In his editorial, Orth [1] raises several concerns about our paper [2]. He criticizes our use of cut-off criteria that maintain 100% test specificity. From a clinical standpoint, the rationale for achieving a specificity of nearly 100% is to avoid misclassifying cases of ectopic adrenocorticortropic hormone (ACTH) syndrome as Cushing disease, an error that would lead to inappropriate pituitary surgery.

Second, Dr. Orth suggests that restricting the study to patients whose 24-hour urine collections varied no more than 10% might have yielded different cut-off criteria and different study results. We have reanalyzed the study using coefficients of variation of 10% or less (Table 1). Diagnosis was based on information obtained at trans-sphenoidal surgery or during follow-up. The reanalysis yielded identical cut-off criteria. Thus, restricting the analysis to patients with less variation in creatinine exertion did not appreciably alter the results.

Dr. Orth also seeks additional information about the patient with primary adrenal disease who was initially thought to have Cushing disease. This 56-year-old woman had mild, intermittent hypercortisolism (urine free cortisol level, 41 to 229 µg/d; normal, <9 µg/d). She had a baseline plasma ACTH level of 22 pg/mL that increased to 105 pg/mL in response to treatment with corticotropin-releasing hormone. Twelve weeks later, her baseline plasma ACTH level at inferior petrosal sinus sampling was 7 pg/mL peripherally and 25 pg/mL in the right petrosal sinus. After administration of corticotropin-releasing hormone, her plasma ACTH levels increased to 1170 pg/mL in the right petrosal sinus, 25 pg/mL in the left petrosal sinus, and 13 pg/mL peripherally. The patient was thought to have Cushing disease, with macronodular adrenal hyperplasia as the explanation for a unilateral left adrenal nodule. However, no tumor was found during trans-sphenoidal surgery, and she did not respond to a right hemihypophysectomy. An iodocholesterol scan showed unilateral uptake in the adrenal nodule. She was cured of hypercortisolism by a left adrenalectomy. Intermittent hypercortisolism, with failed suppression of endogenous ACTH secretion from the normal corticotroph cells, may have caused the error in diagnosis [3, 4]. Responsiveness to corticotropin-releasing hormone and to metyrapone in this patient and in others with intermittent or mild hypercortisolism has taught us to interpret results of tests of differential diagnosis with great caution. The iodocholesterol scan is useful in determining whether hypercortisolism with a unilateral adrenal nodule represents an adrenal adenoma (unilateral uptake) or macronodular hyperplasia associated with Cushing disease (bilateral uptake).

Finally, Dr. Orth noted an inconsistency: The urine collections for the 2-day protocol should begin at 0600 h, as was done in our study, and not at 0800 h.