenge. Dogs and cats can be maintained on insulin CRI indefinitely but the goal is to discharge them from the hospital. The question is whether they are sufficiently stable that they can be weaned from a system of insulin administration that can be constantly adjusted to their hourly needs to method of insulin administration that can only be changed one or twice daily. Clearly the patient must show interest in food and actually start eating before the transition can occur. Whether the clinician has judged correctly can only be determined by making the change. Iatrogenic hypoglycemia, after all, can always be treated.