AACE® 8
Contents lists available at ScienceDirect
Endocrine Practice
journal homepage: www.endocrinepractice.org
AACE. Endocrine Practice™ -
Original articlen
W
-
Original Article
Revisiting Cushing Syndrome, Milder Forms Are Now a Common Occurrence: A Single-Center Cohort of 76 Subjects
Check for updates
Arza Rosset, Medical Student, Yona Greenman, MD, Etty Osher, MD, PHD, Naftali Stern, MD, Karen Tordjman, MD * **
Institute of Endocrinology, Metabolism and Hypertension, Tel Aviv Sourasky Medical Center, Tel Aviv University, Sackler Faculty of Medicine, Tel Aviv, Israel
ARTICLE INFO
Article history: Received 30 November 2020 Accepted 17 February 2021 Available online 27 February 2021
Key words:
Cushing’s syndrome diagnosis guidelines modern era presentation screening
ABSTRACT
Objective: Guidelines recommend initiating testing for Cushing syndrome (CS) based on strong clinical suspicion. Our recent experience suggests the absence of classical stigmata in the majority of patients with CS. We aimed to confirm this premise by revisiting the clinical features of this syndrome in a modern series of patients from a single center.
Methods: Computerized records of subjects with CS diagnosed at Tel Aviv Sourasky Medical Center between 2000 and 2018 were reviewed. A Cushing inventory score, including all clinical components of the syndrome, was computed for each subject. Data were compared between the subtypes and evaluated in light of those in the literature.
Results: Of the 76 subjects with CS (60 women/16 men), 49 (64.5%) had Cushing disease; 16 (21.1%), adrenal adenoma; 7 (9.2%), adrenocortical carcinoma; and 4 (5.3%), ectopic adrenocorticotropic hormone secretion. In only 15 of 74 cases (20.3%), clinical suspicion of CS led to testing. Catabolic signs of CS were present in less than 30% of cases. The most common symptom was weight gain (52/67, 77.6%), and the most common comorbidity was hypertension (47/76, 61.8%). There were no differences in the Cushing inventory score between the subtypes. Signs, symptoms, and comorbidities were all significantly less common than in the classical syndrome.
Conclusions: Modern-day CS presents with subtler features than in the past. Initiating a testing cascade solely based on a strong clinical suspicion may lead to underdiagnosis of milder cases. A concerted effort to devise cost-efficient testing for CS in the current era is needed.
@ 2021 AACE. Published by Elsevier Inc. All rights reserved.
Introduction
Endogenous Cushing syndrome (CS) is a rare and potentially lethal condition whose annual incidence is generally estimated at 2 to 3 per million.1 However, most published figures are derived from a series collected in the 1960s.2,3 More recent investigations, or series stratified by periods, have suggested an increasing annual incidence of up to 4.5 new cases per million.4-7 The true cause of this increasing incidence remains uncertain; however, it has been
suggested that it could be due to the diagnosis of milder cases as a result of clinicians’ heightened awareness to the syndrome.1 Additionally, since the emergence of incidentalomas, particularly those of the adrenal glands, it has been recognized that subtler and even subclinical forms of the syndrome may occur. This is sup- ported by what appears to be a changing distribution between the various forms of the syndrome, with a larger fraction (up to 40% of cases) now taken up by adrenal adenomas (AAs),5 probably due to an increasing trend of abdominal imaging.6 Thus, indolent clinical manifestations of hypercortisolism may simply comprise 1 or more components of the metabolic syndrome.
Studies have repeatedly demonstrated increased mortality in CS of about twice that expected for a comparable population, mostly from cardiovascular complications.8-10 Although it continues to be elevated,9 mortality risk may decrease with successful treatment and return to normal with prolonged remission in Cushing disease (CD).10 It is thus crucial to establish a prompt diagnosis to offer curative treatment with minimum delay. With surging rates of obesity, diabetes, and metabolic syndrome in the population and
Abbreviations: AA, adrenal adenoma; ACC, adrenocortical carcinoma; ACTH, adrenocorticotropic hormone; ANOVA, analysis of variance; BMI, body mass index; CD, Cushing disease; CIS, Cushing inventory score; CS, Cushing syndrome; EAS, ectopic adrenocorticotropic hormone secretion; ERCUSYN, European Registry on Cushing’s syndrome; IQR, interquartile range.
* Address author correspondence and reprint requests to Dr Karen Tordjman, Institute of Endocrinology, Metabolism and Hypertension, Tel Aviv Sourasky Med- ical Center, 6 Weizmann Street, Tel Aviv, 6423909, Israel.
E-mail address: karent@tlvmc.gov.il (K. Tordjman).
possibly milder cases of CS, establishing a diagnosis has become a conundrum. While CS diagnosis may be increasing, it still remains exceedingly rare. Thus, guidelines advocate against screening for CS in obesity, diabetes, or hypertension.11 Rather, they recommend that screening for CS starts with strong clinical suspicion, one that increases the pretest probability of such an investigation resulting in a true diagnosis. Such a suspicion should rely on the presence of multiple and progressive discriminatory signs (ie, facial plethora and catabolic symptoms of cortisol excess) and/or features unex- pected in a young subject, such as osteoporosis and hypertension.11-13
It has been our department’s clinical impression that many established cases of CS in the past 2 decades might have been excluded from the diagnostic process for failing to arouse such stringently defined clinical suspicions. To test this impression, we sought to characterize our recent cohort of subjects with CS. We aimed to determine the circumstances that led to making the diagnosis while potentially redefining the clinical presentation of CS in the modern era.
Methods
Study Design and Subjects
This was a descriptive retrospective cohort study of all subjects diagnosed with CS at Tel Aviv Sourasky Medical Center between January 2000 and April 2018. Subjects were identified by searching the computerized hospital records using ICD-9 diagnoses to encompass all possibilities. The codes used were 255.0; 227.0; 237.0; 194.3; 255.8; 227.3; 253.1 07.63, 07.62, and 07.2; 259.3; and 253.4. These codes comprised all conditions associated with hypercortisolism, pituitary or adrenal neoplasms, hyperfunction, or surgical procedures performed on these glands as well as ectopic hormone secretion that might lead to CS. A diagnosis of CS was established according to standard diagnostic criteria, most commonly the failure to suppress serum cortisol after a low-dose Liddle test and urinary free cortisol excretion repeatedly above twice the upper limit of normal. Ancillary procedures were often added, such as a high-dose Liddle test, high-dose overnight dexa- methasone suppression, corticotropin-releasing hormone stimu- lation, and/or inferior petrosal sinus sampling. Diagnosis confirmation was further validated from the pathologic material obtained at surgery. Thus, after manually searching the charts, only subjects with confirmed CS were retained for our study. This was a convenience sample with no local historical control group (due to the almost impossible task of locating cases that predated the complete computerization of our medical records).
Due to the retrospective nature of our study, no informed con- sent was required. The study received the approval of the Human Studies Committee of our medical center.
Retrieved Data
For each subject, the independent variables extracted from the charts were gender, age at presentation, year of diagnosis, and type of CS. Dependent variables included all anthropometric data, the reason for the investigation, relevant clinical symptoms and signs, comorbidities, and laboratory variables (see Supplementary Methods). A Cushing inventory score (CIS) was constructed by summing, in an unweighted fashion, each relevant clinical feature (symptoms and signs) and concomitant comorbidities, with a total maximum possible score of 26 in women (including menstrual disturbances in women of reproductive age and hirsutism) and 24 in men (see Supplementary Methods). Any item that could be either a sign, symptom, or comorbidity, such as acne, depression, or hirsutism, was counted only once.
Statistical Procedures
Descriptive statistics were used to present continuous variables as means + SD when normally distributed or medians and inter- quartile range (IQR) when otherwise. Categorical variables were presented as proportions. Comparisons between groups were car- ried out using the t test and Mann-Whitney test. For more than 2 groups, analysis of variance (ANOVA) was utilized for normally distributed continuous variables or the Kruskal-Wallis test for nonparametric data, with post hoc comparisons when significant. Proportions between groups were assessed by X2 or Fisher exact tests. Correlations were performed using the Spearman correlation test. Finally, comparisons between rates of clinical features in our sample with those found in the literature were performed by the single-sample binomial test. All computations were done with SPSS v26.0 (IBM Corp). A P value <. 05 was deemed significant.
Results
In total, we located the records of 76 subjects with a confirmed diagnosis of CS. In the entire cohort, there were 60 women and 16 men, a ratio of 3.75:1. Forty-nine subjects had pituitary CD (64.5%), and 16 had an AA (21.1%). There were 7 subjects with adrenocortical carcinoma (ACC; 9.2%), and 4 had ectopic adrenocorticotropic hormone secretion (EAS; 5.3%), all from neuroendocrine tumors (Table 1).
| Variable | All N = 76 | CD n = 49 | AA n =16 | ACC n = 7 | EAS n = 4 | P value |
|---|---|---|---|---|---|---|
| Age (y) | 47.9 (15.3) | 44.6 (14.9) | 55.4 (11.9) | 42.4 (14.3) | 68.5 (9.0) | .002ª |
| Men % | 21.1 | 22.4 | 6.3 | 28.6 | 50 | .16 |
| Men/women | (16/60) | (11/38) | (1/15) | (2/5) | (2/2) | … |
| BMI (kg/m2) | 28.05 (24.7-35.6) | 29.06 (25.13-36.3) | 28.0 (22.9-35.0) | 27.0 (23.5-33.1) | 26.1 (22.7-23.33) | .682 |
| Systolic BP (mm Hg) | 143.7 (27.3) | 141.3 (27.9) | 150.2 (15.5) | 153.9 (23.5) | 128.5 (30.4) | .33 |
| Diastolic BP (mm Hg) | 86.2 (14.4) | 85.5 (13.5) | 86.3 (14.8) | 97.1 (11.2) | 74.8 (20.6) | .08 |
Abbreviations: AA = adrenal adenoma; ACC = adrenocortical carcinoma; BMI = body mass index; BP = blood pressure; CD = Cushing disease; EAS = ectopic adrenocorti- cotropic hormone secretion.
Data are given as mean (± SD) or median with interquartile range for BMI.
Comparisons between groups were made using the Fisher exact test, analysis of variance, or the Kruskal-Wallis test, as appropriate.
a By post hoc analysis. Subjects with CD and ACC were significantly younger than patients with EAS (P < . 01). Both subjects with CD and ACC were younger than subjects with AA (P < . 05).
Demographic and Anthropometric Data of Study Subjects
Gender distribution did not differ among the groups, and women represented almost 80% of all subjects in the entire cohort. The mean age at presentation was 47.9 + 15.3 years old (range, 18- 79); however, there were age differences (P = . 002 by ANOVA). Patients with EAS were the oldest, and CD and ACC patients were generally younger than subjects from the other groups. Table 1 summarizes the subjects’ anthropometric data at the time of pre- sentation. Notably, median body mass index (BMI) was only 28.05 kg/m2 (IQR, 24.7-35.6) for the entire group, and there were no significant differences between the groups with respect to BMI or blood pressure.
Reasons Leading to the Investigation
In only one fifth of the entire cohort was a clinical suspicion of CS, as defined by the Endocrine Society Clinical Practice Guide- lines,11 the motive for initiating the investigation. In subjects with CD, who comprised the largest subtype of CS, the figure was 10/48 (20.8%). In patients with ACC, however, such suspicion was indeed the leading motive for embarking on the investigation (4/7, 57.1%), whereas this was never the case in subjects with an AA. In over 35% of all patients, an incidental imaging or laboratory finding was the only reason for conducting the diagnostic tests although there had been no preceding clinical suspicion (Table 2). The median time between referral and a final diagnosis was 1 year (IQR, 0.5-2.0). A small number of patients (4/76, 5.3%) were referred with an established diagnosis of CS. In all the other cases, the investigation was initiated at our institution in subjects referred for any possible reason to our general endocrine clinic. All patients were evaluated and followed by a senior endocrinologist.
Presenting Symptoms and Signs
The most common symptom among patients with CS was weight gain, which was present in 77.6% of all subjects and at an even greater frequency in subjects with CD and AA (83.7% and 84.6%, respectively). Objective documentation of the ponderal in- crease was available from the charts of 28 of the 52 subjects who reported having gained weight. The median weight gain was 10 kg (IQR, 6.3-20; maximum, 40 kg), which had occurred over a median of 2.5 years (IQR, 1-6). The weight gain pattern was not different between patients with CD, AA, or ACC (not significant by Kruskal- Wallis). Notably, weight gain was never reported in EAS. In fact, 3 of the 4 subjects had lost weight. Among women, the second most common symptom was hirsutism, which was particularly trou- blesome in women with ACC (80%). Menstrual disturbances were present in close to half the women of reproductive age. Proximal muscle weakness, a symptom felt to be highly discriminatory, was
common in CD (44.4%) but not in other types of CS. Upon exami- nation, echoing the complaint of weight gain, abdominal obesity was the most common sign in all types of CS, while purple striae, supraclavicular fullness, facial plethora, and pedal edema were uncommon. In CD, a moon/rounded face and a dorsocervical fat pad (buffalo hump) were each noted in about half of the subjects, but they were far less frequent in the other types of CS. A tabulation of selected symptoms and signs is given in Table 3. It is noteworthy that once a diagnosis of CS was either highly likely or established, in the majority (53.4%) of cases, the endocrinologist following the patient made a point to document the lack of a typical Cushingoid appearance. This was particularly true for subjects with AA and ACC, and this was universally the case in subjects with EAS. Even in CD, over 40% of patients had not impressed an experienced endo- crinologist as having CS.
Comorbidities at the Time of Presentation
Hypertension was undoubtedly the most prevalent comorbidity in all types of CS (61.8%). It was least common in CD but nearly universal in all other types, although the difference did not reach statistical significance (Table 4). New-onset hypertension was noted in 14 of the 47 subjects, while in 11 subjects, it had worsened over a period of 7 ± 5.8 months. Interestingly, in keeping with the unimpressive median BMI reported for the entire cohort and the specific subtypes, despite the leading complaint of weight gain, obesity was noted in less than 50% of subjects. Diabetes was present in 36.8% of all subjects, while impaired fasting glucose added another 17.1%, bringing the prevalence of glucose homeostasis ab- normalities to 53.9% in the entire cohort. For those with diabetes, the diagnosis had been established 6.1 + 5.7 years before present- ing to our department. Only 2 of 28 subjects had recent-onset diabetes, while 4 subjects complained of a recent worsening of glycemic control. Other conditions traditionally reported with an increased frequency in CS were remarkably elusive. There was 1 case of acute psychosis, 1 case of widespread candidiasis, and 3 subjects with a history of nephrolithiasis (3.9%).
Relevant Hormonal and Biochemical Data
A selection of relevant laboratory data is presented in Table 5 (additional data can be found in the Supplementary Material). None of the relevant hormonal variables were normally distributed and are thus presented as medians and IQRs. As expected, adre- nocorticotropic hormone (ACTH) was significantly different be- tween the groups (P < . 0001, by Kruskal-Wallis ANOVA) because subjects with AA and ACC had significantly lower levels of ACTH than patients with CD (P < . 0001 for both pair comparisons). Additionally, basal serum cortisol concentration was also different between the groups (P =. 021) due to the fact that subjects with CD
| Reasonª | All | CD | AA | ACC | EAS |
|---|---|---|---|---|---|
| Incidental imaging or laboratory finding | 26/74 (35.1) | 15/48 (31.3) | 9/15 (60) | 0 | 1/4 (50) |
| Nonspecific complaint or suggestive lab findingb | 29/74 (39.2) | 21/48 (43.8) | 5/15 (33/3) | 2/7 (28.6) | 2/4 (25) |
| Clinical suspicion of CS | 15/74 (20.3) | 10/48 (20.8) | 0 | 4/7 (57.1) | 1/4 (25) |
| Pre-existing diagnosis | 4/74 (5.4) | 2/48 (4.2) | 1/15 (6.7) | 1/7 (14.3) | 0 |
Abbreviations: AA = adrenal adenoma; ACC = adrenocortical carcinoma; CD = Cushing disease; CS = Cushing syndrome; EAS = ectopic adrenocorticotropic hormone secretion.
Numbers expressed as fractions of the total number of subjects (n/N) for whom a reason was mentioned are also expressed in percentages in parenthesis. In 2 subjects, no explanation as to what led to the investigation was given.
a Comprehensive details of all the reasons for the investigation are given in the Supplementary Material.
b For instance, a complaint of weight gain or the finding of an elevated urinary free cortisol on a routine panel sent for evaluation of osteoporosis, all of which were without any clinical suspicion.
| Symptom/signs n/N (%) | All | CD | AA | ACC | EAS |
|---|---|---|---|---|---|
| Weight gain | 52/67 (77.6) | 36/43 (83.7) | 11/13 (84.6) | 4/7 (57.1) | 1/4 (25.0) |
| Menstrual disturbanceª | 15/31 (48.4) | 13/24 (54.2) | 1/3 (33.3) | 1/4 (25.0) | N/Ab |
| Hirsutism | 29/59 (49.2) | 20/38 (52.6) | 4/14 (28.6) | 4/5 (80.0) | 1/2 (50.0) |
| Proximal muscle weakness | 21/71 (29.6) | 20/45 (44.4) | 0/15 (0) | 0/7 (0) | 1/4 (25.0) |
| Bruising | 19/72 (26.4) | 12/36 (26.1) | 4/15 (26.7) | 1/7 (14.3) | 2/4 (50.0) |
| Insomnia | 17/74 (23.0) | 12/48 (25.0) | 4/15 (26.7) | 1/7 (14.3) | 0/4 (0) |
| Acne | 13/59 (18.1) | 9/46 (19.6) | 1/15 (6.7) | 3/7 (42.9) | 0/4 (0) |
| Abdominal obesity | 49/72 (68.1) | 35/46 (76.1) | 8/15 (53.3) | 4/7 (57.1) | 2/4 (50.0) |
| Dorsocervical fat pad | 27/71 (38.0) | 22/45 (48.9) | 3/15 (20.0) | 1/7 (14.3) | 1/4 (25.0) |
| Moon/rounded face | 31/71 (43.7) | 24/45 (53.3) | 4/15 (26.7) | 2/7 (28.6) | 1/4 (25.0) |
| Facial plethora | 20/71 (28.2) | 14/45 (31.1) | 3/15 (20.0) | 1/7 (14.3) | 2/4 (50.0) |
| Pedal edema | 19/71 (26.8) | 11/45 (24.4) | 4/15 (26.7) | 2/7 (28.6) | 2/4 (50.0) |
| Striae | 15/72 (20.8) | 10/46 (21.7) | 3/15 (20.0) | 2/7 (28.6) | 0/4 (0) |
| Supraclavicular fullness | 10/71 (14.1) | 10/48 (22.2) | 1/16 (6.3) | 0/7 (0) | 0/4 (0) |
| Cushingoid appearance | 27/58 (46.6) | 20/46 (55.6) | 5/14 (35.7) | 2/6 (33.3) | 0/2 (0) |
Abbreviations: AA = adrenal adenoma; ACC = adrenocortical carcinoma; CD = Cushing disease; EAS = ectopic adrenocorticotropic hormone secretion; N/A = not applicable. Less frequent presenting symptoms and clinical signs in the different types of CS were recorded and included in the severity score computation. Data are expressed as percentages in parentheses, computed from actual denominators as specified in each ratio.
a This complaint was only recorded for women of reproductive age.
b There were no women of reproductive age in this group.
| Comorbidity n/N (%) | All | CD | AA | ACC | EAS | P value |
|---|---|---|---|---|---|---|
| Hypertension | 47/76 (61.8) | 25/49 (51.0) | 12/16 (75.0) | 6/7 (85.7%) | 4/4 (100) | .06 |
| Obesityª | 30/68 (44.1) | 21/44 (47.7) | 6/14 (42.9) | 2/6 (33.3) | 1/4 (25.0) | .81 |
| Hyperlipidemiab | 36/75 (48.0) | 22/49 (44.9) | 7/15 (46.7) | 5/7 (71.4) | 2/4 (50.0) | .63 |
| Osteoporosis | 29/76 (38.2) | 20/49 (40.8) | 7/16 (43.8) | 1/7 (14.3) | 1/4 (25.0) | .6 |
| Diabetes“ | 28/76 (36.8) | 18/49 (38.8) | 5/16 (31.3) | 1/7 (14.3) | 3/4 (75.0) | .26 |
| Depression | 19/75 (25.3) | 13/48 (27.1) | 4/16 (25.0) | 0/7 (0) | 2/4 (50.0) | .3 |
Abbreviations: AA = adrenal adenoma; ACC = adrenocortical carcinoma; CD = Cushing disease; EAS = ectopic adrenocorticotropic hormone secretion.
Only bona fide diagnosed conditions were included.
Comparisons between subtypes were made using the Fisher exact test.
a Obesity was defined as a body mass index ≥30 kg/m2.
b This comprised all types of diagnosed and/or treated hyperlipidemias.
” Only a documented diagnosis was recorded as diabetes.
had significantly higher serum cortisol concentrations than those with AA (P =. 003 by Mann-Whitney), while none of the other types differed from one another in this respect. All other cortisol-related variables were not different among the groups. The only significant biochemical difference found was serum potassium concentration, which was significantly lower in subjects with adrenal CS (AA and ACC) than in subjects with CD (P < . 05).
Cushing Inventory Score
The CIS averaged 7.29 (+3.4). The highest score was that of a 52- year-old woman with CD who had no evidence of
hyperandrogenism (no acne or hirsutism) and scored 16 out of a possible maximum of 25 (she was postmenopausal). Although subjects with CD appeared to score higher (Figure), a Kruskal- Wallis ANOVA failed to show any association with the type of CS (P = . 065). We further questioned whether this score correlated with any of the specific laboratory variables. The only significant, albeit moderate, correlation found was with the serum cortisol concentration after an overnight 1-mg dexamethasone suppression test (Is = 0.43, P = . 002). However, we found a robust and highly significant correlation between the CIS and the endocrinologist’s subjective assessment of the presence of a typical Cushingoid appearance (Is = 0.56, P < . 001).
| Variable | CD | AA | ACC | EAS | P value |
|---|---|---|---|---|---|
| 8:00 AM SCortisol (ug/dL), (4.2-22.4), n | 27.8 [20.0-30.3] (40) | 20.7 [17.3-22.9] (12) | 23.7 [20.0-28.6] (6) | 20.7 [18.0-28.8] (4) | .07 |
| sCortisol after 1 mg ONDST (µg/dL), (<1.8), n | 15.2 [9.1-19.9] (26) | 10.1 [4.0-21.1] (13) | 23.3 [16.8-23.3] (2) | 24.2 [21.0-27.5] (2) | .38 |
| UFCª (ug/24 h), (<85), n | 192 [136-307] (40) | 124 [55-372] (15) | 354 [128-671] (7) | 324 [99-1085] (4) | .20 |
| Midnight salivary cortisol (µg/dL), (<0.19), n | 0.49 [0.31-0.95] (18) | 0.3 [0.17-0.62] (7) | 0.99 [0.36-3.41] (4) | 0.83 [0.71-0.83] (2) | .081 |
| ACTH (pg/mL), (5-46), n | 50.4 [21.8-77.0] (41) | 2.5 [2.5-2.5] (11) | 2.5 [2.5-2.5] (5) | 28 [2.5-28] (3) | .000b |
| Serum K (mEq/L), (3.5-5.0), n | 4.23 [± 0.59] (40) | 3.73 [±0.7] (14) | 3.43 [±0.53] (7) | 3.71 [±1.11] (4) | .006€ |
Abbreviations: AA = adrenal adenoma; ACC = adrenocortical carcinoma; ACTH = adrenocorticotropic hormone; CD = Cushing disease; EAS = ectopic adrenocorticotropic hormone secretion; K = potassium; ONDST = overnight dexamethasone suppression test; sCortisol = serum cortisol; UFC = urinary free cortisol. With the exception of serum potassium, which was normally distributed (shown as mean + SD), all other variables are presented as median and [ interquartile range or range]. The number of subjects for whom each analyte was available is indicated in parentheses.
The reference range for normal values is indicated for each variable.
a The numbers represent the average excretion of up to 3 urine collections.
b By post hoc analysis. Subjects with AA and ACC had ACTH concentrations that were significantly lower than those with CD (P < . 0001).
” Subjects with AA and ACC had serum potassium levels that were significantly lower than those of CD patients (P < . 05).
20
Cushing Inventory Score
15
10
5-
0
CD
AA
ACC
EAS
CS Type
Specifically, CIS was neither associated with age nor with gender. Gender was associated with only 2 features, namely, a complaint of weight gain, which was present in 86.5% of women and only 13.5% of men (P = . 01), and the presence of striae, which were only noted in women (P = . 03).
Comparative Assessment of Selected Clinical Features
Based on extensive cumulative clinical experience, our premise for this study was that modern era CS patients may often not exhibit many of the clinical features needed a priori to embark on a testing cascade for the syndrome. After thoroughly documenting the various clinical manifestations in our cohort, we compared the rates with those extracted from a compilation of contemporary authori- tative sources11-17 and identified an array of common features. It is noteworthy that the prevalence of clinical hyperandrogenism (hir- sutism and menstrual cycle abnormalities) was significantly lower than traditionally reported. Likewise, clinical features of the cata- bolic effects of hypercortisolism, considered highly discriminatory for CS, such as skin fragility and myopathy, were about half as common as in the literature and were present in less than one third of subjects. Facial plethora, a traditionally universal and compelling feature of CS,18 was remarkably infrequent (28%). Comorbidities were also, by and large, less common than the rates quoted in the available literature. Only the most frequent diseases are included in Table 6; however, it should be emphasized that nephrolithiasis, re- ported to affect up to 50% of patients with CS, was only documented in 3 (3.9%) subjects in our cohort (P < . 0001).
Discussion
In this single tertiary center series, we identified 76 subjects with endogenous CS over 18 years. Compared with traditional figures relating to the different etiologies of endogenous CS, our findings correlate with the European Registry on Cushing’s syndrome (ERCU- SYN) database and with the data of a very large recent Chinese study (1652 subjects from a single center), which both suggested a relative decrease in the contribution of CD and a reciprocal increase in ACTH- independent cases.19,20 Adrenal causes of CS accounted for a little over
30% of the cases in our series. AAs were responsible for 21%, slightly less than the 27% reported in the ERCUSYN, but similar to the Chinese study. Unfortunately, ACC subjects were excluded from the ERCUSYN database, while they represented 9.2% of our cohort, which is almost twice the rate generally ascribed to this pathology in CS but was only reported in 1% of the cases in the Chinese study.19,20 The contribution of EAS was almost identical to that reported in the ERCUSYN database and the Chinese study, being the cause of CS in about 5% of the cases, which was far less often than in the older series.21 All cases were due to neuroendocrine tumors. Notably, there was not a single case of EAS caused by lung cancer. The overall decrease in lung cancer incidence in the past decades might offer a partial explanation.22
The primary aim of this study, to characterize a contemporary cohort of patients with CS, stemmed from a strong clinical impres- sion that milder forms of the disease presently comprise a large fraction of diagnosed patients relative to the severe classical syn- drome. This impression is strongly supported by the analysis of our single-center cohort. The understanding that mild forms of cortisol excess are increasingly being detected has already permeated the literature, mostly, but not exclusively, in connection with subclinical CS encountered in patients with adrenal incidentalomas. This reali- zation prompted a number of efforts to propose diagnostic scores that would improve the pretest probability for CS.16
Current guidelines recommending a strong clinical suspicion as a prerequisite to initiate a diagnostic workup are still engrained in the literature.11-13 However, none of the subjects in our cohort had more than 2 of the typical discriminatory signs, such as purple striae, easy bruising, facial plethora, buffalo hump, edema, and proximal muscle weakness at the time of presentation. This is supported by the fact that all but one of the cases in our current series omitted to report on these classical signs.19,20
Is the increased amount of mild forms in our series simply a reflection of unjustified and excessive testing? Although testing in Israel is at the discretion of the physician and tends to be relatively unrestricted, partly perhaps because patients do not get charged, we do not believe this to be the case. First, a hormonal workup, including the exclusion of cortisol excess, is mandatory once adrenal masses are identified, whether incidentally or in pursuit of some CS-related symptoms.23 Second, as seen in this report, adrenal carcinomas with CS do not often express a highly suspicious Cushingoid pheno- type. Third, at least a subset of patients with mild symptoms consis- tent with but not pathognomonic of CS tends to progress with time, which, in itself, warrants timely identification.24 Last, since the pub- lication of the Endocrine Society’s guidelines in 2008, new evidence
| Symptom, sign, or comorbidity | Frequency in the literature | Frequency in our study | P value |
|---|---|---|---|
| Weight gain | 0.91 | 0.77 | .001 |
| Menstrual disturbances | 0.84 | 0.48 | <. 001 |
| Hirsutism | 0.81 | 0.49 | <. 001 |
| Easy bruising | 0.62 | 0.26 | <. 001 |
| Muscle weakness | 0.56 | 0.30 | <. 001 |
| Facial plethora | 0.94 | 0.28 | <. 001 |
| Moon face | 0.88 | 0.44 | <. 001 |
| Buffalo hump | 0.54 | 0.38 | .005 |
| Striae | 0.56 | 0.21 | <. 001 |
| Edema | 0.5 | 0.27 | <. 001 |
| Obesity | 0.97 | 0.44 | <. 001 |
| Hypertension | 0.74 | 0.62 | .013 |
| Dyslipidemia | 0.71 | 0.48 | <. 001 |
| Depression | 0.62 | 0.25 | <. 001 |
| Diabetes | 0.5 | 0.37 | .027 |
| Osteoporosis | 0.5 | 0.38 | .05 |
Comparisons with rates in the classical literature11-17 were carried out by one sample binomial test.
had emerged that even mild cortisol excess is linked to increased metabolic disease as well as to cardiovascular morbidity and mor- tality.25-28 Therefore, the detection of milder forms of CS whose presenting symptoms would not, by themselves, raise a strong sus- picion of this condition does not appear futile.
Among the metabolic derangements typical of CS, hypertension continues to be the most common and has not decreased much in comparison with earlier reports, hovering around 72%-76% (62% in our series). In accordance with our report, the current series 19,20 indicates that obesity, dyslipidemia, and diabetes are less com- mon in subjects with CS than in the past, although still more prevalent than in the general population.9,15 The same trend seems to apply to metabolic bone disease and fractures. Indeed, we recorded a fracture rate of 18%, similar to that reported in the ERCUSYN database (21%) and in the Chinese study (17.1%), which were all less frequent than in the older series. Taken together, the data indicate that CS presentation in the modern era is mostly devoid of the typical stigmata associated with this syndrome.29
Our findings highlight the clinician’s conundrum of not missing the diagnosis of CS, a potentially lethal condition, while adhering to guidelines that advocate against liberal screening in common metabolic conditions. This topic, which initially spurred special interest with respect to type 2 diabetes mellitus,30,31 has recently gained further support. Indeed, in a prospective study of unselected patients with newly diagnosed type 2 diabetes mellitus who were devoid of the classical stigmata of CS, the prevalence of hyper- cortisolism was 5%.32 Considering the lifetime cost incurred by the care of diabetes, some form of screening for CS, even in the absence of solid clinical suspicion, does not seem excessive.
This study has some limitations. Inherent to its retrospective design, not all typical stigmata of CS were necessarily recorded, possibly leading to an underestimation of some classical features of the syndrome. By carefully reviewing the patients’ charts, we were often able to determine the period over which the weight gain took place or whether hypertension or diabetes had become difficult to control. Such information, however, was almost never available for other features, such as insomnia or muscle weakness. Nevertheless, the fact that a senior endocrinologist aware of the diagnosis made a point to state that the patient did not appear to have CS in more than half of the cases supports a generally milder clinical presentation than in the past. Moreover, the lower than expected rate of comorbidities also attests to a somewhat less severe condition. As alluded to, the liberal screening approach for CS in obesity and diabetes in Israel could be one of the reasons for a wider detection of the syndrome at a lower stage of severity. Although we fully appreciate that typical CS presentation occurs nowadays, the fact that our findings appear to be consistent with contemporary data generated in a variety of European countries and China suggests that this is not only a local phenomenon.19,20
In summary, our findings challenge the premise that screening for CS should only occur if there is a strong clinical suspicion based on stigmata that have become relatively rare nowadays. These findings do not allow us to devise a clinical threshold that would render such an investigation cost-effective. However, a more liberal approach other than that advocated by the current guidelines should be considered, which could lead to the earlier detection of this potentially lethal condition.
Acknowledgments
We are indebted to Dr Tomer Ziv-Baran for his support of the statistical analysis of the data. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Disclosure
The authors have no disclosure to report.
References
1. Barbot M, Zilio M, Scaroni C. Cushing’s syndrome: overview of clinical pre- sentation, diagnostic tools and complications. Best Pract Res Clin Endocrinol Metab. 2020;34(2):101380.
2. Etxabe J, Vazquez JA. Morbidity and mortality in Cushing’s disease: an epide- miological approach. Clin Endocrinol (Oxf). 1994;40(4):479-484.
3. Bolland MJ, Holdaway IM, Berkeley JE, et al. Mortality and morbidity in Cushing’s syndrome in New Zealand. Clin Endocrinol (Oxf). 2011;75(4):436-442.
4. Wengander S, Trimpou P, Papakokkinou E, Ragnarsson O. The incidence of endogenous Cushing’s syndrome in the modern era. Clin Endocrinol (Oxf). 2019;91(2):263-270.
5. Hirsch D, Tsvetov G, Manisterski Y, et al. Incidence of Cushing’s syndrome in patients with significant hypercortisoluria. Eur J Endocrinol. 2017;176(1):41-48.
6. Moreno CC, Hemingway J, Johnson AC, Hughes DR, Mittal PK, Duszak Jr R. Changing abdominal imaging utilization patterns: perspectives from Medicare beneficiaries over two decades. J Am Coll Radiol. 2016;13(8):894-903.
7. Ragnarsson O, Olsson DS, Chantzichristos D, et al. The incidence of Cushing’s disease: a nationwide Swedish study. Pituitary. 2019;22(2):179-186.
8. Papakokkinou E, Olsson DS, Chantzichristos D, et al. Excess morbidity persists in patients with Cushing’s disease during long-term remission: a Swedish nationwide study. J Clin Endocrinol Metab. 2020;105(8):2616-2624.
9. Dekkers OM, Horváth-Puhó E, Jørgensen JOL, et al. Multisystem morbidity and mortality in Cushing’s syndrome: a cohort study. J Clin Endocrinol Metab. 2013;98(6):2277-2284.
10. Clayton RN, Jones PW, Reulen RC, et al. Mortality in patients with Cushing’s disease more than 10 years after remission: a multicentre, multinational, retrospective cohort study. Lancet Diabetes Endocrinol. 2016;4(7):569-576.
11. Nieman LK, Biller BMK, Findling JW, et al. The diagnosis of Cushing’s syndrome: an Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab. 2008;93(5):1526-1540.
12. Nieman LK. Diagnosis of Cushing’s Syndrome in the modern era. Endocrinol Metab Clin North Am. 2018;47(2):259-273.
13. Nieman LK. Diagnostic tests for Cushing’s syndrome. Ann N Y Acad Sci. 2002;970:112-118.
14. Sharma ST, Nieman LK, Feelders RA. Cushing’s syndrome: epidemiology and developments in disease management. Clin Epidemiol. 2015;7:281-293.
15. Sharma ST, Nieman LK, Feelders RA. Comorbidities in Cushing’s disease. Pitu- itary. 2015;18(2):188-194.
16. Braun LT, Riester A, Oßwald-Kopp A, et al. Toward a diagnostic score in Cushing’s syndrome. Front Endocrinol (Lausanne). 2019;10:766.
17. Stewart PM, Newell-Price JDC. The adrenal cortex. In: Melmed S, Polonsky KS, Larsen PR, Kronenberg HM, eds. Williams Textbook of Endocrinology. 13th ed. Elsevier Inc; 2016:507-517.
18. Stratakis CA. Skin manifestations of Cushing’s syndrome. Rev Endocr Metab Disord. 2016;17(3):283-286.
19. Valassi E, Santos A, Yaneva M, et al. The European Registry on Cushing’s syn- drome: 2-year experience. Baseline demographic and clinical characteristics. Eur J Endocrinol. 2011;165(3):383-392.
20. Zhou J, Zhang M, Bai X, et al. Demographic characteristics, etiology, and comorbidities of patients with Cushing’s Syndrome: A 10-year retrospective study at a large general hospital in China. Int J Endocrinol. 2019;2019:7159696.
21. Isidori AM, Kaltsas GA, Pozza C, et al. The ectopic adrenocorticotropin syn- drome: clinical features, diagnosis, management, and long-term follow-up. J Clin Endocrinol Metab. 2006;91(2):371-377.
22. Lu T, Yang X, Huang Y, et al. Trends in the incidence, treatment, and survival of patients with lung cancer in the last four decades. Cancer Manag Res. 2019;11: 943-953.
23. Fassnacht M, Arlt W, Bancos I, et al. Management of adrenal incidentalomas: European Society of Endocrinology Clinical Practice Guideline in collaboration with the European Network for the Study of Adrenal Tumors. Eur J Endocrinol. 2016;175(2):G1-G34.
24. Morelli V, Arosio M, Chiodini I. Cardiovascular mortality in patients with subclinical Cushing. Ann Endocrinol (Paris). 2018;79(3):149-152.
25. Di Dalmazi G, Vicennati V, Rinaldi E, et al. Progressively increased patterns of subclinical cortisol hypersecretion in adrenal incidentalomas differently pre- dict major metabolic and cardiovascular outcomes: a large cross-sectional study. Eur J Endocrinol. 2012; 166(4):669-677.
26. Morelli V, Reimondo G, Giordano R, et al. Long-term follow-up in adrenal incidentalomas: an Italian multicenter study. J Clin Endocrinol Metab. 2014;99(3):827-834.
27. Debono M, Prema A, Hughes TJB, et al. Visceral fat accumulation and post dexamethasone serum cortisol levels in patients with adrenal incidentaloma. J Clin Endocrinol Metab. 2013;98(6):2383-2391.
28. Debono M, Bradburn M, Bull M, Harrison B, Ross RJ, Newell-Price J. Cortisol as a marker for increased mortality in patients with incidental adrenocortical ad- enomas. J Clin Endocrinol Metab. 2014;99(12):4462-4470.
29. Cushing H. The basophil adenomas of the pituitary body and their clinical mani- festations (pituitary basophilism). Bull Johns Hopkins Hospital. 1932:L:137-195.
30. Leibowitz G, Tsur A, Chayen SD, et al. Pre-clinical Cushing’s syndrome: an unexpected frequent cause of poor glycaemic control in obese diabetic patients. Clin Endocrinol (Oxf). 1996;44(6):717-722.
31. Terzolo M, Reimondo G, Chiodini I, et al. Screening of Cushing’s syn- drome in outpatients with type 2 diabetes: results of a prospective
multicentric study in Italy. J Clin Endocrinol Metab. 2012;97(10): 3467-3475.
32. Steffensen C, Dekkers OM, Lyhne J, et al. Hypercortisolism in newly diagnosed type 2 diabetes: a prospective study of 384 newly diagnosed patients. Horm Metab Res. 2019;51(1):62-68.