The Short Synacthen (Corticotropin) Test Can Be Used to Predict Recovery of Hypothalamo-Pituitary-Adrenal Axis Function

The Short Synacthen (Corticotropin) Test Can Be Used to Predict Recovery of... Abstract Context The 250-μg short Synacthen (corticotropin) test (SST) is the most commonly used tool to assess hypothalamo-pituitary-adrenal (HPA) axis function. There are many potentially reversible causes of adrenal insufficiency (AI), but no data to guide clinicians as to the frequency of repeat testing or likelihood of HPA axis recovery. Objective To use the SST results to predict adrenal recovery. Design A retrospective analysis of 1912 SSTs data. Patients Seven hundred seventy-six patients with reversible causes of AI were identified who had at least two SSTs performed. A subgroup analysis was performed on individuals previously treated with suppressive doses of glucocorticoids (n = 110). Main Outcome Measures Recovery of HPA axis function. Results SST 30-minute cortisol levels above or below 350 nmol/L (12.7 μg/dL) best predicted HPA axis recovery [area under the curve (AUC) receiver operating curve (ROC) = 0.85; median recovery time: 334 vs 1368 days, P = 8.5 × 10–13]: 99% of patients with a 30-minute cortisol >350 nmol/L recovered adrenal function within 4 years, compared with 49% in those with cortisol levels <350 nmol/L. In the subgroup analysis, delta cortisol (30-minute–basal) best predicted the recovery (AUC ROC = 0.77; median recovery time: 262 vs 974 days, P = 7.0 × 10–6). No patient with a delta cortisol <100 nmol (3.6 μg/dL) and a subsequent 1-year random cortisol <200 nmol/L (7.3 μg/dL) recovered HPA axis function. Conclusions Cortisol levels across an SST can be used to predict recovery of AI and may guide the frequency of repeat testing and inform both clinicians and patients as to the likelihood of restoration of HPA axis function. Adrenal insufficiency (AI) is a life-threatening condition with an established increase in morbidity and mortality (1–3) that is characterized by the inability of the adrenal cortex to produce sufficient amounts of glucocorticoids and/or mineralocorticoids (4). AI has classically been subdivided into primary (conditions directly impacting upon adrenal function, such as Addison disease) and secondary AI (diseases limiting pituitary ACTH synthesis and secretion). Although many of these underlying conditions are believed to be irreversible, for example, the impact of pituitary radiotherapy, there are many situations in which recovery of hypothalamo-pituitary-adrenal (HPA) axis function may occur. As a specific example, the widespread use of glucocorticoid therapy (5, 6), largely for its anti-inflammatory actions, is associated with many adverse effects, including AI as a result of ACTH suppression (7, 8). Indeed, this is probably the commonest form of AI, although almost certainly underrecognized (1). There are very few studies in the published literature that have approached the assessment of recovery of HPA axis function in a systematic manner (9, 10). In the context of suppressive doses of glucocorticoids, the published studies have examined very small numbers of patients, often in pediatric populations (11, 12). As a result, there has been little consensus as to the components of glucocorticoid exposure (functions of both dose and time) that may contribute to HPA axis suppression and potential recovery (10). Clinicians are therefore often faced with the challenge of trying to predict when, and indeed if, HPA axis function will recover. There are currently no published data with which to guide clinicians as to an appropriate frequency of repeat dynamic testing, or information for both the clinician and patient as to the eventual likelihood of restoration of normal adrenal function. The short Synacthen (corticotropin) test (SST) at the conventional dose of 250 μg has been validated against the “gold standard” insulin tolerance test (ITT) to be a reliable tool in the investigation of patients with suspected AI (13–16). In contrast to the ITT, it is a simple test to perform, is well tolerated with very few adverse effects, and is relatively low cost. We, and others, have described the utility of a morning cortisol level to predict SST outcome as a strategy to rationalize the use of dynamic testing (11, 17); however, the results from the SST have the potential to be far more informative. It is well established that the 30-minute cortisol level is used as the criterion to define adequate or inadequate adrenal cortisol reserve, and is the standard by which decisions are made to instigate (or terminate) glucocorticoid replacement. This result provides a readout as to how the adrenal gland is functioning on that day and whether this is adequate or not. The test results are reliant upon the ability of the adrenal gland to respond to a pharmacological stimulus of synthetic ACTH, and although this may be reflected by the 30-minute cortisol, we speculate that the incremental response (delta cortisol: 30-minute minus 0-minute) might provide a predictive indicator for future adrenal gland recovery of function. We have therefore undertaken a retrospective analysis of repeat SSTs performed in patients with potentially reversible causes of AI to determine if there are features of the SST results (basal, 30-minute, or delta cortisol) that might both guide a strategy for repeat testing and in addition help to identify groups of patients in whom HPA axis function is likely (or unlikely) to be restored. Materials and Methods Patient selection We undertook an observational, retrospective analysis of repeat SSTs performed in three secondary/tertiary care centers across all medical specialties between January 2008 and May 2017 (Oxford University Hospital National Health Service Foundation Trust, United Kingdom; University Hospitals Birmingham National Health Service Foundation Trust, United Kingdom; and Sapienza University of Rome, Italy). During this time, 3942 SSTs were performed in 2535 individuals. Within this cohort, 776 patients were identified as having underlying conditions that could have the possibility of HPA axis function recovery and had undergone at least two SSTs. A total of 1912 SSTs were performed (mean number of tests per patient 2.5 ± 0.8). The mean total duration of follow-up (from first to last SST) was 1182 ± 284 days, and the mean time between the first and second SST was 340 ± 322 days. Patient demographics and indications for the SST are presented in Table 1. A subgroup analysis was performed in a cohort of 110 patients either currently taking or having previously been exposed to suppressive doses of glucocorticoid therapy who underwent a total of 277 SSTs. Patients who failed the SST and no longer required glucocorticoid treatment of their underlying medical condition were commenced on hydrocortisone replacement. Inhaled glucocorticoid therapy was continued between SSTs. Tests performed in patients treated with radiotherapy, Addison disease, congenital adrenal hyperplasia, adrenal metastases, and bilateral adrenalectomy or those in whom the indication for the SST was not clearly identified were excluded from the analysis as the potential for restoration of HPA axis function was thought to be limited or could not be determined. All data were collected from electronic medical records. Table 1. Baseline Characteristics of 776 Patients With Reversible Causes of AI Whole Cohort Glucocorticoid Exposed N 776 110 Age, y 53 ± 18 53 ± 17 Total number of SSTs 1912 277 Number of SSTs per patient 2.5 ± 0.8 2.5 ± 0.9 Total duration of follow-up, d 1182 ± 284 1166 ± 466 Time between first and second SST, d 340 ± 322 268 ± 215 Sex  Male 335 (43%) 50  Female 441 (57%) 60 SST indication  Previous treatment with suppressive doses of glucocorticoids (topical, nasal, inhaled, oral, intramuscular, intravenous) 110 (14%)   Prednisolone 49   Hydrocortisone 18   Dexamethasone 3   Others (budesonide, fluticasone, ciclesonide, beclometasone) 40 Postoperative following pituitary surgery 120 (15%) Pituitary adenoma (nonfunctioning, no surgery, no radiotherapy) 136 (17%) Others central nervous system tumors 154 (20%) Other pituitary conditions (including hypohysitis) 42 (5%) Adrenal disease (including unilateral adrenalectomy, adenoma, carcinoma) 34 (4%) Other indications (autoimmune disease, hyponatremia, vomiting, weight loss, hyperkalemia, hypoglycemia, hypotension, collapse, fatigue) 180 (23%) Whole Cohort Glucocorticoid Exposed N 776 110 Age, y 53 ± 18 53 ± 17 Total number of SSTs 1912 277 Number of SSTs per patient 2.5 ± 0.8 2.5 ± 0.9 Total duration of follow-up, d 1182 ± 284 1166 ± 466 Time between first and second SST, d 340 ± 322 268 ± 215 Sex  Male 335 (43%) 50  Female 441 (57%) 60 SST indication  Previous treatment with suppressive doses of glucocorticoids (topical, nasal, inhaled, oral, intramuscular, intravenous) 110 (14%)   Prednisolone 49   Hydrocortisone 18   Dexamethasone 3   Others (budesonide, fluticasone, ciclesonide, beclometasone) 40 Postoperative following pituitary surgery 120 (15%) Pituitary adenoma (nonfunctioning, no surgery, no radiotherapy) 136 (17%) Others central nervous system tumors 154 (20%) Other pituitary conditions (including hypohysitis) 42 (5%) Adrenal disease (including unilateral adrenalectomy, adenoma, carcinoma) 34 (4%) Other indications (autoimmune disease, hyponatremia, vomiting, weight loss, hyperkalemia, hypoglycemia, hypotension, collapse, fatigue) 180 (23%) Patients treated with radiotherapy, who had been diagnosed with Addison disease, congenital adrenal hyperplasia, adrenal metastases, or who had undergone bilateral adrenalectomy were excluded from the study. Data are expressed as mean ± SD where appropriate. Figures in parentheses reflect percentage of the entire cohort. View Large Table 1. Baseline Characteristics of 776 Patients With Reversible Causes of AI Whole Cohort Glucocorticoid Exposed N 776 110 Age, y 53 ± 18 53 ± 17 Total number of SSTs 1912 277 Number of SSTs per patient 2.5 ± 0.8 2.5 ± 0.9 Total duration of follow-up, d 1182 ± 284 1166 ± 466 Time between first and second SST, d 340 ± 322 268 ± 215 Sex  Male 335 (43%) 50  Female 441 (57%) 60 SST indication  Previous treatment with suppressive doses of glucocorticoids (topical, nasal, inhaled, oral, intramuscular, intravenous) 110 (14%)   Prednisolone 49   Hydrocortisone 18   Dexamethasone 3   Others (budesonide, fluticasone, ciclesonide, beclometasone) 40 Postoperative following pituitary surgery 120 (15%) Pituitary adenoma (nonfunctioning, no surgery, no radiotherapy) 136 (17%) Others central nervous system tumors 154 (20%) Other pituitary conditions (including hypohysitis) 42 (5%) Adrenal disease (including unilateral adrenalectomy, adenoma, carcinoma) 34 (4%) Other indications (autoimmune disease, hyponatremia, vomiting, weight loss, hyperkalemia, hypoglycemia, hypotension, collapse, fatigue) 180 (23%) Whole Cohort Glucocorticoid Exposed N 776 110 Age, y 53 ± 18 53 ± 17 Total number of SSTs 1912 277 Number of SSTs per patient 2.5 ± 0.8 2.5 ± 0.9 Total duration of follow-up, d 1182 ± 284 1166 ± 466 Time between first and second SST, d 340 ± 322 268 ± 215 Sex  Male 335 (43%) 50  Female 441 (57%) 60 SST indication  Previous treatment with suppressive doses of glucocorticoids (topical, nasal, inhaled, oral, intramuscular, intravenous) 110 (14%)   Prednisolone 49   Hydrocortisone 18   Dexamethasone 3   Others (budesonide, fluticasone, ciclesonide, beclometasone) 40 Postoperative following pituitary surgery 120 (15%) Pituitary adenoma (nonfunctioning, no surgery, no radiotherapy) 136 (17%) Others central nervous system tumors 154 (20%) Other pituitary conditions (including hypohysitis) 42 (5%) Adrenal disease (including unilateral adrenalectomy, adenoma, carcinoma) 34 (4%) Other indications (autoimmune disease, hyponatremia, vomiting, weight loss, hyperkalemia, hypoglycemia, hypotension, collapse, fatigue) 180 (23%) Patients treated with radiotherapy, who had been diagnosed with Addison disease, congenital adrenal hyperplasia, adrenal metastases, or who had undergone bilateral adrenalectomy were excluded from the study. Data are expressed as mean ± SD where appropriate. Figures in parentheses reflect percentage of the entire cohort. View Large Cortisol assays Two cortisol assays were used to analyze samples across the SST: Siemens ADVIA Centaur (Siemens Health Care Diagnostics, Frimley, United Kingdom) and the Roche Modular System (Roche, Mannheim, Germany). Of the 1912 SSTs that were included in the analysis, 1368 (from 559 patients) were analyzed on the Roche assay and 544 (from 217 patients) on the Siemens assay. The same assay was used for repeated testing on the same patient. Serum cortisol analyzed by Siemens ADVIA Centaur showed an interassay imprecision of 10.5% at 83 nmol/L, 6.0% at 524 nmol/L, and 7.0% at 904 nmol/L and by the Roche Modular System of <8% for levels between 76 and 925 nmol/L. An analysis directly comparing the results from the Roche and Siemens assays is presented within the Results section. SST protocol and interpretation All SSTs were performed between 9 and 12 am, at least 18 hours after the most recent dose of glucocorticoids. Individual clinicians determined the frequency of repeat testing on a case-by-case basis. Patients taking the oral contraceptive pill or other estrogen replacement were required to stop the treatment at least 6 weeks before the test. Blood was sampled for serum cortisol at baseline and after 30 minutes: baseline serum cortisol levels were measured prior to injection of 250 μg Synacthen (Questcor Operations Limited, Dublin, Ireland, for Siemens assays; Alliance Pharmaceuticals, Chippenham, United Kingdom, and Sigma-tau Pharmaceutical, Rome, Italy, for Roche assays) intramuscularly or intravenously. The 30-minute response to intramuscular or intravenous Synacthen has been shown to be equivalent (18). After administration of Synacthen, the patients were observed for 15 minutes for signs of any allergic reaction. The interpretation of the SST is based on the 30-minute serum cortisol where an adequate response to Synacthen was defined as >450 nmol/L for Siemens ADVIA Centaur (19), as >550 nmol/L for the Roche Generation I Modular System (tests done before February 2016) (19), and as >450 nmol/L for the Roche Generation II Modular System (tests done after February 2016). The incremental response to Synacthen was calculated as: delta cortisol = [30-minute – 0-minute cortisol]. Statistical methods A binomial logistic regression was performed on the whole cohort to ascertain the effects of selected variables on the likelihood that participants will show recovery at the subsequent test. Six variables were inserted into the model: age, sex, 30-minute cortisol, basal cortisol of the subsequent test, use of steroid medication, and different assay used. Linearity of the continuous variables with respect to the logit of the dependent variable was assessed via the Box-Tidwell (1962) procedure. A Bonferroni correction was applied using all six terms in the model resulting in statistical significance being accepted when P < 0.008. Based on this assessment, all continuous independent variables were found to be linearly related to the logit of the dependent variable. Data are expressed as median with 95% CI assuming a normal distribution. Receiver operating curve (ROC) analysis was performed using a uniform threshold for any selected variables [basal cortisol (0-minute), delta cortisol (30-minute minus basal), 30-minute cortisol, basal cortisol of the subsequent test] according to the 95% specificity on the ROC analysis. According to this, a cutoff of 100 nmol/L (3.6 μg/dL), 350 nmol/L (12.7 μg/dL), 100 nmol/L (3.6 μg/dL), and 200 nmol/L (7.3 μg/dL) were identified and selected respectively for basal cortisol (0-minute), 30-minute cortisol, and delta cortisol and for the basal cortisol of the subsequent test. Area under the curve (AUC) analysis was used to express the overall diagnostic accuracy of the index criterion. Kaplan-Meier survival analysis was conducted to assess the median time to recovery for each group, and pairwise log rank comparisons were conducted to determine which group had significantly different survival distributions. A P < 0.05 was considered indicative of a statistically significant difference. Statistical analyses were performed using SPSS version 24 (IBM software, Chicago, IL) and GraphPad Prism 7.0 software package (GraphPad Software, Inc., La Jolla, CA). Results The patients’ characteristics are presented in Table 1, including the relevant clinical indications as well as the number and timing of the SSTs performed. A total of 776 subjects were recruited, all with potentially reversible causes of AI. A subgroup analysis was performed in 110 patients with AI secondary to treatment with suppressive doses of glucocorticoids. Overall, 37% of patients of the whole cohort who initially failed the SST eventually went on to pass, and 57% of those with nonfunctioning pituitary tumors and 44% of those patients who underwent pituitary surgery eventually passed the SST. Logistic regression modeling The binomial logistic regression model was statistically significant (χ2 = 143.8, P < 0.0001) and explained 47.9% of the variance in adrenal recovery, correctly classifying 88.6% of cases. Sensitivity was 72%, specificity was 94.7%, positive predictive value was 83.2%, and negative predictive value was 90.0%. Of the six predictor variables incorporated into the model, two were statistically significant: 30-minute cortisol (P < 0.0001) and the basal cortisol of the subsequent test (P < 0.0001). Lower 30-minute cortisol and basal cortisol of the subsequent test were associated with an increased likelihood of failing the subsequent test. HPA axis recovery All information derived from the SST was examined for its ability to predict whether the subsequent SST would be passed or failed (Fig. 1a–1d). A 30-minute cortisol >350 nmol/L (12.7 μg/dL) was the best indicator for future adrenal recovery (ROC AUC = 0.85, Fig. 1e). In this group, the estimated median time to HPA axis recovery was 334 (95% CI: 300 to 368) days compared with 1368 days (1067 to 1668) in those with a 30-minute cortisol <350 nmol/L (12.7 μg/dL) (χ2 = 51.2, P < 0.0001). There was a clear stratification of response with those individuals with a 30-minute cortisol between 150 and 350 nmol/L (5.4 and 12.7 μg/dL) having a lower chance of recovery of HPA axis function, although recovery was worst in those with a 30-minute cortisol <150 nmol/L (5.4 μg/dL) [median time to recovery: <150 nmol/L = 1174 (979 to 1369) days, χ2 = 40.42, P < 0.0001; 150 to 350 nmol/L = 465 (288 to 641) days, χ2 = 14.50, P < 0.0001] (Fig. 1d). Combining a 30-minute cortisol level with a delta cortisol of above or below 100 nmol/L (3.6 μg/dL) did not improve the diagnostic accuracy of the test (ROC AUC = 0.84) (Fig. 1f). The percentages of patients that recovered HPA axis function at 1, 2, and 4 years after their initial SST are presented in Supplemental Table 1. Figure 1. View largeDownload slide Kaplan-Meier plots estimating time to recovery of HPA axis function in 776 patients with potentially reversible causes of AI stratified by (a) basal (0-min) cortisol of the same test, (b) delta cortisol (30-min – basal cortisol), (c) 30-min cortisol, and (d) 30-min cortisol stratified by cutoffs at >350, 150 to 350, and <150 nmol/L using data from their initial SST. (e) ROC curve analysis to determine the ability of the characteristics of the initial SST to predict eventual recovery of adrenal function. (f) Combining 30-min cortisol with delta cortisol (30-min – basal cortisol) levels does not improve the ability to predict passing a subsequent SST. (g) In those patients with a 30-min cortisol <350 nmol/L on their initial SST, a subsequent random morning cortisol (>18 h after the last replacement dose) of >200 nmol/L significantly increases the likelihood of HPA axis recovery. Figure 1. View largeDownload slide Kaplan-Meier plots estimating time to recovery of HPA axis function in 776 patients with potentially reversible causes of AI stratified by (a) basal (0-min) cortisol of the same test, (b) delta cortisol (30-min – basal cortisol), (c) 30-min cortisol, and (d) 30-min cortisol stratified by cutoffs at >350, 150 to 350, and <150 nmol/L using data from their initial SST. (e) ROC curve analysis to determine the ability of the characteristics of the initial SST to predict eventual recovery of adrenal function. (f) Combining 30-min cortisol with delta cortisol (30-min – basal cortisol) levels does not improve the ability to predict passing a subsequent SST. (g) In those patients with a 30-min cortisol <350 nmol/L on their initial SST, a subsequent random morning cortisol (>18 h after the last replacement dose) of >200 nmol/L significantly increases the likelihood of HPA axis recovery. To improve further the accuracy of the test, we combined the use of the 30-minute cortisol with the baseline cortisol of the subsequent SST, which was assumed to reflect a “random morning cortisol” (assessed between 9 and 12 am) in the absence of replacement therapy (>18 hours). The mean time between the first SST and random cortisol was 340 ± 322 days. In the group with a 30-minute cortisol <350 nmol/L (12.7 μg/dL) and a subsequent random morning cortisol <200 nmolL (7.3 μg/dL), 93% of individuals continued to have AI at 4 years [median time to recovery: 2083 (1928 to 2238) days, χ2 = 17.48, P = 3 × 10–5; ROC AUC = 0.95) (Fig. 1g). The proportion of patients recovering HPA axis over time in the whole cohort is presented in Fig. 2. Figure 2. View largeDownload slide HPA axis recovery rates after SST in 776 patients with potentially reversible causes of AI, including nonfunctioning pituitary tumors and pituitary surgery. Figure 2. View largeDownload slide HPA axis recovery rates after SST in 776 patients with potentially reversible causes of AI, including nonfunctioning pituitary tumors and pituitary surgery. The results of the analysis were not altered when corrected for both age and sex. In addition, a further analysis was undertaken in which only patients that failed their first test were included (n = 248), and the results were not different from those presented above (Supplemental Figs. 1 and 2). HPA axis recovery in glucocorticoid-exposed patients In the majority of cases, patients treated with suppressive doses of glucocorticoids do not have underlying endocrine disease, and therefore a dedicated subgroup analysis on these patients was performed (n = 110). In this group, the delta cortisol was the best predictor for future adrenal recovery: those patients with a delta cortisol >100 nmol/L (3.6 μg/dL) had a shorter estimated median time to recovery [262 (212 to 312) vs 974 (633 to 1314) days, χ2 = 20.233, P = 7.0 × 10–6; ROC AUC = 0.77) (Fig. 3a–3d). Figure 3. View largeDownload slide Kaplan-Meier plots estimating time to recovery of HPA axis function in 110 patients with AI due to exposure to suppressive doses of glucocorticoid therapy stratified by (a) basal (0-min) cortisol of the same test, (b) 30-min cortisol, and (c) delta cortisol (30-min – basal cortisol) of their initial SST. (d) ROC curve analysis to determine the ability of the characteristics of the initial SST to predict eventual recovery of adrenal function. Figure 3. View largeDownload slide Kaplan-Meier plots estimating time to recovery of HPA axis function in 110 patients with AI due to exposure to suppressive doses of glucocorticoid therapy stratified by (a) basal (0-min) cortisol of the same test, (b) 30-min cortisol, and (c) delta cortisol (30-min – basal cortisol) of their initial SST. (d) ROC curve analysis to determine the ability of the characteristics of the initial SST to predict eventual recovery of adrenal function. Combining the delta cortisol <100 nmol/L (3.6 μg/dL) with the subsequent random cortisol (above or below 200 nmol/L) helped to refine the predictive ability of the test; patients who had a random morning cortisol >200 nmol/L (>18 hours after their last replacement dose) following their initial SST had a median time to HPA axis recovery of 322 (134 to 357) days. No patient who had a subsequent random cortisol <200 nmol/L recovered from their iatrogenic AI over the 4-year duration of the study. The results of the analysis were not altered when delta cortisol values were corrected for baseline (0-minute) cortisol levels (data not shown). The proportion of glucocorticoid-exposed patients recovering HPA axis over time are presented in Fig. 4. Figure 4. View largeDownload slide HPA axis recovery rates after SST in 110 patients previously exposed to suppressive doses of glucocorticoid therapy. Figure 4. View largeDownload slide HPA axis recovery rates after SST in 110 patients previously exposed to suppressive doses of glucocorticoid therapy. Comparison between cortisol assays As described above, two different cortisol assays were used in this cohort study. In the main cohort, the logistic regression model as well as the ROC analysis did not show significant differences between assays. However, to confirm these findings, an analysis was performed for each assay individually using the value of the 30-minute cortisol and the subsequent random morning cortisol. Surprisingly, a 30-minute cortisol of 350 nmol/L (12.7 μg/dL) and a random morning cortisol of 200 nmol/L (7.3 μg/dL) represented the 95% specificity in the ROC analysis for both the assays. In addition, the proportion of patients that subsequently went on to pass the SST at 1, 2, and 4 years was entirely comparable between assays (1 year: 55% vs 57%; year 2: 88% vs 89%; year 4: 98% vs 99%; Roche vs Siemens) (Supplemental Table 2). Discussion We have demonstrated that the stimulated 30-minute cortisol from the SST can be used to predict the likelihood of HPA axis recovery. Furthermore, combining these measurements with assessment of a 1-year random morning cortisol measurement (between 9 and 12 am and >18 hours after the last replacement dose) provides an accurate prediction of those individuals who are likely to recover adrenal function and those in whom it may not. Notably, 57% of our patients with nonfunctioning pituitary tumors and 44% of those patients who underwent pituitary surgery recovered HPA axis function at subsequent testing. This implies that there is realistic potential for reversibility of secondary AI in these patients. We (17) and others (10) have previously described the use of a morning cortisol to assess adrenal reserve, but to date, there have been very little attempts to use the SST to inform a strategy for repeating testing that in addition might serve as a guide as to the likelihood of restoration of HPA axis function. There is currently a lack of consensus as to how and when to assess patients with pituitary disease for recovery of HPA axis function after surgery. Different approaches have been proposed including the use of a baseline morning cortisol (16, 17, 20, 21), SST (22), or the “gold standard” ITT (23, 24). The timing of postoperative testing is also not uniform, with marked variability across centers (24 hours, 48 hours, 1 to 6 weeks) (20). Further testing beyond this time point (6 weeks) is not regarded as routine and performed in some centers, but not in others. If a patient fails a SST postoperatively, they may end up on lifelong glucocorticoid replacement therapy. However, cases where recovery of HPA axis function is delayed are reported (23). More recently, a study has suggested that a significant number of patients with underlying pituitary tumors, including those that have had surgery, may recover over a median duration of 20 months (range: 8 to 51) (21). Taken together with our data, this would suggest that repeat assessments of the HPA axis are clearly justified. We have therefore proposed that in those individuals in whom there is the potential to recover from AI, the strategy for repeat testing could be guided by their SST results. In our cohort, those individuals who had a 30-minute cortisol level >350 nmol/L (12.7 μg/dL) were likely to recover more quickly, and therefore 6-month testing could be advocated. In contrast, the rate of recovery in those with a 30-minute cortisol <350 nmol/L (12.7 μg/dL) was much slower, and annual testing may be more appropriate in this group. Furthermore, we found that the use of a random morning cortisol could help to identify the group of patients who were most likely to recover HPA axis function and allow further rationalization of dynamic testing (Fig. 5). Figure 5. View largeDownload slide Proposed flow chart for the use of SST in patients with potentially reversible causes of AI. *Random morning cortisol was measured between 9 and 12 am and at least 18 h after the last dose of glucocorticoid. Figure 5. View largeDownload slide Proposed flow chart for the use of SST in patients with potentially reversible causes of AI. *Random morning cortisol was measured between 9 and 12 am and at least 18 h after the last dose of glucocorticoid. The increasing use of glucocorticoid therapy has led to a dramatic increase in the awareness and diagnoses of secondary AI in the context of “iatrogenic Cushing syndrome.” It is estimated that 2% to 3% of the population of the United Kingdom and United States are taking prescribed glucocorticoids, and these are most commonly used in the more elderly populations and at doses that are known to suppress the HPA axis (25). Data on recovery of the HPA axis after exposure to a suppressive dose of glucocorticoids are limited and often in studies that have recruited small numbers of patients (ranging from n = 1 to 49) (10, 26–35). In addition, much of the published literature has been in pediatric populations (35–40), and therefore simple extrapolation into the adult setting is not straightforward. These studies are limited not only in the variability of the populations that they have studied, but also in the differences in duration and dose exposure to glucocorticoids. As a result, the published literature that is currently available does not allow us to determine whether the duration, dose, or cumulative glucocorticoid exposure are the main drivers to the development of secondary AI in this context (9). There is substantial variability between individuals in their susceptibility to the development of the adverse metabolic effects associated with glucocorticoid use, and it is highly likely that the same will apply to the development (and subsequent potential for recovery) of HPA axis suppression. In total, 43% of our patients who at the time of their SST were either currently or previously exposed to suppressive doses of glucocorticoids regained HPA axis function. Based upon the results of the initial SST undertaken at steroid withdrawal, in this subgroup of patients, the delta cortisol was the best predictor of future recovery of HPA axis function, such that 95% of patients who had an initial delta cortisol >100 nmol/L (3.6 μg/dL) passed the SST within 4 years. Although 4-year recovery rates were lower in the delta cortisol <100 nmol/L (67%), combining this with a subsequent random morning cortisol identified a group of patients who were unlikely to recover from the AI (0% recover at 4 years for those with a subsequent random cortisol <200 nmol/L, 7.3 μg/dL). However, we acknowledge that this analysis was only undertaken in a small subcohort (n = 32) (Fig. 4) and requires confirmation in larger studies. As with the whole cohort discussed above, this stratification could be used to guide the frequency of repeat testing, and an algorithm is presented in Fig. 5. Our analysis did not shown differences in recovery rates between patients taking different steroid formulations. The majority of our patients had been taking prednisolone (49.5%), but the small numbers of individuals taking other preparations, as well as limitations in the availability of data relating to the exact dose and timing, do not allow us to make detailed comments or analysis with regards to the precise nature of the glucocorticoid exposure that might dictate the SST response. An additional limitation is the continuation of exposure to suppressive doses of glucocorticoid therapy between SSTs, which is likely to have occurred in the majority of patients taking inhaled stored treatment. Surprisingly, even though we have used assay-specific thresholds for defining a pass or fail of the SST, the logistic regression model demonstrated the independence of the analysis from the different assay methods used. The results for each assay (assessed independently) suggest that the same threshold of a 30-minute cortisol of >350 nmol/L and a 1-year random morning cortisol of >200 nmol/L (after 18 hours of steroid withdrawal) can be used. The reasons underpinning this are not entirely clear, although clearly results of the SST in this analysis are being used in a different context in this analysis (i.e., predicting recovery in future tests as opposed to assessing current HPA axis integrity). Further analysis across larger cohorts and including additional assays is clearly warranted. Importantly, in our study there are some limitations. This is a retrospective analysis of clinical data obtained from electronic patient records, and therefore there is the potential for selection bias in those individuals undergoing short Synacthen testing. We do not have accurate data on the dose and duration of exposure to glucocorticoids, and therefore we are unable to answer questions that relate to cumulative exposure that might predict AI. However, this does represent the largest analysis of this kind published to date. Our data extend to approximately 4 years of follow-up and therefore it is not possible to say whether further recovery of the HPA axis may occur after this time point. The algorithm that we have proposed is designed to guide and help the clinician, but is only derived from this data set and has not been tested or validated in other cohorts. As a result, it is important that individual clinicians use their discretion and clinical judgment in the frequency and interpretation of dynamic assessments of HPA axis function. In conclusion, the results generated from the 250 μg SST can provide useful information for both the clinician and patient to guide the frequency of repeat testing as well as the likelihood of HPA axis recovery. When combined with a subsequent random morning cortisol, the diagnostic accuracy improves significantly. Finally, this analysis has the potential to rationalize the use of the SST; with the cutoffs proposed in the algorithm (Fig. 5), 27% of SSTs could have been avoided, limiting both inconvenience for the patient and reducing medical and nursing costs. Abbreviations: Abbreviations: AI adrenal insufficiency AUC area under the curve HPA hypothalamo-pituitary-adrenal ITT insulin tolerance test ROC receiver operating curve SST short Synacthen (corticotropin) test Acknowledgments Financial Support: This work was supported by the Medical Research Council (MR/P011462/1 to J.W.T.), a National Institute for Health Research Oxford Biomedical Research Centre Principal Investigator Award (to J.W.T.), an Irish Endocrine Society Traveling Fellowship Grant (to C.P.W.), and the Exchange in Endocrinology Expertise Programme of the European Union of Medical Specialists 3E Fellowship (to R.P.). Disclosure Summary: The authors have nothing to disclose. References 1. Shulman DI , Palmert MR , Kemp SF ; Lawson Wilkins Drug and Therapeutics Committee . Adrenal insufficiency: still a cause of morbidity and death in childhood . Pediatrics . 2007 ; 119 ( 2 ): e484 – e494 . Google Scholar CrossRef Search ADS PubMed 2. Bergthorsdottir R , Leonsson-Zachrisson M , Odén A , Johannsson G . Premature mortality in patients with Addison’s disease: a population-based study . J Clin Endocrinol Metab . 2006 ; 91 ( 12 ): 4849 – 4853 . Google Scholar CrossRef Search ADS PubMed 3. Tomlinson JW , Holden N , Hills RK , Wheatley K , Clayton RN , Bates AS , Sheppard MC , Stewart PM ; West Midlands Prospective Hypopituitary Study Group . Association between premature mortality and hypopituitarism . Lancet . 2001 ; 357 ( 9254 ): 425 – 431 . Google Scholar CrossRef Search ADS PubMed 4. Bornstein SR , Allolio B , Arlt W , Barthel A , Don-Wauchope A , Hammer GD , Husebye ES , Merke DP , Murad MH , Stratakis CA , Torpy DJ . Diagnosis and treatment of primary adrenal insufficiency: an Endocrine Society clinical practice guideline . J Clin Endocrinol Metab . 2016 ; 101 ( 2 ): 364 – 389 . Google Scholar CrossRef Search ADS PubMed 5. Adcock IM , Mumby S . Glucocorticoids . Handb Exp Pharmacol . 2017 ; 237 : 171 – 196 . Google Scholar CrossRef Search ADS PubMed 6. Paragliola RM , Papi G , Pontecorvi A , Corsello SM . Treatment with synthetic glucocorticoids and the hypothalamus-pituitary-adrenal axis . Int J Mol Sci . 2017 ; 18 ( 10 ): 18 . Google Scholar CrossRef Search ADS 7. Henzen C , Suter A , Lerch E , Urbinelli R , Schorno XH , Briner VA . Suppression and recovery of adrenal response after short-term, high-dose glucocorticoid treatment . Lancet . 2000 ; 355 ( 9203 ): 542 – 545 . Google Scholar CrossRef Search ADS PubMed 8. Krasner AS . Glucocorticoid-induced adrenal insufficiency . JAMA . 1999 ; 282 ( 7 ): 671 – 676 . Google Scholar CrossRef Search ADS PubMed 9. Joseph RM , Hunter AL , Ray DW , Dixon WG . Systemic glucocorticoid therapy and adrenal insufficiency in adults: a systematic review . Semin Arthritis Rheum . 2016 ; 46 ( 1 ): 133 – 141 . Google Scholar CrossRef Search ADS PubMed 10. Broersen LH , Pereira AM , Jørgensen JOL , Dekkers OM . Adrenal insufficiency in corticosteroids use: systematic review and meta-analysis . J Clin Endocrinol Metab . 2015 ; 100 ( 6 ): 2171 – 2180 . Google Scholar CrossRef Search ADS PubMed 11. Woods CP , Argese N , Chapman M , Boot C , Webster R , Dabhi V , Grossman AB , Toogood AA , Arlt W , Stewart PM , Crowley RK , Tomlinson JW . Adrenal suppression in patients taking inhaled glucocorticoids is highly prevalent and management can be guided by morning cortisol . Eur J Endocrinol . 2015 ; 173 ( 5 ): 633 – 642 . Google Scholar CrossRef Search ADS PubMed 12. Gordijn MS , Rensen N , Gemke RJ , van Dalen EC , Rotteveel J , Kaspers GJ . Hypothalamic-pituitary-adrenal (HPA) axis suppression after treatment with glucocorticoid therapy for childhood acute lymphoblastic leukaemia . Cochrane Database Syst Rev . 2015 ; ( 8 ): CD008727 . 13. Hurel SJ , Thompson CJ , Watson MJ , Harris MM , Baylis PH , Kendall-Taylor P . The short Synacthen and insulin stress tests in the assessment of the hypothalamic-pituitary-adrenal axis . Clin Endocrinol (Oxf) . 1996 ; 44 ( 2 ): 141 – 146 . Google Scholar CrossRef Search ADS PubMed 14. Agha A , Tomlinson JW , Clark PM , Holder G , Stewart PM . The long-term predictive accuracy of the short synacthen (corticotropin) stimulation test for assessment of the hypothalamic-pituitary-adrenal axis . J Clin Endocrinol Metab . 2006 ; 91 ( 1 ): 43 – 47 . Google Scholar CrossRef Search ADS PubMed 15. Lindholm J , Kehlet H , Blichert-Toft M , Dinesen B , Riishede J . Reliability of the 30-minute ACTH test in assessing hypothalamic-pituitary-adrenal function . J Clin Endocrinol Metab . 1978 ; 47 ( 2 ): 272 – 274 . Google Scholar CrossRef Search ADS PubMed 16. Abdu TA , Elhadd TA , Neary R , Clayton RN . Comparison of the low dose short synacthen test (1 microg), the conventional dose short synacthen test (250 microg), and the insulin tolerance test for assessment of the hypothalamo-pituitary-adrenal axis in patients with pituitary disease . J Clin Endocrinol Metab . 1999 ; 84 ( 3 ): 838 – 843 . Google Scholar PubMed 17. Sbardella E , Isidori AM , Woods CP , Argese N , Tomlinson JW , Shine B , Jafar-Mohammadi B , Grossman AB . Baseline morning cortisol level as a predictor of pituitary-adrenal reserve: a comparison across three assays . Clin Endocrinol (Oxf) . 2017 ; 86 ( 2 ): 177 – 184 . Google Scholar CrossRef Search ADS PubMed 18. Wang TW , Wong MS , Smith JF , Howlett TA . The use of the short tetracosactrin test for the investigation of suspected pituitary hypofunction . Ann Clin Biochem . 1996 ; 33 ( Pt 2 ): 112 – 118 . Google Scholar CrossRef Search ADS PubMed 19. El-Farhan N , Pickett A , Ducroq D , Bailey C , Mitchem K , Morgan N , Armston A , Jones L , Evans C , Rees DA . Method-specific serum cortisol responses to the adrenocorticotrophin test: comparison of gas chromatography-mass spectrometry and five automated immunoassays . Clin Endocrinol (Oxf) . 2013 ; 78 ( 5 ): 673 – 680 . Google Scholar CrossRef Search ADS PubMed 20. Inder WJ , Hunt PJ . Glucocorticoid replacement in pituitary surgery: guidelines for perioperative assessment and management . J Clin Endocrinol Metab . 2002 ; 87 ( 6 ): 2745 – 2750 . Google Scholar CrossRef Search ADS PubMed 21. Munro V , Tugwell B , Doucette S , Clarke DB , Lacroix A , Imran SA . Recovery of adrenal function after chronic secondary adrenal insufficiency in patients with hypopituitarism . Clin Endocrinol (Oxf) . 2016 ; 85 ( 2 ): 216 – 222 . Google Scholar CrossRef Search ADS PubMed 22. Honegger J , Buchfelder M , Fahlbusch R . Surgical treatment of craniopharyngiomas: endocrinological results . J Neurosurg . 1999 ; 90 ( 2 ): 251 – 257 . Google Scholar CrossRef Search ADS PubMed 23. Arafah BM , Kailani SH , Nekl KE , Gold RS , Selman WR . Immediate recovery of pituitary function after transsphenoidal resection of pituitary macroadenomas . J Clin Endocrinol Metab . 1994 ; 79 ( 2 ): 348 – 354 . Google Scholar PubMed 24. Auchus RJ , Shewbridge RK , Shepherd MD . Which patients benefit from provocative adrenal testing after transsphenoidal pituitary surgery ? Clin Endocrinol (Oxf) . 1997 ; 46 ( 1 ): 21 – 27 . Google Scholar CrossRef Search ADS PubMed 25. van Staa TP , Leufkens HG , Abenhaim L , Begaud B , Zhang B , Cooper C . Use of oral corticosteroids in the United Kingdom . QJM . 2000 ; 93 ( 2 ): 105 – 111 . Google Scholar CrossRef Search ADS PubMed 26. Cydulka RK , Emerman CL . A pilot study of steroid therapy after emergency department treatment of acute asthma: is a taper needed ? J Emerg Med . 1998 ; 16 ( 1 ): 15 – 19 . Google Scholar CrossRef Search ADS PubMed 27. Miró J , Amado JA , Pesquera C , López-Cordovilla JJ , Berciano J . Assessment of the hypothalamic-pituitary-adrenal axis function after corticosteroid therapy for MS relapses . Acta Neurol Scand . 1990 ; 81 ( 6 ): 524 – 528 . Google Scholar CrossRef Search ADS PubMed 28. Barrier JH , Larrose C , Magadur/Joly G , Hamidou M . Potential risks of the withdrawal of corticoids in Horton’s disease. Prospective study of 22 patients . Rev Med Interne . 1993 ; 14 : 976 . Google Scholar CrossRef Search ADS PubMed 29. Boots JM , van den Ham EC , Christiaans MH , van Hooff JP . Risk of adrenal insufficiency with steroid maintenance therapy in renal transplantation . Transplant Proc . 2002 ; 34 ( 5 ): 1696 – 1697 . Google Scholar CrossRef Search ADS PubMed 30. Bacon PA , Myles AB , Beardwell CG , Daly JR . Corticosteroid withdrawal in rheumatoid arthritis . Lancet . 1966 ; 2 ( 7470 ): 935 – 937 . Google Scholar CrossRef Search ADS PubMed 31. Campieri M , Ferguson A , Doe W , Persson T , Nilsson LG ; The Global Budesonide Study Group . Oral budesonide is as effective as oral prednisolone in active Crohn’s disease . Gut . 1997 ; 41 ( 2 ): 209 – 214 . Google Scholar CrossRef Search ADS PubMed 32. Rodger RS , Watson MJ , Sellars L , Wilkinson R , Ward MK , Kerr DN . Hypothalamic-pituitary-adrenocortical suppression and recovery in renal transplant patients returning to maintenance dialysis . Q J Med . 1986 ; 61 ( 235 ): 1039 – 1046 . Google Scholar PubMed 33. Desramé J , Sabaté JM , Agher R , Bremont C , Gaudric M , Couturier D , Chaussade S . Assessment of hypothalamic-pituitary-adrenal axis function after corticosteroid therapy in inflammatory bowel disease . Am J Gastroenterol . 2002 ; 97 ( 7 ): 1785 – 1791 . Google Scholar CrossRef Search ADS PubMed 34. Jamilloux Y , Liozon E , Pugnet G , Nadalon S , Heang Ly K , Dumonteil S , Gondran G , Fauchais AL , Vidal E . Recovery of adrenal function after long-term glucocorticoid therapy for giant cell arteritis: a cohort study . PLoS One . 2013 ; 8 ( 7 ): e68713 . Google Scholar CrossRef Search ADS PubMed 35. Pugnet G , Sailler L , Astudillo L , Vernet J , Bennet A , Arlet P . Frequency and risk factors of biological adrenal insufficiency screened by the 250 microg Synacthene stimulation test after a prolonged course of systemic glucocorticoid therapy. A study of 100 patients . Rev Med Interne . 2010 ; 31 ( 5 ): 332 – 336 . Google Scholar CrossRef Search ADS PubMed 36. Mendoza-Cruz AC , Wargon O , Adams S , Tran H , Verge CF . Hypothalamic-pituitary-adrenal axis recovery following prolonged prednisolone therapy in infants . J Clin Endocrinol Metab . 2013 ; 98 ( 12 ): E1936 – E1940 . Google Scholar CrossRef Search ADS PubMed 37. Graber AL , Ney RL , Nicholson WE , Island DP , Liddle GW . Natural history of pituitary-adrenal recovery following long-term suppression with corticosteroids . J Clin Endocrinol Metab . 1965 ; 25 ( 1 ): 11 – 16 . Google Scholar CrossRef Search ADS PubMed 38. Cunha CF , Silva IN , Finch FL . Early adrenocortical recovery after glucocorticoid therapy in children with leukemia . J Clin Endocrinol Metab . 2004 ; 89 ( 6 ): 2797 – 2802 . Google Scholar CrossRef Search ADS PubMed 39. Tominaga A , Uozumi T , Arita K , Kurisu K , Yano T , Hirohata T . Anterior pituitary function in patients with nonfunctioning pituitary adenoma: results of longitudinal follow-up . Endocr J . 1995 ; 42 ( 3 ): 421 – 427 . Google Scholar CrossRef Search ADS PubMed 40. Yedinak C , Hameed N , Gassner M , Brzana J , McCartney S , Fleseriu M . Recovery rate of adrenal function after surgery in patients with acromegaly is higher than in those with non-functioning pituitary tumors: a large single center study . Pituitary . 2015 ; 18 ( 5 ): 701 – 709 . Google Scholar CrossRef Search ADS PubMed Copyright © 2018 Endocrine Society http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Clinical Endocrinology and Metabolism Oxford University Press

The Short Synacthen (Corticotropin) Test Can Be Used to Predict Recovery of Hypothalamo-Pituitary-Adrenal Axis Function

Loading next page...
 
/lp/ou_press/the-short-synacthen-corticotropin-test-can-be-used-to-predict-recovery-W3GiY2qbUt
Publisher
Endocrine Society
Copyright
Copyright © 2018 Endocrine Society
ISSN
0021-972X
eISSN
1945-7197
D.O.I.
10.1210/jc.2018-00529
Publisher site
See Article on Publisher Site

Abstract

Abstract Context The 250-μg short Synacthen (corticotropin) test (SST) is the most commonly used tool to assess hypothalamo-pituitary-adrenal (HPA) axis function. There are many potentially reversible causes of adrenal insufficiency (AI), but no data to guide clinicians as to the frequency of repeat testing or likelihood of HPA axis recovery. Objective To use the SST results to predict adrenal recovery. Design A retrospective analysis of 1912 SSTs data. Patients Seven hundred seventy-six patients with reversible causes of AI were identified who had at least two SSTs performed. A subgroup analysis was performed on individuals previously treated with suppressive doses of glucocorticoids (n = 110). Main Outcome Measures Recovery of HPA axis function. Results SST 30-minute cortisol levels above or below 350 nmol/L (12.7 μg/dL) best predicted HPA axis recovery [area under the curve (AUC) receiver operating curve (ROC) = 0.85; median recovery time: 334 vs 1368 days, P = 8.5 × 10–13]: 99% of patients with a 30-minute cortisol >350 nmol/L recovered adrenal function within 4 years, compared with 49% in those with cortisol levels <350 nmol/L. In the subgroup analysis, delta cortisol (30-minute–basal) best predicted the recovery (AUC ROC = 0.77; median recovery time: 262 vs 974 days, P = 7.0 × 10–6). No patient with a delta cortisol <100 nmol (3.6 μg/dL) and a subsequent 1-year random cortisol <200 nmol/L (7.3 μg/dL) recovered HPA axis function. Conclusions Cortisol levels across an SST can be used to predict recovery of AI and may guide the frequency of repeat testing and inform both clinicians and patients as to the likelihood of restoration of HPA axis function. Adrenal insufficiency (AI) is a life-threatening condition with an established increase in morbidity and mortality (1–3) that is characterized by the inability of the adrenal cortex to produce sufficient amounts of glucocorticoids and/or mineralocorticoids (4). AI has classically been subdivided into primary (conditions directly impacting upon adrenal function, such as Addison disease) and secondary AI (diseases limiting pituitary ACTH synthesis and secretion). Although many of these underlying conditions are believed to be irreversible, for example, the impact of pituitary radiotherapy, there are many situations in which recovery of hypothalamo-pituitary-adrenal (HPA) axis function may occur. As a specific example, the widespread use of glucocorticoid therapy (5, 6), largely for its anti-inflammatory actions, is associated with many adverse effects, including AI as a result of ACTH suppression (7, 8). Indeed, this is probably the commonest form of AI, although almost certainly underrecognized (1). There are very few studies in the published literature that have approached the assessment of recovery of HPA axis function in a systematic manner (9, 10). In the context of suppressive doses of glucocorticoids, the published studies have examined very small numbers of patients, often in pediatric populations (11, 12). As a result, there has been little consensus as to the components of glucocorticoid exposure (functions of both dose and time) that may contribute to HPA axis suppression and potential recovery (10). Clinicians are therefore often faced with the challenge of trying to predict when, and indeed if, HPA axis function will recover. There are currently no published data with which to guide clinicians as to an appropriate frequency of repeat dynamic testing, or information for both the clinician and patient as to the eventual likelihood of restoration of normal adrenal function. The short Synacthen (corticotropin) test (SST) at the conventional dose of 250 μg has been validated against the “gold standard” insulin tolerance test (ITT) to be a reliable tool in the investigation of patients with suspected AI (13–16). In contrast to the ITT, it is a simple test to perform, is well tolerated with very few adverse effects, and is relatively low cost. We, and others, have described the utility of a morning cortisol level to predict SST outcome as a strategy to rationalize the use of dynamic testing (11, 17); however, the results from the SST have the potential to be far more informative. It is well established that the 30-minute cortisol level is used as the criterion to define adequate or inadequate adrenal cortisol reserve, and is the standard by which decisions are made to instigate (or terminate) glucocorticoid replacement. This result provides a readout as to how the adrenal gland is functioning on that day and whether this is adequate or not. The test results are reliant upon the ability of the adrenal gland to respond to a pharmacological stimulus of synthetic ACTH, and although this may be reflected by the 30-minute cortisol, we speculate that the incremental response (delta cortisol: 30-minute minus 0-minute) might provide a predictive indicator for future adrenal gland recovery of function. We have therefore undertaken a retrospective analysis of repeat SSTs performed in patients with potentially reversible causes of AI to determine if there are features of the SST results (basal, 30-minute, or delta cortisol) that might both guide a strategy for repeat testing and in addition help to identify groups of patients in whom HPA axis function is likely (or unlikely) to be restored. Materials and Methods Patient selection We undertook an observational, retrospective analysis of repeat SSTs performed in three secondary/tertiary care centers across all medical specialties between January 2008 and May 2017 (Oxford University Hospital National Health Service Foundation Trust, United Kingdom; University Hospitals Birmingham National Health Service Foundation Trust, United Kingdom; and Sapienza University of Rome, Italy). During this time, 3942 SSTs were performed in 2535 individuals. Within this cohort, 776 patients were identified as having underlying conditions that could have the possibility of HPA axis function recovery and had undergone at least two SSTs. A total of 1912 SSTs were performed (mean number of tests per patient 2.5 ± 0.8). The mean total duration of follow-up (from first to last SST) was 1182 ± 284 days, and the mean time between the first and second SST was 340 ± 322 days. Patient demographics and indications for the SST are presented in Table 1. A subgroup analysis was performed in a cohort of 110 patients either currently taking or having previously been exposed to suppressive doses of glucocorticoid therapy who underwent a total of 277 SSTs. Patients who failed the SST and no longer required glucocorticoid treatment of their underlying medical condition were commenced on hydrocortisone replacement. Inhaled glucocorticoid therapy was continued between SSTs. Tests performed in patients treated with radiotherapy, Addison disease, congenital adrenal hyperplasia, adrenal metastases, and bilateral adrenalectomy or those in whom the indication for the SST was not clearly identified were excluded from the analysis as the potential for restoration of HPA axis function was thought to be limited or could not be determined. All data were collected from electronic medical records. Table 1. Baseline Characteristics of 776 Patients With Reversible Causes of AI Whole Cohort Glucocorticoid Exposed N 776 110 Age, y 53 ± 18 53 ± 17 Total number of SSTs 1912 277 Number of SSTs per patient 2.5 ± 0.8 2.5 ± 0.9 Total duration of follow-up, d 1182 ± 284 1166 ± 466 Time between first and second SST, d 340 ± 322 268 ± 215 Sex  Male 335 (43%) 50  Female 441 (57%) 60 SST indication  Previous treatment with suppressive doses of glucocorticoids (topical, nasal, inhaled, oral, intramuscular, intravenous) 110 (14%)   Prednisolone 49   Hydrocortisone 18   Dexamethasone 3   Others (budesonide, fluticasone, ciclesonide, beclometasone) 40 Postoperative following pituitary surgery 120 (15%) Pituitary adenoma (nonfunctioning, no surgery, no radiotherapy) 136 (17%) Others central nervous system tumors 154 (20%) Other pituitary conditions (including hypohysitis) 42 (5%) Adrenal disease (including unilateral adrenalectomy, adenoma, carcinoma) 34 (4%) Other indications (autoimmune disease, hyponatremia, vomiting, weight loss, hyperkalemia, hypoglycemia, hypotension, collapse, fatigue) 180 (23%) Whole Cohort Glucocorticoid Exposed N 776 110 Age, y 53 ± 18 53 ± 17 Total number of SSTs 1912 277 Number of SSTs per patient 2.5 ± 0.8 2.5 ± 0.9 Total duration of follow-up, d 1182 ± 284 1166 ± 466 Time between first and second SST, d 340 ± 322 268 ± 215 Sex  Male 335 (43%) 50  Female 441 (57%) 60 SST indication  Previous treatment with suppressive doses of glucocorticoids (topical, nasal, inhaled, oral, intramuscular, intravenous) 110 (14%)   Prednisolone 49   Hydrocortisone 18   Dexamethasone 3   Others (budesonide, fluticasone, ciclesonide, beclometasone) 40 Postoperative following pituitary surgery 120 (15%) Pituitary adenoma (nonfunctioning, no surgery, no radiotherapy) 136 (17%) Others central nervous system tumors 154 (20%) Other pituitary conditions (including hypohysitis) 42 (5%) Adrenal disease (including unilateral adrenalectomy, adenoma, carcinoma) 34 (4%) Other indications (autoimmune disease, hyponatremia, vomiting, weight loss, hyperkalemia, hypoglycemia, hypotension, collapse, fatigue) 180 (23%) Patients treated with radiotherapy, who had been diagnosed with Addison disease, congenital adrenal hyperplasia, adrenal metastases, or who had undergone bilateral adrenalectomy were excluded from the study. Data are expressed as mean ± SD where appropriate. Figures in parentheses reflect percentage of the entire cohort. View Large Table 1. Baseline Characteristics of 776 Patients With Reversible Causes of AI Whole Cohort Glucocorticoid Exposed N 776 110 Age, y 53 ± 18 53 ± 17 Total number of SSTs 1912 277 Number of SSTs per patient 2.5 ± 0.8 2.5 ± 0.9 Total duration of follow-up, d 1182 ± 284 1166 ± 466 Time between first and second SST, d 340 ± 322 268 ± 215 Sex  Male 335 (43%) 50  Female 441 (57%) 60 SST indication  Previous treatment with suppressive doses of glucocorticoids (topical, nasal, inhaled, oral, intramuscular, intravenous) 110 (14%)   Prednisolone 49   Hydrocortisone 18   Dexamethasone 3   Others (budesonide, fluticasone, ciclesonide, beclometasone) 40 Postoperative following pituitary surgery 120 (15%) Pituitary adenoma (nonfunctioning, no surgery, no radiotherapy) 136 (17%) Others central nervous system tumors 154 (20%) Other pituitary conditions (including hypohysitis) 42 (5%) Adrenal disease (including unilateral adrenalectomy, adenoma, carcinoma) 34 (4%) Other indications (autoimmune disease, hyponatremia, vomiting, weight loss, hyperkalemia, hypoglycemia, hypotension, collapse, fatigue) 180 (23%) Whole Cohort Glucocorticoid Exposed N 776 110 Age, y 53 ± 18 53 ± 17 Total number of SSTs 1912 277 Number of SSTs per patient 2.5 ± 0.8 2.5 ± 0.9 Total duration of follow-up, d 1182 ± 284 1166 ± 466 Time between first and second SST, d 340 ± 322 268 ± 215 Sex  Male 335 (43%) 50  Female 441 (57%) 60 SST indication  Previous treatment with suppressive doses of glucocorticoids (topical, nasal, inhaled, oral, intramuscular, intravenous) 110 (14%)   Prednisolone 49   Hydrocortisone 18   Dexamethasone 3   Others (budesonide, fluticasone, ciclesonide, beclometasone) 40 Postoperative following pituitary surgery 120 (15%) Pituitary adenoma (nonfunctioning, no surgery, no radiotherapy) 136 (17%) Others central nervous system tumors 154 (20%) Other pituitary conditions (including hypohysitis) 42 (5%) Adrenal disease (including unilateral adrenalectomy, adenoma, carcinoma) 34 (4%) Other indications (autoimmune disease, hyponatremia, vomiting, weight loss, hyperkalemia, hypoglycemia, hypotension, collapse, fatigue) 180 (23%) Patients treated with radiotherapy, who had been diagnosed with Addison disease, congenital adrenal hyperplasia, adrenal metastases, or who had undergone bilateral adrenalectomy were excluded from the study. Data are expressed as mean ± SD where appropriate. Figures in parentheses reflect percentage of the entire cohort. View Large Cortisol assays Two cortisol assays were used to analyze samples across the SST: Siemens ADVIA Centaur (Siemens Health Care Diagnostics, Frimley, United Kingdom) and the Roche Modular System (Roche, Mannheim, Germany). Of the 1912 SSTs that were included in the analysis, 1368 (from 559 patients) were analyzed on the Roche assay and 544 (from 217 patients) on the Siemens assay. The same assay was used for repeated testing on the same patient. Serum cortisol analyzed by Siemens ADVIA Centaur showed an interassay imprecision of 10.5% at 83 nmol/L, 6.0% at 524 nmol/L, and 7.0% at 904 nmol/L and by the Roche Modular System of <8% for levels between 76 and 925 nmol/L. An analysis directly comparing the results from the Roche and Siemens assays is presented within the Results section. SST protocol and interpretation All SSTs were performed between 9 and 12 am, at least 18 hours after the most recent dose of glucocorticoids. Individual clinicians determined the frequency of repeat testing on a case-by-case basis. Patients taking the oral contraceptive pill or other estrogen replacement were required to stop the treatment at least 6 weeks before the test. Blood was sampled for serum cortisol at baseline and after 30 minutes: baseline serum cortisol levels were measured prior to injection of 250 μg Synacthen (Questcor Operations Limited, Dublin, Ireland, for Siemens assays; Alliance Pharmaceuticals, Chippenham, United Kingdom, and Sigma-tau Pharmaceutical, Rome, Italy, for Roche assays) intramuscularly or intravenously. The 30-minute response to intramuscular or intravenous Synacthen has been shown to be equivalent (18). After administration of Synacthen, the patients were observed for 15 minutes for signs of any allergic reaction. The interpretation of the SST is based on the 30-minute serum cortisol where an adequate response to Synacthen was defined as >450 nmol/L for Siemens ADVIA Centaur (19), as >550 nmol/L for the Roche Generation I Modular System (tests done before February 2016) (19), and as >450 nmol/L for the Roche Generation II Modular System (tests done after February 2016). The incremental response to Synacthen was calculated as: delta cortisol = [30-minute – 0-minute cortisol]. Statistical methods A binomial logistic regression was performed on the whole cohort to ascertain the effects of selected variables on the likelihood that participants will show recovery at the subsequent test. Six variables were inserted into the model: age, sex, 30-minute cortisol, basal cortisol of the subsequent test, use of steroid medication, and different assay used. Linearity of the continuous variables with respect to the logit of the dependent variable was assessed via the Box-Tidwell (1962) procedure. A Bonferroni correction was applied using all six terms in the model resulting in statistical significance being accepted when P < 0.008. Based on this assessment, all continuous independent variables were found to be linearly related to the logit of the dependent variable. Data are expressed as median with 95% CI assuming a normal distribution. Receiver operating curve (ROC) analysis was performed using a uniform threshold for any selected variables [basal cortisol (0-minute), delta cortisol (30-minute minus basal), 30-minute cortisol, basal cortisol of the subsequent test] according to the 95% specificity on the ROC analysis. According to this, a cutoff of 100 nmol/L (3.6 μg/dL), 350 nmol/L (12.7 μg/dL), 100 nmol/L (3.6 μg/dL), and 200 nmol/L (7.3 μg/dL) were identified and selected respectively for basal cortisol (0-minute), 30-minute cortisol, and delta cortisol and for the basal cortisol of the subsequent test. Area under the curve (AUC) analysis was used to express the overall diagnostic accuracy of the index criterion. Kaplan-Meier survival analysis was conducted to assess the median time to recovery for each group, and pairwise log rank comparisons were conducted to determine which group had significantly different survival distributions. A P < 0.05 was considered indicative of a statistically significant difference. Statistical analyses were performed using SPSS version 24 (IBM software, Chicago, IL) and GraphPad Prism 7.0 software package (GraphPad Software, Inc., La Jolla, CA). Results The patients’ characteristics are presented in Table 1, including the relevant clinical indications as well as the number and timing of the SSTs performed. A total of 776 subjects were recruited, all with potentially reversible causes of AI. A subgroup analysis was performed in 110 patients with AI secondary to treatment with suppressive doses of glucocorticoids. Overall, 37% of patients of the whole cohort who initially failed the SST eventually went on to pass, and 57% of those with nonfunctioning pituitary tumors and 44% of those patients who underwent pituitary surgery eventually passed the SST. Logistic regression modeling The binomial logistic regression model was statistically significant (χ2 = 143.8, P < 0.0001) and explained 47.9% of the variance in adrenal recovery, correctly classifying 88.6% of cases. Sensitivity was 72%, specificity was 94.7%, positive predictive value was 83.2%, and negative predictive value was 90.0%. Of the six predictor variables incorporated into the model, two were statistically significant: 30-minute cortisol (P < 0.0001) and the basal cortisol of the subsequent test (P < 0.0001). Lower 30-minute cortisol and basal cortisol of the subsequent test were associated with an increased likelihood of failing the subsequent test. HPA axis recovery All information derived from the SST was examined for its ability to predict whether the subsequent SST would be passed or failed (Fig. 1a–1d). A 30-minute cortisol >350 nmol/L (12.7 μg/dL) was the best indicator for future adrenal recovery (ROC AUC = 0.85, Fig. 1e). In this group, the estimated median time to HPA axis recovery was 334 (95% CI: 300 to 368) days compared with 1368 days (1067 to 1668) in those with a 30-minute cortisol <350 nmol/L (12.7 μg/dL) (χ2 = 51.2, P < 0.0001). There was a clear stratification of response with those individuals with a 30-minute cortisol between 150 and 350 nmol/L (5.4 and 12.7 μg/dL) having a lower chance of recovery of HPA axis function, although recovery was worst in those with a 30-minute cortisol <150 nmol/L (5.4 μg/dL) [median time to recovery: <150 nmol/L = 1174 (979 to 1369) days, χ2 = 40.42, P < 0.0001; 150 to 350 nmol/L = 465 (288 to 641) days, χ2 = 14.50, P < 0.0001] (Fig. 1d). Combining a 30-minute cortisol level with a delta cortisol of above or below 100 nmol/L (3.6 μg/dL) did not improve the diagnostic accuracy of the test (ROC AUC = 0.84) (Fig. 1f). The percentages of patients that recovered HPA axis function at 1, 2, and 4 years after their initial SST are presented in Supplemental Table 1. Figure 1. View largeDownload slide Kaplan-Meier plots estimating time to recovery of HPA axis function in 776 patients with potentially reversible causes of AI stratified by (a) basal (0-min) cortisol of the same test, (b) delta cortisol (30-min – basal cortisol), (c) 30-min cortisol, and (d) 30-min cortisol stratified by cutoffs at >350, 150 to 350, and <150 nmol/L using data from their initial SST. (e) ROC curve analysis to determine the ability of the characteristics of the initial SST to predict eventual recovery of adrenal function. (f) Combining 30-min cortisol with delta cortisol (30-min – basal cortisol) levels does not improve the ability to predict passing a subsequent SST. (g) In those patients with a 30-min cortisol <350 nmol/L on their initial SST, a subsequent random morning cortisol (>18 h after the last replacement dose) of >200 nmol/L significantly increases the likelihood of HPA axis recovery. Figure 1. View largeDownload slide Kaplan-Meier plots estimating time to recovery of HPA axis function in 776 patients with potentially reversible causes of AI stratified by (a) basal (0-min) cortisol of the same test, (b) delta cortisol (30-min – basal cortisol), (c) 30-min cortisol, and (d) 30-min cortisol stratified by cutoffs at >350, 150 to 350, and <150 nmol/L using data from their initial SST. (e) ROC curve analysis to determine the ability of the characteristics of the initial SST to predict eventual recovery of adrenal function. (f) Combining 30-min cortisol with delta cortisol (30-min – basal cortisol) levels does not improve the ability to predict passing a subsequent SST. (g) In those patients with a 30-min cortisol <350 nmol/L on their initial SST, a subsequent random morning cortisol (>18 h after the last replacement dose) of >200 nmol/L significantly increases the likelihood of HPA axis recovery. To improve further the accuracy of the test, we combined the use of the 30-minute cortisol with the baseline cortisol of the subsequent SST, which was assumed to reflect a “random morning cortisol” (assessed between 9 and 12 am) in the absence of replacement therapy (>18 hours). The mean time between the first SST and random cortisol was 340 ± 322 days. In the group with a 30-minute cortisol <350 nmol/L (12.7 μg/dL) and a subsequent random morning cortisol <200 nmolL (7.3 μg/dL), 93% of individuals continued to have AI at 4 years [median time to recovery: 2083 (1928 to 2238) days, χ2 = 17.48, P = 3 × 10–5; ROC AUC = 0.95) (Fig. 1g). The proportion of patients recovering HPA axis over time in the whole cohort is presented in Fig. 2. Figure 2. View largeDownload slide HPA axis recovery rates after SST in 776 patients with potentially reversible causes of AI, including nonfunctioning pituitary tumors and pituitary surgery. Figure 2. View largeDownload slide HPA axis recovery rates after SST in 776 patients with potentially reversible causes of AI, including nonfunctioning pituitary tumors and pituitary surgery. The results of the analysis were not altered when corrected for both age and sex. In addition, a further analysis was undertaken in which only patients that failed their first test were included (n = 248), and the results were not different from those presented above (Supplemental Figs. 1 and 2). HPA axis recovery in glucocorticoid-exposed patients In the majority of cases, patients treated with suppressive doses of glucocorticoids do not have underlying endocrine disease, and therefore a dedicated subgroup analysis on these patients was performed (n = 110). In this group, the delta cortisol was the best predictor for future adrenal recovery: those patients with a delta cortisol >100 nmol/L (3.6 μg/dL) had a shorter estimated median time to recovery [262 (212 to 312) vs 974 (633 to 1314) days, χ2 = 20.233, P = 7.0 × 10–6; ROC AUC = 0.77) (Fig. 3a–3d). Figure 3. View largeDownload slide Kaplan-Meier plots estimating time to recovery of HPA axis function in 110 patients with AI due to exposure to suppressive doses of glucocorticoid therapy stratified by (a) basal (0-min) cortisol of the same test, (b) 30-min cortisol, and (c) delta cortisol (30-min – basal cortisol) of their initial SST. (d) ROC curve analysis to determine the ability of the characteristics of the initial SST to predict eventual recovery of adrenal function. Figure 3. View largeDownload slide Kaplan-Meier plots estimating time to recovery of HPA axis function in 110 patients with AI due to exposure to suppressive doses of glucocorticoid therapy stratified by (a) basal (0-min) cortisol of the same test, (b) 30-min cortisol, and (c) delta cortisol (30-min – basal cortisol) of their initial SST. (d) ROC curve analysis to determine the ability of the characteristics of the initial SST to predict eventual recovery of adrenal function. Combining the delta cortisol <100 nmol/L (3.6 μg/dL) with the subsequent random cortisol (above or below 200 nmol/L) helped to refine the predictive ability of the test; patients who had a random morning cortisol >200 nmol/L (>18 hours after their last replacement dose) following their initial SST had a median time to HPA axis recovery of 322 (134 to 357) days. No patient who had a subsequent random cortisol <200 nmol/L recovered from their iatrogenic AI over the 4-year duration of the study. The results of the analysis were not altered when delta cortisol values were corrected for baseline (0-minute) cortisol levels (data not shown). The proportion of glucocorticoid-exposed patients recovering HPA axis over time are presented in Fig. 4. Figure 4. View largeDownload slide HPA axis recovery rates after SST in 110 patients previously exposed to suppressive doses of glucocorticoid therapy. Figure 4. View largeDownload slide HPA axis recovery rates after SST in 110 patients previously exposed to suppressive doses of glucocorticoid therapy. Comparison between cortisol assays As described above, two different cortisol assays were used in this cohort study. In the main cohort, the logistic regression model as well as the ROC analysis did not show significant differences between assays. However, to confirm these findings, an analysis was performed for each assay individually using the value of the 30-minute cortisol and the subsequent random morning cortisol. Surprisingly, a 30-minute cortisol of 350 nmol/L (12.7 μg/dL) and a random morning cortisol of 200 nmol/L (7.3 μg/dL) represented the 95% specificity in the ROC analysis for both the assays. In addition, the proportion of patients that subsequently went on to pass the SST at 1, 2, and 4 years was entirely comparable between assays (1 year: 55% vs 57%; year 2: 88% vs 89%; year 4: 98% vs 99%; Roche vs Siemens) (Supplemental Table 2). Discussion We have demonstrated that the stimulated 30-minute cortisol from the SST can be used to predict the likelihood of HPA axis recovery. Furthermore, combining these measurements with assessment of a 1-year random morning cortisol measurement (between 9 and 12 am and >18 hours after the last replacement dose) provides an accurate prediction of those individuals who are likely to recover adrenal function and those in whom it may not. Notably, 57% of our patients with nonfunctioning pituitary tumors and 44% of those patients who underwent pituitary surgery recovered HPA axis function at subsequent testing. This implies that there is realistic potential for reversibility of secondary AI in these patients. We (17) and others (10) have previously described the use of a morning cortisol to assess adrenal reserve, but to date, there have been very little attempts to use the SST to inform a strategy for repeating testing that in addition might serve as a guide as to the likelihood of restoration of HPA axis function. There is currently a lack of consensus as to how and when to assess patients with pituitary disease for recovery of HPA axis function after surgery. Different approaches have been proposed including the use of a baseline morning cortisol (16, 17, 20, 21), SST (22), or the “gold standard” ITT (23, 24). The timing of postoperative testing is also not uniform, with marked variability across centers (24 hours, 48 hours, 1 to 6 weeks) (20). Further testing beyond this time point (6 weeks) is not regarded as routine and performed in some centers, but not in others. If a patient fails a SST postoperatively, they may end up on lifelong glucocorticoid replacement therapy. However, cases where recovery of HPA axis function is delayed are reported (23). More recently, a study has suggested that a significant number of patients with underlying pituitary tumors, including those that have had surgery, may recover over a median duration of 20 months (range: 8 to 51) (21). Taken together with our data, this would suggest that repeat assessments of the HPA axis are clearly justified. We have therefore proposed that in those individuals in whom there is the potential to recover from AI, the strategy for repeat testing could be guided by their SST results. In our cohort, those individuals who had a 30-minute cortisol level >350 nmol/L (12.7 μg/dL) were likely to recover more quickly, and therefore 6-month testing could be advocated. In contrast, the rate of recovery in those with a 30-minute cortisol <350 nmol/L (12.7 μg/dL) was much slower, and annual testing may be more appropriate in this group. Furthermore, we found that the use of a random morning cortisol could help to identify the group of patients who were most likely to recover HPA axis function and allow further rationalization of dynamic testing (Fig. 5). Figure 5. View largeDownload slide Proposed flow chart for the use of SST in patients with potentially reversible causes of AI. *Random morning cortisol was measured between 9 and 12 am and at least 18 h after the last dose of glucocorticoid. Figure 5. View largeDownload slide Proposed flow chart for the use of SST in patients with potentially reversible causes of AI. *Random morning cortisol was measured between 9 and 12 am and at least 18 h after the last dose of glucocorticoid. The increasing use of glucocorticoid therapy has led to a dramatic increase in the awareness and diagnoses of secondary AI in the context of “iatrogenic Cushing syndrome.” It is estimated that 2% to 3% of the population of the United Kingdom and United States are taking prescribed glucocorticoids, and these are most commonly used in the more elderly populations and at doses that are known to suppress the HPA axis (25). Data on recovery of the HPA axis after exposure to a suppressive dose of glucocorticoids are limited and often in studies that have recruited small numbers of patients (ranging from n = 1 to 49) (10, 26–35). In addition, much of the published literature has been in pediatric populations (35–40), and therefore simple extrapolation into the adult setting is not straightforward. These studies are limited not only in the variability of the populations that they have studied, but also in the differences in duration and dose exposure to glucocorticoids. As a result, the published literature that is currently available does not allow us to determine whether the duration, dose, or cumulative glucocorticoid exposure are the main drivers to the development of secondary AI in this context (9). There is substantial variability between individuals in their susceptibility to the development of the adverse metabolic effects associated with glucocorticoid use, and it is highly likely that the same will apply to the development (and subsequent potential for recovery) of HPA axis suppression. In total, 43% of our patients who at the time of their SST were either currently or previously exposed to suppressive doses of glucocorticoids regained HPA axis function. Based upon the results of the initial SST undertaken at steroid withdrawal, in this subgroup of patients, the delta cortisol was the best predictor of future recovery of HPA axis function, such that 95% of patients who had an initial delta cortisol >100 nmol/L (3.6 μg/dL) passed the SST within 4 years. Although 4-year recovery rates were lower in the delta cortisol <100 nmol/L (67%), combining this with a subsequent random morning cortisol identified a group of patients who were unlikely to recover from the AI (0% recover at 4 years for those with a subsequent random cortisol <200 nmol/L, 7.3 μg/dL). However, we acknowledge that this analysis was only undertaken in a small subcohort (n = 32) (Fig. 4) and requires confirmation in larger studies. As with the whole cohort discussed above, this stratification could be used to guide the frequency of repeat testing, and an algorithm is presented in Fig. 5. Our analysis did not shown differences in recovery rates between patients taking different steroid formulations. The majority of our patients had been taking prednisolone (49.5%), but the small numbers of individuals taking other preparations, as well as limitations in the availability of data relating to the exact dose and timing, do not allow us to make detailed comments or analysis with regards to the precise nature of the glucocorticoid exposure that might dictate the SST response. An additional limitation is the continuation of exposure to suppressive doses of glucocorticoid therapy between SSTs, which is likely to have occurred in the majority of patients taking inhaled stored treatment. Surprisingly, even though we have used assay-specific thresholds for defining a pass or fail of the SST, the logistic regression model demonstrated the independence of the analysis from the different assay methods used. The results for each assay (assessed independently) suggest that the same threshold of a 30-minute cortisol of >350 nmol/L and a 1-year random morning cortisol of >200 nmol/L (after 18 hours of steroid withdrawal) can be used. The reasons underpinning this are not entirely clear, although clearly results of the SST in this analysis are being used in a different context in this analysis (i.e., predicting recovery in future tests as opposed to assessing current HPA axis integrity). Further analysis across larger cohorts and including additional assays is clearly warranted. Importantly, in our study there are some limitations. This is a retrospective analysis of clinical data obtained from electronic patient records, and therefore there is the potential for selection bias in those individuals undergoing short Synacthen testing. We do not have accurate data on the dose and duration of exposure to glucocorticoids, and therefore we are unable to answer questions that relate to cumulative exposure that might predict AI. However, this does represent the largest analysis of this kind published to date. Our data extend to approximately 4 years of follow-up and therefore it is not possible to say whether further recovery of the HPA axis may occur after this time point. The algorithm that we have proposed is designed to guide and help the clinician, but is only derived from this data set and has not been tested or validated in other cohorts. As a result, it is important that individual clinicians use their discretion and clinical judgment in the frequency and interpretation of dynamic assessments of HPA axis function. In conclusion, the results generated from the 250 μg SST can provide useful information for both the clinician and patient to guide the frequency of repeat testing as well as the likelihood of HPA axis recovery. When combined with a subsequent random morning cortisol, the diagnostic accuracy improves significantly. Finally, this analysis has the potential to rationalize the use of the SST; with the cutoffs proposed in the algorithm (Fig. 5), 27% of SSTs could have been avoided, limiting both inconvenience for the patient and reducing medical and nursing costs. Abbreviations: Abbreviations: AI adrenal insufficiency AUC area under the curve HPA hypothalamo-pituitary-adrenal ITT insulin tolerance test ROC receiver operating curve SST short Synacthen (corticotropin) test Acknowledgments Financial Support: This work was supported by the Medical Research Council (MR/P011462/1 to J.W.T.), a National Institute for Health Research Oxford Biomedical Research Centre Principal Investigator Award (to J.W.T.), an Irish Endocrine Society Traveling Fellowship Grant (to C.P.W.), and the Exchange in Endocrinology Expertise Programme of the European Union of Medical Specialists 3E Fellowship (to R.P.). Disclosure Summary: The authors have nothing to disclose. References 1. Shulman DI , Palmert MR , Kemp SF ; Lawson Wilkins Drug and Therapeutics Committee . Adrenal insufficiency: still a cause of morbidity and death in childhood . Pediatrics . 2007 ; 119 ( 2 ): e484 – e494 . Google Scholar CrossRef Search ADS PubMed 2. Bergthorsdottir R , Leonsson-Zachrisson M , Odén A , Johannsson G . Premature mortality in patients with Addison’s disease: a population-based study . J Clin Endocrinol Metab . 2006 ; 91 ( 12 ): 4849 – 4853 . Google Scholar CrossRef Search ADS PubMed 3. Tomlinson JW , Holden N , Hills RK , Wheatley K , Clayton RN , Bates AS , Sheppard MC , Stewart PM ; West Midlands Prospective Hypopituitary Study Group . Association between premature mortality and hypopituitarism . Lancet . 2001 ; 357 ( 9254 ): 425 – 431 . Google Scholar CrossRef Search ADS PubMed 4. Bornstein SR , Allolio B , Arlt W , Barthel A , Don-Wauchope A , Hammer GD , Husebye ES , Merke DP , Murad MH , Stratakis CA , Torpy DJ . Diagnosis and treatment of primary adrenal insufficiency: an Endocrine Society clinical practice guideline . J Clin Endocrinol Metab . 2016 ; 101 ( 2 ): 364 – 389 . Google Scholar CrossRef Search ADS PubMed 5. Adcock IM , Mumby S . Glucocorticoids . Handb Exp Pharmacol . 2017 ; 237 : 171 – 196 . Google Scholar CrossRef Search ADS PubMed 6. Paragliola RM , Papi G , Pontecorvi A , Corsello SM . Treatment with synthetic glucocorticoids and the hypothalamus-pituitary-adrenal axis . Int J Mol Sci . 2017 ; 18 ( 10 ): 18 . Google Scholar CrossRef Search ADS 7. Henzen C , Suter A , Lerch E , Urbinelli R , Schorno XH , Briner VA . Suppression and recovery of adrenal response after short-term, high-dose glucocorticoid treatment . Lancet . 2000 ; 355 ( 9203 ): 542 – 545 . Google Scholar CrossRef Search ADS PubMed 8. Krasner AS . Glucocorticoid-induced adrenal insufficiency . JAMA . 1999 ; 282 ( 7 ): 671 – 676 . Google Scholar CrossRef Search ADS PubMed 9. Joseph RM , Hunter AL , Ray DW , Dixon WG . Systemic glucocorticoid therapy and adrenal insufficiency in adults: a systematic review . Semin Arthritis Rheum . 2016 ; 46 ( 1 ): 133 – 141 . Google Scholar CrossRef Search ADS PubMed 10. Broersen LH , Pereira AM , Jørgensen JOL , Dekkers OM . Adrenal insufficiency in corticosteroids use: systematic review and meta-analysis . J Clin Endocrinol Metab . 2015 ; 100 ( 6 ): 2171 – 2180 . Google Scholar CrossRef Search ADS PubMed 11. Woods CP , Argese N , Chapman M , Boot C , Webster R , Dabhi V , Grossman AB , Toogood AA , Arlt W , Stewart PM , Crowley RK , Tomlinson JW . Adrenal suppression in patients taking inhaled glucocorticoids is highly prevalent and management can be guided by morning cortisol . Eur J Endocrinol . 2015 ; 173 ( 5 ): 633 – 642 . Google Scholar CrossRef Search ADS PubMed 12. Gordijn MS , Rensen N , Gemke RJ , van Dalen EC , Rotteveel J , Kaspers GJ . Hypothalamic-pituitary-adrenal (HPA) axis suppression after treatment with glucocorticoid therapy for childhood acute lymphoblastic leukaemia . Cochrane Database Syst Rev . 2015 ; ( 8 ): CD008727 . 13. Hurel SJ , Thompson CJ , Watson MJ , Harris MM , Baylis PH , Kendall-Taylor P . The short Synacthen and insulin stress tests in the assessment of the hypothalamic-pituitary-adrenal axis . Clin Endocrinol (Oxf) . 1996 ; 44 ( 2 ): 141 – 146 . Google Scholar CrossRef Search ADS PubMed 14. Agha A , Tomlinson JW , Clark PM , Holder G , Stewart PM . The long-term predictive accuracy of the short synacthen (corticotropin) stimulation test for assessment of the hypothalamic-pituitary-adrenal axis . J Clin Endocrinol Metab . 2006 ; 91 ( 1 ): 43 – 47 . Google Scholar CrossRef Search ADS PubMed 15. Lindholm J , Kehlet H , Blichert-Toft M , Dinesen B , Riishede J . Reliability of the 30-minute ACTH test in assessing hypothalamic-pituitary-adrenal function . J Clin Endocrinol Metab . 1978 ; 47 ( 2 ): 272 – 274 . Google Scholar CrossRef Search ADS PubMed 16. Abdu TA , Elhadd TA , Neary R , Clayton RN . Comparison of the low dose short synacthen test (1 microg), the conventional dose short synacthen test (250 microg), and the insulin tolerance test for assessment of the hypothalamo-pituitary-adrenal axis in patients with pituitary disease . J Clin Endocrinol Metab . 1999 ; 84 ( 3 ): 838 – 843 . Google Scholar PubMed 17. Sbardella E , Isidori AM , Woods CP , Argese N , Tomlinson JW , Shine B , Jafar-Mohammadi B , Grossman AB . Baseline morning cortisol level as a predictor of pituitary-adrenal reserve: a comparison across three assays . Clin Endocrinol (Oxf) . 2017 ; 86 ( 2 ): 177 – 184 . Google Scholar CrossRef Search ADS PubMed 18. Wang TW , Wong MS , Smith JF , Howlett TA . The use of the short tetracosactrin test for the investigation of suspected pituitary hypofunction . Ann Clin Biochem . 1996 ; 33 ( Pt 2 ): 112 – 118 . Google Scholar CrossRef Search ADS PubMed 19. El-Farhan N , Pickett A , Ducroq D , Bailey C , Mitchem K , Morgan N , Armston A , Jones L , Evans C , Rees DA . Method-specific serum cortisol responses to the adrenocorticotrophin test: comparison of gas chromatography-mass spectrometry and five automated immunoassays . Clin Endocrinol (Oxf) . 2013 ; 78 ( 5 ): 673 – 680 . Google Scholar CrossRef Search ADS PubMed 20. Inder WJ , Hunt PJ . Glucocorticoid replacement in pituitary surgery: guidelines for perioperative assessment and management . J Clin Endocrinol Metab . 2002 ; 87 ( 6 ): 2745 – 2750 . Google Scholar CrossRef Search ADS PubMed 21. Munro V , Tugwell B , Doucette S , Clarke DB , Lacroix A , Imran SA . Recovery of adrenal function after chronic secondary adrenal insufficiency in patients with hypopituitarism . Clin Endocrinol (Oxf) . 2016 ; 85 ( 2 ): 216 – 222 . Google Scholar CrossRef Search ADS PubMed 22. Honegger J , Buchfelder M , Fahlbusch R . Surgical treatment of craniopharyngiomas: endocrinological results . J Neurosurg . 1999 ; 90 ( 2 ): 251 – 257 . Google Scholar CrossRef Search ADS PubMed 23. Arafah BM , Kailani SH , Nekl KE , Gold RS , Selman WR . Immediate recovery of pituitary function after transsphenoidal resection of pituitary macroadenomas . J Clin Endocrinol Metab . 1994 ; 79 ( 2 ): 348 – 354 . Google Scholar PubMed 24. Auchus RJ , Shewbridge RK , Shepherd MD . Which patients benefit from provocative adrenal testing after transsphenoidal pituitary surgery ? Clin Endocrinol (Oxf) . 1997 ; 46 ( 1 ): 21 – 27 . Google Scholar CrossRef Search ADS PubMed 25. van Staa TP , Leufkens HG , Abenhaim L , Begaud B , Zhang B , Cooper C . Use of oral corticosteroids in the United Kingdom . QJM . 2000 ; 93 ( 2 ): 105 – 111 . Google Scholar CrossRef Search ADS PubMed 26. Cydulka RK , Emerman CL . A pilot study of steroid therapy after emergency department treatment of acute asthma: is a taper needed ? J Emerg Med . 1998 ; 16 ( 1 ): 15 – 19 . Google Scholar CrossRef Search ADS PubMed 27. Miró J , Amado JA , Pesquera C , López-Cordovilla JJ , Berciano J . Assessment of the hypothalamic-pituitary-adrenal axis function after corticosteroid therapy for MS relapses . Acta Neurol Scand . 1990 ; 81 ( 6 ): 524 – 528 . Google Scholar CrossRef Search ADS PubMed 28. Barrier JH , Larrose C , Magadur/Joly G , Hamidou M . Potential risks of the withdrawal of corticoids in Horton’s disease. Prospective study of 22 patients . Rev Med Interne . 1993 ; 14 : 976 . Google Scholar CrossRef Search ADS PubMed 29. Boots JM , van den Ham EC , Christiaans MH , van Hooff JP . Risk of adrenal insufficiency with steroid maintenance therapy in renal transplantation . Transplant Proc . 2002 ; 34 ( 5 ): 1696 – 1697 . Google Scholar CrossRef Search ADS PubMed 30. Bacon PA , Myles AB , Beardwell CG , Daly JR . Corticosteroid withdrawal in rheumatoid arthritis . Lancet . 1966 ; 2 ( 7470 ): 935 – 937 . Google Scholar CrossRef Search ADS PubMed 31. Campieri M , Ferguson A , Doe W , Persson T , Nilsson LG ; The Global Budesonide Study Group . Oral budesonide is as effective as oral prednisolone in active Crohn’s disease . Gut . 1997 ; 41 ( 2 ): 209 – 214 . Google Scholar CrossRef Search ADS PubMed 32. Rodger RS , Watson MJ , Sellars L , Wilkinson R , Ward MK , Kerr DN . Hypothalamic-pituitary-adrenocortical suppression and recovery in renal transplant patients returning to maintenance dialysis . Q J Med . 1986 ; 61 ( 235 ): 1039 – 1046 . Google Scholar PubMed 33. Desramé J , Sabaté JM , Agher R , Bremont C , Gaudric M , Couturier D , Chaussade S . Assessment of hypothalamic-pituitary-adrenal axis function after corticosteroid therapy in inflammatory bowel disease . Am J Gastroenterol . 2002 ; 97 ( 7 ): 1785 – 1791 . Google Scholar CrossRef Search ADS PubMed 34. Jamilloux Y , Liozon E , Pugnet G , Nadalon S , Heang Ly K , Dumonteil S , Gondran G , Fauchais AL , Vidal E . Recovery of adrenal function after long-term glucocorticoid therapy for giant cell arteritis: a cohort study . PLoS One . 2013 ; 8 ( 7 ): e68713 . Google Scholar CrossRef Search ADS PubMed 35. Pugnet G , Sailler L , Astudillo L , Vernet J , Bennet A , Arlet P . Frequency and risk factors of biological adrenal insufficiency screened by the 250 microg Synacthene stimulation test after a prolonged course of systemic glucocorticoid therapy. A study of 100 patients . Rev Med Interne . 2010 ; 31 ( 5 ): 332 – 336 . Google Scholar CrossRef Search ADS PubMed 36. Mendoza-Cruz AC , Wargon O , Adams S , Tran H , Verge CF . Hypothalamic-pituitary-adrenal axis recovery following prolonged prednisolone therapy in infants . J Clin Endocrinol Metab . 2013 ; 98 ( 12 ): E1936 – E1940 . Google Scholar CrossRef Search ADS PubMed 37. Graber AL , Ney RL , Nicholson WE , Island DP , Liddle GW . Natural history of pituitary-adrenal recovery following long-term suppression with corticosteroids . J Clin Endocrinol Metab . 1965 ; 25 ( 1 ): 11 – 16 . Google Scholar CrossRef Search ADS PubMed 38. Cunha CF , Silva IN , Finch FL . Early adrenocortical recovery after glucocorticoid therapy in children with leukemia . J Clin Endocrinol Metab . 2004 ; 89 ( 6 ): 2797 – 2802 . Google Scholar CrossRef Search ADS PubMed 39. Tominaga A , Uozumi T , Arita K , Kurisu K , Yano T , Hirohata T . Anterior pituitary function in patients with nonfunctioning pituitary adenoma: results of longitudinal follow-up . Endocr J . 1995 ; 42 ( 3 ): 421 – 427 . Google Scholar CrossRef Search ADS PubMed 40. Yedinak C , Hameed N , Gassner M , Brzana J , McCartney S , Fleseriu M . Recovery rate of adrenal function after surgery in patients with acromegaly is higher than in those with non-functioning pituitary tumors: a large single center study . Pituitary . 2015 ; 18 ( 5 ): 701 – 709 . Google Scholar CrossRef Search ADS PubMed Copyright © 2018 Endocrine Society

Journal

Journal of Clinical Endocrinology and MetabolismOxford University Press

Published: Aug 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off