SUMMARY OF OBSERVATIONS.

Owner observations were available from 62 cats diagnosed as having FCS (Table 7-5). The most common concerns noted in the histories obtained from these owners was PU/PD in 51 cats (82%), polyphagia in 42 cats (68%), weight loss or failure to gain weight in 32 cats (52%), and lethargy or reports of “sleeps more” in 29 cats (47%). Concerns about skin problems were noted 61 times. These concerns ranged from the extremely worrisome fragile or torn skin noted in six cats, to a failure to grow hair after it had been shaved (usually for venipuncture or abdominal ultrasonography examination). There were also concerns over grooming excessively or not enough and concerns about a hair coat that had become unusually coarse. Multiple skin problems were noted in some cats, whereas others had none. The owners of 21 cats with PDH (from the total of 50, or 42%) and the owners of 3 cats with ATH (from the total of 12, or 25%) did not observe any problem relative to the skin or hair coat. Similar owner observations in cats with PDH as compared with those with ATH underscores the final common denominator of disease in these cats, that is, chronic exposure to excesses in circulating cortisol.

TABLE 7-5 OWNER OBSERVATIONS IN CATS WITH NATURALLY OCCURRING HYPERADRENOCORTICISM (TOTAL OF 62 CATS)

It has been suggested that clinical signs of hyperadrenocorticism are not commonly detected by owners or veterinarians until cats with FCS develop diabetes mellitus. The results of this review suggest that this is true for a majority of afflicted cats. However, some of the signs in these cats did precede development of diabetes mellitus, and other cats never developed diabetes. Thus a suspicion of FCS may result from a history and physical examination in nondiabetic cats. Such abnormalities as thin skin, skin fragility, “potbelly” appearance, muscle atrophy, weakness, PU/PD, and various hair coat disorders might lead to the diagnosis of FCS in nondiabetic cats.

LIVER ENZYMES.

The introduction to this section discusses the common routine biochemical test result abnormalities from dogs with naturally occurring or iatrogenic Cushing’s syndrome. The results from cats with this endocrine disorder are strikingly different (Table 7-8). Although dramatic increases in serum alkaline phosphatase (SAP) activities (average >1,000 IU/L) are the most common biochemical abnormality in dogs, only 3 of 43 cats (7%) with naturally occurring Cushing’s syndrome had an increase in serum alkaline phosphatase activity. Another common abnormality in dogs with Cushing’s syndrome is mild to moderate increase in serum alanine aminotransferase (ALT) activity. This abnormality is identified in more than 50% of dogs, but occurred in only 12 of 43 cats (28%). Several of the cats with increases in ALT activity had significant hepatopathies that were not related to their Cushing’s syndrome. Dogs with iatrogenic or naturally occurring excesses in serum cortisol concentration commonly develop “steroid hepatopathy,” as well as significant circulating concentrations of the “steroid-induced” isoenzyme of SAP. Neither of these typical canine features of Cushing’s syndrome is believed to occur in cats or human beings. However, “steroid hepatopathy” that included liver enlargement and a vacuolar appearance to hepatocytes has been occasionally described in cats with FCS.

TABLE 7-8 SERUM BIOCHEMICAL AND URINALYSIS RESULTS FROM CATS WITH NATURALLY OCCURRING HYPERADRENOCORTICISM (43 CATS)

SERUM CHOLESTEROL AND T₄.

Increases in serum cholesterol concentration are identified in 60% to 70% of dogs with Cushing’s syndrome. Fewer than 5% of those dogs have diabetes mellitus. In contrast, the serum cholesterol concentration was increased in only 13 of the 43 cats with FCS (30%), and all 13 had concurrent diabetes mellitus. Diabetes mellitus and hyperadrenocorticism are two of the classic causes for increases in serum cholesterol concentration. Excess serum cortisol concentrations also cause negative feedback to the pituitary, thereby decreasing TSH secretion and resulting in secondary hypothyroidism. Hypothyroidism is another classic cause of hypercholesterolemia. Hypothyroidism, together with the direct lipolytic actions of glucocorticoids exaggerated in Cushing’s syndrome, are common explanations for the increases in serum cholesterol concentration typically identified in nondiabetic dogs with hyperadrenocorticism. However, serum T₄ concentrations were lower than the reference range in only 4 of the 43 cats (9%) with FCS, making this explanation for hypercholesterolemia unlikely. Cats with FCS were consistently euthyroid, and evidence of thyroid disease was identified in only three cats (all were hyperthyroid) at necropsy. Although thyroid disease remains possible in any cat, it would not be considered typical of cats with FCS.

BUN, SERUM CREATININE, URINE SPECIFIC GRAVITY.

Perhaps one of the more consistent group of abnormalities seen in dogs that have Cushing’s syndrome are those related to polyuria and polydipsia. A tremendous majority of these dogs have urine specific gravities <1.020 (especially on samples obtained by owners from dogs in their home environment), many are <1.014 and about 30% are <1.008. In other words, polyuria in this population of dogs is common and represents a frequent owner concern. Polyuria is one of the two most common reasons that owners seek veterinary care for their dog (the other is alopecia). If not resolved, polyuria is a frequent reason for owners to elect euthanasia. In addition to the PU/PD and the previously discussed isosthenuria/hyposthenuria in dogs with Cushing’s syndrome are low-normal-to-low BUN and serum creatinine concentrations. These are classic features of canine Cushing’s syndrome. Emphasizing this concept, increases in BUN or creatinine not only are rare in dogs with Cushing’s syndrome but are a serious indication for avoiding treatment. Appetite in such dogs may be dependent on cortisol excess, and the concern in treating such dogs for Cushing’s syndrome is unmasking both renal failure and its clinical signs (including poor appetite).

A review of the serum biochemical data presented in Table 7-8 from cats with naturally occurring FCS demonstrates remarkable differences in test results as compared with those from dogs that have Cushing’s syndrome. None of the 43 cats had a BUN concentration below the reference range, and only 1 of the 43 had a serum creatinine concentration below the reference range. Only 1 of 43 cats had a randomly obtained urine specific gravity <1.020. That cat had both isosthenuria and renal failure. Furthermore, 23 of the 43 cats (53%) had abnormally increased BUN concentrations at the time of diagnosis, and 14 of those 23 cats (14 of the total 43 = 32%) also had abnormally increased serum creatinine concentrations. Thus it is likely that the polyuria and polydipsia recognized in most cats with FCS is either due to diabetes mellitus, chronic renal failure, or a combination of both conditions.

There have been descriptions of cats with FCS that have PU/PD but no clinical or laboratory evidence supporting concurrent renal failure or diabetes mellitus. The number of such cats is unknown, but the concept is that hypercortisolemia does occasionally cause PU/PD in some cats. On the one hand, it is possible that this does occur. On the other hand, we remain suspicious that some of these cats may have periodic hyperglycemia leading to episodic glycosuria and, thereby, episodic secondary polyuria. Stress-induced hyperglycemia is one explanation for hyperglycemia noted in some nondiabetic cats. This stress-induced condition interferes with our ability to distinguish those cats with persistent euglycemia except at times of stress from those with periodic physiologic hyperglycemia due to their hypercortisolemia. Even use of glycosylated hemoglobin or fructosamine concentrations would not consistently identify cats with periodic hyperglycemia. There is no evidence, however, to suggest that any cat with FCS has a physiologic syndrome similar to the diabetes insipidus–like condition that occurs commonly in dogs with Cushing’s syndrome, since such a situation would cause dilute urine. Cats with FCS rarely have dilute urine.

SERUM POTASSIUM AND SODIUM.

Although cats with FCS are often described by their owners as being weak, the weakness is likely a direct result of glucocorticoid-induced catabolic effects on muscle. One suggested concern was that these cats may be predisposed to hypokalemia. However, only 1 of 43 cats had a serum potassium concentration lower than the reference range, and only 2 had serum sodium concentrations below the reference range.

PROTEINURIA, SERUM CALCIUM, ALBUMIN, GLOBULINS, AND TOTAL PROTEIN.

Perhaps related to the frequent incidence of chronic renal failure in cats with hyperadrenocorticism, 7 of the 43 cats (16%) were mildly hypocalcemic (based on total serum calcium concentrations). However, each hypocalcemic cat also had mild decreases in serum albumin concentration. Proteinuria was common in this population of cats (26 of 43 cats, or 60%), and this is the most likely explanation for both their hypoalbuminemia and hypocalcemia. However, only 8 of 43 cats (17%) were hypoalbuminemic, and all 8 had serum albumin concentrations >2.2 g/dl. It is also of interest to point out that only 2 of 43 cats with FCS had serum albumin concentrations greater than their simultaneous serum globulin concentration. Furthermore, 22 of the 43 cats (52%) had serum globulin concentrations above the reference range. This hyperglobulinemia could be explained by a normal response to chronic antigen exposure, which occurs in any older individual. However, this degree of hyperglobulinemia was such that 9 of the 43 cats (22%) had hyperproteinemia.

URINARY TRACT INFECTION.

Urinary tract infection was uncommon among the cats with FCS (2 of 43 cats, or 3%). Since many had been referred after being given antibiotics, perhaps the failure to identify infection was a result of this treatment. Once again, however, this is in contrast to our experience with dogs that have naturally occurring hyperadrenocorticism. In those dogs the incidence of infection is much greater, and most are also referred after being treated with antibiotics.

BLOOD GLUCOSE CONCENTRATIONS.

Perhaps the most unpredictable routine biochemical test result, as compared with expectations, was the random blood glucose concentration. Since a majority of cats diagnosed with FCS have diabetes mellitus, increases in blood glucose concentration are expected. However, of the 43 cats with serum biochemical test results available at the approximate time of diagnosis, 10 cats were not diabetic. Of the 33 cats with diabetes mellitus whose serum chemistry results were available for inclusion in Table 7-8, 32 were being treated with exogenous insulin at the time of FCS diagnosis. It is not surprising, then, that six of the insulin-treated cats demonstrated hypoglycemia on random blood glucose testing. It is somewhat surprising, however, that only 2 of all 43 cats with FCS had blood glucose concentrations within reference limits. Furthermore, one of those two cats was diabetic and had received insulin prior to being tested that day. Ten of 11 FCS cats that had not been treated with insulin, therefore, demonstrated hyperglycemia when blood chemistry results were randomly obtained during the hospitalization in which the FCS was diagnosed. Therefore it does seem reasonable to suggest that finding a blood glucose concentration within reference limits is unusual in cats diagnosed with Cushing’s syndrome.

HYPOGLYCEMIA VERSUS INSULIN DOSE.

Six of 43 FCS cats evaluated with a routine biochemical profile were hypoglycemic at the time that blood was obtained for the chemistry profile. All six hypoglycemic cats were diabetic and had received insulin prior to testing that day. Four of those six cats were described by their owners and referring veterinarians as “insulin-resistant.” Insulin dose (per cat or per kilogram of body weight) varied tremendously. The occasional finding of hypoglycemia in any population of insulin-treated diabetic dogs or cats should never be viewed as unexpected. However, we believe that repeatedly increasing insulin dosage (as recommended by some veterinarians) in an attempt to “control” hyperglycemia in diabetic cats is dangerous. It is our experience that individual insulin doses in excess of 2.2 U/kg of body weight are unsafe. Although we do not doubt that some cats respond to extremely high insulin doses after appearing resistant to lower doses, “resistance” (if it exists in any particular cat) is not a continuum. In other words, insulin resistance seems to fluctuate, or “wax and wane.” If this is true, the dose given when a cat appears resistant will be an overdose in that same cat at another time. We have had cats with FCS that had severe, life-threatening, hypoglycemic reactions to insulin. It is not recommended that any cat be given insulin doses in excess of 2.2 U/kg body weight.

HYPERGLYCEMIA.

Of the 43 cats with FCS whose laboratory data were included in Table 7-8, 35 were hyperglycemic at the time that blood was obtained for the chemistry profile. Twenty-six of the 35 hyperglycemic cats were diabetic, and 25 of the 26 were being treated with insulin and had received insulin that day. Since cats with FCS are commonly believed to be insulin-resistant, the finding of hyperglycemia would be expected in some of these cats. Resistance is not an acceptable explanation for all these hyperglycemic cats, however. An additional reason for the hyperglycemia could be inadequate control of the diabetes (with numerous causes ranging from owner error to FCS). Another possible explanation is that blood was obtained before or after the period of insulin effectiveness. Finally, any of these cats may have had stress-induced hyperglycemia in addition to their diabetes mellitus.

Nine of the 35 hyperglycemic cats were not diabetic. How many of the nine cats had simple stress-induced hyperglycemia at the time that blood was obtained? Of equal interest, how many of the nine hyperglycemic cats had some degree of glucose intolerance that had not been recognized as representing diabetes mellitus? The answers to these questions are unknown, but it is likely that at least some cats in this population fit into one of these two categories.

DIABETES MELLITUS.

Diabetes mellitus is among the more common conditions diagnosed in small animal practice, and it is, along with hyperthyroidism, one of the two most frequently diagnosed endocrine disorders in cats. In contrast, the diagnosis of feline hyperadrenocorticism, or FCS, is considered uncommon-to-rare. We cannot state that FCS is diagnosed only in diabetic cats with insulin resistance. To the contrary, FCS is diagnosed in nondiabetic as well as diabetic cats. It seems easier to suspect FCS in an insulin-resistant cat, however, than in a cat that is simply lethargic or pot-bellied. However, clinical signs of thin skin, skin fragility, alopecia, failure to regrow hair after it has been clipped, or PU/PD are also problems that could reasonably be explained should the cat be confirmed to have FCS.

One problem facing the veterinary practitioner is determining which poorly controlled diabetic cats are truly insulin-resistant and which of those truly insulin-resistant cats have FCS. Many of our diabetic cats have stress-induced hyperglycemia when examined at veterinary hospitals. Most of our poorly controlled diabetic cats have problems that can be traced to one or more of many potential owner errors. Although monitoring serum glycosylated hemoglobin or fructosamine concentrations may aid in identification of such cats, these tests are not perfect.

An additional group of poorly controlled diabetic cats are given insulin that has an extremely brief duration of action; these cats continue to exhibit clinical signs and have frequent hyperglycemia, since it is more likely to detect hyperglycemia than euglycemia at any given time of day in such cats. Another category of causes for poor control, or insulin resistance, in a diabetic cat includes many concurrent conditions that interfere with insulin action. These include pancreatitis, other nonseptic inflammatory conditions, infections, and other disorders (such as neoplasia and heart disease). Finally, problems such as acromegaly, anti-insulin antibodies, and hyperadrenocorticism are uncommon but popular conditions to diagnose.

We encourage veterinarians having difficulty controlling diabetes in any cat to prioritize the potential explanations for difficult control. The top of this priority list should be the fact that in-hospital testing may not reflect what is happening in the home environment. Owner history is the single most important monitoring tool in the long-term management of diabetes in cats, and if the owner is satisfied with the cat’s response to therapy, in-hospital test results should not supersede this opinion unless hypoglycemia is documented. In-hospital stress-induced hyperglycemia is common, and this differential diagnosis should not be limited to “mean” (or obviously stressed) cats.

Next, veterinarians must consider all the potential owner errors that commonly take place. These errors include, but are not restricted to, using the incorrect insulin, mixing the insulin improperly, drawing insulin into the syringe incorrectly, using incorrect syringes, using out-dated insulin, and practicing improper administration techniques. Every time a diabetic cat is evaluated in the hospital, owners should be watched as they handle and administer their insulin. Once these potential problems have been dismissed, the veterinarian should consider the possibility that the insulin being used is either too weak (such as sometimes occurs with ultralente or PZI insulins) or too potent (which sometimes occurs with NPH or lente insulins) for the cat being evaluated. Additionally, one should consider the possibility that a concurrent disorder exists and may be interfering with insulin sensitivity. This could include any infection (urinary tract and skin infections are only two of the numerous potential infections that may or may not be obvious) or any source of inflammation (pancreatitis is just one of numerous potential inflammatory conditions that may or may not be obvious). Additional conditions that could cause insulin resistance include neoplasias, heart disease, and others. Last among the considerations of causes for insulin resistance should be anti-insulin antibodies, progesterone excess, acromegaly, and FCS.

HYPERTENSION.

Blood pressure assessment is not commonly obtained from cats, and we have little data regarding the incidence of hypertension among cats with iatrogenic or naturally occurring hypercortisolism. However, hypertension is quite common among human beings with naturally occurring hyperadrenocorticism. Since there are more similarities between human beings and cats with this condition than between dogs and cats, we suspect that hypertension may be a common sequela to FCS. Long-term effects of hypertension in cats is also not well studied, but it is possible that chronic excesses in serum cortisol may have deleterious effects on the kidneys of cats.

BACKGROUND.

The urine cortisol-to-creatinine ratio (UC:CR) has been recommended as a screening test to aid in separating patients with Cushing’s syndrome from those without the condition. A thorough explanation of the test and the indications for its use are provided in Chapter 6. In summary, the “gold standard” screening test for Cushing’s syndrome in human beings has been to assess total cortisol excreted in urine over a 24-hour period. Repeating this test several times further ensures validity of results. Cortisol excreted in urine reflects that secreted by the adrenal glands over time and negates concern of minute-to-minute pulsatile fluctuations in plasma concentrations. It is considered a reliable test for understanding whether the adrenal glands are producing normal or excessive quantities of cortisol.

The use of a randomly collected single urine sample, assessing it for cortisol concentration, and then creating a ratio of that cortisol to urine creatinine (UC:CR) is a short-cut to the more cumbersome 24-hour collection procedure. Although the UC:CR is much easier for patients and physicians, most physicians continue to use the 24-hour protocol. The UC:CR has been well studied in dogs. It is highly recommended by some veterinary endocrinologists but not by others because the test has both positive and negative traits. Those who recommend the UC:CR test usually point to several attributes: it is inexpensive, easy to perform and interpret, and sensitive (there is a high predictive value of a negative test result). Virtually all veterinarians who have critically evaluated the UC:CR in dogs with naturally occurring Cushing’s syndrome agree that it is highly sensitive. In other words, a tremendous percentage (in the range of 97% to 99%) of dogs with the disease have an abnormal test result. Therefore if one is suspicious of Cushing’s syndrome in a dog and the UC:CR test result is abnormal, there is a possibility that the dog has Cushing’s syndrome.

However, the problem that others note with the UC:CR is that finding an abnormal result is not specific for dogs with Cushing’s syndrome. For example, the UC:CR result is frequently abnormal in dogs with polyuria and polydipsia of any cause, such as those with diabetes mellitus, diabetes insipidus, renal failure, hypercalcemia, pyometra, and liver failure. Therefore, results of the UC:CR test may be consistent with a diagnosis of Cushing’s syndrome in animals that do not have the disease.

PUBLISHED DATA.

In the initial study evaluating the UC:CR in cats, it was suggested that there was a paradox in comparing reference ranges for dogs versus those for cats because of relative rates of urinary cortisol excretion when comparing the two species. The paradox is related to the relatively high reference UC:CR range in cats versus their low urinary excretion of administered radiolabeled cortisol. Healthy dogs excrete larger amounts of cortisol into urine than do healthy cats. It was suggested that the higher UC:CR in healthy cats, despite their lower overall excretion, may be explained by a higher glomerular filtration rate and/or a lower renal reabsorption of free cortisol. Thus the ratio between unaltered cortisol to conjugates and metabolites of cortisol in urine would be higher in cats than dogs (Goossens et al, 1995).

It is also wise to review the concerns that many veterinary clinical researchers have regarding specificity of UC:CR testing when it is used as the sole diagnostic aid for confirming Cushing’s syndrome in cats. The results from one study indicate that health status of cats affects UC:CR (Henry et al, 1996). Test results in healthy cats were not affected by age, gender, or neuter status. Ill cats have higher UC:CR values than healthy cats, implying that specificity may be a problem (as in dogs). The UC:CR from 16 ill cats were evaluated. Three cats had test results consistent with hyperadrenocorticism (≥3.6 × 10⁻⁵), and another seven had results in the range we would consider “borderline” (i.e., between 1.0 and 3.6 × 10⁻⁵. Thus 10 of 16 ill cats had UC:CR results that could be considered consistent with a diagnosis of hyperadrenocorticism. However, although the study demonstrated that sick cats had significantly higher UC:CR results than healthy cats, it did not include any results from cats with naturally occurring Cushing’s syndrome, negating direct comparison of data from each of the three groups in question (healthy, ill, and FCS cats; Henry et al, 1996).

We are not aware of any additional studies evaluating specificity of the UC:CR in cats. Thus it is premature to suggest that the test should not be used. To the contrary, we do recommend use of the test, especially if owners can bring urine samples from home (perhaps using small quantities of litter or nonabsorbable litter to make urine easier to collect). Home-collected urine negates concern of spurious results due to the stress associated with in-hospital urine collection. Also, there is less concern about UC:CR results from ill cats than results from cats confirmed as not having hyperadrenocorticism but having some clinical signs suggestive of the disease. It is this group of cats that must be evaluated before concrete suggestions are made regarding use of the UC:CR.

TEST INTERPRETATION.

Two studies have been published suggesting reference ranges for feline UC:CR. The reference range from the first report was <3.6 × 10⁻⁵ (Goossens et al, 1995), and from the second report it was <2.8 × 10⁻⁵ (Henry et al, 1996). If the initial reference range for cats (<3.6 × 10⁻⁵) is used, it should be remembered that this value is slightly higher than that for dogs (<1.0 or <1.3 × 10⁻⁵). It seems prudent to suggest, from the information published in these two reports coupled with our experience, that a UC:CR value ≥3.6 × 10⁻⁵ in a cat with clinical signs consistent with hyperadrenocorticism (FCS) is consistent with that diagnosis. We also recommend that UC:CR values between 1.3 and 3.6 × 10⁻⁵ in cats with appropriate clinical signs be considered “borderline.” It is possible for a cat with test results in this range to have naturally occurring hyperadrenocorticism.

RESULTS.

As can be seen from the data presented in Tables 7-9 and 7-10, we were able to collate UC:CR data from a total of 37 cats that had confirmation of their naturally occurring hyperadrenocorticism. Twenty-four of the cats were from our series, and 13 were from cats from the literature with well-documented FCS (Goossens et al, 1995 [6 cats]; Meij et al, 2001 [7 cats]). Of these 37 cats, 29 had PDH and 8 had functioning adrenocortical tumor hyperadrenocorticism (ATH). Twenty-nine of the total of 37 cats (78%; 24 from the PDH group and 5 from the ATH group) had results above the reference range of 3.6 × 10⁻⁵. An additional 7 cats (19%) had results in the borderline range (4 from the PDH group and 3 from the ATH group). This now accounts for results from all 8 cats with ATH and for 28 of the 29 cats with PDH. Borderline values were defined as those above the approximate canine reference range (<1.3 × 10⁻⁵) but lower than the higher range established for cats in the previously mentioned study (<3.6 × 10⁻⁵). Only one cat had a UC:CR test result that was considered normal (that cat had PDH). Thus the UC:CR appears to be a sensitive test for confirming the diagnosis of Cushing’s syndrome in cats, with a sensitivity approaching that demonstrated for dogs. We have not evaluated the specificity of the UC:CR but suspect that it may have similar problems to those noted in dogs.

TABLE 7-9 ENDOCRINE SCREENING TEST RESULTS FROM CATS WITH NATURALLY OCCURRING PITUITARY-DEPENDENT CUSHING’S SYNDROME