Monitoring diabetes mellitus in diabetic cats (Sponsored by Intervet Schering-Plough Animal Health)

Protocols for monitoring diabetic cats have evolved over the years. Random spot-checking a blood glucose concentration, although easy to perform, is unacceptable as the sole means of monitoring because it cannot accurately represent the cat's response to insulin. Therefore other methods of monitoring diabetic cats need to be employed.

The goals of diabetic regulation include the following:

1. Maintain the blood glucose level between 100 to 300 mg/dl over 24 hours. 1

2. Ameliorate the clinical signs of diabetes mellitus; polydipsia, polyuria, and polyphagia.1

3. Attain and maintain a healthy body weight.2

Before or during the initiation of insulin therapy, you must identify and resolve or control any other medical problems. I routinely evaluate retroviral status, serum biochemical and electrolyte panels, total T4 concentrations, complete blood count with differential, complete urinalysis, and urine culture. In addition, thoracic and abdominal radiographs are useful to determine any additional abnormalities that need to be addressed. I also find it useful to assess the pancreas via abdominal ultrasound and feline pancreatic lipase immunoreactivity (fPLI) testing. Establishing the baseline serum fructosamine concentration is also helpful for long-term monitoring.

Blood glucose curves

In my opinion the blood glucose curve remains the gold standard for determining effective diabetic regulation.1 Veterinarians have learned the importance of performing a blood glucose curve to establish the nadir, the duration of insulin response, and the variation of blood glucose concentration. In addition, the blood glucose curve is the best way to detect the Somogyi effect.3,4 The advent of glucometers requiring an extremely small blood sample, along with the ease of using the lateral ear vein for sample collection, have facilitated performing blood glucose curves in cats.

Traditionally, the curve begins at the time of insulin administration (time 0), and additional blood samples are taken every two hours for the next 10 to 12 hours. The ideal blood glucose curve starts and ends with blood glucose concentrations of 250 mg/dl with the nadir at 100 mg/dl and occurring five to six hours after insulin administration, based on a 12-hour dosing interval.4 In reality, all of these values can vary widely, underscoring the importance of performing a blood glucose curve—rather than spot-checking. Optimally, the difference between the nadir and the starting and ending blood glucose values should be as small as possible.4,5 If the difference remains large, this may indicate the need for a different type of insulin.

Hospitalized blood glucose curves

When I perform curves with a hospitalized cat, I prefer to have the client feed the cat at home and to administer the insulin at the hospital (after the time 0 glucose concentration is determined). Offer the cat additional food throughout the day, and obtain blood glucose samples two, four, six, eight, 10, and 12 hours later, if time permits. Alternatively, the client feeds and administers the insulin at home following the routine schedule, then drops off both the cat and its food for the day. Then take blood glucose samples two, four, six, eight, 10, and 12 hours after insulin administration. The latter scenario is occasionally what happens in my hospital if the client routinely administers the insulin one to two hours prior to the hospital opening. To keep things simple, I strive for minimal disruption of the client's and cat's routine. Figures 1a to 1d explain how to take a sample for a blood glucose curve.

Figure 1. How to take a sample for a blood glucose curve.

Utilizing the lateral ear vein has dramatically reduced the stress my in-clinic patients undergo. My staff and I can obtain a sample in the cage, while the cat is eating or sleeping, or while it is held on a lap. I use a 25-ga needle and find that warming and shaving the ear first are usually not necessary. I do not use an alcohol swab because the cats seem to dislike the odor and it adds additional time to the sampling.

The advantages of hospitalized blood glucose curves include:

1. You can conduct a physical exam, including determining the cat's weight.

2. You can observe the owner administering the insulin (this is of paramount importance early in the course of treatment).

3. Trained personnel obtain the samples with minimal disturbance to the cat.

4. You can observe the cat during its stay, administering glucose intravenously or Karo syrup orally, if necessary.

In some cases, accuracy can be the major drawback of the hospitalized blood glucose curve. The blood glucose values in some cats will always be falsely elevated because of excitement and stress. During other curves, blood glucose concentrations may be deceivingly low, presumably because the cat would not eat while hospitalized. These cats need glucose or dextrose supplementation and leave us questioning the reason for the low blood glucose concentrations—is it because they did not eat, or is the insulin dose too high?

In-home blood glucose curves

In-home blood glucose curves have become more popular in recent years because of the premise that they are less stressful to cats. The validity of these blood glucose curves is solely dependent on the client's ability to accurately and gently obtain the blood sample. In one study, the blood glucose concentrations for the same group of cats were actually higher at home than in the hospital, possibly because of client-induced stress or the cats' improved appetite at home.3

Careful client selection and training are crucial to the success of in-home monitoring.3 You must ensure that the client understands the optimal blood glucose range and how to respond when values are outside the range (they will know how to respond after consulting with a veterinarian). I believe that in-home monitoring offers the best opportunity to determine the true status of a diabetic cat, especially if you make it fun and educational for the client. However, remember that this process can become overwhelming, even for carefully chosen clients, and you must be sensitive to their needs and the cats' needs.

Other monitoring methods

Additional methods of monitoring diabetic cats include client observation and the testing of urine glucose, serum fructosamine, and glycosylated hemoglobin levels.

Client observation. Teaching clients to observe notable parameters is but one of the many tools used to monitor diabetes. It is the most important tool if you use glargine insulin because of the rapid (within one week) and dramatic decrease in water consumption, possibly correlating to hypoglycemia.1 As the cat becomes regulated, the client should notice a return to normal water consumption (less than 80 ml/kg every 24 hours),2 less urine in the litter box, less hunger, and an increase in—or stabilization of—body weight. In my experience, the cat also acts more like its old self and is more even-tempered, both at home and in the hospital. It is often helpful for the client to record these parameters in a daily journal. By stressing the importance of observation, clients can quickly identify changes and alert their veterinarian.

Urine glucose. It's important for clients to follow urine glucose and ketone concentrations at home every one to two weeks. The renal glucose threshold in cats ranges from 220 to 300 mg/dl.1 Most well-controlled diabetic cats will have trace or 1+ glucose in their urine. If urine glucose is negative for 48 hours, the cat may be in remission.1 In my opinion, if ketones, at any level, are present, the client should seek medical attention. The dose of insulin should never be adjusted based solely on the results of urine glucose testing.5,6,7

Serum fructosamine. Fructosamine is a glycosylated serum protein and a more specific indicator of pancreatic endocrine activity than glucose. The reaction between glucose and this protein is irreversible; therefore, it represents the average blood glucose concentrations over a one- to three-week period.6,8 While reference ranges for serum fructosamine concentrations vary by laboratory, it is common that values less than 450 μmol/L represent moderate to good long-term glycemic control, while values less than 400 μmol/L indicate good to excellent control.4 You should evaluate relative changes in serum fructosamine concentrations, as well as absolute values. Because fasting and stress do not influence the serum fructosamine concentrations, monitoring them is a viable option for evaluating excited cats.1 The hazard of adjusting the insulin dose based only on serum fructosamine concentrations (with or without urine glucose concentrations) is that both of these parameters will remain elevated in the face of a Somogyi effect (i.e., hypoglycemia-induced rebound hyperglycemia).

Elevated serum fructosamine may also be caused by concurrent illness, especially infection, inflammation, and pancreatitis; however, the most common cause is client error (i.e., the inability to correctly administer insulin).9

Some veterinarians prefer utilizing serum fructosamine concentrations, owner observations, and midday blood glucose concentrations, to utilizing blood glucose curves in established, stable diabetic cats.4

Glycosylated hemoglobin. Glycosylated hemoglobin is an indicator of the average blood glucose concentration that erythrocytes were exposed to in their lifetime.8 It is utilized in human medicine to monitor glycemic control over a three- to four-month period, but its use in cats is questionable because it is more difficult to measure and interpret.4

Persistently hyperglycemic cats

The underlying reasons that some diabetic cats remain persistently hyperglycemic fall into three categories:1,9,10

1. Operator or case management error

2. Insulin resistance

3. The Somogyi effect.

Veterinarians must first explore category one. Is the insulin outdated or has it been inappropriately handled? Is the client giving injections correctly—or at all? Is the client using the correct syringe? (The client's insulin administration technique should be re-evaluated in-clinic at this time.) If the answers to these questions are satisfactory, then move to category two.

The Somogyi effect is due to insulin overdose. This can be tricky to differentiate from insulin resistance, because blood glucose values can remain consistently high for two to three days after the hypoglycemic episodes. In the classic Somogyi curve, the time 0 blood glucose concentration is very high, the nadir blood glucose concentration is generally less than 50 mg/dl (2.7 μmol/L), and the 12-hour blood glucose is very elevated, often higher than the time 0 measurement.11 If the Somogyi effect is confirmed by the blood glucose curve, you can decrease the insulin dose to 0.3 to 0.5 U/kg for 10 to 14 days, and then gradually increase it while monitoring the blood glucose curve.2

Other differential diagnoses of insulin resistance include concurrent disease or medications, such as corticosteroids or progestational substances.1,11,12,13 The most common causes of insulin resistance I find are infections of the urinary tract, oral cavity, and skin. Hyperthyroidism, renal insufficiency, pancreatitis, neoplasia, hyperadrenocorticism, acromegaly, and hyperlipidemia are also recognized as causing insulin resistance.1,4,11 In addition, obesity can be the sole cause for insulin resistance.4 It is essential that you perform the necessary diagnostic tests. In one study, 12% of diabetic cats (7 out of 57) had bacterial urinary tract infections, based on a urine culture.14 Interestingly, only two of those seven cats showed signs consistent with lower urinary tract disease, and only five had bacteria in the urine sediment. This emphasizes the importance of performing routine urine cultures, regardless of glycemic control.

Regulating newly diabetic cats can be challenging for many reasons, especially with concurrent inflammatory disease. Other common reasons include poor patient cooperation and client compliance. One of my most challenging cases involved all of these factors.

Case study

Scamper, a 14-year-old spayed female domestic shorthaired cat, was presented to my hospital for her routine annual physical exam. The client casually mentioned that she had noticed Scamper drinking more water, compared with her other cats. The physical exam was unremarkable except for grade 3 out of 4 dental disease—which I later expected would present a challenge in attaining glycemic control—as well as a dull coat with dander. This formerly obese cat (body weight of 13.4 lbs one year earlier) was now thin and weighed only 10.7 lbs.

I performed a serum biochemical profile, an electrolyte panel, a complete blood count with a differential, and a complete urinalysis. Heartworm antibody, feline leukemia virus antigen, and feline immunodeficiency virus antibody tests were negative. I did not obtain a pretherapy serum fructosamine concentration or radiographs. Abnormal biochemical results are listed in Table 1

Table 1. Abnormal laboratory results before initiating insulin

Scamper was known to be fractious. In addition, the client was unable to perform an ear prick at home. Faced with these challenges, I initiated treatment with insulin glargine, hoping that Scamper would quickly go into remission. I started insulin glargine at 0.5 U/kg (two units) subcutaneously every 12 hours. I instructed the client to watch the cat closely over the next two weeks for a decrease in water consumption and to schedule a blood glucose curve at that time.

With eight other cats, the client admitted difficulty knowing what volume of urine was produced or water consumed by any individual cat. The only reason she noticed polydipsia initially was because she began to see Scamper at the water bowl multiple times daily, which she had never previously witnessed. Scamper also drank a larger volume of water relative to the other cats.

Three weeks after initiating insulin, the client, paying closer attention to Scamper, had not observed any difference in her thirst, or perceived urine production (based on the collective amount of urine). On my recommendation, the client brought Scamper into the clinic for a blood glucose curve. Scamper was given two units of insulin at 6 a.m., so a time 0 blood glucose value was not obtained. Blood glucose values were obtained over the next 12 hours (Figure 2).

Figure 2. Initial blood glucose curve

The values in the curve startled me, and I administered 5 cc of Karo syrup orally after evaluating each blood glucose test result. This likely contributed to the 12 hour value of 117 mg/dl. Was the insulin dose too high, or had Scamper gone into remission? I discharged her and discontinued the insulin. Three days later, Scamper's blood glucose during her curve was between 250 to 300 mg/dl. Consequently, I restarted insulin at one unit every 12 hours.

Two weeks later, a blood glucose curve was performed over a 12.5-hour period, minus a time 0 blood glucose value. (Figure 3). Scamper did not eat well in the hospital and was given 5 cc of Karo syrup orally at time 12 hours. Based on the glucose values obtained, I felt certain that Scamper was now in remission and discontinued the insulin.

Figure 3. Follow-up blood glucose curve

One month later, the client again witnessed Scamper drinking a large volume of water. Two blood glucose samples, at 10 a.m. and 1 p.m., were obtained. The values were 388 mg/dl and 571 mg/dl, respectively. Her serum fructosamine concentration was 489 μmol/L. Scamper was started back on 0.5 units of insulin glargine every 12 hours. Two weeks later, the cat's blood glucose curve remained relatively flat, ranging from 453 mg/dl to 520 mg/dl. Consequently, I increased the dose to one unit every 12 hours. Three weeks later, another blood glucose curve was performed. The glucose concentration never decreased below 300 mg/dl and was 464 mg/dl at its highest. Scamper's insulin dosage was increased to two units every 12 hours.

Three weeks after the increase to 2 U twice a day, the client wasn't seeing Scamper at the water bowl, and the cat's behavior seemed normal. A blood glucose curve was performed at that time (Figure 4). We concluded that the blood glucose concentration nadir (60 mg/dl) was falsely low because the cat stopped eating while in the hospital. We also concluded that the length of time to the nadir (eight hours) was accurate because insulin glargine is known to have a slower onset of action than short- or intermediate-acting insulins.1 Three weeks after this curve, Scamper's serum fructosamine concentration was 401 μmol/L, which suggested fair glycemic control.

Figure 4. Follow-up blood glucose curve

It took four months to attain minimally acceptable glycemic control in this patient. In retrospect, after two months without approaching optimal regulation, I should have changed the insulin to protamine zinc insulin or lente insulin (Vetsulin—Intervet) to attain better glycemic control. In addition, the cat's concurrent dental disease, which was later eliminated, most likely contributed to the difficulty in interpreting the monitoring parameters and in achieving regulation.

Scamper subsequently became more fractious—and the client more anxious—with each visit to the hospital, even for routine exams. Because of this and my concern for both staff and patient safety, blood glucose curves were no longer attempted. Client observations along with a serum fructosamine concentration and physical exam every three to four months, while not optimal, have allowed me to adequately monitor Scamper.

Summary

In my opinion, the blood glucose curve remains the gold standard for monitoring diabetic cats and is critical during the first one to four months for newly diagnosed diabetic cats. Training clients to observe pertinent home behaviors and to monitor their cat for changes in these behaviors is also an integral part of monitoring.9 It can take one to four months to stabilize a newly diabetic cat, so be patient and avoid over-adjusting the insulin.2 Checking the serum fructosamine concentration every three months, along with a physical examination (including determination of body weight), and midday blood glucose concentration checks may be adequate for monitoring purposes in established, stable diabetic cats.4

For successful diabetic management, you must identify—and resolve —any concurrent diseases. Even in the absence of clinical signs, urine cultures are advised to semiannually identify occult urinary tract infections.

The best method of monitoring a diabetic cat depends on both the cat and the client. Not every cat can be monitored in an identical fashion. Tailor the method to the cat and the client. You must also use good clinical judgment when interpreting any test. Do the numbers fit with what you are seeing and how the patient is behaving?

Remember: Keep it simple. Even though diabetes is a complicated disease, most clients need simplicity. While this principle may not result in ideal regulation and maintenance, it may be adequate and result in less patient loss due to owner concerns about the ongoing management of diabetes.

References

1. Zoran D. Feline diabetes: What's up with insulin, in Proceedings. American Association of Feline Practitioners, Fall 2005 Conference.

2. Rand J. Understanding feline diabetes and its management: Part II, in Proceedings. ACVIM, 2002.

3. Reusch C, Kley S, Casella M, et al. Home monitoring of the diabetic cat. J Feline Med Surg 2006;8:119-127.

4. Greco D. Methods for monitoring diabetic cats, in Proceedings. Western Veterinary Conference, 2002.

5. Greco D. Monitoring diabetics, in Proceedings. Atlantic Coast Veterinary Conference, 2002.

6. Sirois M. Monitoring insulin therapy, in Proceedings. ACVIM, 2003.

7. Schaer M. Monitoring glycosuria in diabetic cats, in Proceedings. Western Veterinary Conference, 2003.

8. Sirois M. Advances in clinical chemistry, in Proceedings. ACVIM, 2002.

9. Feldman E. Successful management of non-responsive diabetic cats, in Proceedings. Waltham Feline Medicine Symposium, 2002.

10. Schermerhorn T. Treatment of diabetes mellitus in dogs and cats. Clinician's Brief 2008;6:35-39.

11. Wilson S. Management of feline diabetes and the Somogyi phenomenon. VIN AAFP Rounds, July 10, 2005.

12. Melendez L, Lorenz M. Feline insulin resistance and diabetes mellitus, in Proceedings. Western Veterinary Conference, 2002.

13. Nelson R. Poorly controlled diabetes in a cat. Clinician's Brief 2007;5:68-71.

14. Mayer-Roenne B, Goldstein RE, Erb HN, et al. Urinary tract infections in cats with hyperthyroidism, diabetes mellitus and chronic kidney disease. J Feline Med Surg 2007;9:124-132.

Additional Resource:

markNtime. How to test your diabetic cat's blood sugar at home. Available at: www.youtube.com/watch?v=_zE12-4fVn8. Accessed March 17, 2008.