Characterization of a paradoxical post-operative increase in serum cortisol in Cushing disease.

Abstract

In Cushing disease, early post-operative serum cortisol fluctuations have not been adequately characterized, and their association with initial remission and recurrence is unclear. A retrospective cohort study of patients with Cushing disease was conducted at two institutions. A "riser" was defined a priori as a paradoxical increase in serum cortisol with an immediate incremental increase in serum cortisol over three consecutive cortisol draws separated by roughly 6-h (definition 1). Post hoc analyses used a definition of two consecutive increases (definition 2). Risers were compared to non-risers for initial remission and time-to-recurrence. A total of 505 patients with Cushing disease were screened, and 469 had adequate data for group assignment. Analysis of post-operative cortisol showed a subgroup of "risers" with a frequency of 3.6% for definition 1 and 42.6% for definition 2. In these patients, cortisol levels were significantly higher until approximately 36h post-operatively, and cortisol had a significantly longer mean serum half-life. In the post hoc analysis, definition 2 risers had a lower remission rate compared to non-risers (162/196, 82.7%, versus 243/264, 92.0%) with an odds ratio of 0.41 (0.23-0.73; p = 0.003). For both definitions, recurrence was similar between groups. We found that almost half of Cushing disease patients experienced a temporary increase in serum cortisol level during the early post-operative period. Serum cortisol half-life was longer, and the remission rates were lower, however, recurrence rates were similar to non-risers.