The study was presented as an abstract at the 22nd ECVIM-CA Congress, Maastricht, Netherlands, September 6–8, 2012
Corresponding author: Nadja Sieber-Ruckstuhl, Clinic for Small Animal Internal Medicine, Vetsuisse Faculty University of Zurich, Winterthurerstrasse 260, Zurich 8057, Switzerland; e-mail: email@example.com.
The effectiveness of trilostane treatment is currently monitored by regular ACTH stimulation tests, which are time-consuming and expensive. Therefore, a monitoring system without a stimulation protocol and with less client expense would be preferable.
The aim of our study was to evaluate if baseline cortisol, endogenous ACTH (ACTH) concentration or the baseline cortisol to ACTH ratio (cortisol/ACTH ratio) could replace the ACTH stimulation test.
Forty trilostane-treated dogs with pituitary-dependent hypercortisolism (PDH) were included in this prospective study.
A total of 148 ACTH stimulation tests and 77 ACTH concentrations and cortisol/ACTH ratios were analyzed. Control of cortisol release was classified according to cortisol concentration after ACTH administration as excessive (<1.5 μg/dL; group 1), adequate (1.5–5.4 μg/dL; group 2), or inadequate (>5.4 μg/dL; group 3).
Baseline cortisol concentrations had considerable overlap between excessively, adequately, and inadequately controlled dogs. Only baseline cortisol >4.4 μg/dL (in 12% of tests) was a reliable diagnosis of inadequate control. Endogenous ACTH concentrations did not differ between groups. The overlap of the cortisol/ACTH ratio between groups was large. Correct classification was only possible if the cortisol/ACTH ratio was >15, which occurred in 4% of tests.
Conclusions and Clinical Importance
To monitor trilostane treatment the ACTH stimulation test cannot be replaced by baseline cortisol, ACTH concentration, or the cortisol/ACTH ratio.
Over the last decade, trilostane, a PO-administered competitive inhibitor of the 3β-hydroxysteroid dehydrogenase enzyme system, has become the treatment of choice for pituitary-dependent hypercortisolism (PDH) in dogs. It leads to a significant reduction in baseline and post-ACTH cortisol concentrations and induces a substantial improvement in clinical signs in 70–96% of cases.[1-5] Because of the loss of negative feedback inhibition, endogenous ACTH concentrations increase during trilostane treatment.[6-8]
Treatment response is monitored by assessing resolution of clinical signs of hypercortisolism (HC) and by measuring adrenocortical reserve capacity with an ACTH stimulation test at the time of maximal trilostane action.[1, 2] However, ACTH stimulation is time-consuming and expensive. Furthermore, concerns have been raised that ACTH in excessive amounts could have deleterious effects on adrenal glands. Therefore, a monitoring system without the need for ACTH stimulation would be preferable.
The baseline cortisol concentration has recently been investigated for this purpose. The authors of that study defined a range for baseline cortisol concentrations that could be used to reliably indicate acceptable control of adrenal gland function. With a value within that range excessive suppression and inadequate suppression seemed accurately excluded. However, in 63% of tests (214 of 342), baseline cortisol concentration did not fall into the defined range and ACTH stimulation had to be performed. Further studies are needed to resolve this uncertainty.
Finally, although ACTH concentrations increase and cortisol to ACTH ratios (cortisol/ACTH ratio) decrease during trilostane treatment, their usefulness as monitoring tools has not been addressed.[6-8]
Therefore, the aim of our study was to evaluate whether baseline cortisol, ACTH concentration, or the cortisol/ACTH ratio could replace the ACTH stimulation test to monitor trilostane treatment in dogs. We hypothesized that in many cases determination of a single baseline cortisol concentration alone is not helpful and that ACTH stimulation is finally needed. In addition, we proposed that dogs treated with trilostane and showing excessive control of cortisol release would have a higher ACTH concentration and a lower cortisol/ACTH ratio than dogs with adequate or inadequate control.
Materials and Methods
All procedures were conducted in accordance with guidelines established by the Animal Welfare Act of Switzerland.
Client-owned dogs with PDH diagnosed between April 2006 and December 2011 were prospectively enrolled in the study. Informed consent was obtained from the owners of all dogs. All dogs underwent a thorough clinical examination. Blood and urine samples were collected for a CBC, biochemical profile, urinalysis, and urine culture. Further work-up included a low-dose dexamethasone suppression test (LDDS), measurement of the UCCR in urine samples collected at home, determination of the endogenous ACTH concentration, and ultrasonographic examination of the adrenal glands. PDH was diagnosed on the basis of the dog's concentration of endogenous ACTH, a symmetrical ultrasonographic appearance (with or without enlargement) of the adrenal glands, or both. Dogs were included in the study when consistent clinical and laboratory findings for HC were present, the LDDS, the urinary UCCR, or both were positive, treatment with trilostane induced an adequate response, and at least three rechecks during trilostane treatment at our institution were conducted.
The ACTH stimulation tests were performed by collecting blood samples for the determination of serum cortisol before and 1 hour after intravenous injection of 0.25 mg of synthetic ACTH.1 Serum cortisol concentrations were determined by chemiluminescence assay.2 The LDDS test was performed by collecting blood samples before, 4 and 8 hours after an intravenous injection of 0.01 mg/kg dexamethasone.3 A cortisol concentration of ≥1 μg/dL 8 hours after dexamethasone administration was considered consistent with HC. UCCR measurements were performed on urine collected at home by the owners. A UCCR of more than 10 × 10−6 was considered abnormal. Endogenous ACTH was determined before ACTH stimulation by collecting blood into chilled EDTA-coated tubes placed on ice. After centrifugation at 4°C, plasma was stored at −80°C until assayed by a chemiluminescence assay.2
The initial dose of trilostane was 2–5 mg/kg body weight PO q24h. Efficacy of trilostane treatment was assessed by clinical signs and results of ACTH stimulation tests. Re-evaluations after starting trilostane treatment were scheduled after 1–2, 3–6, and 7–15 weeks, and after 4–7 and 8–12 months, and then every 6 months. At each re-evaluation, the test was performed 2–3 hours after trilostane administration in the morning. Treatment goal was to achieve a post-ACTH cortisol concentration of 1.5–5.4 μg/dL.4 In dogs with serum cortisol concentrations <1.5 μg/dL or >5.4 μg/dL after ACTH stimulation, the dose of trilostane was decreased or increased, respectively. Adjustments were made in increments of 5–20 mg/dog depending on the size of the dog.
Control of cortisol release was classified according to the post-ACTH cortisol concentrations as excessive (<1.5 μg/dL; group 1), adequate (1.5–5.4 μg/dL; group 2), or inadequate (>5.4 μg/dL; group 3).
Results were analyzed by means of nonparametric statistical methods.5, 6 Ranges and median values are reported. Differences between the groups were tested by the Kruskal-Wallis test and Dunn's post-test. Linear correlations were calculated by Spearman's nonparametric correlation. Differences were considered significant at values of P < .05. For values below the detection limit, the mean between 0 and the detection limit was entered for statistical analysis.
Forty dogs met the inclusion criteria. Age ranged from 6 to 16 years (median, 11 years) and body weight from 3.8 to 38.6 kg (median, 10.4 kg). There were 26 females (21 spayed) and 14 males (7 castrated). Breeds represented included crossbreed (n = 16), Dachshund (4), Yorkshire Terrier (3), Parson Jack Russell Terrier (2), Shih-Tzu (2), West Highland White Terrier (2), Australian Terrier (1), Bichon Frisé (1), Border Collie (1), Cairn Terrier (1), Cocker Spaniel (1), Galgo Español (1), German Shepherd (1), Pinscher (1), Tibetan Terrier (1), Toy Poodle (1), and Magyar Vizsla (1).
The LDDS test and ultrasonographic examination of the adrenal glands were performed in all dogs. Measurements of the UCCR and that of endogenous ACTH were performed in 16 and 27 dogs, respectively. In 38 dogs the 8-hour cortisol concentration in the LDDS test was >1 μg/dL and in 2 dogs <1 μg/dL. Of the last 2 dogs, one had one positive UCCR and the other had 2 positive UCCRs.
A total of 148 ACTH stimulation tests during trilostane treatment were analyzed. In relation to the start of trilostane treatment, the tests were distributed as follows: 38 tests were carried out after 1–2 weeks, 37 tests after 3–6 weeks, 32 tests after 7–15 weeks, 24 tests after 4–7 months, 10 tests after 8–12 months, and 7 tests after >1 year.
Baseline cortisol concentrations ranged from <0.2 to 8.6 μg/dL (median, 1.8 μg/dL) and post-ACTH concentrations from <0.2 to 22.6 μg/dL (median, 4.1 μg/dL). In 11/148 (7.4%) ACTH stimulation tests, the post-ACTH cortisol concentration was lower than the baseline concentration. There was a significant correlation between the baseline and the post-ACTH cortisol concentration (r = 0.80; P < .0001).
In 31 (21%) tests, the control of cortisol release was excessive (group 1), in 58 (39%) tests it was adequate (group 2), and in 59 (40%) tests it was inadequate (group 3).
Baseline Cortisol Concentrations of Different Groups
Baseline cortisol concentrations of group 1 ranged from <0.2 to 2.7 μg/dL (median, 0.4 μg/dL), that of group 2 from <0.2 to 4.4 μg/dL (median, 1.3 μg/dL), and that of group 3 from 0.6 to 8.6 μg/dL (median, 3.3 μg/dL) (Fig. 1). Baseline cortisol concentrations differed significantly between the 3 groups (P < 0.0001). However, because the overlap between the groups was large, it was not possible to define a target range which would reliably exclude excessive or inadequate suppression.
A baseline cortisol concentration >1 μg/dL excluded excessive suppression in 97% of dogs, but a value <1 μg/dL could not distinguish between tests with excessive, adequate or inadequate suppression. Similarly, a baseline cortisol concentration >4.4 μg/dL correctly identified inadequate suppression, but values below did not distinguish between excessive, adequate or inadequate suppression.
Correct classification was only possible in tests with baseline cortisol concentrations >4.4 μg/dL, which was seen in 18 (12%) tests.
Endogenous ACTH and Cortisol/ACTH Ratio
A total of 77 ACTH values were analyzed and 77 cortisol/ACTH ratios calculated.
There was no correlation between ACTH and baseline cortisol (P = .84, r = 0.02) or post-ACTH cortisol (P = .6, r = 0.07). ACTH concentrations of group 1, group 2, and group 3 ranged from 29.4 to 211 pg/mL (median, 52.9 pg/mL), 10 to 837 pg/mL (median, 86.3 pg/mL), and 25.9 to 1250 pg/mL (median, 76 pg/mL), respectively. ACTH concentrations were not significantly different between the 3 groups (P = .5) (Fig. 2).
The cortisol/ACTH ratio of groups 1, 2, and 3 ranged from 0.2 to 3.7 (median, 0.29), 0.3 to 14 (median, 2.3), and 0.3 to 41.6 (median, 5.0), respectively (Fig. 3). The cortisol/ACTH ratio differed significantly between groups 1 and 3 (P < .0001) and between groups 2 and 3 (P = .0026); however, the overlap was large. Correct classification was only possible in tests with a cortisol/ACTH ratio >15, which was seen in 6 (4%) tests.
The results of this study indicate that ACTH stimulation cannot be replaced by determination of baseline cortisol to monitor trilostane treatment, as baseline cortisol concentrations overlapped considerably between excessively, adequately, and inadequately controlled dogs.
Recently, baseline cortisol was evaluated as a monitoring tool during trilostane treatment. A baseline cortisol concentration >1.3 μg/dL excluded excessive control in 98% of tests. Similarly in our study, exclusion of excessive control was possible in 97% of tests, if baseline cortisol concentrations were >1 μg/dL. Although, these numbers are high, the question remains whether they are high enough in a clinical situation with a potentially life-threatening complication (hypocortisolism), as in 5/342 tests in the study above and in 3/148 tests in our study an overdosed dog would not have been recognized. Furthermore, although baseline cortisol values >1 μg/dL may readily exclude excessive suppression, low-baseline cortisol concentrations are not a reliable indicator for excessive control. Several of our dogs had low (<1 μg/dL) or undetectable baseline cortisol concentrations, but showed a good stimulation after ACTH application.
A good monitoring parameter should not only exclude excessive suppression, but also and more importantly, aid in the fine adjustment of trilostane treatment. Therefore, an optimal range for baseline cortisol was defined in the study above; if a test result was within that range, excessive control and inadequate suppression of cortisol release seemed excluded. However, in using such a range, 2 important points have to be considered: first, the actual number of tests with baseline cortisol concentrations that fall into this range. In 37% of their tests, the baseline cortisol concentrations were within that range; however, in 63% of tests ACTH stimulation was needed for the final classification. As the measurement of cortisol is usually not readily available, these dogs would have to return to the veterinary hospital. Second, if baseline cortisol is within the range, the question arises how correct the final classification will be. The percentage of correct classification if baseline cortisol was within the defined target was 88%. The reason for this unreliability of baseline cortisol to predict correct classification appears to be the considerable overlap of baseline cortisol values between excessively, adequately, and inadequately controlled dogs. In our study, the overlap of baseline cortisol between the three groups was even larger. To define an optimal range for baseline cortisol seemed therefore inappropriate. Consistent classification was only possible in tests with baseline cortisol concentrations >4.4 μg/dL. We therefore strongly believe that measurement of baseline cortisol concentrations cannot be used as an alternative to ACTH stimulation to determine the optimal dose of trilostane.
As cortisol concentrations decrease during trilostane treatment, the negative feedback to the pituitary is lost, which leads to an increase in ACTH.[6-8],7 In 2 studies, ACTH concentrations were found to be significantly elevated in overdosed animals compared with other trilostane-treated dogs.,8 Consequently, the authors of these studies hypothesized that ACTH concentrations and cortisol/ACTH ratios could be used to monitor treatment response.,8 In our study, no correlation between ACTH and baseline or post-ACTH cortisol concentration existed; in addition, no difference in ACTH concentrations in dogs treated with trilostane in relation to their control of cortisol release was documented. Increased values were seen in all groups and were not associated with trilostane overdosage. Cortisol/ACTH ratios differed significantly between groups, but showed a large overlap and were thus not helpful in discriminating between the different groups. Serum cortisol levels decreased with minimum levels 3 hours after drug dosing, ACTH concentrations increased after trilostane application and reached peak levels after 4–8 hours.7 The interindividual variations of ACTH levels over 24 hours were high, most likely because of the different duration of action of trilostane and different sensitivities of each dog to the negative feedback response.
A limitation of all the studies about trilostane treatment is the lack of consensus on the optimal time point for blood sampling during trilostane treatment and on the optimal treatment goal for post-ACTH cortisol concentrations. In this study, the ACTH stimulation test was performed 2–3 hours after trilostane administration. By contrast, the ACTH stimulation test was run 4–6 hours after trilostane application in the earlier study. Maximum effect of trilostane on glucocorticoid production is reached 2–4 hours after application of the drug.7 The timing of the ACTH stimulation test in relation to the trilostane administration must therefore be standardized to obtain comparable results. As clinicians schedule the ACTH stimulation test during trilostane treatment differently, the cortisol concentrations in different studies and the cortisol target ranges are difficult to compare. In our study, the time point of blood sampling and the target range of post-ACTH cortisol defined at the Vetoryl® consensus meeting in Amsterdam April 2006 were used. These are based on the time of maximal effect of trilostane and the clinical experiences of different specialists in veterinary endocrinology. More importantly, however, the target range used seems to reflect the clinical response to treatment in our population of dogs very accurately; during long-term therapy, dogs with post-ACTH cortisol concentrations above the target range are usually badly controlled and dogs with a post-ACTH cortisol concentration within the target range are well-controlled. In some of our dogs, however, clinical signs only disappear when post-ACTH cortisol concentrations are at the lower end of the target range. This is a phenomenon that has already been observed by other authors and has been attributed to interindividual differences in sensitivity to cortisol. Using the clinical picture as a guideline for trilostane treatment has been proposed by some endocrinologists. Many owners are highly satisfied and document a substantial clinical improvement, although the post-ACTH cortisol concentration is still increased. However, during long-term treatment these dogs, whereas much more active and showing much less polyuria and polydipsia, are not well controlled (e.g, they show poor hair growth). Therefore, we strongly believe that the clinical picture alone is not a good monitoring parameter, but should be used in combination with the post-ACTH cortisol concentration.
Many clinicians have started to use trilostane twice daily.[4, 5, 12] Results of our study cannot be extrapolated to dogs treated twice daily. Further studies are needed to evaluate the different hormone levels as monitoring parameters in these patients.
In summary, baseline cortisol, ACTH concentrations, and cortisol/ACTH ratio showed a considerable overlap between dogs with excessive, adequate, and inadequate control of cortisol release. Therefore, for the present, trilostane treatment still has to be monitored by regular ACTH stimulation tests and determination of the post-ACTH cortisol concentration.
The authors gratefully acknowledge all colleagues at the Clinic for Small Animal Internal Medicine, Vetsuisse Faculty University of Zurich for their contribution of cases.
Conflict of Interest: Authors disclose no conflict of interest.
Lehnert C, Neiger R. 24 hours hormone and electrolyte levels of dogs with pituitary-dependent hyperadrenocorticism treated with trilostane. Inaugural-Dissertation, Giessen 2007
Buijtels JJCWM, Galac S, Kooistra HS. Measurement of plasma ACTH concentration to determine the optimal dose of trilostane in dogs with pituitary-dependent hyperadrenocorticism. 13th ECVIM-CA Congress, Uppsala, Sweden (abstract)