Current concepts in canine adrenal disease (Proceedings)

Hyperadrenocorticism

ACTH stimulation test

Unfortunately, the cost of cosyntropin, the most reliable agent with which to perform an ACTH stimulation test, is expensive. When cosyntropin was a fraction of its current cost, the dosing for an ACTH stimulation test was simple: one 250-µg vial per dog. A dose of 5 µg/kg IV or IM has been established as resulting in equivalent results (Kerl ME, et al. JAVMA 1999;214:1497-1501, Behrend EN, et al. JAVMA 2006;229:528-530). Once reconstituted, however, cosyntropin is not stable and must be frozen to maintain potency. For freezing and storage, it should be reconstituted with 1-ml sterile saline, aliquoted into 50 µg doses (0.2 ml) in TB or insulin syringes and frozen at -20C. At anything warmer, it is likely only stable for 21 days and should be stable for up to 6 months at -20C in a dedicated freezer (one that is not constantly opened and closed). Unexpected ACTH stimulation test results should prompt the user to question the potency of stored cosyntropin.

Desmopressin stimulation test

The DST is intended to be used as a differentiation test for pituitary-dependent hyperadrenocorticism and adrenal tumor. DDAVP stimulates V3 receptors in the pituitary, specifically expressed in pituitary corticotropes. In pituitary corticotrope tumors, this receptor is over-expressed. Stimulation of the V3 receptor results in marked ACTH release and subsequent marked increase in cortisol, a response that would not be seen with adrenal tumors. In one study, 80 dogs suspected of having Cushing's disease were evaluated with this test (27 non-adrenal disease, 46 PDH, 7 AT). The peak cortisol rise occurred at 30 minutes and increased by a median of 51% in dogs with PDH (range -24 to +563%). Dogs with AT had a change in baseline cortisol of -12% (range -44 to +5%). Using a %rise cut-off of 10%, the test had an 87% sensitivity for identifying PDH with 100% specificity. If the 20% cut-off was used, the sensitivity dropped to 76%. As a comparison, both abdominal US and the endogenous ACTH test had a sensitivity of 84% for differentiating disease (although the abdominal US probably more accurately had a sensitivity of 100% in this study, as they used enlarged adrenals as a marker, not just symmetrical adrenals).

Atypical hyperadrenocorticism in dogs: 17a-OH-progesterone

The diagnosis of hyperadrenocorticism is usually straight-forward, however a few dogs will have equivocal testing on the ACTH stimulation test or low-dose dexamethasone suppression test. In one study of 40 dogs with PDH and 12 with AT, the ACTH stimulation test was accurate in 79%, the LDDS in 93%. The measurement of 17a-OH-progesterone was only abnormal in 69% of these dogs, but was abnormal in 2 dogs that had normal cortisol but other findings suggestive of hyperadrenocorticism. The suggestion in the literature is that some dogs may have defects in the pathway of conversion of cholesterol to cortisol, resulting in excess production of intermediate metabolites, such as 17a-OH-progesterone. In most cases of Cushing's disease, these levels are proportionate with cortisol, even after the ACTH stimulation test, but in a small number of cases, the cortisol levels may not be diagnostic of Cushing's disease

Author note: Although uncommon, we have recognized a few cases of atypical hyperadrenocorticism. We do not routinely run a sex hormone profile with every dog, but currently recommend drawing an extra serum sample with the pre- and post-ACTH samples. This extra serum sample can be frozen and submitted at a later date for 17a-OH-progesterone or other sex hormone assays if needed. Drawing the sample with the original ACTH stimulation test saves the client the expense of cosyntropin a second time. Melatonin is often suggested as therapy for dogs with adrenal sex hormone excess, but there is no evidence that it reduces levels of these hormones, even if haircoat changes (alopecia X) resolves in some of these dogs.

Novel therapies for canine Cushing's disease

Retinoic acid

13-cis-retinoic acid (isotretinoin, Accutane®) was described by Castillo V, et al (Endocrinology 2006;147:4438-4444) for the treatment of pituitary-dependent hyperadrenocorticism. Isotretinoin (and, possibly, aliretinoin and acitretin (Soriatane®)) has been shown to inhibit proopiomelanocortin gene transcription (POMC gives rise to ACTH and a-MSH) and inhibit corticotrophinoma development and proliferation. Retinoic acid may induce cell-cycle arrest and apoptosis in corticotrophinoma cells, arrest tumor growth, and inhibit ACTH secretion from pituitary adenomas.

In their study, 42 dogs with pituitary-dependent hyperadrenocorticism were treated either with isotretinoin (2 mg/kg/day; 22 dogs) or ketoconazole (20 mg/kg/d; 20 dogs) for 180 days. All dogs had a pituitary adenoma detected on MRI prior to the study. The urine cortisol:creatinine ratios dropped in both groups throughout the study, although this drop was greater in isotretinoin treated dogs. Plasma ACTH levels dropped significantly in the isotretinoin treated dogs but not in the ketoconazole treated dogs. Pituitary adenoma size was also reduced significantly in the isotretinoin group but they were unchanged in the ketoconazole group. Survival in the ISO group was good: 2 dogs died due to complications of Cushing's, one died from cardiac failure, and 2 were removed but did not die. Survival in the KTZ group was poor: 10 died due to adverse effects of Cushing's, one died of pituitary apoplexy, and one left the study but did not die. Clinical signs resolved in all dogs in ISO. Signs had remained in remission 6-12 months after completing therapy.

Commentary: This is a potentially exciting therapy for pituitary-dependent Cushing's disease, especially for dogs with pituitary macroadenomas. There were no adverse events reported with isotretinoin in this study. The drug is currently available in the US, although it is teratogenic in the fetus, so caution should be taken by the client administering the medication. Available in 10-, 20-, and 40-mg tablet sizes for $3-5/pill.

Cabergoline

Cabergoline is a dopamine D2 receptor agonist used in the treatment of prolactinomas and growth-hormone secreting adenomas in humans. It has also been shown to be effective in aggressive corticotroph adenoma and non-functional corticotroph adenomas in humans. Cabergoline does directly what L-deprenyl is supposed to do indirectly: increase the relative dopaminergic influence on the corticotrophs to reverse hyperplasia associated with dopamine depletion. L-deprenyl was shown to be effective in only 20% of dogs with PDH, which correlated well with previous studies that found only 20% of pituitary tumors of the dog arose in the pars intermedia.

Castillo VA, et al (Res Vet Sci 2007 epress ahead of publication) treated 40 dogs with PDH with cabergoline at 0.07 mg/kg/week (divided into 3 doses to be administered over the span of 7 days). The only side effect of therapy was vomiting, which occurred in 90% of the dogs after taking the first dose. This was reduced to 10% after the second dose, and 0% with subsequent doses. Initially 24 of 40 dogs (60%) had a good response, with resolution of clinical signs. Of these, 7 dogs became symptomatic again while on cabergoline (after 4-6 months). Overall, 17 of 40 dogs (42.5%) had good long-term resolution of signs. Tumor size was a good predictor of failure, suggesting that this was an ineffective therapy for dogs with pituitary macroadenomas. For the responders, tumor size reduced as did plasma ACTH.

Commentary: The results of this study were less impressive than treatment with isotretinoin and there appeared to be no significant advantages with this drug. The results were more encouraging than what has been reported with l-deprenyl, likely owing to the direct dopaminergic effect of cabergoline. Given that cabergoline would likely be ineffective with pituitary macroadenomas, I doubt that this drug will see much use in the future. This drug is available in the US (Dostinex®: 0.25 and 0.5 mg tablets for approximately $15/tablet).

Trilostane

Trilostane is competitive inhibitor of 3 -hydroxysteroid dehydrogenase that blocks the conversion of cholesterol to cortisol, but may also inhibit aldosterone production. Although the drug has been available in England and Europe for several years, there is no pharmacokinetic information available for dogs, so dosing (amount, frequency) is imprecise. Trilostane has been the treatment of choice in Europe for canine Cushing's syndrome since its introduction. Several recent studies have evaluated the efficacy and are summarized below:

 • Neiger R, et al. (Vet Rec 2002;150:799-804) evaluated trilostane in 78 dogs with PDH. Their treatment protocol was as follows: dogs < 5 kg received 30 mg once daily or every other day, dogs 5-20 kg received 60 mg once daily, dogs >20 received 120 mg once daily. The dose was increased to twice daily if control was not evident at 4 weeks. The mean starting dose was 5.9 mg/kg (range 1.8-20 mg/kg). A dose change was subsequently required in 35/78 dogs: 23 had the dose increased (mean final dose of 11.4 mg/kg; one dog was increased from 3 to 27.3 mg/kg over 6 months), and 12 dogs had the dose decreased to a mean final dose of 3.2 mg/kg. The clinical signs of PU/PD and polyphagia resolved in 70% of those affected. Within 10 days of starting therapy, basal cortisol had decreased significantly. Side effects generally were mild and limited to transient changes on the serum biochemistry profile. Two dogs died, one in heart failure and one unknown. Addison's disease developed in two dogs, one of which died.

 • Barker EN, et al. (JVIM 2005;19:810-815) compared the survival times of dogs with PDH treated with either mitotane or trilostane. This study is out of the UK, which is reflected in the treatment bias. There were 123 dogs treated with trilostane and 25 treated with mitotane. There was no significant difference in survival times of the two groups (662 days for T, 708 days for M).

 • Braddock JA, et al. (Australian Vet J 2003;81:600-607) evaluated the use of trilostane in 30 dogs with PDH. The therapeutic regimen was the same as used by Neiger et al. The dogs were monitored by UCCR and ACTH-stimulation testing. The final median trilostane dose was 17 mg/kg with a range of 5-50 mg/kg (larger dogs need lower doses compared to smaller dogs). Successful treatment was achieved in 29/30 dogs. The one other dog had incomplete control at 30 days but died of cervical vertebral neoplasia before control could be established. The UCCR had poor correlation in this study to ACTH stimulation test results, agreeing in only 43-47% of cases at 90 and 180 days. Four dogs developed clinical signs of hypoadrenocorticism, although all after the study period was completed.

 • Alenza DP, et al. (JAAHA 2006;42:269-276) reported on the long-term efficacy of trilostane administered twice daily to dogs with PDH. There were 44 dogs in this study and the mean starting dose of trilostane was 3.1 mg/kg PO q 12 hours (range 1.2-7.5 mg/kg). The dosing schedule used was the following: dogs <5 kg were given 15 mg PO q12h, dogs 5-20 were given 30 mg PO q12h, dogs 20-40 kg received 60 mg AM and 30 mg PM, and dogs >40 kg received 60 mg PO q12h. The goal was to maintain a post-ACTH cortisol between 1 and 5 µg/dl. At one week PU/PD had improved in 38 (86%) dogs. Five dogs required dose reduction at one week due to GI upset, 8 had doses increased due to high cortisol levels post-ACTH. All dogs had resolution of PU/PD at 1 month, but 10 dogs were still polyphagic and their cortisols were high, so they had dose increase. At 3 months, 3 dogs were PU/PD and 10 were still polyphagic. In the latter 10 dogs, the dose was increased again. The median dose of trilostane after two years was 3.25 mg/kg PO q12h (range 1.1 to 10.2 mg/kg). Adverse events were reported in 11 dogs related to hypocortisolemia. In 6 of these dogs, adrenal suppression was apparently permanent, with one dog requiring prednisone and DOCP. In the remaining dogs, adrenal recovery took several weeks.

 •Vaughan MA, et al. (JAVMA 2008;232:1321-1328) evaluated the effects of twice-daily oral administration of low-dose trilostane in 28 dogs with naturally occurring hyperadrenocorticism. They also evaluated the duration of trilostane effects. Dogs were started at somewhere between 0.5 and 2.5 mg/kg PO q 12 hours (based on using 30, 60, and 120 mg capsules). Dogs were evaluated at 1-2 weeks, 4-8 weeks, and 8-16 weeks after each prior evaluation and an ACTH stimulation test was done 3-4 hours after administration of the morning dose. In the low-dose study, there were 22 dogs, 18 with PDH and 4 with AT. The mean starting dose was 1.4 mg/kg (range 0.7-2.4) and that increased to 1.9 mg/kg (range 1.1-2.8) by the end of the study. By the end of the study, 16/22 dogs were reported to have good control, with mean dose in those 16 dogs of 1.7 mg/kg. The urine cortisol-to-creatinine ratio was significantly lower in the good responders (mean 14.8: range 5.4-36) as compared to the poor responders (mean 47.4: range 16-154) (their cut-off for normal UCCR is >13.5). They also concluded that inhibition of cortisol production persists for 8-9 hours in most dogs.

All of these studies support the use of trilostane in dogs as an effective therapy for PDH. These and other studies also support that twice daily administration is preferable over once daily administration.

Dosing: The recommended dosing for trilostane is based on weight ranges (dogs <10 kg receive 30 mg daily, 10-20 kg get 60 mg daily, etc...)providing a dose range of 3-6 mg/kg/day. A flow chart is available to help manage dosing decisions. Although this dosing scheme should work well, I currently recommend starting the dose around 1-1.5 mg/kg PO q 12 hours. As long as the dog does not demonstrate signs of hypoadrencorticism or other adverse events associated with trilostane, I typically wait 30 days and then re-evaluate the dog. If they are still PU/PD, I double the dose to 2-3 mg/kg PO q 12 hours and reassess in another month. I don't do an ACTH stimulation test at that time. I evaluate the patient on a monthly basis and raise the dose 1-1.5 mg/kg PO q 21 hours each time if necessary. If the dog appears to be well-controlled clinically, then I evaluate with an ACTH stimulation test. My goals are to give the drug long enough to be effective without subsequent overdosing and to minimize expense associated with repeated ACTH stimulation testing.

Monitoring: Several studies have shown that trilostane does not affect aldosterone levels, but surprisingly there are several reports of complete adrenal failure on trilostane therapy. Careful monitoring of the patients is warranted, although laboratory work may not be required if the patient is clinically normal. The recommended monitoring is at 7-14 days, one month, 3 months, and then every 3 months after that. The recommendation is to perform an ACTH stimulation test 4-6 hours after administration of trilostane in the morning, which is when peak activity occurs. A post-ACTH cortisol of 1-5 µg/dl (or 26-110 nmol/L) suggests that no change is required. If the cortisol is below this, the dose should be reduced after a 2-3 day discontinuation. If the cortisol is >5 and <9.1, dosage changes should correspond to clinical signs. If the cortisol exceeds 9.1, a dose increase is warranted.

Trilostane was shown to affect other steroid hormones in the following manner: there was an increase in 17a-OH-pregnenolone and dehydroepiandrosterone, aldosterone increased on basal levels but not post-ACTH, 17a-OH-progesterone and androstenedione levels did not change, post-ACTH levels of 21-deoxycortisol decreased and baseline concentrations of 11-deoxycortisol increased.

Author's note: The major problem and expense, in addition to the cost of trilostane, is the cost of repeated ACTH stimulation tests. Given that clinical signs fairly reliably predict successful therapy, I think it is appropriate to bypass the ACTH stimulation test and increase the dose if the dog remains PU/PD and polyphagic. Once clinical signs are controlled, the clinician can decide the value of the ACTH stimulation test and may choose not to perform one. If GI signs are present, a single low basal cortisol may provide sufficient evidence to discontinue therapy. The goal here is to minimize expense but still provide an appropriate level of care. The UCCR may have some value in monitoring, but isn't necessarily inexpensive at most laboratories.

Melatonin

Melatonin is frequently recommended for the management of pituitary disease that results in sex hormone excess. Alopecia X is frequently associated with abnormalities of adrenal steroid intermediates and sex hormones in miniature poodles and certain Nordic and plush-coated breeds. This disease has been previously labeled growth hormone-responsive dermatitis and pseudo-Cushings syndrome and has shown response to mitotane therapy. Melatonin is anecdotally reported to be effective in treating this disease, possibly by altering gonadotropin release and sex hormone concentrations. Frank LA, et al. (Vet Dermatology 2004;15:278-284) evaluated the response of 29 dogs with alopecia X. In their study, 62% of the dogs had partial to complete hair regrowth, but they failed to identify significant alterations in sex hormones or 17-hydroxyprogesterone. The dose in these dogs was 3-6 mg/dog PO q 12 hours.

Author's note: Success or failure with melatonin may be related to the formulation administered, and that information was not available in this study. I recommend using the capsule formulation and not the tablets, but if the tablets are used they should be pulverized prior to use. The tablets may not dissolve adequately, resulting in a negligible amount of absorbed drug. I have no experience with melatonin implants, but they may prove to be useful for treatment of this condition. Trilostane was also shown to be effective in treating 3 Alaskan malamutes with alopecia X. Unfortunately in that study the authors failed to follow 17-OH P4 levels, which were mildly elevated at the beginning of therapy.

Hypophysectomy

Transsphenoidal hypophysectomy is the treatment of choice for pituitary tumors in man, but it has traditionally been considered technically more complicated in dogs due to anatomic differences. Hanson et al (JVIM 2005;19:687-694) described the outcome of this procedure in 150 dogs with PDH. Although 12 dogs died post-operatively, 127 dogs went into remission and only 32 of these dogs had recurrence. The median time to recurrence was 18.3 months (range 6-56 months). The 2-year survival rate was 76% and 71% at 3-years. Side effects included keratoconjunctivitis sicca (31%) and central diabetes insipidus (53%). Hanson et al (J Neurosurg 2007;107:830-840) also reported on prognostic factors for outcome after this procedure. They determined that old age, large pituitary size, and high pre-operative plasma ACTH levels were associated with an increased risk of death. Large pituitary size, thick sphenoid bone, high plasma alpha-MSH levels, and high urinary cortisol-to-creatinine ratios (all pre-operative measures) were associated with an increased risk of recurrence. Although this is an attractive option for therapy, the surgical skill necessary to perform this procedure precludes its routine availability. The mortality rate would be substantially higher in the early stages of the learning curve.

Hypoadrenocorticism

Testing - basal serum or plasma cortisol

Lennon et al (JAVMA 2007;231:413-416) wanted to determine whether basal serum or plasma cortisol concentrations alone could be used as a screening test to rule out hypoadrenocorticism. They looked at 110 dogs with nonadrenal illnesses and 13 dogs with hypoadrenocorticism. Six of the 13 dogs with hypoadrenocorticism did not have electrolyte abnormalities. Interestingly, the control group had serum potassium ranging from 3.2 to 8.2 and Na:K ranging from 16-43, which made for a good comparison group. All dogs with hypoadrenocorticism had a basal cortisol concentration of ≤ 1 µg/dl, whereas on 2/110 dogs with nonadrenal illness had a basal cortisol that low. This gave basal cortisol measurement a sensitivity of 100% and specificity of 98.2% if the ≤ 1 µg/dl cut-off was used. The authors point out that in 11 published studies, only 6 of 561 dogs had a basal cortisol of > 2 µg/dL. Although the negative predictive value is very high (100%), making this a good screening test to decide whether ACTH stimulation testing is required, the positive predictive value is relatively low given the low prevalence of this disease (2.3% PPV at 0.5% prevalence rate). It would have been helpful if the authors described the diagnoses in the two dogs with nonadrenal disease that had very low cortisol levels.

Testing - aldosterone-to-renin ratio and cortisol-to-ACTH ratio

Javadi et al (JVIM 2006;20:556-561) looked at ARR and CAR in an attempt to identify single-blood sample tests that would help establish a diagnosis of hypoadrenocorticism. In hypoadrenocorticism, cortisol deficiency removes the negative feedback for ACTH release, so the CAR should be very low. Likewise, renin release is stimulated by hyponatremia and hypovolemia, so the ARR should be very low in hypoadrenocorticism. The reference range for normal ARR and CAR in healthy dogs was 0.1-1.5 and 1.1-2.6, respectively. In dogs with hypoadrenocorticism, the ranges for ARR and CAR were 0.002-0.008 and 0.003-0.17, respectively. These differences were significant (P < 0.001) with no overlap. These appear to be good tests for hypoadrenocorticism, although the ARR would probably not perform well in dogs with cortisol-deficient atypical Addison's disease. It would have been ideal, however, to evaluate this test on dogs with non-adrenal disease similar to the study by Lennon et al in their evaluation of basal cortisol.