Endocrine Abstracts

Introduction: Metyrapone is used as holding treatment in patients with Cushing’s syndrome awaiting definitive treatment or as adjunctive treatment following pituitary irradiation. The metyrapone day curve is used to assess the adequacy of suppression of cortisol synthesis, and in our institution involves taking four blood cortisol samples at 9 am, 12 pm, 3 pm and 6 pm respectively. A mean result between 200 and 300 nmol/l is deemed satisfactory.

Aims: We investigated if assessment of treatment efficacy could be undertaken using one; the mean of the first two; or the mean of the first three samples, with implications for patients and resources.

Methods: Consecutive patients attending for metyrapone day curves since November 2005 were identified from the admissions book. Sample times and results were obtained from computerised laboratory records. Statistical analysis was undertaken using ROC curves with SPSS software.

Results: Eighteen patients (7 male/11 female; mean age 48.9 years; 9 Cushing’s disease/6 adrenal Cushing’s/3 ectopic ACTH syndrome) underwent 54 metyrapone day curves in the stipulated time. Five tests were discounted due to sampling errors. No patients were on medications known to interact with metyrapone. The areas under the ROC curve were .854 for the one-, .948 for the two- and 1.00 for the three-sample test. Using a cortisol value of 292 nmol/l as the goal the sensitivity and specificity of the three-sample test mirror that of the four-sample test. For the two-sample test, a target value of 306 nmol/l gives a 91.9% sensitivity and an 83.3% specificity.

Conclusions: The three-sample test is an adequate substitute for the four-sample test in identifying whether mean serum cortisol levels of 300 nmol/l have been achieved and would have cumulatively saved 9 out of 54 patient working days and obviated the need for 54 blood tests. The two-sample test is not an acceptable alternative in our view. It would have led to over-treatment in one and under-treatment in three cases.