Quarterly Journal of Medicine, New Series 77, No. 283, pp.

1113-1133, November 1990

An Evaluation of the Distinction of Ectopic and Pituitary ACTH Dependent Cushing's Syndrome by Clinical Features, Biochemical Tests and Radiological Findings
S. B. BLUNT, L. M. SANDLER, J. M. BURRIN and G. F. JOPLIN From the Endocrine Unit, Department of Medicine, Royal Postgraduate Medical School, Hammersmith Hospital, London W12 ONN, U.K.
Accepted 3 April 1990

SUMMARY The efficiency of various laboratory and radiological investigations in the differentiation of ectopic from pituitary dependent Cushing's syndrome was studied, based on findings in 23 patients with verified Cushing's disease and seven patients with the ectopic ACTH syndrome. Clinical features strongly favouring the ectopic type were male sex and history for less than 18 months. Basal biochemical features strongly indicating the ectopic syndrome included plasma K+ < 3.0 mmol/1 and HCO3 > 30 mmol/1; serum cortisol at 9 a.m. or midnight of > 800 nmol/1; urine free cortisol > 1300 nmol/24 hours; plasma ACTH > 100 ng/1. In the high-dose dexamethasone suppression test, suppression by < SO per cent of 9 a.m. serum cortisol, urine free cortisol or 17-oxogenic steroids was usually indicative of an ectopic source of ACTH. A mean suppressed value of > 450 nmol/1 for the 9 a.m. and midnight cortisol combined occurred in all of those with the ectopic syndrome, but in none of the 23 patients with Cushing's disease. For urine free cortisol, a mean suppressed value of < 1000 nmol/24 hours was found in all patients with Cushing's disease, but in none of those in the ectopic group. In the metyrapone test, there was an increase of < 3-fold in 11-deoxycortisol at 24 hours in patients with ectopic ACTH; the increase was > 3-fold in all but one of the patients with Cushing's disease. Failure to respond to either dexamethasone or metyrapone was found in only one of the patients with Cushing's disease (Patient 16); in the ectopic group, all patients except Patient D failed to respond to either test. It is concluded that patients presenting with clinically obvious Cushing's syndrome along with measurable plasma ACTH can be reliably divided by conventional tests into those that are driven from the pituitary and those driven by ectopic ACTH.

Address correspondence to Dr. S. B. Blunt, Department of Neurology, Institute of Psychiatry, De Crespigny Park, Denmark Hill, London SE5 8AF, U.K. Oxford University Press 1990

1114

S. B. Blunt and others

INTRODUCTION Following the identification of Cushing's syndrome, the differential diagnosis of the broad group that are ACTH driven from the smaller group caused by a cortisol-secreting adrenal tumour is relatively easy [1]. Real difficulty arises in the distinction between an ACTHsecreting pituitary adenoma (Cushing's disease) and an ACTH secreting tumour which is located ectopically. The results of surgical management of an ectopic group of patients have been discussed by our team in an initial report [2]. Subsequently, Howlett et al. [3] presented a systematic account of a series which compared the results of investigations obtained from a pituitary-driven group in contrast to an ectopic group, to which they added further studies later [4]. We now report ourfindingsfrom conventional tests, based on 23 patients with pituitarydriven Cushing's disease, all confirmed by cure of the disease by pituitary ablative treatment, in comparison with seven patients with histologically confirmed ectopic sources of ACTH. PATIENTS AND METHODS Two groups of adult patients were studied. The first comprised 23 patients with hitherto untreated Cushing's disease who were investigated and treated consecutively in our unit with interstitial pituitary irradiation during the period 1971 to 1982, and followed up by ourselves into full biochemical and clinical remission [5]. The second group comprised seven patients also presenting with obvious Cushing's syndrome, and who were found to have histologically identifiable ectopic ACTH-producing tumours; they were studied in the period 1967 to 1987. All patients in both groups had previously been shown to have elevated serum cortisol concentrations with a loss of diurnal variation, measurable plasma ACTH levels, and high urine free cortisol concentrations. All but one were ambulant when studied. CLINICAL FEATURES, RADIOLOGY, AND HISTOPATHOLOGY Cushing's Disease Group (Table 1) All but one of the 23 patients were female and were aged 18-60 years. The length of history, based on the time when symptoms were first noticed up to the time of presentation to our unit, ranged from 12 to 72 months (mean 40 months). Pituitary fossa radiology by plain coned views and by lateral hypocycloidal tomography (3 mm cuts) was performed in all 23 patients; 10 were abnormal. Pituitary CT scan was performed in the later cases (Patients 6,13 and 14): this showed the presence of a tumour in Patient 13; Patients 6 and 14 had normal CT scans. The diagnosis of adenoma was confirmed by pituitary histology in all 11 patients with adequate biopsies. The period of follow-up since referral has ranged from 3 to 13 years (159 patient-years) with no recurrence of disease in any patient. Ectopic ACTH Group (Table 2) Seven patients with histologically proven ectopic sources of ACTH were aged 23 to 59 years (mean 38 years) at presentation to us. Only one patient was female and her case has been reported in detail [6]. The length of history at presentation to us ranged from 2 to 12 months (mean 7.4 months). There were already clinical signs of an underlying tumour in four: Patients B, C and E had irregular hepatomegaly consistent with metastases, and this was

Ectopic and Pituitary-dependent Cushing's Syndrome TABLE 1. Clinical, radiological and pathological details in the group of 23 patients with Cushing's disease Patient Age Length of Length of Pituitary fossa Pituitary no. (years)* history follow-up radiology histology (months) (years)b
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16d 17 18 19 20 21 22 23 37 51 31 52 45 44 28 18 33 53 53 56 30 32 28 52 52 60 32 46 26 30 47 37 48 36 36 72 24 72 36 24 24 24 24 24 36 36 60 18 72 48 72 12 36 24 60 40 4 3 5 10 5 3 3 4 7 5 4 7 3 5 13 8 8 12 11 7 10 10 12 7

1115

Normal Abnormal Abnormal Normal Abnormal Normal Normal Normal Normal Abnormal Normal Abnormal Abnormal Normal Normal Normal Normal Abnormal Abnormal Abnormal Normal Abnormal Normal

NP

Adenoma
NP

Adenoma
NP NP NP * * NP NP

Adenoma Adenoma Adenoma Adenoma Adenoma

*

Adenoma
*

Adenoma Adenoma Adenoma

Mean Range
/*

18-60

12-72 <0.0001

3-13

NP, not performed;*, no, or insufficient, pituitary material. Age at referral to HH. b From the time when first assessed by no. c VS. ectopic group in Table 2. 6 The only male. confirmed radiologically by liver scan or angiogram, while Patient G had clinical and radiological evidence of metastatic spread of a previously treated medullary carcinoma of the thyroid. In Patients A and D, plain chest radiographs and tomograms showed a lung mass consistent with a bronchial carcinoid. In Patient F, an ectopic nature was suspected on the grounds of his biochemical tests, but there was no clinical or radiological clue as to the site of the ACTH source, and the appendicular carcinoid was found only at abdominal exploration. Some data pertaining to six of our ectopic group have been reported by Davies et al. [2]. Our Patients A.B.C, D, E, F, are those referred to before as Patients2,3,4,10, 9,8. Our Patient E was reported by Ghatei et al. [7] in considerable biochemical detail. BIOCHEMICAL METHODS Blood was taken for plasma potassium from each patient on several occasions before any diuretic or potassium therapy, and the mean is given.

1116

5. B. Blunt and others

TABLE 2. Clinical, radiological and pathological details in the group of seven patients with ectopic A CTH Patient Age Length of Length of (years) history follow-up (months) (years)*
A B C D E F Gb 27 59 29 28 49 23 49 38 2 2 6 12 12 6 12 7.4

Radiology Chest Other

Histology

14 RLL mass _ Death within Normal Liver scan: 1 year metastases Death within Normal Liver scan: 2 years metastases 8 RLL mass CT thorax: RLL mass Death within Normal Angiography: 1 year liver metasts 8 Normal Barium enema: Normal 1 5

Lung carcinoid
MCT

metastases Pancreatic apudoma Lung carcinoid Metastatic apudoma Benign appendicular carcinoid Metastatic MCT

Metastases -

Mean Range

23-59

2-12

< 1-14

* From time when first assessed. "The only female. Blood glucose was measured on venous whole blood by the ferricyanide autoanalyser method in the first 14 patients, and as plasma glucose with the glucose oxidase method in the remaining 16. Fasting samples were taken from all patients, the upper limit of normal adopted being 90 mg/dl for blood glucose or 6 mmol/1 for plasma glucose. Oral glucose tolerance testing was performed using 50 g glucose, with an abnormal test being said to occur when the peak level attained exceeded 160 mg/dl for blood glucose or 10 mmol/1 for plasma glucose, together with a 2 hour value exceeding 120 mg/dl blood glucose or 7.1 mmol/1 plasma glucose [8]. Any patient showing an elevated fasting glucose and/or an abnormal glucose tolerance test has been reported in Tables 3 and 4 as having impaired glucose tolerance [9]. Serum cortisol concentrations were obtained at 9.00 a.m. and midnight. A competitive protein binding method [10, 11] was used in all but Patient G, with an upper limit of normal at 9.00 a.m. of 700 nmol/1 and at midnight of 200 nmol/1. Details of how the normal ranges were established have been reported earlier [11, 12]. All of our patients, except Patient G, were ambulant at the time of testing. The effect of corticosterone on the assay was quantitated and found to be small, therefore its contribution to cortisol readings was not taken into account. Nevertheless, the cortisol reading is within 12 per cent of true cortisol levels as shown by chromatographic methods [10]. Serum cortisol levels in Patient G were measured by a solid phase first antibody radioimmunoassay [13]. Intra- and interassay coefficients of variation were 6.5 per cent and 9.8 per cent respectively at a cortisol value of 447 nmol (NR<700 nmol/1 at 9.00 a.m.; <250 nmol/1 at midnight). No samples were measured by the Mattingly method in this series. Plasma for determination of ACTH levels was taken at 9.00 a.m. and midnight and was measured by double antibody radioimmunoassay after extraction onto vycor glass [14]. The normal range at 9.00 a.m. is 10-80 ng/1 and at midnight < 10 ng/1. The interassay coefficient of variation is 10.9 percent at 123 ng/1.

Ectopic and Pituitary-dependent Cushing's Syndrome

1117

Urinary free (unconjugated) cortisol was measured by competitive protein binding [10], the upper limit of normal being 270 nmol/24 hours [15] with a mean of 100 nmol/24 hours. The method is also sensitive to 11 -deoxycortisol and corticosterone which give an effect that is 90 per cent and 70 per cent of that of cortisol respectively. The quantities of these two steroids are usually small. The average deviation from values obtained by chromatographic methods for cortisol was 13 per cent (p< 0.001) [10]. Urine 17-oxogenic steroids and 17-oxosteroids were determined by the method of Appleby et al. [16] in all cases. The upper limits of normal are 60 /*mol/24 hours, and 77 ^mol/24 hours, respectively [16, 17]. Serum 11-deoxycortisol ('compound S') measurements were made by a competitive protein binding method in all cases except Patient G. The CPB method for 11-deoxycortisol was an adaptation of the method used for determination of unbound cortisol [10], and involved an additional extraction onto carbon tetrachloride, which was then dried and assayed as for urine free cortisol. The cross-reactivity of cortisol in this assay was 1.3 per cent and our basal range is 28-84 nmol/1. Serum 11-deoxycortisol measurements in Patient G were made by a direct non-extraction double antibody radioimmunoassay (Radioimmunoassay Systems Laboratories Inc., California, USA). The intra- and interassay coefficients of variation were 5.9 per cent and 13.7 per cent respectively, at a level of 8.4 nmol/1; crossreactivity with cortisol is 1.4% The normal range (basal) is <23 nmol/1. DYNAMIC TESTS High Dose Dexamethasone Suppression Test Dexamethasone 2 mg was administered 6 hourly for 72 hours. Basal collections of blood at 9 a.m. and midnight for cortisol, and urine for 17-oxogenic steroids, 17-oxosteroids and unbound cortisol were made on two consecutive 24-hour periods immediately prior to the test, and on the second and third days on dexamethasone. The extent of suppression of each parameter was calculated as the percentage change of the mean values found on days 2 and 3 on dexamethasone, compared with the mean basal value. Suppression by 50 per cent or more of urine 17-hydroxy corticosteroids levels (equivalent to 17-oxogenic steroids) is said to be diagnostic of Cushing's disease rather than an adrenal tumour [18, 19], and will be adopted here as a criterion of suppression. Metyrapone Stimulation Test Metyrapone was given at a dose of 750 mg 4-hourly for 24 hours, commencing at 9 a.m. Twenty-four hour urine collections of 17-oxogenic steroids (and 17-oxosteroids) were made on the day of metyrapone administration and on the day after metyrapone. Blood for 11deoxycortisol levels was taken at 9 a.m. on the morning just before starting metyrapone, and at 9 a.m. the next day. A positive response to metyrapone was recorded when there was an increase of 100 per cent or more of 11 -deoxycortisol or urine 17-oxogenic steroids [ 18] either on the day of metyrapone, or the day after, compared with their basal values. STATISTICAL METHODS Student's unpaired /-test was used for the comparisons of all values between the two groups. The x1 test was used for comparison of the blood sugar data, and for the analysis of the responses to combined dexamethasone suppression and metyrapone stimulation.

1118 RESULTS

S. B. Blunt and others

BASAL BIOCHEMICAL DATA (Tables 3 and 4) Plasma Potassium In the 23 patients with Cushing's disease, levels ranged from 3.5 to 4.5 mmol/1 with a mean of 3.9 mmol/1; four patients (17 per cent) showed marginally subnormal levels ( < 3.8 mmol/1). All seven patients with ectopic ACTH production had hypokalaemia (range 1.7-2.8 mmol/1, mean 2.2 mmol/1). The difference between the two groups was highly significant (p < 0.0001). If the two groups are compared in relation to a reference level of 3.0 mmol/1, there is no overlap between the groups. Plasma Bicarbonate Levels ranged from 24 to 35 mmol/1 in the Cushing's disease group (mean 29 mmol/1), compared with a range of 32-44 mmol/1 (mean 37 mmol/1) in the ectopic group (p < 0.0001). In relation to a reference level of 30 mmol/1, a distinction, with two marginal exceptions, (Patients 1 and 77, Table 3) is seen between the two groups. Glucose Tolerance This was abnormal in five of the 23 patients with Cushing's disease (21 per cent) compared with four of seven in the ectopic group (57 per cent) (/} = 3.18; p = 0.087). Basal 9 a.m. Serum Cortisol Values obtained under basal conditions were above the upper limit of normal (700 nmol/1) in only two patients with Cushing's disease (Patients 2 and 18). The range of values found in this group (303-1173 nmol/1), was almost entirely outside the range of values found in patients with the ectopic ACTH syndrome (828-2159 nmol/1). The overlap was due to one patient (Patient 18) in the Cushing's disease group who had unusually high 9 a.m. (and midnight) values. The other 22 patients with Cushing's disease (96 per cent) had 9 a.m. serum cortisol values below a reference level of 800 nmol/1, whereas all seven patients with ectopic ACTH production had values greater than this. The difference between the mean values of the two groups (576 nmol/1 and 1460 nmoi/1) was highly significant (p< 0.001). Basal Midnight Serum Cortisol Values were elevated ( > 200 nmol/1) in all patients in both groups. The range found in the Cushing's disease group (321-1048 nmol/1) was almost completely outside that found in the ectopic group (800-1515 nmol/1). The overlap was again due to the high value in Patient 18 in the Cushing's disease group. The remaining 22 patients in this group had midnight values below a reference level of 800 nmol/1, whereas all seven with the ectopic ACTH syndrome had values greater than this. The mean values for the two groups of 519 nmol/1 and 1117 nmol/1 were significantly different (/?<0.0001). Combined Basal 9 a.m. and Midnight Serum Cortisol Elevation of serum cortisol at both 9 a.m. and midnight to above a reference level of 800

TABLE 3. Basal biochemical data" at time of referral to Endocrine Unit in the group of 23 patients with Cushing's disease.
Impaired glucose tolerance 9 a.m. serum cortisol (nmol/1) [200-700] Midnight serum cortisol (nmol/1) [<200] 18 69 34 9 a.m. plasma ACTH (ng/1) [10-80] Midnight plasma ACTH (ng/1) [<10] 17-OGSC 17OSC UFC (nmol/24 h) Omol/24 h) (/(mol/24 h) [<270] [<60] [<77|

Patient no.

Plasma K+ (mmol/l) [3.8-5.2]

Plasma HCO 3 (mmol/l) [24-30]

5?

5
j? a

57 38 89 36 24 49 30 32

a
3

76 81 33 38 14 40 70 35 38 19 43 39 38 53 34 9 33 -

P I'

No No No No No Yes No No No Yes No Yes No Yes No No No No No No No Yes No

f
1
39 15-206 22 0.02

85 124 98 41 70 133 206 105 105 147 142 59 117 78 72 46 71 _ 50 125 102 91 63 96 118 <I5 91 98 45 30 24 51 30 15-118 18 < 0.008 50 33 44 66 110 45 23 3-81 22 0 02

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Mean S.D. Range n /* 21%; yes 0.087 26 94 38 _ 64 59 202 46 47 92 30 56 109 24 130 30 49 98 15 51 11 64 46 11-202 20 <0001 1453 505 527 717 375 848 608 1000 1200 1184 496 356 1000 1488 759 886 295 1446 644 786 587 1305 600 839 373 295-1488 23 < 0.001

4.1 3.8 3.4 3.9 3.8 34 3.9 3.6 42 4.0 4.3 3.8 4.0 4.5 3.9 4.3 3.9 4.0 3.9 4.0 4.5 4.2 3.5 3.9 0.3 3.5-4.5 23 < 0.0001 321 765 329 580 423 581 552 639 479 741 513 412 516 492 476 330 321 1048 386 507 427 331 772 519 180 321-1048 23 < 0.0001

31 27 27 30 29 29 29 29 30 30 31 25 28 24 25 30 29 27 29 29 30 29 29 2.23 24-34 22 < 0.0001

454 799 469 529 695 432 690 698 660 631 518 565 516 612 360 483 495 1173 303 625 441 437 662 576 178 303-1173 23 < 0.001

* Normal limits shown in square brackets. b us. ectopic group in Table 4. UFC. urine free cortisol. 17-OGS. 17-oxogenic steroids; 17-OS, 17-oxosteroids

TABLE 4. Basal biochemical data" at time of referral to Endocrine Unit in the group of seven patients with ectopic A CTH
9 a.m. serum cortisol (nmol/1) [200-700] Midnight serum cortisol (nmol/1) [<200] 9 a.m. plasma ACTH (ng/1) [10-80] Midnight UFC b (nmol/24 h) plasma ACTH [<270] (ng/1) [<10] 17-OGS b 17OSb (/<mol/24 h) (//mol/24 h) [<60] [<77]

Patient

Plasma K+ (mmol/1) [3.8-5.2]

Plasma Impaired glucose HCO 3 (mmol/1) tolerance [24-30]

2.3

37

No

1.9

38

Yes

Yes No No Yes Yes

A B C D E F G Mean S.D. Range n 1850 1667 828 912 1490 2159 1316 1460 485 828-2159 7 1160 1490 800 950 1515 854 1009 1111 291 800-1515 7

1.7 2.8 2.1 2.3 2.5 2.2 0.3 1.7-2.8 7

44 32 34 40 34 37 4.4 32-44 7

57%; yes 7

_ 513 96 321 138 170 247 170 96-513 5

_ _ 150 _ 69 106 108 40.5 69-150

_ 27 887 6637 1360 >6000 4303 2000 8031 9950 1360-27 837 6

330 151 _ 209 94 164 117 28-336 4

141 64 _ 78 35 80 45 35-141 4

Normal limits shown in brackets. UFC, urinary free cortisol; 17-OGS, 17-oxogenic steroids; 17-OS, 17-oxosteroids.

Ectopic and Pituitary-dependent Cushing's Syndrome

1121

nmol/1 was found in only one of the 23 patients with Cushing's disease, but in all seven with ectopic ACTH. The means of the combined values for the two groups were 547 nmol/1 and 1286 nmol/1 respectively (p< 0.001). Basal 9 a.m. plasma ACTH Twenty patients were assessed in the Cushing's disease group (range 11-202 ng/1) and 30 per cent had elevated concentrations, while all five of those assessed in the ectopic group had elevated values (range 96-513 ng/1). There was clearly a substantial overlap between the two groups; the mean value was, however, higher in the ectopic group (247 ng/1 compared with 64 ng/1; p< 0.001). Only three of the 20 pituitary-driven cases exceeded a reference level of 100 ng/1, while four of the five ectopic cases did so. Basal midnight plasma ACTH ACTH was detectable in 17 of the 18 assessed patients with Cushing's disease (range 10-118 ng/1, mean 51 ng/1) and in all of the three assessed patients with ectopic ACTH (range 69-150 ng/1, mean 108 ng/1). Again there was a substantial overlap between the two groups. Basal UFC Values were above the upper limit of normal (270 nmol 24 hours) in all patients assessed in the two groups. In Cushing's disease, values ranged from 295 to 1488 nmol/24 hours (mean 839 nmol/24 hours) and four of the 23 patients (17 per cent) had values greater than a reference value of 1300 nmol/24 hours. In the ectopic group, values ranged from 1360 nmol/ 24 hours to 27 887 nmol/24 hours, (mean 8031 nmol/24 hours), and were significantly higher (p< 0.001), with only one ectopic patient coming into the range found in Cushing's disease, and all six exceeding 1300 nmol/24 hours. Basal 17-oxogenic Steroids and 17-oxosteroids Tables 3 and 4 show that both of these values showed little distinction between the two groups. The mean of 17-oxogenic steroids was somewhat higher in the ectopic group (164 /imol/24 hours compared with 96 /^mol/24 hours, /> = 0.02). For 17-oxosteroids, the mean value in the ectopic group was 80 /^mol/24 hours, compared with 45 ^mol/24 hours in those with Cushing's disease (p = 0.02). In addition, four (18 per cent) of the patients with Cushing's disease showed a basal value of 17-oxogenic steroids excretion within the normal range ( < 60 /imol/24 hours) compared with none of the four patients with ectopic ACTH. For 17-oxosteroids, 90 per cent of patients with Cushing's disease had values within the normal range ( < 77 /zmol/24 hours) compared with 2 of the 4 patients with ectopic ACTH.

DYNAMIC TESTS High Dose Dexamethasone Suppression Test This was performed on all 30 patients, with evaluations available of the responses of at least two of the following four measurements: 9 a.m. and midnight serum cortisol (taken as a single parameter), urinary free cortisol, 17-oxogenic steroids and 17-oxosteroids. The

TABLE 5. High dose dexamethasone suppression test in the group of 23 patients with Cushing's disease 1122 UFO 17-OGS" Mean Change suppressed (%) value /imol/24 h
,<mol/l _

Patient 9 a.m. serum cortisol 17-OS' Mean Change suppressed (%) value Mean Change suppressed (%) value nmol/24 h

Midnight serum cortisol

no.

Mean Change suppressed (%) value nmol/l _ -68

Mean Change suppressed (%) value nmol/l

C/0
CB

_ 236 _ _ 126 96 262 628 195 _ -70 -83 -52 -2 -60

-56 682 -42 65 -50 10

760 129 _ 502 208 150 601 243 316 -49 -74 _ -30 -44 -82 0 -83 -73 32 9 9 24 31 97 16 39 -90
-91

_ -74 -96 -75 -65 -76 -52 -84 -62 55 23 _ 13 5 20 39 10 16

-27 -71 _ -65 -64 -75 -44 -71 -57

-42

326

to

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Mean SD Range -76 -42 _ -85 -16 -84 -94 -74 -99 >-75 -50 -64 26 -299 17 0.09 -78 -86 -92 -59 -91 -74 -83 -94 -95 -88 -71 25 0-95 21 <0.0001 46 32 220 197 55 358 27 367 104 26 27 152 246 223 26-760 21 < 0.009 21 22 77 12 15 20 9 45 10 10 23 33 30 24 3-97 21 <0.001 -85 -62 -34 -83 -79 -38 -88 -65 -91 -89 -74 -43 -70 18 -34100 21 < 0.0001 10 9 77 18 12 16 7 12 10 10 20 25 20 18 2-77 21 <0.001

Si.

f*

_ 218 124 110 460 220 195 392 193 309 486 41 179 290 55 380 165 203 110 368 233 126 41-136 19 < 0.0001

_ -72 _ _ -82 -74 -33 -68 -59 -37 -62 -49 -16 -92 -50 -40 -88 -67 -45 -74 >-75 -18 -58 22 -1692 19 < 0.002

122 296 _ 69 276 50 60 97 <30 <U0 166 186 147 30-628 17 <0.001

-76 -76 + 100 -66 -54 + 87 -80 -76 -69 -75 -71 -77 -50 49 -77-+100 21 0.4

" UFC, urinary free cortisol; 17-OGS, 17-oxogenic steroids; 17-OS, 17-oxosteroids. b os. ectopic group.

TABLE 6. High dose dexamethasone suppression test in the group of seven patients with the ectopic ACTH syndrome UFO Change (%) 17-OGS' 17-OS1

Patient 9 a.m. serum cortisol Change (%)

Midnight serum cortisol

TJ
*

Mean Change suppressed (%) value nmol/1 Mean suppressed value nmol/24 h 173 160 Mean Change suppressed (%) value //mol/24 h

Mean suppressed value nmol/1

Mean Change suppressed (%) value /imol/24 h

itar c

t 1 a 5 !
Co"

A B C D E F G Mean SD Range n P -64 _ -20 -63 -41 -15 -41 23 -1563 5 0.09 152 75 140 44 -66- + 5 4 <0.001

552 1876 634 692 1743 952 926 1070 587 552-1876 7 <0.0001

-70 414 _ + 12 -23 636 -24 347 + 16 -55 500 -29 851 -28 550 31 200 - 7 0 - + 1 6 347-851 7 5 < 0.002 <0.001

_ 38651 1394 1142 5419 >1600 9641 16311 11421-38 5 < 0.009

_ + 38 -79 -16 + 26 0 -6 46 -79-+173 5 <0.0001

-47 +5 _ -27 -20 -22 21 -66- + 5 4 <0.0001

69 72 _ 61 25 57 21 25-72 4 <0.001

-49 -10 _ -21 -28 -28 16 -1049 4 0.4

Co

UFC, urinary free cortisol; 17-OGS, 17-oxogenic steroids; 17-OS, 17-oxosteroids. b vs. pituitary-dependent group in Table 5.

1124

S. B. Blunt and others

response of 9.00 a.m. and midnight serum cortisol, and of urinary steroids has been assessed by taking the means of the values obtained on days 2 and 3 (i.e. at 48 and 72 h) on dexamethasone and calculating the percentage change when compared with the basal value prior to administration of dexamethasone. The absolute values obtained are listed in Tables 5 and 6. The mean suppression of 9 a.m. serum cortisol was 58 per cent for Cushing's disease, and 28 per cent for the ectopic group (p < 0.002). Five of the seven patients in the ectopic group showed less than 50 per cent suppression (an empirically chosen reference point), compared with eight of the 19 patients with Cushing's disease. The mean suppression of midnight serum cortisol was 64 per cent for the Cushing's disease group, and 41 per cent for the ectopic group (p = 0.09); there was substantial overlap. Three of the five patients in the ectopic group suppressed by less than 50 per cent compared with four of the 17 patients tested in the Cushing's disease group. When the extent of suppression was expressed as a mean of the 9 a.m. and midnight percentage changes from basal values, there was some overlap in range of values found in the two groups, and the difference between their means was not significant (p = 0.063). Suppression of the mean value by 50 per cent occurred in nine of the 20 patients with Cushing's disease, and four of the five with ectopic ACTH. The mean of the absolute values for the 9 a.m. and midnight cortisol was calculated for each patient, and then the means for each group were compared. A mean suppressed value of 222 nmol/1 in the Cushing's disease group was considerably lower (p < 0.0001) than the value of 651 nmol/1 in the ectopic group (Fig. 1). There was no overlap in their ranges. None of 23 patients with Cushing's disease exceeded a reference value of 450 nmol/1 on dexamethasone, whereas all of five ectopic cases did so. The suppression of urinary free cortisol in the Cushing's disease group was 71 per cent compared with a mean suppression of only 6 per cent in the ectopic group (p< 0.0001). All but five of the 21 patients with Cushing's disease achieved 50 per cent reduction in unbound cortisol while only one of the ectopic group did so. The importance of the apparent suppression of urinary free cortisol found in this patient (Patient C) is difficult to evaluate, as his basal unbound cortisol levels had shown marked spontaneous fluctuation (from 1394 to 6637 nmol/24 hours). The mean suppressed value (Fig. 1) in the Cushing's disease group was 246 nmol/24 hours, compared with 9641 nmol/24 hours in the ectopic group (p< 0.009). None of 21 patients with Cushing's disease exceeded a reference value of 1000 nmol/24 hour while on dexamethasone, while all five ectopic cases did so. There was a highly significant difference (p< 0.0001) in the mean suppression of urinary 17-oxogenic steroids between the two groups: 70 per cent suppression for the Cushing's disease group compared with 22 per cent for the ectopic group. All 18 patients with over 50 per cent suppression had Cushing's disease. The one patient (Patient A) from the ectopic group who showed impressive suppression for 17-oxogenic steroids (but < 50 per cent) also showed > 50 per cent suppression of both the 9 a.m. and the midnight serum cortisol. He had otherwise performed as expected, with severe hypokalaemic alkalosis and extremely high basal levels of serum cortisol and urinary steroids. Three of the Cushing's disease group failed to show suppression of 50 per cent. There was marked overlap between the two groups in extent of suppression of urinary 17oxogenic steroids: mean suppression for the ectopic group was 28 per cent, compared with 50 per cent for the Cushing's disease group (p = QA). However, all 17 patients showing a reduction of > 50 per cent had Cushing's disease.

Ectopic and Pituitary-dependent Cushing's Syndrome CD 3=1200

r

1125

Ectopic

CD

Ectopic

1000

800

600

400

200

Se. cortisol 9 am & MN mean

UFC/24h

FIG. 1. Results of mean absolute values of 9 a.m. and midnight serum cortisol and urinary free cortisol in the high-dose dexamethasone suppression test in 21 patients with Cushing's disease andfivewith the ectopic ACTH syndrome. Metyrapone Test This was performed in 22 of the 23 patients with Cushing's disease and in all seven with ectopic ACTH. The absolute values obtained are shown in Table 7 and 8. There was almost complete overlap between the ranges of percentage change of urine 17oxogenic steroids seen in the two groups; comparison of their mean values showed no significant difference (p = 0.08). When the 21 patients with Cushing's disease are considered, failure of a ^100 per cent rise occurred in nine, compared with four of the five with ectopic ACTH. The two groups of patients showed complete overlap in urine 17-oxosteroids. When changes in levels of serum 11-deoxycortisol (measured 24 hours after the first dose of metyrapone) were compared, a difference between the two groups emerged: in the 16 Cushing's disease patients studied, concentrations rose by a mean of 530 per cent, compared with a mean increase of only 80 per cent in the ectopic group (p = 0.06). Values did not at least treble in four of the 16 Cushing's disease patients, or in any of the four patients in the ectopic group. The absolute values of 11-deoxycortisol did not differ between the two groups. Combination of Dynamic Test Criteria The response to both dexamethasone and metyrapone for each patient was considered. A patient was deemed to have shown overall suppression by dexamethasone if there was at

TABLE 7. Metyrapone test in the group of 23 patients with Cushing's disease After metyrapone Change (%) Compound S (9 a.m.)

Patient Before metyrapone

no.

17-OGS Compound S 17-OS" nmol/1 17-OS* 17-OGS1 Oimol/24 h) Oimol/24 h) (9 a.m.) (nmol/1) /zmol/24 h Change(%) /imol/24 h Change(%) 37 129 <70 59 71 <100 59 315 68 85 52 25 26 92 106 71 68 50 > 195 1109 400 >385 1168
300 90 143 117

i 2 3 4 5 6 7 8 9 10 685 564 300 >500 337 328 >230 1400 >300 >552 270
138 274 124

85 192 171 41 143 78 135 140 222 316 47 46 23 31 35 83 63 -

65 50 32 38 18 43 69 25 41 29

232 484 456 88 316 333 370 310 260 370

172 152 166 98 120 320 174 120 14 17

4 70 62 -32 44 113 53 184 65 72

11

122

37

Co

S"
ft
Sr

19 86 63

118 335 73 21 49 147 _ 50 123 102 207 137 124 116 154 39 60 97 37-315 16 60 27 19-126 19 < 0.003

51 55 55 15 24 _ 32 44 76 39 _

12 13 14 15 16 17 18 19 20 21 22 23 Mean SD Range n /*

21-335 21

14-69 19

110 522 164 99 66 70 118 111 273 130 245 143 65-522 21 0.07

-6 55 124 370 33 31 136 -9 167 -37 112 103 37-326 21 0.08

-62 56 14 56 28 9 87 -140 43 72 140-184 19 0.8

>340 809 402 469 461 179 582 522 292 195-1168 16 0.5

620 1658 192 309 193 364 750 530 436 117-1658 16 0.06

17-OGS, 17-oxogenic steroids; 17-OS, 17-oxosteroids. b os. ectopic group in Table 8.

TABLE 8. Metyrapone test in the group of seven patients with ectopic ACTH production Ecto\ After Metyrapone 17-OS /imol/24 h Change nmol/1 Compound S
3

Patient

Before Metyrapone

Compound S 17-OGS 17-OGS 17-OS (/imol/24 h) (//mol/24 h ) (9 a.m.) (nmol/1) /jmol/24 h Change

4 Pituitar

Change

_ -

386 171 288 440 750 -

96 113 81 104 -

1 " 3
8-

407 46-643 5

99

s -*

1
1 "
to*

_ 414 480 497 643 173 660 441 179 165 591 275 815 405 286 41-815 5 0.07 -53 55 -3 75 186 105 -37 8 188 25 126 71 55 -37-105 24-188 5 4 0.08 < 0.003 -45 -34 129 80 33 85 -45-129 4 0.8

43-750 5

25-113 4

A B C D E F G Mean SD Range n /*

225 1017 642 659 635 323 21-1017 4 0.5

-53 104 270 0 80 142 - 54-270 0.06

to

17-OGS, 17-oxogenic steroids; 17-OS, 17-oxosteroids. b vs. pituitary-dependent group in Table 7.

1128

S. B. Blunt and others

least a 50 per cent suppression of at least two of the four measurements (9.00 a.m. and midnight serum cortisol; urinary free cortisol, 17-oxogenic steroids, and 17-oxosteroids) in the test; overall stimulation in response to metyrapone was deemed to have occurred if there was a ^ 100 per cent increase in both measurements (serum 11-deoxycortisol and urine 17oxogenic steroids). In the Cushing's disease group, only one {Patient 16) of the 19 patients (4 per cent) with complete tests showed neither overall suppression in response to dexamethasone, nor overall stimulation to metyrapone. All but one of the seven patients in the ectopic ACTH group (14 per cent), (Patient D, who failed to halve serum cortisol or unbound cortisol, but doubled both 17-oxogenic steroids and 11-deoxycortisol) showed failure of responses to both tests. The two groups clearly differed (j2= 15.57, p< 0.0001). Ten of the 19 patients with Cushing's disease in whom both tests were performed responded appropriately to both tests. This contrasts with the ectopic group, where no patient showed overall suppression to dexamethasone according to the above criteria, and only one patient (Patient D) showed overall stimulation to metyrapone. RADIOLOGY Radiographs of the Pituitary Fossa Plain coned radiographs or lateral hypocycloidal tomography of the pituitary fossa showed an abnormal appearance in 10 of the 23 patients (48 per cent) with Cushing's disease. No abnormality was found in the seven patients with ectopic ACTH. Computerized Tomography CT scans of the pituitary were also performed in the last three patients (Patients 6,13, and 14) in the Cushing's disease group, one of whom (Patient 13) had appearances, both on CT and plain films, consistent with the presence of a tumour; the two patients whose CT scans were normal also had normal plain radiographs, although case 14 had an adenoma shown on biopsy. Other Radiography In six of the seven patients with ectopic ACTH, the location of the primary tumour, or of its metastases was established by relatively non-invasive radiology. Thus, plain chest films revealed abnormal shadows in Patients A, D and G, confirmed on tomography; in Patients A and D the masses represented a bronchial carcinoid, and in Patient G, metastatic deposits from a known medullary carcinoma of the thyroid. In Patients B, CandE the clinical finding of hepatomegaly was investigated by liver scan or angiogram, and this confirmed the presence of metastatic deposits. Liver biopsy in Patients C, D and E established the histological diagnoses shown in Table 2. Patient Fhad no clinical or radiological features to suggest the site of the ACTH-producing tumour. An anterior approach for bilateral adrenalectomy was planned. At operation the appendicular carcinoid responsible for his Cushing's syndrome was revealed and removed. Barium studies had previously been normal. Complete and permanent cure was obtained. VENOUS SAMPLING FOR ACTH LEVELS Venous sampling for ACTH gradients at various sites, including high internal jugular veins,

Ectopic and Pituitary-dependent Cushing's Syndrome

1129

was performed on Patients C, D, and E with ectopic ACTH production. In Patient C (pancreatic apudoma and liver metastases), venous sampling at laparotomy revealed a minimal step-up in ACTH concentrations from the aortic blood to the venous blood draining the pancreatic tumour (878 pg/ml compared with 1070 pg/ml) whereas a 1.5 times increase is expected for it to be significant [20]. In Patient D (bronchial carcinoid), no ACTH gradient was found despite sampling from high and low internal jugular vein sites, right and left subclavian veins, SVC above the right atrium and FVC below it. In Patient E (apudoma with liver metastases, primary unknown), total body venous sampling for ACTH revealed the highest concentration in the right hepatic vein, which was consistent with metastases, but unhelpful regarding the site of the primary tumour. CONSIDERATION OF REFERENCE CRITERIA Table 9 summarizes the results of all the criteria proposed from this study to distinguish between Cushing's disease and the ectopic ACTH syndrome. Criteria which are most in favour of an ectopic cause (p< 0.001 at least) are: male sex, history < 18 months; hypokalaemic alkalosis, basal serum cortisol (at both 9 a.m. and midnight) > 800 nmol/1; midnight ACTH > 100 ng/1; unbound cortisol > 1300 nmol/24 hours; suppression on high dose dexamethasone suppression test of 17-oxogenic steroids by < 50 per cent and urinary free cortisol on dexamethasone of > 1000 nmol/24 hours; and a mean of 9 a.m. and midnight serum cortisol on dexamethasone of >450 nmol/1. Failure to respond to both dexamethasone or metyrapone is also a strong indication of ectopic disease. DISCUSSION In this investigation of patients showing obvious Cushing's syndrome, a number of clinical features, basal and dynamic biochemical tests (summarized in Table 9), radiology and histology have been assessed as discriminants between pituitary-driven or ectopic ACTH syndrome. Of the clinical features, the length of history (mean 40 months in Cushing's disease compared with 7.4 months in the ectopic group; p < 0.0001) and the sex of the patient (96 per cent of Cushing's disease patients were female compared with 14 per cent in the ectopic group) appeared to be good discriminants. In confirmation of the results of others [3], the plasma potassium concentration was found to be a reliable pointer, with a value < 3.0 mmol/1 being highly suggestive of an ectopic cause. Glucose intolerance was a poor discriminant. A comparison between the two groups in respect of the basal 9 a.m. or midnight serum cortisol levels revealed an almost complete demarcation; a value at either or both sampling times greater than 800 nmol/1 is highly suggestive (only one error) of an ectopic source of ACTH, and also the difference in mean values for the two groups was highly significant (p< 0.001). However, the usefulness of basal serum cortisol levels has been questioned by Howlett el al. [3], who found considerable overlap between ectopic patients and those with Cushing's disease using mean values of plasma cortisol obtained from six samples over two days. In our study there was a substantial overlap between the two groups in respect of the ACTH level at either 9 a.m. or midnight. In agreement with thefindingsreportedelsewhere [3,21] we suggest that plasma ACTH measurements are of limited discriminating use, unless values greater than 200 ng/1 at 0900 are found, in which case an ectopic source is highly likely [22-24].

1130

S. B. Blunt and others

TABLE 9. Summary of criteria distinguishing between Cushing's disease and ectopic ACTH
Observation' Positive in Cushing's disease % Basal observations 5 Male Length of history < 18 months 5 Plasma K+ <3.0 mmol/1 0 Plasma HCO > 30 MMOL/L 5 21 Glucose intolerance Serum cortisol at: 9 a.m. >800nmol/l 5 m.n. > 800 nmol/1 5 9 a.m. & m.n. each > 800 nmol/1 5 Plasma ACTH at: 9 a.m. >100ng/l 15 UFC >1300nmol/24h 17 On dexamethasone Dexamethasone suppression test shows fall of <. 50% in: 9 a.m. serum cortisol m.n. serum cortisol mean of 9 a.m. & m.n. UFC Urine 17-OGS Mean of 9 a.m. & m.n. serum cortisol on dexa. is ;>450 nmol/1 UFC on dexa. > 1000 nmol/24 h Mean of 9 a.m. & m.n. serum cortisol on dexa. > 450 nmol/1 Number 1/23 1/23 0/23 1/23 5/23 1/23 1/23 1/23 3/20 4/23 Positive in ectopic ACTH % 85 100 100 100 57 Number 6/7 7/7 7/7 7/7 4/7 20.00 25.1 30.00 39.3 3.18 25.1 25.1 25.1 8.38 14.3 <0.001 <0.001 <0.001 < 0.001 < 0.087 < 0.001 < 0.001 < 0.001 <0.01 <0.001 X1 P

100 7/7 100 in 100 7/7 80 4/5 100 6/6

40 24 30 25 19 0 0 0

8/19 4/17 6/17 5/21 4/21 0/17 0/21 0/23 9/21 3/16

70 60 80 80 100

5/7 3/5 4/5 4/5 5/5

1.75 2.3 3.46 5.6 11.6 22.00 26.00 28.00 2.2 9.2

0.18 0.15 0.07 0.017 <0.001 <0.001 <0.001 <0.001 0.13 <0.01

100 5/5 100 5/5 100 5/5 80 4/5 100 4/4

On metyrapone: urine 17-OGS failed to double 45 serum compound S increased less than 3-fold 25 On dexamethasone and metyrapone Failed to respond to dexa. & metyrapone

1/19

100 6/7

15.57

<0.0001

* Drafted in terms of positivity favouring the ectopic ACTH type.

Basal measurements of UFC showed little overlap between the two groups, values greater than 1300 nmol/24 hours being highly suggestive of an ectopic source of ACTH. In contrast, basal measurements of urine 17-oxogenic steroids and 17-oxogenic steroids were unhelpful, in accord with findings elsewhere [15], and could be abandoned in this context. The high-dose dexamethasone suppression test was fairly effective for aetiological diagnosis. Thus, suppression by < 50 per cent in 9 a.m. or midnight serum cortisol, urine free cortisol and 17-oxogenic steroids all pointed to a certain extent to an ectopic source of ACTH. Changes in midnight levels of serum cortisol were not more discriminatory than

Ectopic and Pituitary-dependent Cushing's Syndrome

1131

those at 9.00 a.m. However, combination of the 9.00 a.m. and midnight absolute values produced a mean value which strongly discriminated between the two groups: <450 nmol/1 was invariably found in Cushing's disease, and > 450 n mol/1 was always found in the ectopic syndrome. We can compare our findings from the dexamethasone suppression procedure with those of some other groups. Howlett et al. [3] found that while 91 per cent of their 44 patients with Cushing's disease showed suppression of urine 11-OHCS (measured from 0 to 48 hours and defined as suppression by ^ 50 per cent of the basal value), 33 per cent of nine patients with occult ectopic ACTH also responded in this way; they found that fewer ectopic, but also fewer pituitary cases showed significant suppression of urine 17-oxogenic steroids. Suppression of plasma cortisol by 50 per cent was found in 78 per cent of 46 patients with Cushing's disease compared with 11 per cent of those nine with ectopic ACTH. Grossman et al. [4] report that four of 20 patients with Cushing's disease failed to suppress their serum cortisol by ^ 50 per cent, while none of three ectopic cases did so, which closely accords with our series. Any differences may be due to the fact that we use an extended form of dexamethasone suppression test, using 8 mg per day over 72 hours, and taking samples only on the second and third days, whereas the groups mentioned above performed the test over 48 hours with collection during the first and second days on dexamethasone. Liddle [18] found that it was occasionally necessary to extend the dexamethasone test to 72 hours in order to unveil suppressibility of 50 per cent in patients with Cushing's disease. Nicholls [25] reviewed the literature between 1940 and 1966 concerning steroid tests in the differential diagnosis of Cushing's syndrome and found the dexamethasone procedure to be the most useful in identifying cause, when this was either a pituitary or adrenal tumour. However, as has been the case with many of the series addressing this problem, the number of patients reported with a proven ectopic syndrome are few (for reviews see Crapo [1] and Gold [26]. Subsequently, reports of an unexpected lack of suppressibility by high-dose dexamethasone in patients with Cushing's disease have been largely anecdotal [27, 28] and none have used the extended form of the test. However, Findling and Tyrrell [29] describe 10 patients with the ectopic syndrome (only five of whom had identifiable tumours) and found that four showed 50 per cent suppression of urinary or plasma steroid levels on high-dose dexamethasone. With the metyrapone test, we found that changes in serum 11-deoxycortisol levels yielded a more useful discrimination than changes in 17-oxogenic steroids or 17-oxosteroids, as was also noted by Perry et al. [30], though the difference between the mean percentage responses of 11-deoxycortisol in our two groups was not quite significant (p = 0.06). Alternatively, it is noted that the highest response of 11-deoxycortisol in the ectopic group was less than 3-fold, while 11 of the 16 patients with Cushing's disease had a > 3-fold rise. Howlett et al. [3] found an increment of 100 per cent or more in urine 17-oxogenic steroids over basal levels in 64 per cent of their patients with Cushing's disease compared with 50 per cent of those with the ectopic syndrome; our findings are in agreement with this. They found that serum 11deoxycortisol increments peaked at 24 hours in all cases, and the responses (as measured by absolute values) seen in the three ectopic patients studied were entirely within the range seen in their Cushing's disease group; we also found an overlap in absolute values reached by the two groups, but when the response was assessed by percentage increase in 11-deoxycortisol, there was reasonably good discrimination between the two groups. When the responses to both dexamethasone and to metyrapone were considered together, we obtained a powerful discrimination: among the 19 patients with Cushing's disease, all except one (Patient 16) showed either suppression in response to dexamethasone or stimulation in response to metyrapone, whereas only one (Patient D) of the seven ectopic patients responded in this way.

1132

S. B. Blunt and others

Venous sampling for ACTH gradients was undertaken in three patients in this series, but in no case did this yield information that had not already been obtained by other means. The usefulness of petrosal vein sampling and the corticotrophin-releasing factor test in enhancing the ability to distinguish between Cushing's disease and the ectopic ACTH syndrome was not assessed in this series. Hermus et al. [31] suggest that the diagnostic accuracy of the corticotrophin-releasing factor test is comparable to that of the high-dose dexamethasone test; others have found variable patterns of response which have limited its usefulness in certain situations [32]. In further evaluation of these tests, Grossman et al. [4] showed that either test on its own may give 20-25 per cent misleading results in patients with Cushing's disease, whereas three patients with ectopic ACTH responded appropriately in both tests; in agreement with others [31,33] they found that the combination of the high-dose dexamethasone test and the corticotrophin-releasing factor test is superior to either test alone. We would suggest that the certainty of diagnosis may be further increased by the additional use of the metyrapone test. Although radiology was of particular use in ascertaining the site of an ectopic tumour, CT scanning of the pituitary is still insufficiently sensitive to detect or exclude the smaller ACTHsecreting adenomas reliably. In conclusion we suggest that, provided care is taken in the collection and handling of samples and analyses, the aetiology of ACTH-dependent Cushing's syndrome can usually be established by standard procedures without resorting to invasive procedures. It would now be of interest to ascertain the diagnostic efficiency of the criteria listed in Table 9 from independent series.

ACKNOWLEDGEMENTS It is a pleasure to express our indebtedness to our present and former colleagues whose skill and care enabled this series to be formed: Professor Russell Fraser in Clinical Endocrinology, Prof. R. B. Welbourn and Mr. J. Lynn in Endocrine Surgery, Prof. F. H. Doyle and Prof. D. J. Allison in Diagnostic Radiology, Drs. P. Harsoulis, K. Mashiterand K. Fotherby in the Hormone Laboratories, Miss Rosemary Arnot in the Hospital Medical Physics Department, and the Sisters and Nursing Staffof the Metabolic Unit where all the tests were performed. REFERENCES 1. Crapo L. Cushing's syndrome: a review of diagnostic tests. Metabolism 1979; 28: 955-975. 2. Davies CJ, Joplin GF, Welbourn RB. Surgical management of the ectopic ACTH syndrome. Ann Surg 1982; 196: 246-256. 3. Howlett TA, Drury PL, Perry L, Doniach I, Rees LH, Besser GM. Diagnosis and management of ACTH dependent Cushing's syndrome: a comparison of the features in ectopic and pituitary ACTH production. Clin Endocrinol 1986; 24: 699-713. 4. Grossman AB, Howlett TA, Perry L, et al. CRF in the differential diagnosis of Cushing's syndrome: a comparison with the dexamethasone suppression test. Clin Endocrinol 1988; 19:167178. 5. Sandier LM, Richards NT, Carr DH, Mashiter K, Joplin GF. Long term follow-up of patients with Cushing's disease treated by interstitial irradiation. J Clin Endocrinol Metab 1987; 65: 441447. 6. Blunt SB, Pirmohammed M, Chatterjee VKK, Burrin JM, Allison DJ, Joplin GF. Use of adrenal arterial embolization in severe ACTH-dependent Cushing's syndrome. Postgrad Med J 1989; 65: 575-579. 7. Ghatei MA, Stratton MR, Allen JM, Joplin GF, Polak JM, Bloom SR. Co-secretion of calcitonin

Ectopic and Pituitary-dependent

Cushing's Syndrome

1133

8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.

gene-related peptide, gastrin-rcleasing peptide and ACTH by a carcinoid tumour metastasizing to cerebellum. Postgrad Med J 1987; 63: 123-130. Wright AD, Dixon HG, Joplin GF. Diabetes and latent diabetes in pregnancy. Br Med Bull 1968; 24:25-31. Fitzgerald MG, Keen H. Diagnostic classification of diabetes. Br Med J 1964; 2: 1568. Beardwell CG, Burke CW, Cope CL. Urinary free cortisol measured by competitive protein binding. J Endocrinol 1968; 42: 79-98. Burke CW. Biologically active cortisol in plasma of oestrogen-treated and normal subjects. Br Med J 1969; 2: 798-800. Galvao-Teles A, Graves L, Burke CW, Fotherby K, Fraser R. Free cortisol in obesity; effect of fasting. Acta Endocrinol 1976; 81: 321-329. Seth J, Brown LM. A simple RIA for cortisol. Clin Chim Acta 1978; 86: 109-112. Rees LH, Cook DM, Kendall JW, Allen CF, Kramer RM, Ratcliffe JG. A radioimmunoassay for rat plasma ACTH. Endocrinology 1972; 89: 254-261. Burke CW, Beardwell CG. Cushing's syndrome. An evaluation of the clinical usefulness of urinary free cortisol and other urinary steroid measurements in diagnosis. Q J Med 1973; 42: 175-204. Appleby JA, Gibson G, Norymberski JK, Stubbs RD. Indirect analysis of corticosteroids. Biochem J 1955; 60: 453^67. Gaddum JH. A standard method of estimating 17-oxosteroids and total 17-oxogenic steroids. Lancet 1963; ii: 1415. Liddle GW. Tests of pituitary-adrenal suppressibility in the diagnosis of Cushing's syndrome. J Clin Endocrinol Metab 1960; 20: 1539-1552. Liddle GW. Cushing's syndrome. In: Eisenstein AB, ed. The Adrenal Cortex. Boston: Brown and Little, 1967: pp. 523-551. Corrigan DF, Marcus Schaaf MD, Whaley RA, Czerwinski CL, Earl JM. New Engl J Med 1977; 296: 861-862. Newton RW, D'A Semple P, Brownings MCK, Gunn A. Misleading corticotrophin levels in Cushing's disease. JAMA 1978; 240: 110-11 \. Besser GM, Edwards CRW. Cushing's syndrome. Clin Endocrinol Metab 1972; 1: 451-490. Ratcliffe JG, Knight RA, Besser GM, Landon J, Stansfield AG. Tumour and plasma ACTH concentrations in patients with and without the ectopic ACTH syndrome. Clin Endocrinol 1972; 1: 27-44. Rees LH. ACTH lipoprotein and MSH in health and disease. Clinics Endocrinol Metab 1977; 6: 137-153. Nicholls T, Nugent CA, Tyler FH. Steroid laboratory tests in the diagnosis of Cushing's syndrome. Am J Med 1968; 45: 116-28. Gold EM. The Cushing syndromes: Changing views of diagnosis and treatment. Ann Intern Med 1979; 90: 829-844. Fachnie JD, Zafar MS, Mellinger RC, el al. Pituitary carcinoma mimics the ectopic ACTH syndrome. Clin Endocrinol Metab 1980; 50: 1062-1065. Hale AC, Millar BG, Ratter SJ, et al. A case of pituitary dependent Cushing's disease with the clinical and biochemical features of the ectopic ACTH syndrome. Clin Endocrinol 1985; 22: 479488. FindlingJW, Tyrrell JB. Occult ectopic secretion of corticotrophin. Arch Int Med 1986; 146:929933. Perry LA, Drury PL, Rees LH, Besser GM. Measurement of 11-deoxycortisol during the metyrapone test in the differential diagnosis of Cushing's syndrome. Joint Meeting British Endocrine Societies, 1984; Abstract 184. Hermus AR, Pieters GF, Pesman GJ, Smals AG, Benraad TJ, Klopenberg PW. The CRH test versus the high dose dexamethasone test in the differential diagnosis of Cushing's syndrome. Lancet 1986; 2: 540-544. Boscaro M, Rampazzo A, Sonino N, Merola G, Scanerini M, Mantero F. CRH stimulation test: diagnostic aspects in Cushing's syndrome. J Endocrinol Invest 1987; 10: 297-301. Nieman LK, Chrousos GP, Oldfield EH, Augerinos PC, Cutler PB, Loriaux DL. The ovine corticotrophin-releasing hormone stimulation test and dexamethasone suppression test in the differential diagnosis of Cushing's syndrome. Ann Int Med 1986; 105: 862-866.