- Email: [email protected]
Basal Circulating Cortisol Concentrations and Hypoadrenocorticism in Dogs
PA G E 4
A D VA N C E S
BASAL CIRCULATING CORTISOL CONCENTRATIONS AND HYPOADRENOCORTICISM IN DOGS
Background Naturally occurring hypoadrenocorticism is a deficiency of glucocorticoids with or without mineralocorticoid deficiency. Typical clinical signs are nonspecific and include lethargy, vomiting, diarrhea, weakness, tremors, collapse, polyuria, and polydipsia. Because of vague clinical signs, hypoadrenocorticism may be mistaken for renal failure, gastrointestinal tract disease, or neurologic disease, among others. Failure to correctly recognize hypoadrenocorticism before a sudden collapse of the patient occurs can risk the life of the patient. Therefore, adrenal gland function testing for hypoadrenocorticism in suspect cases is commonly performed. The gold standard for the diagnosis of hypoadrenocorticism is the adrenocorticotropic hormone (ACTH) stimulation test. However, the cost of the test has been escalating, and ACTH preparations are not always available. Several ACTH products have recently been withdrawn from the market in the Unites States. The only remaining FDA-approved product is cosyntropin which has become quite expensive. Compounded ACTH products are available from compounding pharmacies, but
A D VA N C E S
compounded drugs are not regulated by the FDA. Although also expensive, these products may vary among pharmacies with respect to potency, stability, purity, quality, and duration of activity. Basal serum or plasma cortisol concentrations are generally not considered clinically useful as a diagnostic test. This is based on cortisol being secreted episodically, and it periodically decreases to less than the reference limit in clinically normal dogs.
Objectives To determine whether basal serum or plasma cortisol concentration can be used as a screening test to rule out hypoadrenocorticism in dogs.
Procedure The medical records of 110 dogs with nonadrenal gland illnesses and 13 dogs with hypoadrenocorticism were examined retrospectively. The sensitivity and specificity of basal serum or plasma cortisol concentrations of either 1 µg/dl or less, or 2 µg/d or less to detect dogs with hypoadrenocorticism were estimated by use of the ACTH stimulation test as the gold standard.
Results Basal cortisol concentrations of 1 µg/dl or less had excellent sensitivity (100%) and specificity (98.2%) for detecting dogs with hypoadrenocorticism. For basal cortisol concentrations of 2 µg/dl or less, sensitivity was 100%, but specificity was 78.2%.
Author Conclusion On the basis of sensitivity and specificity, basal serum or plasma cortisol concentrations had high negative predictive values over a wide range of prevalence rates and can be used to rule out a diagnosis of hypoadrenocorticism. Dogs with basal cortisol concentrations of more than 2 µg/dl that are not receiving corticosteroids that could interfere with the cortisol assay are highly unlikely to have hypoadrenocorticism. However, if the basal cortisol concentration is 2 µg/dl or less, an ACTH stimulation test should be performed.
Inclusions One figure, 1 table, 22 references.
Editor Annotation The authors of this paper have provided clinicians with a new tool for assessing dogs for hypoadrenocorticism. Be careful,
though, because this new tool will pinch your fingers if you do not use it properly. The study results do not show that basal cortisol measurement are useful for the diagnosis of hypoadrenocorticism, just for ruling it out. This is a point so important for understanding this paper that the authors chose to put it in the article’s title. So, if you canceled your back ordered cosyntropin after reading this paper, you had better reconsider. I liked this study for several reasons. First, I think it provides information useful to clinicians. Second, the authors evaluated test performance in a group that included dogs that had hypoadrenocorticism and dogs that did not but looked clinically as if they might. This makes the study setting more like the real life testing situation. Third, they used relevant estimates of disease prevalence to evaluate test performance. It is the latter exercise that makes the study results most useful to clinicians. Disease prevalence affects the predictive value of the test. As prevalence increases, the positive predictive value (PPV, the probability of disease in a patient with a positive test) increases and the negative predictive value (NPV, the probability of not having disease in a patient with a negative test) decreases. A decrease in prevalence has the opposite effects on predictive values. Hypoadrenocorticism is not a common disease in dogs and the authors estimated prevalence at just 0.5% in the general population. At a prevalence of 0.5%, the basal cortisol level (using a cutoff of 2.0 µg/dl or less to indicate a positive test) has a PPV of 2.3% and a NPV of 100%. In a population of sick dogs in which the authors estimated a 15% prevalence of hypoadrenocorticism, the PPV was 45% and NPV was 100%. A NPV of 100% at low and high prevalence means that a negative test result (basal cortisol or more than 2.0 µg/dl) essentially rules out hypoadrenocorticism 100% of the time. But — this is when you pinch your fingers if you are not careful — the low PPV, even when the prevalence is high, means that the probability that a positive test (basal cortisol 2.0 µg/dl or less) is correct is only 45% — less accurate than a coin flip. The low PPV is the reason basal cortisol measurement cannot be used to diagnose (i.e., rule-in) hypoadrenocorticism and why an ACTH stimulation test is still needed for definitive diagnosis. (TS)
PAG E 5
Lennon EM, Boyle TE, Hutchins RG, et al. Use of basal serum or plasma cortisol concentrations to rule out a diagnosis of hypoadrenocorticism in dogs: 123 cases (2000-2005). J Am Vet Med Assoc 2007;231:413-416.
A D VA N C E S
BASAL CIRCULATING CORTISOL CONCENTRATIONS AND HYPOADRENOCORTICISM IN DOGS
Background Naturally occurring hypoadrenocorticism is a deficiency of glucocorticoids with or without mineralocorticoid deficiency. Typical clinical signs are nonspecific and include lethargy, vomiting, diarrhea, weakness, tremors, collapse, polyuria, and polydipsia. Because of vague clinical signs, hypoadrenocorticism may be mistaken for renal failure, gastrointestinal tract disease, or neurologic disease, among others. Failure to correctly recognize hypoadrenocorticism before a sudden collapse of the patient occurs can risk the life of the patient. Therefore, adrenal gland function testing for hypoadrenocorticism in suspect cases is commonly performed. The gold standard for the diagnosis of hypoadrenocorticism is the adrenocorticotropic hormone (ACTH) stimulation test. However, the cost of the test has been escalating, and ACTH preparations are not always available. Several ACTH products have recently been withdrawn from the market in the Unites States. The only remaining FDA-approved product is cosyntropin which has become quite expensive. Compounded ACTH products are available from compounding pharmacies, but
A D VA N C E S
compounded drugs are not regulated by the FDA. Although also expensive, these products may vary among pharmacies with respect to potency, stability, purity, quality, and duration of activity. Basal serum or plasma cortisol concentrations are generally not considered clinically useful as a diagnostic test. This is based on cortisol being secreted episodically, and it periodically decreases to less than the reference limit in clinically normal dogs.
Objectives To determine whether basal serum or plasma cortisol concentration can be used as a screening test to rule out hypoadrenocorticism in dogs.
Procedure The medical records of 110 dogs with nonadrenal gland illnesses and 13 dogs with hypoadrenocorticism were examined retrospectively. The sensitivity and specificity of basal serum or plasma cortisol concentrations of either 1 µg/dl or less, or 2 µg/d or less to detect dogs with hypoadrenocorticism were estimated by use of the ACTH stimulation test as the gold standard.
Results Basal cortisol concentrations of 1 µg/dl or less had excellent sensitivity (100%) and specificity (98.2%) for detecting dogs with hypoadrenocorticism. For basal cortisol concentrations of 2 µg/dl or less, sensitivity was 100%, but specificity was 78.2%.
Author Conclusion On the basis of sensitivity and specificity, basal serum or plasma cortisol concentrations had high negative predictive values over a wide range of prevalence rates and can be used to rule out a diagnosis of hypoadrenocorticism. Dogs with basal cortisol concentrations of more than 2 µg/dl that are not receiving corticosteroids that could interfere with the cortisol assay are highly unlikely to have hypoadrenocorticism. However, if the basal cortisol concentration is 2 µg/dl or less, an ACTH stimulation test should be performed.
Inclusions One figure, 1 table, 22 references.
Editor Annotation The authors of this paper have provided clinicians with a new tool for assessing dogs for hypoadrenocorticism. Be careful,
though, because this new tool will pinch your fingers if you do not use it properly. The study results do not show that basal cortisol measurement are useful for the diagnosis of hypoadrenocorticism, just for ruling it out. This is a point so important for understanding this paper that the authors chose to put it in the article’s title. So, if you canceled your back ordered cosyntropin after reading this paper, you had better reconsider. I liked this study for several reasons. First, I think it provides information useful to clinicians. Second, the authors evaluated test performance in a group that included dogs that had hypoadrenocorticism and dogs that did not but looked clinically as if they might. This makes the study setting more like the real life testing situation. Third, they used relevant estimates of disease prevalence to evaluate test performance. It is the latter exercise that makes the study results most useful to clinicians. Disease prevalence affects the predictive value of the test. As prevalence increases, the positive predictive value (PPV, the probability of disease in a patient with a positive test) increases and the negative predictive value (NPV, the probability of not having disease in a patient with a negative test) decreases. A decrease in prevalence has the opposite effects on predictive values. Hypoadrenocorticism is not a common disease in dogs and the authors estimated prevalence at just 0.5% in the general population. At a prevalence of 0.5%, the basal cortisol level (using a cutoff of 2.0 µg/dl or less to indicate a positive test) has a PPV of 2.3% and a NPV of 100%. In a population of sick dogs in which the authors estimated a 15% prevalence of hypoadrenocorticism, the PPV was 45% and NPV was 100%. A NPV of 100% at low and high prevalence means that a negative test result (basal cortisol or more than 2.0 µg/dl) essentially rules out hypoadrenocorticism 100% of the time. But — this is when you pinch your fingers if you are not careful — the low PPV, even when the prevalence is high, means that the probability that a positive test (basal cortisol 2.0 µg/dl or less) is correct is only 45% — less accurate than a coin flip. The low PPV is the reason basal cortisol measurement cannot be used to diagnose (i.e., rule-in) hypoadrenocorticism and why an ACTH stimulation test is still needed for definitive diagnosis. (TS)
PAG E 5
Lennon EM, Boyle TE, Hutchins RG, et al. Use of basal serum or plasma cortisol concentrations to rule out a diagnosis of hypoadrenocorticism in dogs: 123 cases (2000-2005). J Am Vet Med Assoc 2007;231:413-416.