| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Clinical Studies |
Department of Laboratory Medicine and Pathology, Mayo Clinic and Foundation, Rochester, Minnesota 55905
Address all correspondence and requests for reprints to: Pai C. Kao, Ph.D., Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In conclusion, urinary free cortisol plus cortisone determined simultaneously by HPLC added a new dimension to the diagnosis of Cushing’s syndrome. It should be considered when exogenous Cushing’s syndrome is suspected or when only one urinary cortisol test is allowed to be ordered.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The efficacy of urinary free cortisol determined by HPLC is unclear
compared to that of the competitive binding assay. In the past 2 yr, 29
patients with Cushing’s syndrome (20 with Cushing’s disease, 5 with
ectopic ACTH syndrome, and 4 with Cushing’s syndrome) plus 6 patients
with exogenous Cushing’s syndrome had both HPLC and competitive
binding urinary cortisol tests ordered by their treating physicians.
Retrospectively, their histories were reviewed, and the results were
compared. Another purpose of this manuscript is to illustrate the
usefulness of HPLC as the sole assay instead of a backup assay for
urinary free cortisol in complying with the current concept that less
testing is better. Specifically, the cost to order a HPLC assay is only
a fraction more expensive (
10%) than the RIA and competitive
binding assays of urinary free cortisol, even though HPLC is
technically more complicated to perform than RIA or competitive binding
assays.
In 1989 we began recording cortisone in addition to cortisol levels in a single urine specimen measured by HPLC. We suspected that urinary cortisone may assist in the diagnosis of Cushing’s syndrome. Cortisone, a downstream metabolite of cortisol, is converted from cortisol by 11ß-hydroxysteroid dehydrogenase, which deactivates the cortisol activity. Investigators suspected that the severity of the ectopic ACTH syndrome occurred from either overload with a high concentration of cortisol (4) or inhibition (5) of this enzymatic deactivation mechanism. In this study, we report a saturation pattern of the cortisol and cortisone relationship and the use of the cortisone/cortisol ratio with plasma potassium in an attempt to separate the ectopic ACTH syndrome from Cushing’s disease. We found that there is a separation; however, we did not have enough patients with the ectopic ACTH syndrome to prove that this observation will always be true. Currently, the highly invasive procedure of petrosal sinus sampling for measuring ACTH after ovine CRH stimulation still is the preferred test in the differential diagnosis of ectopic ACTH syndrome and Cushing’s disease (6, 7, 8).
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
During a 2-yr period, 35 patients had urinary free cortisol tests performed by both HPLC and competitive binding methods. Retrospectively, their histories were reviewed for this study. Twenty-nine patients had surgically proven Cushing’s syndrome; 20 had pituitary ACTH-dependent Cushing’s disease, 5 had ectopic ACTH-dependent Cushing’s syndrome, and 4 had glucocorticoid-producing adrenal tumors, including 1 patient with adrenal cortical carcinoma. None of these patients had evidence of renal function impairment. Five of the patients with Cushing’s disease had no histopathological evidence of pituitary adenoma; however, because they all showed improvement in their disease after pituitary surgery or surgery plus irradiation, the final diagnosis of Cushing’s disease remained. In the 5 patients with ectopic ACTH syndrome, 3 had carcinoid tumor of the lung, 1 had small cell lung cancer, and 1 had metastatic prostate cancer. In addition, 6 patients with exogenous Cushing’s syndrome were included in this retrospective study. All test results summarized in this manuscript were ordered by the patients’ treating physicians. Healthy individuals were volunteers from the institutional normal pool, whose ages ranged from 24–67 yr. In the first group of 41 healthy individuals, there were 21 males and 20 females; in the second group of 60, there were 26 males and 34 females. They were all Caucasian, apparently healthy, and free of disorders, and all had given the proper consent.
Urinary free cortisol and cortisone by HPLC
A 5-mL aliquot of 24-h urine was washed with 5% ethyl acetate
in cyclohexane and then extracted with 15 mL methylene chloride. The
chromatography was performed with a Beckman Ultrasphere IP 5-µm,
250 x 4.6-mm, C18 reverse phase column (Beckman, Palo
Alto, CA) tandemed with a Hewlett-Packard Hypersil 3-µm, 60 x
4.6-mm, C18 column (Hewlett-Packard, Palo Alto, CA). An
isocratic mobile phase consisting of 30% acetonitrile and 70% 0.05
mol/L potassium phosphate buffer (vol/vol) at pH 6.2 was used at a
controlled temperature of 40 C. The absorbance of the column eluent was
monitored at 248 nm, and the sensitivity was set at 0.02 absorbance
units full scale. 6
-Methylprednisolone was used as the internal
standard. Urinary creatinine levels were determined by a Hitachi
multichannel analyzer (Hitachi, Hialeah, FL) to ensure completeness of
urine collection. An entire range (up to the 100th percentile) of
urinary free cortisol from 5–55 µg/24 h determined 10 yr ago (3) was
routinely used as the reference range. The monthly average interassay
coefficients of variation (CVs) were 5.7% and 5.6% for cortisol and
cortisone, respectively.
Sixty healthy volunteers from the institutional normal pool had given proper consent. Their 24-h urine specimens were collected and analyzed for reference ranges of cortisol and cortisone. Urinary free cortisol levels ranged from 8–51 µg/24 h (mean ± SD, 23 ± 8), whereas cortisone levels ranged from 16–128 µg/24 h (mean ± SD, 73 ± 22). To increase the accuracy of the reference range by including a larger group of normal subjects, the cortisol levels of these 60 healthy individuals were combined with those of the previous 41 normal subjects studied 10 yr previously by the same method. The cortisol results of the 41 normal subjects were 25 ± 8 µg/24 h and ranged from 9–55 µg/24 h (3). Cortisone levels, however, were not recorded because the value of cortisone measurment for diagnosis was unknown during that time. The combined mean ± SD urinary free cortisol of the 101 normal subjects was 24 ± 9 µg/24 h and ranged from 8–55 µg/24 h. By including the previous 41 normal subjects, the upper reference range of urinary free cortisol increased from 51 to 55 µg/24 h. The 100th percentile reference ranges of urinary free cortisol (5–55 µg/24 h) and urinary free cortisone (16–128 µg/24 h) were used to analyze the test sensitivity in this study. The correlation of cortisone and cortisol in the 60 normal subjects was r = 0.51 (y = 1.4 x + 39, where y is cortisone; P < 0.00001). The ratio of cortisone/cortisol in the 60 normal subjects ranged from 1.65–9.06 (mean ± SD, 3.35 ± 1.28).
Urinary free cortisol determined by competitive protein binding assay
Urinary free cortisol was measured by competitive protein binding assay after solvent extraction (1, 2). In brief, an organic solvent extract of urine was dried and then mixed with a binding solution that contained [3H]corticosterone. After incubation, free and bound fractions were separated by florisil. Measurement of the radioactivity remaining in the supernatant represented the bound fraction, which allows calculation of the urinary free cortisol level with a standard curve. The full reference range was 24–108 µg/24 h. Inter- and intraassay CVs were 6.1% and 6.0%, respectively.
Plasma cortisol
Both morning and evening plasma cortisol levels were measured by a double antibody RIA kit without prior extraction (Diagnostic Products Corp., Los Angeles, CA). The reference range of morning plasma cortisol was 7–25 µg/dL, and the evening reference range was 2–14 µg/dL. Specimens were drawn between 0800–1000 and 1600–1800 h, respectively. Inter- and intraassay CVs were 6.4% and 4.3%, respectively.
Plasma corticotropin (ACTH) was measured by a RIA that makes use of an antibody directed to a synthetic amino-terminal segment of corticotropin-(1–24). An extraction procedure was performed before the assay to remove inactive precursors of corticotropin and interference. The reference range of morning (between 0800–1000 h) plasma ACTH was below 60 pg/mL (8, 9). Inter- and intraassay CVs were 12% and 8%, respectively.
Serum potassium levels were determined by a Hitachi multichannel analyzer. The reference range was 3.6–4.8 mEq/dL.
Statistical methods
All data are presented as the mean ± SD by descriptive statistics. Results between patient groups were compared with two-tailed nonpaired Student’s t tests with unequal variance. Linear regression was used to evaluate the correlation between variables. The relationship between urinary free cortisone and cortisol was examined by logarithmic curve fit in both controls and patients with Cushing’s syndrome.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The urinary free cortisol level of the 60 healthy subjects ranged
from 8–51 µg/24 h (mean ± SD, 23 ± 8).
Another 41 healthy subjects whose urinary free cortisol was determined
10 yr previously by the same procedure had a range from 9 to 55 µg/24
h (mean ± SD, 25 ± 8). Statistically by
nonpaired Student’s t test, the 2 groups (n = 41 and
the current n = 60) showed no significant difference
(P = 0.22). These 2 groups were combined to set a more
conservative and accurate reference range. The combined mean ±
SD of cortisol is 24 ± 9, which is similar to
previously reported values (10, 11, 12). By combining these two groups, the
100th percentile upper reference range increased from 51 to 55 µg/24
h. The full reference range upper limit of 55 µg/24 h was used for
the analysis of test efficacy. The HPLC urinary free cortisol results
of the 29 patients with a variety of Cushing’s syndromes are presented
in Fig. 1
(middle panel).
|
Cortisone simultaneously determined by HPLC
We have long suspected that cortisone, the deactivated metabolite
of cortisol by 11ß-hydroxysteroid dehydrogenase, has clinical
diagnostic value. Since 1989, we have been recording the value of
measuring urinary free cortisone, which is simultaneously determined
with cortisol by HPLC in a single assay without additional cost. Here
we report its value alone and with cortisol in the diagnosis of
Cushing’s syndrome. Cortisone in 60 healthy subjects ranged from
16–128 µg/24 h (mean ± SD, 73 ± 22). Among
the 29 patients with Cushing’s syndrome, 4 had urinary free cortisone
levels less than the upper reference range of 128 µg/24 h. The 4 were
all Cushing’s disease patients (n = 20); none of the patients
with ectopic ACTH syndrome (n = 5) or adrenal Cushing’s syndrome
(n = 4) had urinary free cortisone below the upper reference
range. A total of 25 patients had elevated cortisone levels. The test
efficacy (or sensitivity) by cortisone alone was 86% (25 of 29). The
urinary free cortisone levels of normal subjects and those with
Cushing’s disease are shown in Fig. 1
(lower
panel). When using both parameters (urinary free cortisol and
cortisone), only 2 patients were within the reference range. In other
words, 27 of the 29 (93%) patients with Cushing’s syndrome had either
both urinary free cortisol and urinary free cortisone elevated (n
= 19) or one of them (cortisol or cortisone) elevated (n = 8).
This gives an undisputed advantage to the HPLC method over competitive
binding assays (RIA and competitive protein binding).
An additional advantage of HPLC is in the detection of patients with exogenous Cushing’s syndrome; in the 6 patients with exogenous Cushing’s taking prednisone, all of them had undetectable urinary free cortisol and cortisone plus highly elevated urinary prednisone (mean ± SD, 788 ± 652 µg/24 h) and prednisolone (mean ± SD, 1337 ± 1457 µg/24 h). On the other hand, by the competitive binding method, 5 had increased urinary free cortisol results (689 ± 689 µg/24 h), ranging from 176-1849 µg/24 h. One had a urinary free cortisol measurement of 100 µg/24 h; this is close to the upper limit of reference range (108 µg/24 h).
Ratio of cortisone/cortisol
The ratios of urinary free cortisone/cortisol of 60 normal
subjects and 29 patients with Cushing’s syndrome are presented in Fig. 1 (upper panel). Even though the differences between healthy
controls and patients with Cushing’s disease or those with ectopic
ACTH syndrome were both statistically significant, there were a number
of overlaps between healthy individuals and patients with Cushing’s
disease. The 5 patients with ectopic ACTH syndrome had suppressed
ratios. There was no overlap with healthy controls, but there was
overlap with Cushing’s disease patients, of whom a few also had
suppressed ratios. Patients with Cushing’s syndrome and adrenal tumor
had ratios in the same range as the healthy controls (Fig. 1, upper panel). Therefore, the ratio used alone has little
value in the diagnosis of Cushing’s syndrome. However, by using the
ratio combined with plasma potassium measurements, we observed a
separation of ACTH syndrome patients (n = 5) from Cushing’s
disease patients (n = 20; Fig. 2). Currently, we only had 5
patients with ectopic ACTH syndrome. This observation should not be
used for the differential diagnosis of Cushing’s. Further study is
required.
|
], which may indicate that
the enzyme converting cortisol to cortisone is saturated.
|
The patients with elevated plasma cortisol and ACTH with
ACTH-producing tumors (Cushing’s disease and ectopic ACTH syndrome)
and the patients with elevated plasma cortisol plus suppressed ACTH
with cortisol-producing adrenal tumors (Cushing’s syndrome) are
summarized in Table 1.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
To determine test efficacy when a single parameter, urinary free cortisol, is used, there is little difference between HPLC and competitive protein binding assays. The HPLC method is certainly more complicated to perform than the simple method of competitive binding. However, when using two parameters, urinary free cortisol plus urinary free cortisone, the usefulness of the HPLC method is immediately increased. Only 2 of the 29 patients with Cushing’s syndrome had both cortisol and cortisone levels within the reference range. Twenty-seven of them (93%) had at least 1 elevated parameter; either both urinary free cortisol and urinary free cortisone were elevated (n = 19), or 1 of them was elevated (n = 8). This indicates that the efficacy of the HPLC assay is better than that of single parameter competitive binding assays, which can only report cortisol alone.
We could use the 95th percentile reference range for cortisol and cortisone. In that case, cortisol or cortisone alone will reach a test sensitivity of 90%. However, Cushing’s syndrome is rare. A test needs to be highly specific to screen out most normal individuals without Cushing’s. For that reason we set the reference range at the 100th percentile. The combined test efficacy of cortisol plus cortisone was still 93%. It did not reduce the efficacy of the HPLC test. An additional advantage in setting the specificity of a test at the 100% (or the 100th percentile) level as the reference range is that the number of false positives for healthy individuals approached zero. When the false positive value is zero, the test sensitivity (or efficacy) is equal to the positive predictive value. This will eliminate an additional cumbersome number of positive predictive values that a clinician has to remember when using a test.
The HPLC method has a unique advantage in identifying patients with
exogenous Cushing’s. In this study, six patients with exogenous
Cushing’s taking prednisone had undetectable urinary free cortisol and
cortisone by the HPLC method and the presence of high levels of
prednisone and prednisolone, the metabolite of prednisone. In
comparison, five of the six patients with exogenous Cushing’s syndrome
had elevated urinary free cortisol by competitive protein binding
assay; one had urinary free cortisol close to the upper limit of
reference range at 100 µg/24 h. We do not believe that the HPLC
method can determine exogenous Cushing’s syndrome caused by
hydrocortisone administration. In cases of suspicious hydrocortisone
administration, assays of upstream metabolites of cortisol should be
performed, such as 11-deoxycortisol, 11-deoxycorticosteroids,
17-hydroxyprogesterone, and free dehydroepiandrosterone. These
compounds should be suppressed because the pituitary secretion of ACTH
is suppressed by hydrocortisone.
One question to ask is when should HPLC detemination of cortisol be ordered. In our opinion, any time a suspicion is raised by RIA or competitive binding assay of urinary free cortisol, the HPLC method should be considered. Another condition would be if only one urinary corticosteroid test is allowed to be ordered, then the choice should be HPLC instead of competitive binding assay. Economically, the two-parameter assay of urinary free cortisol and cortisone by HPLC is only about 10% more expensive than the single parameter competitive binding method of cortisol, even though technically HPLC is more complicated for a laboratory to perform.
We do not suggest using the ratio of cortisone/cortisol and plasma potassium as the final diagnostic test to differentiate ectopic ACTH syndrome and Cushing’s disease, because we only had five ectopic ACTH syndrome patients in this study. The preferred test for the differential diagnosis of ectopic ACTH syndrome and pituitary disease is still petrosal sinus sampling for the measurement of ACTH after ovine CRH stimulation (6, 7, 8).
An additional contribution of simultaneously measuring cortisol and cortisone was that the high concentration of cortisol reached (1000 µg/24 h) inhibited the further conversion of cortisone, which reached a plateau. This phenomenon, however, was not restricted to only ectopic ACTH syndrome, but was present in Cushing’s disease patients as well as adrenal Cushing’s syndrome patients who had high concentrations of cortisol. The data suggest an overload of enzyme 11ß-hydroxysteroid dehydrogenase, which converts cortisol to cortisone (4), instead of specific inhibition of the enzyme by ectopic ACTH-producing tumors.
Conclusion
In conclusion, measurement of urinary free cortisol and cortisone by HPLC has added new value in the diagnosis of Cushing’s syndrome. It is more valuable than measuring cortisol alone by either HPLC or competitive binding assays. Patients with exogenous Cushing’s syndrome caused by steroid drugs other than hydrocortisone will have suppressed levels of cortisol and cortisone. If only one urinary corticosteroid test is allowed to be ordered for economic reasons, then urinary free cortisol plus cortisone determination by HPLC is the test of choice.
|
Footnotes |
|---|
2 Visiting clinician, Department of Laboratory Medicine and
Pathology.
3 Current address: Cathay General Hospital, Taipei, Taiwan.
4 Current address: National Cheng Kung University Hospital, Tainan,
Taiwan.
Received February 14, 1996.
Revised August 21, 1996.
Accepted September 21, 1996.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  L. Wood, D. H Ducroq, H. L Fraser, S. Gillingwater, C. Evans, A. J Pickett, D. W Rees, R. John, and A. Turkes Measurement of urinary free cortisol by tandem mass spectrometry and comparison with results obtained by gas chromatography-mass spectrometry and two commercial immunoassays Ann Clin Biochem, July 1, 2008; 45(4): 380 - 388. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. K. Nieman, B. M. K. Biller, J. W. Findling, J. Newell-Price, M. O. Savage, P. M. Stewart, and V. M. Montori The Diagnosis of Cushing's Syndrome: An Endocrine Society Clinical Practice Guideline J. Clin. Endocrinol. Metab., May 1, 2008; 93(5): 1526 - 1540. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Remer, T. Dimitriou, and C. Maser-Gluth Renal Net Acid Excretion and Plasma Leptin Are Associated with Potentially Bioactive Free Glucocorticoids in Healthy Lean Women J. Nutr., February 1, 2008; 138(2): 426S - 430S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Remer and C. Maser-Gluth Simultaneous Measurements of Urinary Free Cortisol and Cortisone for the Assessment of Functional Glucocorticoid Activity Clin. Chem., October 1, 2007; 53(10): 1870 - 1871. [Full Text] [PDF]
|
|
|
|
|
|
J. W. Findling and H. Raff Cushing's Syndrome: Important Issues in Diagnosis and Management J. Clin. Endocrinol. Metab., October 1, 2006; 91(10): 3746 - 3753. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Fenske How Much "Urinary Free Cortisol" Is Really Cortisol during Water Diuresis in Healthy Individuals? Clin. Chem., June 1, 2004; 50(6): 1102 - 1104. [Full Text] [PDF]
|
|
|
|
|
|
M. N. Kerstens, F. G. H. van der Kleij, A. H. Boonstra, W. J. Sluiter, J. Koerts, G. Navis, and R. P. F. Dullaart Salt Loading Affects Cortisol Metabolism in Normotensive Subjects: Relationships with Salt Sensitivity J. Clin. Endocrinol. Metab., September 1, 2003; 88(9): 4180 - 4185. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Raff and J. W. Findling A Physiologic Approach to Diagnosis of the Cushing Syndrome Ann Intern Med, June 17, 2003; 138(12): 980 - 991. [Full Text] [PDF]
|
|
|
|
|
|
R. L. Taylor, D. Machacek, and R. J. Singh Validation of a High-Throughput Liquid Chromatography-Tandem Mass Spectrometry Method for Urinary Cortisol and Cortisone Clin. Chem., September 1, 2002; 48(9): 1511 - 1519. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Remer, A. Neubert, and C. Maser-Gluth Anthropometry-based reference values for 24-h urinary creatinine excretion during growth and their use in endocrine and nutritional research Am. J. Clinical Nutrition, March 1, 2002; 75(3): 561 - 569. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. C. Kao, D. A. Machacek, M. J. Magera, J. M. Lacey, and P. Rinaldo Diagnosis of Adrenal Cortical Dysfunction by Liquid Chromatography-Tandem Mass Spectrometry Ann. Clin. Lab. Sci., April 1, 2001; 31(2): 199 - 204. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. E. P. Murphy How Much ""UFC"" Is Really Cortisol? Clin. Chem., June 1, 2000; 46(6): 793 - 794. [Full Text] [PDF]
|
|
|
|
|
|
B. E. P. Murphy Lack of Specificity of Urinary Free Cortisol Determinations: Why Does It Continue? J. Clin. Endocrinol. Metab., June 1, 1999; 84(6): 2258a - 2259. [Full Text]
|
|
|
|
 |
|
 J. Newell-Price, P. Trainer, M. Besser, and A. Grossman The Diagnosis and Differential Diagnosis of Cushing's Syndrome and Pseudo-Cushing's States Endocr. Rev., October 1, 1998; 19(5): 647 - 672. [Abstract] [Full Text]
|
|
|
|
|
|
M. V. Mericq and G. B. Cutler Jr. High Fluid Intake Increases Urine Free Cortisol Excretion in Normal Subjects J. Clin. Endocrinol. Metab., February 1, 1998; 83(2): 682 - 684. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
adrenal; • pituitary) on the basis of a serum potassium level less
than 3.5 mEq/dL and a 24-h urinary free cortisone/cortisol ratio less
than 1.5 for cut-off points. These were the same patients as those
reported in Fig. 1