Corticotropin-releasing hormone (CRH) is a 41-amino acid peptide with an important physiologic role in the regulation of the hypothalamic-pituitary-adrenal (HPA) axis. As a diagnostic agent, CRH has been extensively used in the diagnosis and differential diagnosis of Cushing’s syndrome (CS). In the past, ovine CRH (oCRH) was primarily used in the USA, whereas human CRH (hCRH) was primarily used in European countries. Currently, CRH is not commercially available in the USA and several other countries.

The present chapter aims at reviewing data on the performance of diagnostic tests for CS in adults that have used CRH, as well as summarizing information on diagnostic tests that use desmopressin as an alternative to CRH. To compile the reference list, electronic literature searches (PubMed) were conducted, using the keywords CRH, CS, Cushing’s disease (CD), desmopressin, diagnosis, and inferior petrosal sinus sampling (IPSS). Articles were cited at the author’s discretion.

CRH is a 41-amino acid peptide that was isolated from the ovine hypothalamus by Wylie Vale and his colleagues [ , ]. The human CRH sequence is similar to that of the ovine CRH peptide, bearing seven amino acid substitutions [ ]. This peptide is derived by posttranslational processing from a larger precursor protein. Of note, CRH is structurally related to three other peptides (urocortin 1, 2, and 3) [ , ]. Although CRH expression is widespread in the central nervous system (CNS), available data suggest that it is the CRH expressed in the paraventricular nucleus (PVN) of the hypothalamus that has a central role in the regulation of the HPA axis, wherein it potently stimulates corticotropin (ACTH) secretion and proopiomelanocortin (POMC) expression, actions that are subject to negative feedback regulation by glucocorticoids [ ]. In essence, CRH has a critical role in orchestrating endocrine, autonomic, and behavioral responses to stress in mammals [ ].

CRH exerts its actions by binding to one of two cognate receptors, which are G-protein-coupled, seven-transmembrane domain proteins [ , ]. Of the two receptors, CRH-R1 is involved in the regulation of ACTH secretion in pituitary corticotrophs, whereas CRH-R2 is expressed in the heart and mediates cardiovascular responses [ , , ]. The canonical pathway for CRH-mediated ACTH secretion involves stimulation of Gsα, activation of adenylyl cyclase and protein kinase A, leading to exocytosis of ACTH-containing secretory granules and an increase in POMC transcription ( Fig. 12.1 ) [ ]. In addition to this canonical pathway, CRH receptor activation may also involve Gi- and Gq-mediated pathways [ ].

Canonical pathway of CRH signaling in pituitary corticotroph cells. Upon binding of CRH to its receptor (CRH-R1), activation of Gsα occurs, leading to activation of adenylyl cyclase and cAMP generation. This results in closure of the TREK-1 channel, cell depolarization, calcium influx through voltage-gated calcium channels, and ACTH exocytosis. In addition, cAMP activates protein kinase A, which in turn activates the BRAF-MEK1/2-ERK1/2 cascade (details omitted for clarity), resulting in increased POMC transcription. Besides this canonical pathway, CRH receptor activation may involve Gi- and Gq-mediated pathways (not shown). AC , adenylyl cyclase; ACTH , corticotropin; cAMP , cyclic adenosine monophosphate; CRH , corticotropin-releasing hormone; PKA , protein kinase A; POMC , proopiomelanocortin; TREK-1 , TWIK-related potassium channel; VGCC , voltage-gated calcium channel.

Of note, vasopressin has a physiologic role in stimulating ACTH secretion while acting in synergy with CRH [ ]. However, vasopressin does not increase POMC expression [ ]. Vasopressin effects on ACTH secretion are mediated via the arginine-vasopressin 1b receptor, which activates phospholipase C, leading to protein kinase C activation [ ].

The dexamethasone–CRH test has been used as a second-line test in the diagnosis of CS and can be particularly helpful in distinguishing between patients with CS and those with nonneoplastic hypercortisolism (“pseudo-Cushing’s states”) [ ] ( Table 12.1 ). The latter group comprises patients with conditions such as major depression, severe obesity, and chronic alcoholism, who often exhibit some clinical and laboratory findings suggestive of cortisol excess in the absence of tumorous hypercortisolism [ ]. These patients have some symptoms and signs suggestive of hypercortisolism as well as abnormalities in laboratory tests, such as the 1 mg dexamethasone suppression test (DST) or 24-hour urinary free cortisol, making it particularly difficult to distinguish them from patients with endogenous CS [ , ].

To conduct the dexamethasone–CRH test, the patient takes dexamethasone 0.5 mg by mouth every 6 hours for eight doses (first dose at 12:00 p.m.), followed by an (ovine or human) CRH infusion (1 mcg/kg, up to 100 mcg) at 8:00 a.m. (2 hours after the last dexamethasone dose). Blood specimens are collected for serum cortisol and plasma ACTH assays at baseline and 15 minutes after CRH administration. In the original study, a serum cortisol cutpoint of 1.4 mcg/dL at 15 minutes after CRH administration distinguished between 39 patients with CS and 19 patients with “pseudo-Cushing’s states” with 100% sensitivity and 100% specificity [ ]. The findings of subsequent studies evaluating the diagnostic accuracy of the dexamethasone–CRH test confirmed that the test has excellent diagnostic performance, albeit with less than perfect sensitivity and specificity. In a meta-analysis of nine studies, including 258 patients with CS and 210 individuals without CS, the diagnostic characteristics of the dexamethasone–CRH test (95% confidence interval) were as follows: sensitivity: 98.6% (90.4, 99.8), specificity: 85.9% (67.6, 94.7), positive likelihood ratio: 7.01 (2.80, 17.6), and negative likelihood ratio: 0.016 (0.002, 0.118) [ ]. The very low negative likelihood ratio suggests that normal results in the dexamethasone–CRH test have a high negative predictive value (i.e., excluding CS), with the exception of rare patients with cyclic hypercortisolism. The test is well-tolerated; transient flushing and palpitations may occur but are mild; rarely, transient dyspnea or hypotension may also occur.

Desmopressin (1-deamino-8- d -arginine vasopressin) is a vasopressin analog that is essentially devoid of pressor activity (since it does not activate arginine-vasopressin 1a receptors) [ ]. Desmopressin primarily engages arginine-vasopressin two receptors, resulting in antidiuresis, and has much lower affinity for the arginine-vasopressin 1b receptors (previously known as V3), stimulating ACTH secretion. Pituitary corticotroph adenomas usually express vasopressin and CRH receptors [ ]. However, there is controversy regarding the vasopressin receptor subtype expression profile in corticotroph adenomas [ , ].

Nonetheless, desmopressin has been used in dynamic testing to distinguish between CS and nonneoplastic hypercortisolism [ ]. The desmopressin test has gained popularity in light of the lack of commercially available CRH preparations in the USA and other countries, as well as the lower cost of desmopressin. The desmopressin test is performed in the morning. To conduct this test, an intravenous cannula is placed in an arm vein, and peripheral blood specimens are obtained for plasma ACTH and serum cortisol assays before (−15 minutes, 0 minutes) and after (15 minutes, 30 minutes, 45 minutes, 60 minutes, 90 minutes, 120 minutes) the intravenous administration of vasopressin, 10 mcg. The test is generally well tolerated except for occasional flushing and rare hypotension. To minimize the risk of possible hyponatremia after desmopressin administration, moderate fluid restriction (1.2 L/24 hours) is advisable for 24 hours after the test.

It should be noted that there is no consensus on the diagnostic cutpoints for plasma ACTH and serum cortisol used to interpret the results of the desmopressin test [ , , , ]. Criteria for a positive response include an increase in plasma ACTH exceeding 30 pg/mL above baseline or a peak ACTH above 70 pg/mL; an increase in serum cortisol over 6 mcg/dL or a peak cortisol over 18 mcg/dL. However, percentage increases in ACTH (35%–150%) or cortisol (20%–40%) above baseline values have also been proposed as diagnostic cutpoints by some investigators [ , , , ].

A recent meta-analysis of studies examining the diagnostic performance of the dexamethasone–CRH test and the desmopressin test reported that the desmopressin test had 88% sensitivity (95% confidence interval: 77%, 95%) and 94% specificity (95% confidence interval: 83%, 98%) in the diagnosis of CS [ ]. In the same meta-analysis, the dexamethasone–CRH test had 97% sensitivity (95% confidence interval: 88%, 99%) and 92% specificity (95% confidence interval: 84%, 96%) in the diagnosis of CS. It appears that both the dexamethasone–CRH test and the desmopressin test have good diagnostic performance as second-line tests in patients with suspected CS. However, the sensitivity of the desmopressin test appears to be slightly lower than that of the dexamethasone–CRH test, and the cutpoints used to interpret the desmopressin test vary widely across pertinent publications, thereby pointing to the need for larger studies in order to identify optimal diagnostic cutpoints.

Approximately 80% of adult patients with endogenous CS have ACTH-dependent hypercortisolism, which can be either of pituitary origin (CD) (70%) or may occur secondary to ectopic, ACTH-secreting, neuroendocrine tumors (10%) [ , ]. Accurate tumor localization is critical for patient management; however, this can be challenging in light of the typically small size of these neoplasms, which are often radiographically occult, and the presence of incidental sellar masses in 10% of the general population [ , , ].

The majority of pituitary corticotropinomas and a minority of ectopic, ACTH-secreting, neuroendocrine tumors express CRH and/or vasopressin receptors, whereas adrenal tumors do not [ , ]. These observations suggest that the in vivo responses to CRH or vasopressin administration may be of diagnostic utility in the differential diagnosis of patients with CS.

The CRH test has been studied in the differential diagnosis of patients with ACTH-dependent CS ( Table 12.1 ) [ , ]. The test is performed in the morning. To conduct this test, an intravenous cannula is placed in an arm vein, and peripheral blood specimens are obtained for plasma ACTH and serum cortisol assays before (−15 minutes, 0 minutes) and after (15 minutes, 30 minutes, 45 minutes, 60 minutes, 90 minutes, 120 minutes) the intravenous administration of (ovine or human) CRH (1 mcg/kg, up to 100 mcg). A 35%–40% increase in ACTH or a 20%–40% increase in cortisol over baseline has been reported to distinguish between patients with CD, who exhibit an increase in plasma ACTH and/or serum cortisol, and those with ectopic, ACTH-secreting tumors, who generally do not manifest such ACTH or cortisol responses [ , ]. In particular, the cortisol response to CRH is more specific for the diagnosis of CD than the ACTH response [ ]. The exact reason for this observation is not clear; however, it is possible that some ectopic, ACTH-secreting tumors synthesize relatively unprocessed ACTH precursors that may be less efficient than native ACTH in eliciting cortisol secretion from the adrenals.

Whether ovine CRH performs better than human CRH in distinguishing between CD and ectopic ACTH-secreting tumors during the CRH test is not entirely clear. The sensitivity and specificity of the ACTH response during the CRH test have been reported to be similar after ovine CRH versus human CRH administration [ ]. However, the sensitivity of the cortisol response during the CRH test has been reported to be lower after human CRH versus ovine CRH administration [ ].

In light of the limited availability of CRH, other dynamic tests have been used to distinguish between CD and ectopic ACTH secretion. The desmopressin test has been studied in the differential diagnosis of patients with ACTH-dependent CS. A 30%–50% increase in plasma ACTH or a 20% increase in serum cortisol over baseline has been proposed as a diagnostic cutpoint to distinguish between patients with CD, who generally respond to desmopressin, and those with ectopic ACTH-secreting tumors, who generally do not [ , ]. The overnight 8 mg (“high dose”) DST has also been used to distinguish between patients with CD, who generally exhibit a greater than 50% suppression in morning (9:00 a.m.) serum cortisol over baseline, and those with ectopic, ACTH-secreting tumors, who generally do not show such a response [ ]. It should be noted that concordant responses to the CRH test (ACTH and/or cortisol rise) and the 8 mg DST (cortisol suppression) predict CD with high specificity [ ].

Of note, a combined CRH–desmopressin test was also developed to improve the discriminatory ability of the CRH test in distinguishing between CD and ectopic ACTH-secreting tumors. An early study yielded promising observations and suggested that the combined CRH-desmopressin test had higher diagnostic accuracy than the CRH test [ ]. However, a subsequent study did not confirm the diagnostic superiority of the combined CRH-desmopressin test over the CRH test [ ]. As a corollary, the combined CRH-desmopressin test cannot be recommended at present.

A recent meta-analysis summarized published data pertaining to the diagnostic accuracy of the CRH test, the desmopressin test, and the 8 mg DST, used in the differential diagnosis of CS [ ]. Data on the sensitivity (95% confidence interval) of the three tests were as follows: CRH test: 86.9% (82.1%, 90.6%) and 86.2% (78.3%, 91.5%) (the endpoint being plasma ACTH and serum cortisol response, respectively); 8 mg DST: 80.8% (75.1%, 85.5%) (the endpoint being serum cortisol response); desmopressin test: 84.8% (81.4%, 87.8%) and 82.1% (75.3%, 87.4%) (the endpoint being plasma ACTH and serum cortisol response, respectively). In the same meta-analysis, data on the specificity (95% confidence interval) of the three tests were as follows: CRH test: 93.9% (87%, 98.3%) and 89.4% (82.8%, 93.7%) (endpoint being plasma ACTH and serum cortisol response, respectively); 8 mg DST: 84.5% (79.6%, 88.4%) (endpoint being serum cortisol response); desmopressin test: 69.9% (61.4%, 77.3%) and 73% (58.4%, 83.9%) (endpoint being plasma ACTH and serum cortisol response, respectively). In meta-regression, the diagnostic odds ratio for the ACTH response during the CRH test was significantly higher than that of all other tests ( P < .0001) [ ]. In summary, the diagnostic performance of the CRH test appears to be superior to that of the desmopressin test or the 8 mg DST in the differential diagnosis of CS. Concordant responses to the CRH test and the 8 mg DST are highly specific for CD.