Evaluation of Iatrogenic Hypocortisolemia Following Trilostane Therapy in 48 Dogs with Pituitary-Dependent Hyperadrenocorticism
ABSTRACT
This study aimed to retrospectively describe the clinical progression following diagnosis of iatrogenic hypocortisolemia (iHC) in 48 dogs receiving trilostane for pituitary-dependent hyperadrenocorticism. Cortisol concentrations were ≥1.5 mg/dL within 6 mo following diagnosis of iHC in 76.3% of dogs (95% confidence interval [CI] 59.8–88.6%). At the time of study completion, 25% of dogs (95% CI 13.6–39.6%) were receiving either glucocorticoids or mineralocorticoids or both; 42% of dogs (95% CI 27.6–56.8%) were on no adrenal-related medications; and the remaining 33% of dogs (95% CI 20.4–48.4%) were receiving trilostane. No patient-, clinicopathologic-, or trilostane-associated factors were identified to influence adrenal recovery following diagnosis of iHC, and it remains difficult to predict the clinical progression in this population of dogs.
Introduction
Trilostane is a synthetic steroid analogue commonly used in the treatment of pituitary-dependent hyperadrenocorticism (PDH).1 It acts to competitively inhibit the enzyme 3-beta-hydroxy-steroid dehydrogenase and subsequently prevents synthesis of cortisol and, to a lesser extent, aldosterone.1 Dosing recommendations vary widely and range in the veterinary literature from 0.2 mg/kg twice daily to 20 mg/kg once daily.2–7 Therapeutic monitoring for dogs on trilostane typically involves a timed adrenocorticotropic hormone (ACTH) stimulation test, but there is a lack of consensus regarding target ranges for preand post-ACTH–stimulated cortisol, timing of cortisol sampling, and therapeutic adjustment based on adrenal axis testing.1,2 Alternative testing, including baseline cortisol concentrations measured before and/or after trilostane and urine cortisol–creatinine ratios, have been used because of reduced availability of synthetic ACTH and cost concerns.8–10 Anecdotally, clinicians may rely more on clinical signs than cortisol concentrations to assess therapeutic response. Adverse effects from trilostane therapy have been observed at a wide variety of doses and frequency of administration.2–7
Trilostane-associated iatrogenic hypocortisolemia (iHC) has been infrequently documented with varying trilostane dosages and treatment durations in the veterinary literature,11–14 including short-term trilostane usage and ultra-low-dose trilostane (0.21–1.1 mg/kg twice daily).3,5 The largest study to date, a retrospective cohort from Australia in 2017, identified 19 dogs diagnosed with iHC.15 Cumulative incidence of iHC was 15% by 2 yr, and 26% by 4.3 yr. Higher trilostane dose rate and body weight did not increase risk for development of iHC, and 74% of dogs had transient hypocortisolism. This study included both adrenal- and pituitary-dependent forms of hyperadrenocorticism (HAC). Cortisol concentrations prior to development of iHC were not reported, nor was the clinical progression following diagnosis of iHC described in detail. Because trilostane inhibits hormone production, the drug is not expected to cause persistent reduction of adrenal hormone concentrations once it has been metabolized and eliminated from the body.1,2 Maximum plasma concentration of trilostane typically occurs within 1.7–3.8 hr, and plasma concentrations return to baseline within 12 hr in healthy dogs,2 although duration of action may be even shorter in dogs with HAC.2 Risk factors for developing permanent iHC and the mechanism by which it occurs following trilostane therapy remain unknown. Adrenal necrosis has been documented in the literature and is postulated as a cause for sustained iHC after trilostane therapy is withdrawn.13,16,17
The primary objective of this study was to retrospectively describe the clinical progression following diagnosis of iHC in a population of 48 dogs receiving trilostane therapy for PDH, including description of trilostane dosing and cortisol concentrations before and after iHC diagnosis, presence and type of clinical signs exhibited at the time of iHC diagnosis, and time to adrenal recovery (defined as pre- and/or post-ACTH–stimulated cortisol concentrations ≥1.5 μg/dL). A secondary objective was to screen for any patient-, clinicopathologic-, or trilostane-associated factors that would influence adrenal recovery following iHC. We hypothesized that adrenal recovery would occur within 6 mo following iHC diagnosis in >50% of dogs following reduction or cessation of trilostane dosing.
Materials and Methods
Patient Enrollment
The medical databases at The Animal Medical Center, New York, New York; The Ohio State University College of Veterinary Medicine, Columbus, Ohio; and Med Vet (locations: Columbus, Ohio; Indianapolis, Indiana; and Chicago, Illinois) were searched. Cases were also obtained from six additional private specialty practices and two general practices. Cases had to meet the following inclusion criteria: (1) previous diagnosis of PDH, (2) treatment with trilostane for any length of time, and (3) a pre- and post-ACTH–stimulated cortisol of ≤1.1 μg/dL while receiving trilostane monotherapy. Dogs were excluded if they had received any steroid therapy (topical, oral, or inhaled) in the month prior to documented iHC or had been previously treated with mitotane or other drugs with known efficacy against PDH (specifically ketoconazole, l-deprenyl, melatonin, lignans, retinoic acids) within 3 mo prior to documentation of iHC.
Diagnosis of PDH
HAC was diagnosed based on the presence of clinical signs and endocrine testing,18 namely, a low-dose dexamethasone suppression test (LDDST) and/or an ACTH stimulation test. HAC was further differentiated into PDH based on one of the following criteria: (1) suppression to below the standard laboratory cutoff or <50% of basal cortisol concentration at 4 hr after dexamethasone administration; (2) suggestive adrenal imaging, that is, normal sized adrenal glands, symmetric bilaterally enlarged adrenal glands, or lack of an adrenal mass; or (3) an elevated endogenous ACTH concentration.18 Signalment and presence of concurrent disease at the time of PDH diagnosis were recorded.
Trilostane Data
Details of trilostane therapy were recorded, including trilostane dosages at the initial diagnosis of PDH, iHC, and at the last recorded follow-up, use of compounded or commercial trilostane formulations, and duration of trilostane therapy before diagnosis of iHC. Clinicopathological control of PDH was defined as a post-ACTH–stimulated target cortisol ranging from 1.45 to 9.1 μg/dL as based on guidelines from the Vetoryl package insert, where cortisol concentrations ranging from 5.4 to 9.1 μg/dL were classified as acceptable clinicopathologic control when coupled with resolution of clinical signs.2,6 Timing of the ACTH stimulation test in relation to trilostane administration was recorded when available. Only data from dogs who had clinical assessments and ACTH stimulation tests within 6 mo before iHC diagnosis were analyzed, because testing beyond this interval was not considered to provide clinically relevant information. In this subset of dogs, the last recorded ACTH stimulation test before development of iHC was evaluated with emphasis on the post-ACTH–stimulated cortisol concentration; the Δ cortisol value (defined as the difference between pre- and post-ACTH–stimulated cortisol concentrations) was also calculated.
Diagnosis of iHC
Dogs were diagnosed with iHC if pre- and post-ACTH–stimulated cortisol concentrations were ≤1.1 μg/dL while receiving trilostane therapy. In order to perform statistical comparisons among cortisol concentrations tested at different laboratories with three varying lower limits of detection, the cortisol concentrations were adjusted as an average of the possible lower values, that is, cortisol concentrations recorded as <1.0, <0.7, and <0.2 μg/dL were adjusted to 0.5, 0.35, and 0.1 μg/dL, respectively, for analysis. Subsequent cortisol concentrations following iHC were recorded only if glucocorticoids had been discontinued a minimum of 3 days before cortisol testing and if initial patient follow-up was performed within 6 mo following iHC diagnosis. Recorded data included presence, absence, and type of clinical signs consistent with hypocortisolemia, concurrent disease identified at the time of iHC diagnosis, clinicopathological data, adjustment of trilostane dosage, prescription and duration of gluco- and/or mineralocorticoid supplementation, time to adrenal recovery (defined as pre- and/or post-ACTH–stimulated cortisol concentrations ≥1.5 μg/dL following diagnosis of iHC), and lifespan following diagnosis of iHC. Date of death or date of last patient contact was recorded. Survival times were calculated until patient death or last recorded follow-up.
Statistical Methods
Quantitative data were summarized as mean and standard deviation, whereas nonnormally distributed data were presented as median and range. Normally distributed continuous variables in patient characteristics and laboratory tests were analyzed by analysis of variance for independent and dependent models as appropriate. Categorical and qualitative data were described as percentages, with exact binomial confidence intervals (CIs) when appropriate, using χ2 analyses to determine proportional differences, where a Fisher exact test was used to test for differences with a cell size of less than five. χ2 analyses were employed to compare selected patient factors, trilostane dosing regimens, and clinicopathologic data with cortisol concentrations ≥1.5 μg/dL at any time point following diagnosis of iHC, and specifically within 6 mo of iHC diagnosis. Additional proportions over time between group analyses were carried out using a Cochran-Armitage test for trend. Survival analyses were carried out using the Kaplan-Meier product limit estimates and were carried out in univariate and censored when the patient did not meet the primary endpoint or was lost to follow-up. The median of the survival time is defined as the time beyond which 50% of the subjects in the population under study are expected to survive. All analyses were performed with SAS version 9.4a, where P < .05 was deemed significant.
Results
Signalment
Forty-eight dogs met inclusion criteria for this study, obtained from a total of 13 different hospitals. Median age of dogs at diagnosis of PDH was 10.9 yr (range 4–13.5 yr). There were 25 castrated males, 21 spayed females, 1 intact male, and 1 intact female. Thirty breeds were identified, and breeds with more than 1 dog included mixed breed (6), miniature schnauzer (4), Yorkshire terrier (3), Maltese (3), Boston terrier (3), dachshund (2), bichon frise (2), boxer (2), and beagle (2). The median body weight at diagnosis of PDH was 9.8 kg (range 3.36–33.2 kg).
Diagnosis, Treatment, and Monitoring of PDH
Clinical signs consistent with HAC were present at the time of PDH diagnosis in all 48 dogs. Initial diagnosis of HAC was based on an LDDST (36/48, 75%) or a positive ACTH stimulation test (12/48, 25%). Differentiation into PDH was based on suppression during LDDST (27/36 dogs, 75%); in the remaining 21 dogs, PDH was diagnosed based on compatible adrenal gland imaging (20 dogs) or an elevated endogenous ACTH concentration in 1 dog.
Approximately half of the dogs (26/48, 54.2%) were evaluated within 6 mo before diagnosis of iHC. In those 26 dogs, clinical control of PDH, as perceived by the owners, was documented in 23/26 dogs (88.5%). Of the remaining 3 dogs, polydipsia and polyuria were noted in 2 dogs and polyphagia in 1 dog. Clinicopathologic control of PDH was achieved in 25/26 (96%) dogs. The remaining dog had a post-ACTH–stimulated cortisol of 11, although clinical control was noted. The median post-ACTH–stimulated cortisol concentration was 3.4 μg/dL (range 1.5–11 μg/dL) and the median Δ cortisol was 0.92 μg/dL (range −1.8 to 4.1 μg/dL). Timing of the ACTH stimulation test in relation to trilostane administration was recorded in 20 dogs and ranged from 2 to 8 hr (median 4.5 hr).
Trilostane Dosing and Preparation
At the time of PDH diagnosis, the median initial trilostane daily dosage was 3.5 mg/kg/day (range 1.2–11.1 mg/kg/day); 24 dogs received once-daily dosing, and 23 dogs received twice-daily dosing (the dosing frequency of 1 dog was unknown). At the time of iHC diagnosis, the median trilostane daily dosage was 4 mg/kg/day (range 1.17–18 mg/kg/day); 18 dogs received once-daily dosing, and 30 dogs received twice-daily dosing. Immediately prior to diagnosis of iHC, 12 dogs (25%) received compounded trilostane and 36 (75%) dogs received commercially available trilostane; 3/12 dogs on compounded trilostane had been previously switched from the commercially available formulation.
Diagnosis and Treatment of iHC
Clinical signs consistent with hypocortisolemia and/or mineralocorticoid deficiency at the time of iHC diagnosis were documented in 32 dogs (66.7%; Figure 1), and 16 patients were asymptomatic. A new disease process or complication of a previously diagnosed disease was documented at the time of iHC diagnosis in 12/32 symptomatic dogs (37.5%). Specific diseases included hypoglycemia with previously diagnosed diabetes mellitus (2), supraventricular tachycardia (1), congestive heart failure (1), azotemia (2), a lytic lesion of the proximal humerus (1), suspected cholangitis (1), uveitis (1), left forebrain lesion (1), vestibular disease (1), and a urinary tract infection with uroliths (1). New diseases were not documented in any of the 16 asymptomatic patients. Clinicopathologic data with significant changes between time of initial PDH and subsequent iHC diagnosis are highlighted in Table 1.
Citation: Journal of the American Animal Hospital Association 57, 5; 10.5326/JAAHA-MS-7076
At the time of iHC diagnosis, trilostane dosing was adjusted for all dogs. Trilostane was discontinued in 32/48 dogs (66.7%, 95% CI 51.6–79.6%), of whom 29/32 had clinical signs. Trilostane dosing was reduced in the remaining 16 dogs, 13 of whom were asymptomatic. The median percentage of dose reduction following diagnosis of iHC was 33.3% (range 12–50%), excluding dogs who had trilostane discontinued.
Glucocorticoids and mineralocorticoids were prescribed only in symptomatic dogs. Prednisone was initially prescribed in 22/48 dogs (45.8%); no other type of glucocorticoid was prescribed. No dog was prescribed a mineralocorticoid without concurrent glucocorticoid supplementation. Mineralocorticoids were initially prescribed in 15/48 dogs (31.3%); 2 dogs received fludrocortisone, and the remaining 13 received desoxycorticosterone pivalate (DOCP). Baseline aldosterone concentrations were measured in 5 dogs at two hospitals and were decreased in 4/5 dogs (aldosterone concentrations of these 4 dogs were ≤0.04 ng/dL). The decision to measure aldosterone concentrations was based on a decreased Na:K ratio, which ranged from 16 to 25 in these 5 dogs. All 4 dogs with low aldosterone concentrations received ongoing supplementation with DOCP; the 1 dog with a normal aldosterone concentration (3.93 ng/dL) was restarted on trilostane 5 wk following diagnosis of iHC. All dogs prescribed mineralo- and/or glucocorticoids immediately following iHC diagnosis had trilostane discontinued.
Cortisol Concentrations Following Diagnosis of iHC
Cortisol concentrations were measured in 38/48 dogs (79.2%) within 6 mo following iHC. For these 38 dogs, median baseline, post-ACTH–stimulated, and Δ cortisol concentrations at the first testing interval following diagnosis of iHC were 2.5 μg/dL (range 0.3–17.2 μg/dL), 4.7 μg/dL (range 0.35–29.7 μg/dL), and 2.3 (range −2.2 to 19.7), respectively. Persistent hypocortisolemia (i.e., cortisol ≤1.1 μg/dL) was documented in 15/38 dogs (39.5%) at this initial testing interval, and 8/15 dogs were receiving trilostane at this time. Repeat cortisol concentrations were not recorded in 10/48 dogs at any time point following iHC diagnosis because of the following reasons: patient death (3), testing not elected (2), and prednisone therapy preventing interpretation of cortisol concentrations (5).
The majority of dogs (29/38 dogs, 76.3%, 95% CI 59.8–88.6%) achieved adrenal recovery within 6 mo following iHC diagnosis (Figure 2). Of the 9 dogs with persistently low cortisol concentrations 6 mo after diagnosis of iHC, 5 had a subsequent increase in cortisol to >1.5 μg/dL within 8–16 mo of iHC diagnosis. Trilostane had been further lowered or discontinued in these dogs to address their ongoing hypocortisolemia. Adrenal recovery was not documented in the remaining 4 dogs at any recorded time point following iHC; 1 dog continued to receive trilostane at a reduced dose despite the low cortisol concentrations, 2 dogs were subsequently treated with gluco- and mineralocorticoids and cortisol concentrations could not be further interpreted, and the remaining dog was on no adrenal-related medications at 24 mo following iHC diagnosis.
Citation: Journal of the American Animal Hospital Association 57, 5; 10.5326/JAAHA-MS-7076
Patient Follow-Up
Median survival time following diagnosis of PDH was 34 mo (range 1–100 mo). The median length of trilostane therapy before diagnosis of iHC was 12 mo (range 0.5–68 mo). Median survival time following diagnosis of iHC was 15 mo (range 0–70 mo; Figure 3); 17 dogs were alive, 21 deceased, and 10 lost to follow-up at the time of study completion. One dog died acutely following diagnosis of iHC, and the cause of death did not appear to be associated with the recent iHC diagnosis; thromboembolic disease was suspected. This dog was receiving 3.1 mg/kg/day of trilostane when diagnosed with iHC and had been prescribed trilostane for 1 mo before iHC diagnosis. No patient deaths were attributed to iHC.
Citation: Journal of the American Animal Hospital Association 57, 5; 10.5326/JAAHA-MS-7076
Trilostane was prescribed in 16/48 dogs (33%, 95% CI 20.4–48.4%) at the last recorded follow-up or until time of death. Of these 16 dogs, 6 were deceased, 6 were alive at study completion, and 4 had been previously lost to follow-up. The median trilostane dosage was 1.0 mg/kg/day (range 0.25 mg/kg every other day to 5.4 mg/kg/day); 8 dogs received once-daily dosing, 6 dogs received twice-daily dosing, and 2 dogs received every-other-day dosing. Trilostane was never discontinued following iHC in 6/16 dogs; the remaining 10 dogs received intermittent trilostane therapy based on recurrent clinical signs of PDH. For these 10 dogs, the median time frame off trilostane was 6 wk (range 0.7–20 wk).
Gluco- and/or mineralocorticoids had been prescribed in 12/48 dogs (25%, 95% CI 13.6–39.6%) at the last recorded follow-up or until time of death. Of these 12 dogs, 7 were deceased, 4 were alive at study completion, and 1 was previously lost to follow-up. Hormone supplementation was started in 6/12 dogs immediately following diagnosis of iHC, and the remaining 6 were prescribed hormone supplementation at a later date because of clinical signs consistent with hypocortisolemia or electrolyte changes suggestive of a mineralocorticoid deficiency. Prescribed medications included prednisone and DOCP (9 dogs), prednisone and fludrocortisone (1), prednisone alone (1), and fludrocortisone alone (1).
Neither trilostane nor adrenal hormone supplementation was prescribed in 20/48 dogs (42%, 95% CI 27.6–56.8%). Of these 20 dogs, 8 received trilostane on an intermittent basis until it was discontinued before the last recorded follow-up, 8 initially received hormone supplementation in the form of gluco- and/or mineralocorticoids that were later discontinued, and 4 had trilostane immediately discontinued and not restarted following iHC diagnosis. At the time of study completion, 7 dogs were dead, 7 dogs were alive, and 6 dogs were lost to follow-up.
Of the 38 dogs with follow-up extending at least 6 mo after iHC diagnosis, 13 (34%) had one or more changes in their prescribed adrenal-related medications, not including the initial adjustment in trilostane dosing. The medication changes were as follows: (1) 4 dogs had trilostane initially discontinued, glucocorticoids were then prescribed and later discontinued, and trilostane was restarted at a later date; (2) 2 dogs had trilostane dosing initially reduced, trilostane was then discontinued, and glucocorticoids were prescribed at a later date; (3) 6 dogs had trilostane dosing initially reduced, and the trilostane was later stopped and restarted at a later date; (4) 1 dog was initially switched from trilostane to glucocorticoids, and glucocorticoids were then discontinued at a later date.
Factors Associated with Adrenal Recovery
During the first 6 mo following diagnosis of iHC, 29/38 dogs (76.3%) achieved adrenal recovery whereas cortisol concentrations remained persistently low in 9 dogs. Numerous patient-, clinicopathologic-, or trilostane-associated factors were compared between these two groups, and no factors were found to be significant (P ≥ .06 for all analyses). Analyzed factors included patient characteristics, concurrent disease and administration of other medications at the time of iHC diagnosis, clinical and clinicopathologic control of PDH before iHC diagnosis, use of commercial or compounded trilostane products, trilostane dosing before and after diagnosis of iHC, supplementation of gluco- and/or mineralocorticoids following diagnosis of iHC, presence and type of clinical signs at the time of iHC diagnosis, and magnitude of cortisol suppression at the time of iHC diagnosis.
Discussion
In this study, iHC was identified in 48 dogs with pre- and post-ACTH–stimulated cortisol concentrations of ≤1.1 μg/dL following trilostane therapy for PDH. The clinical term of hypoadrenocorticism was avoided, because 33% of dogs exhibited no abnormal symptoms in association with their hypocortisolemia, and 37.5% of the 32 dogs with suggestive clinical signs had a newly developed disease or complication of a prior disease process documented at the time of iHC diagnosis that may have contributed to their clinical presentation. This study demonstrated that a wide variety of trilostane dosing regimens and treatment durations were associated with iHC, including doses typically considered low (<2 mg/kg/day).5,6 Administration of trilostane or adrenal hormone replacement therapy following iHC was variable, and no factors were identified to predict adrenal recovery. iHC was not attributed as a cause of death in any dog, reinforcing the clinical impression that iHC either is reversible or can be effectively managed following diagnosis.
Trilostane was discontinued or dose-reduced in all dogs at the time of iHC diagnosis, regardless of whether clinical signs were present. This consistent response from a variety of clinicians likely reflects the inclusion criteria of a post-ACTH–stimulated cortisol of ≤1.1 μg/dL; this cortisol concentration was deliberately chosen to increase the likelihood that trilostane was contributing to patient morbidity. Although a previous study19 has demonstrated that cortisol concentrations may increase 9–12 hr after peak trilostane activity, there remains concern that these patients can still have clinical repercussions from hypocortisolemia earlier in the day. Because of the retrospective nature of the study, the exact timing of the ACTH stimulation tests in relation to trilostane administration was not consistently recorded, although a time range of 2–8 hr after trilostane dosing was most commonly observed when the timing was documented.
One surprising finding of the study was the intrapatient adjustments in prescribed medications following diagnosis of iHC, specifically whether and how long patients received trilostane or adrenal hormone supplementation. Again, because of its retrospective design, strict criteria were not used to determine what, if any, medication was prescribed, but trilostane was stopped and restarted in a number of dogs based on their subsequent ACTH stimulation tests and associated clinical signs, as were glucocorticoids. Trilostane was ultimately discontinued in 8/16 dogs who were initially asymptomatic at the time of iHC diagnosis, a treatment decision that appeared to be based on subsequent clinical signs consistent with hypocortisolemia. Conversely, 9/32 dogs who were initially symptomatic for hypocortisolemia at the time of iHC diagnosis were receiving trilostane at the last recorded follow-up. Following initial reduction or cessation of trilostane after diagnosis of iHC, 34% of dogs had further changes in their adrenal-related medications, some of which occurred more than 6 mo after diagnosis of iHC. Repeated monitoring of cortisol concentrations and clinical signs may be indicated in these dogs.
Adrenal recovery, defined as pre- and/or post-ACTH–stimulated cortisol concentrations ≥1.5 μg/dL, was achieved in 76.3% dogs within 6 mo following iHC diagnosis. Our definition for adrenal recovery was extrapolated from the 2017 study by King et al.,15 in which dogs were classified as having transient hypoadrenocorticism if post-ACTH–stimulated cortisol concentrations reached ≥1.45 μg/dL at any time point following iHC. Similar to the results reported here, King et al. also documented adrenal recovery in the majority (14/19, 74%) of dogs diagnosed with iHC. In our study, no variable was identified to affect adrenal recovery, although potential factors such as adjustments in trilostane dosing following iHC diagnosis (specifically discontinuation versus dose reduction), presence and type of clinical signs at the time of iHC diagnosis, and the magnitude of cortisol suppression were evaluated. These conclusions are likely influenced by both the small study population and the observation that adrenal recovery occurred in the majority of dogs. When clinicopathologic data were compared between initial diagnosis of PDH and subsequent development of iHC, numerous values were statistically different. Some of these findings, specifically the decreases noted in alkaline phosphatase and alanine aminotransferase and the minor changes in sodium, potassium, and the Na:K ratio, are consistent with previously documented biochemical control of PDH and the expected low-level inhibition of aldosterone with trilostane therapy, respectively.20–23 Although the increases in blood urea nitrogen and creatinine and the drop in albumin are suggestive of hypocortisolemia, risk factors for development of iHC could not be evaluated in our study population.
The mechanism through which trilostane might cause persistent iHC is unknown and cannot be explained by the drug’s mechanism of action, which typically induces diffuse adrenal cortical hyperplasia.1,2 Some dogs may instead experience adrenal necrosis following trilostane therapy.13,16,17 Several theories have been suggested as the cause of adrenal necrosis, including hypersecretion of ACTH,24 accumulation of toxic levels of trilostane or its metabolites in adrenocortical cells, an idiosyncratic reaction, or induction of vascular changes.11–13,16 ACTH administration has been associated with an increased risk of bilateral adrenal hemorrhage in humans and rats.11,17 Trilostane decreases secretion of serum cortisol, which results in a loss of negative feedback at the level of the anterior pituitary gland and hypothalamus, with subsequent increases in endogenous plasma ACTH concentrations.9,20,21,23 Therefore, it is possible that elevated ACTH concentrations led to adrenal hemorrhagic necrosis in the subset of dogs in whom adrenal recovery was not documented, as has been observed in humans and rats.24
There are several notable limitations to this study, mainly because of its retrospective nature. Although each record was reviewed in depth, it is possible that some of these patients were misdiagnosed with either HAC or PDH, because no patient had pituitary imaging performed, and the LDDST, considered to be the screening test of choice, was not performed in 25% of patients. By coupling a dynamic adrenal hormone assay with clinical signs and laboratory abnormalities suggestive of HAC, attempts were made to eliminate this type of error. Initial trilostane regimens were nonuniform and included both once-and twice-daily dosing, along with compounded and commercially available trilostane products. Prescription and dosing of adrenal medications following diagnosis of iHC, along with frequency and timing of cortisol testing, were determined by individual clinicians instead of a standardized protocol, and this likely affected the magnitude and time to adrenal recovery. Patients were diagnosed with iHC over a range of 9 yr (from 2009 to 2018), during which time trilostane has undergone considerable changes in dosing recommendations. Finally, the incidence of iHC following trilostane therapy could not be determined because dogs with this diagnosis were included from other hospitals.
Conclusion
In this study, iHC was documented in 48 dogs with PDH following a wide range of trilostane dosing and length of therapy. No factors were identified to predict adrenal recovery, although 76.3% of the dogs in this study had cortisol concentrations ≥1.5 μg/dL within 6 mo of iHC diagnosis. Adrenocortical hormone status may fluctuate following initial diagnosis of iHC, so periodic cortisol monitoring and reassessment of adrenal-related medications is recommended.
Clinical signs compatible with hypocortisolemia were recorded in 32 dogs at diagnosis of iatrogenic hypocortisolemia; 4 dogs had one of the above listed clinical signs, 20 dogs had two clinical signs, 6 dogs had three clinical signs, and 2 dogs had four clinical signs. Gastrointestinal signs consisted of vomiting (9 dogs) and/or diarrhea (3 dogs). Neurological signs consisted of twitching (3 dogs), shaking (1 dog), suspected seizure (2 dogs), and ataxia (3 dogs). Urinary signs consisted of urinary accidents (3 dogs), decreased urination (1 dog), and pigmenturia (1 dog). Other signs consisted of stiffness (1 dog), vocalizing (1 dog), and tachypnea (1 dog).
Adrenal recovery (i.e., cortisol ≥1.5 μg/dL) was achieved in 29 dogs within selected time periods following diagnosis of iatrogenic hypocortisolemia. Blue boxes represent dogs receiving trilostane therapy, and red boxes represent dogs not receiving trilostane.
Kaplan-Meier survival curve (in months) for 48 dogs following diagnosis of iHC. Red line represents dogs who achieved adrenal recovery (cortisol ≥1.5 μg/dL) within 6 mo of iHC diagnosis. Blue line represents dogs whose cortisol remained <1.5 μg/dL within 6 mo of iHC diagnosis. iHC, iatrogenic hypocortisolemia.
Contributor Notes
From the Department of Internal Medicine, The Animal Medical Center, New York, New York (E.A.); Med Vet Indianapolis, Carmel, Indiana (A.S.); Lamb Statistical Consulting, West St. Paul, Minnesota (K.E.L.); and The Ohio State University College of Veterinary Medicine, Columbus, Ohio (C.L.).
