A Re-Examination of the oCRH Stimulation Test for the Diagnosis of Ectopic ACTH Secretion

Background: Previous studies of the oCRH test for the differential diagnosis of ACTH-dependent Cushing’s syndrome gave imprecise point estimates of specificity because of small numbers (n<20) of patients with ectopic ACTH secretion (EAS). We examined a large EAS population to re-assess the test’s performance and whether benign tumors were associated with a positive response (i.e. Cushing’s disease (CD). Methods: We evaluated 103 EAS patients, including 58 females, with median age of 38 years (range 10 - 83); 99 had surgically-proven tumors and four had biochemical testing consistent with EAS. ACTH and cortisol were measured 5 and 0 minutes before and 15, 30 and 45 minutes after administration of oCRH, 1 ug/kg up to 100 ug, iv. Previous criteria were used to determine EAS responses: mean ACTH increase of <34% at 15 and 30 min or mean cortisol increase of <20% at 30 and 45 min. Mean responses and UFC (expressed as fold increase above upper limit of normal, ULN) were compared with t-tests; a p-value of < 0.05 was considered significant. Results: Tumor types included NET (pulmonary n=67; with 4 spindle type and 5 tumorlets, thymic n=9, pancreas n=7, appendix n=1, prostate n=1), SCLC (N=3), pheochromocytoma/paraganglioma (n=3/1), metastatic without known primary (n=3), teratoma with corticotroph elements (n=1) and occult (n=4). 23 patients had “CD” responses that were significantly higher than those of non-responders: 11 with ACTH (mean+SD: 83.7 + 49.8% vs 1.2 + 18.8%, p=00025), 15 with cortisol (mean+SD: 34.7 + 8.9% vs -1.7 + 14.2%, p< 0.00001). Three patients responded to both ACTH and cortisol. Among the ACTH (only) responders, 6 had cortisol increases of < 10% (range 0-2-8.5%); a similar discordant response was seen among cortisol (only) responders, of whom 7 had ACTH increases of <20% (range -18 - 13.4%). 70% of responders, including 3 with both ACTH and cortisol responses, had pulmonary NET (n=13), spindle cell NET (n=2) or tumorlets (n=4). Other responders had thymic NET (n=3), appendiceal NET n=1, teratoma (n=1) or were occult (n=2). The patient with corticotroph elements in a teratoma had an ACTH response only. UFC fold-increases above ULN were similar in responders and non-responders (20.20 + 26.9 xULN vs 31.4 + 38.4 xULN, p =0.21). Four of 10 patients with UFC cyclicity responded to CRH. Conclusions: Among patients with EAS the rate of false positive responses to CRH was similar to that of initial reports, 11–15%, and was associated with hormonal cyclicity and presence of a thymic carcinoid. The marked discordance of ACTH and cortisol responses may result from reduced sensitivity to ACTH, less biologically active ACTH, or autonomous cortisol release.