CushieBlog

ACTH-secreting pituitary adenomas (PAs) causing Cushing’s disease (CD) represent the majority (~ 70–80%) of cases of endogenous Cushing’s syndrome (CS) in children older than 5 years of age [1, 2]. The diagnosis of CS can be complex and often requires multiple dynamic tests [3]. The Pituitary and Endocrine Societies recommend a stepwise diagnostic approach for suspected CS, starting with screening tests and proceeding to localization studies if hypercortisolism is confirmed [4, 5]. However, variability in assay performance, limited test availability, and the high incidence of incidental findings (such as pituitary incidentalomas), continue to pose challenges in selecting the most appropriate diagnostic tools and interpreting their results.

For patients with ACTH-dependent hypercortisolism, the diagnostic workup focuses on localizing the source of ACTH excess, most commonly a PA, though ectopic ACTH-secreting neuroendocrine tumors are also possible [5]. This becomes particularly challenging when pituitary MRI fails to reveal a visible tumor, which may occur in up to one third of patients [6]. ACTH-secreting PAs express receptors for corticotropin-releasing hormone (CRH) and administration of ovine-CRH (oCRH) stimulates ACTH and cortisol release [7]. The oCRH stimulation test, performed either in peripheral sampling or during bilateral inferior petrosal sinus sampling (BIPSS), has long been validated as a minimally invasive method to differentiate CD from ectopic sources of ACTH [8, 9]. However, the discontinuation of manufacturing of oCRH in the United States and the lower cost of alternative stimulants such as desmopressin (DDAVP, a synthetic arginine vasopressin analogue) have led to increased use of the DDAVP stimulation test.

DDAVP stimulates ACTH release via AVPR1b receptors found in corticotroph PAs, but not typically expressed in normal pituitary tissue or ectopic ACTH-secreting tumors [10]. Therefore, a rise in ACTH and cortisol following DDAVP is suggestive of CD [11, 12]. However, the lack of pediatric specific safety and accuracy data limit its use in the pediatric population.

In this study we describe the procedure of DDAVP stimulation test performed with peripheral sampling or in the context of BIPSS, and we compare its performance with the CRH test in a pediatric cohort.

Study design and patient selection

This was a single-center, retrospective study conducted at the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) in Bethesda, Maryland. All patients were enrolled under an IRB-approved protocol (Protocol ID: NCT 00001595) and they were evaluated at the National Institutes of Health (NIH) Clinical Center (CC). Written informed consent was provided by all parents and assent by pediatric patients if developmentally appropriate for all research procedures.

We identified patients with a final diagnosis of CD who have undergone peripheral DDAVP stimulation test (n = 32) or BIPSS with DDAVP stimulation (n = 15) between 2021 and 2025 (DDAVP group). We then reviewed our historic cohort and identified patients who have undergone peripheral oCRH stimulation test or BIPSS with oCRH stimulation during their diagnostic workup, matched 1:1 for age, sex and tumor size (CRH group).

The diagnosis of CS was based on clinical features and standard biochemical testing, including a 1 mg (or weight-based adjusted dose) overnight oral dexamethasone suppression test, late-night serum cortisol, and/or 24-hour (24h) urinary free cortisol (UFC), in accordance with current guidelines and adjusted for the pediatric population [3, 5]. All patients were eventually diagnosed with CD either by histologic confirmation of the diagnosis on the resected tumor, or clinical and biochemical remission after transsphenoidal surgery (TSS). Demographic, clinical, biochemical, imaging, surgical, and histopathology data were collected for analysis. Tumor size was recorded based on the MRI report, or if no adenoma was reported at the MRI, tumor size was recorded as 0.5 mm since the thinnest slice of the images we obtain are 1 mm, acknowledging that this assumption could underestimate the size of a larger tumor which lacked radiographic characteristics to be distinguished in the MR images. Cortisol was measured with solid-phase, competitive chemiluminescent enzyme immunoassay (CMIA) on Siemens Immulite 2500 analyzer (Malvern, PA) until 2020 and on Abbott Architect from 2020 until 2025. ACTH was measured with CMIA on Siemens Immulite 2500 analyzer until 2012 and on Immulite 200 XPi analyzer from 2012 until 2025. UFC was measured with chemiluminescent enzyme immunoassay until 2011 and with High Performance Liquid Chromatography/Tandem Mass Spectrometry since 2011 (LC-MS/MS). UFC is reported as both absolute values (mcg/24h) and as the fold change from the upper limit of normal (ULN), to account for variable reference range per age and assay.

Peripheral stimulation test

Patients were admitted at the inpatient pediatric floor of NIH CC at least one day prior to the procedure. An intravenous (IV) catheter was placed in the forearm at least one hour prior to the test initiation (most commonly 1–2 days prior to testing). Patients were fasting and remained lying in bed for the duration of the test. Samples were collected at times − 15 and 0 min prior to administration of stimulant at approximately 8:00am.

In the DDAVP group, 10mcg of DDAVP (2.5mL of 4mcg/mL solution) was administered via IV push over 30 s, followed by a 2mL normal saline flush. Samples were then collected at additional timepoints after administration of DDAVP at + 15, +30, + 45, and + 60 min. In a subset of patients, samples were collected at + 10, +20, + 30, +45, and + 60 min but results were not considered significantly different and eventually protocol was adjusted to sampling every 15 min. For this subset of patients (n = 7) the highest value of samples at + 10 and + 20 min was used as the + 15 min value. Patients were advised to follow moderate fluid restriction after DDAVP administration (max 40oz/1.2 L) for 24 h post-procedure, unless otherwise indicated by the treating physician. Intake/output monitoring was recommended for 24 h, and a repeat basic metabolic panel was obtained the following day.

In the CRH group, after baseline samples were obtained, patients received 1mcg/kg, max 100mcg, of oCRH via IV push, and samples were collected at times + 15, +30, and + 45 min after administration [3].

Samples were analyzed for cortisol and ACTH and the percentage change from baseline was calculated as: [(peak level after stimulation – baseline level)/baseline level]*100. The DDAVP test was considered consistent with CD based on previously published criteria: >18% increase in cortisol and >33% increase in ACTH [12]. The CRH test was considered consistent with CD if there was >20% increase in cortisol and >35% increase in ACTH [3].

Bilateral inferior petrosal sinus sampling (BIPSS)

BIPSS was performed based on standard protocols by an interventional radiologist under anesthesia as previously described [13]. Briefly, catheters were advanced to bilateral petrosal sinuses via radiological guidance through femoral veins. Blood samples were collected at all timepoints simultaneously from each of the petrosal sinus catheter (right, left) and peripheral samples drawn from a vascular catheter introducer sheath in a femoral vein. Baseline samples were collected at − 5 and 0 min. In the DDAVP group, after collecting samples at time 0 min, 10mcg of DDAVP (2.5mL of 4mcg/mL solution) was administered as an IV push over 30 s, followed by a 2mL normal saline flush. In the CRH group, after collecting samples at time 0 min, 1mcg/kg, max 100mcg, of oCRH was administered as an IV push over 30 s via peripheral IV catheter. Post-stimulation blood samples were collected at + 3, +5, and + 10 min. Patients who received DDAVP were advised to follow moderate fluid restriction as described above. Test results were considered consistent with CD if the baseline central:peripheral (C:P) ACTH ratio was >2 and/or the stimulated C:P ratio >3 [14,15,16].

Statistical analysis

Baseline characteristics were summarized using descriptive statistics. Non-normally distributed data are shown as median [Q1, Q3] and were compared between groups with the Wilcoxon rank-sum test. Normally distributed data are shown as mean (standard deviation, SD) and were compared between groups with student’s t-test. Categorical data are shown as counts and proportions and were compared between groups using χ² test or Fisher’s exact test as appropriate. To assess whether hormone levels or ratios changed significantly over time and differed between the CRH and DDAVP groups, variables were log-transformed to achieve approximately normal distribution, and two-way repeated measures analysis of variance (ANOVA) was performed with time and stimulation group as fixed effects. Area under the curve (AUC) was calculated for timepoints from 0 min to 45 min for the peripheral stimulation test, and from 0 min to 10 min for BIPSS. Sensitivity was calculated using predefined criteria, and Fisher’s exact test was employed to compare sensitivity between the CRH and DDAVP groups. Missing data were considered as missing by chance and were not replaced. A p-value < 0.05 was considered statistically significant. Analyses were conducted using R/RStudio software.

Cohort characteristics

In the peripheral stimulation test cohort, a total of 64 patients were included, consisting of 34 males (53%) and 30 females (47%). In accordance with the selection criteria, the two groups (DDAVP and CRH) were similar in age at time of testing, tumor size, and proportions of negative MRI at diagnosis. Markers of hypercortisolemia were also similar between the two groups, including late night serum cortisol, 24h UFC, and morning ACTH levels.

In the BIPSS analysis, a total of 30 patients were included, consisting of 18 males (60%) and 12 females (40%). The two groups were similar in age at time of testing, tumor size, and proportions of negative MRI at diagnosis. Markers of hypercortisolemia were also similar between the two groups. All corresponding results have been summarized in Table 1.

When looking into the percentage change from baseline, the median percentage change in cortisol was 58.2% [40.6, 85.7] after DDAVP and 82.3% [44.1, 110.0] after oCRH stimulation (p = 0.21), while the median percentage change in ACTH was 122% [53.5, 206.0] and 188.0% [117.0, 342.0] after DDAVP and oCRH stimulation respectively (p = 0.037, Fig. 2).

In the BIPSS stimulation test, the repeated measures analysis noted that in both groups there was significant effect of time (p < 0.001), but group and time-by-group interaction did not differ (p > 0.05), suggesting that overall, there was a similar change of the C:P ACTH ratios between the two groups post-stimulation (Fig. 1). AUCs for C:P ratios did not reach statistical significance between the two groups (p = 0.21). Peak ratios occurred more frequently at 3 min post-stimulation in both groups. Baseline C:P ACTH ratios were similar between the two groups (5.6 [2.1, 10.6] in the DDAVP group vs. 7.5 [4.2, 20.2] in the CRH group, p = 0.20) and the maximum C:P ratios post-stimulation remained similar (17.1 [8.9, 22.1] in the DDAVP group vs. 18.5 [12.3, 65.4] in the CRH group, p = 0.37, Fig. 2).

Sensitivity analysis

Using the diagnostic thresholds specified above for the peripheral stimulation test, sensitivity in the DDAVP group was 96.9% for cortisol and 81.3% for ACTH. In comparison, the CRH group showed 93.8% sensitivity for cortisol and higher sensitivity for ACTH (96.9%, Fig. 3), which however did not differ statistically between the two groups (p > 0.05).

Eight patients (12.5%) had a false negative response in either cortisol (range of percentage change of cortisol from baseline: 1–13%) or ACTH (range of percentage change of ACTH from baseline: −31-30%). Of these, six patients underwent DDAVP and two patients had oCRH stimulation test. Only two patients had inadequate response in both cortisol and ACTH and would have been misclassified (one in each of the DDAVP and CRH groups). Of the seven patients with false negative stimulation tests, only one was documented to have a hemolyzed blood sample. All patients who underwent DDAVP stimulation had an ACTH response of > 19%.

In the BIPSS analysis, using the thresholds above, sensitivity was 73.3% and 93.3% in the DDAVP and CRH group respectively (p = 0.33) for the baseline ratios, and 80% and 100% in for the stimulated ratios (p = 0.22, Fig. 3). Five (5) patients showed an inadequate response at baseline and three of them also had a false low response post-stimulation. All three of the patients who failed both the baseline and the post-stimulation cutoffs, and who would have been misclassified as possible ectopic CS, were in the DDAVP group and were previously described in a study of negative BIPSS results [13]. Two of these three patients were noted on venograms to have poorly developed inferior petrosal veins that suggested BIPSS might not yield reliable results.

Side effects

Overall patients tolerated the procedures without significant adverse events or complications. One patient in the CRH group reported mild headache, which resolved spontaneously within a few hours. No episodes of hyponatremia or venous thrombosis were recorded.