| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Article |
Quest Diagnostics’ Nichols Institute (M.P.C., T.L., R.E.R., D.A.F.), San Juan Capistrano, California 92690; Department of Pediatrics, University of California, San Diego (M.E.G., K.L.J.), La Jolla, California 92123; Department of Clinical Biochemistry, King’s College and School of Medicine (N.F.T.), London SE5 8RX, United Kingdom; Department of Laboratory Diagnostics, Children’s Memorial Health Institute (E.M.M.), Warsaw, Poland; and Children’s Hospital Oakland Research Institute (C.H.L.S.), Oakland, California 94609
Address all correspondence and requests for reprints to: Cedric H. L. Shackleton, Ph.D., Children’s Hospital Oakland Research Institute, 5700 Martin Luther King, Jr. Way, Oakland, California 94609. E-mail: . cshackleton{at}chori.org
Abstract
Definitive neonatal diagnosis of congenital adrenal hyperplasia (CAH) is frequently complicated by normal 17-hydroxyprogesterone levels in 21-hydroxylase-deficient patients, residual maternal steroids, and other interfering substances in neonatal blood. In an effort to improve the diagnosis, we developed a gas chromatography/mass spectrometry method for simultaneous measurement of 15 urinary steroid metabolites as early as the first day of life. Furthermore, we developed 11 precursor/product ratios that diagnose and clearly differentiate the four enzymatic deficiencies that cause CAH. Random urine samples from 31 neonatal 21-hydroxylase-deficient patients and 59 age-matched normal newborns were used in the development. Additionally, samples from two 11ß-hydroxylase-deficient patients and one patient each for 17 
-hydroxylase and 3ß-hydroxysteroid dehydrogenase deficiencies were used. The throughput for one bench-top gas chromatography/mass spectrometry instrument is 20 samples per day. Thus, this method affords an accurate, rapid, noninvasive means for the differential diagnosis of CAH in the newborn period without the need for invasive testing and ACTH stimulation.
CONGENITAL ADRENAL HYPERPLASIA (CAH) is an autosomal recessive disorder involving deficiency of one of the four enzymes, or a cholesterol-trafficking protein, necessary for adrenal cortisol biosynthesis (1, 2, 3, 4, 5, 6). Diagnosis and treatment of affected infants in the neonatal period is necessary to minimize the morbidity and mortality associated with adrenal insufficiency. The diagnosis is usually accomplished by measuring blood levels of adrenal hormones and precursor steroids (1). However, definitive neonatal diagnosis is complicated by nonelevated 17-hydroxyprogesterone (17OHP) levels in some infants with 21-hydroxylase deficiency and by the presence of residual maternal-placental and fetal steroid products, leading to misdiagnoses (7, 8, 9, 10). Even when very specific RIAs are used, interfering substances in neonatal plasma increase the apparent 17OHP in the samples and reduce the discrimination between normal individuals and those with CAH. Reliable neonatal diagnosis requires extraction and chromatography before serum steroid immunoassay to assure accurate differential diagnosis. In an earlier study, one of the authors (C.H.L.S.) identified 17 -hydroxypregnenolone and its sulfate as major contributing factors to the 17-OHP overestimation (11).
An alternative diagnostic method uses gas chromatography/mass spectrometry (GC/MS), which has been used for the diagnosis of 21-hydroxylase deficiency for 25 yr (12, 13). However, it is only recently that instrumentation has become reliable and inexpensive enough for routine use. There are multiple reports of GC/MS being used for diagnosis of all forms of CAH in the newborn period (12, 13, 14, 15, 16, 17), and straightforward and comprehensive methods that allow diagnosis of each form have been reported (18, 19). There are several advantages of urine analysis for CAH diagnosis. The sample collection is noninvasive, a random sample is suitable, and a full spectrum of steroid hormone metabolites and precursor/product ratio assessments can be measured simultaneously, increasing accuracy and confidence in the diagnosis. Universal methodology can be developed for distinguishing the separate enzymatic forms of the disorder.
Materials and Methods
Normal subjects
Using premature infant diapers, random urine samples were collected from 59 infants in a newborn nursery and outpatient clinics at Children’s Hospital, San Diego, and the University of California, San Diego. The top liner of the diaper was removed and discarded. After removing the plunger of a 20-ml syringe, the wetted diaper padding was placed into the syringe barrel. The plunger was replaced, and 3–5 ml urine was extracted by compressing the diaper padding. Studies were approved by the University of California San Diego Human Subjects Institutional Review Board.
Patients with CAH
Urine samples were obtained from 36 neonates with CAH before starting glucocorticoid therapy.
21-Hydroxylase deficiency. Samples from 31 patients with 21-hydroxylase deficiency were stored at -20 C. Of these, 28 were from patients aged 1–11 d, and three were from older patients (14, 18, and 23 d, respectively). Creatinine measurements were performed on 21 of these samples.
11 ß-Hydroxylase deficiency. Samples from two patients aged 14 d and 7 wk were obtained.
17-Hydroxylase deficiency. One urine sample from a 7-d-old patient was obtained.
3ß-Hydroxysteroid dehydrogenase (3ß-HSD) deficiency. Samples were obtained at 7 and 15 d of age from one individual in whom 3ß-HSD deficiency was later confirmed by mutation analysis.
Methodology
The methodology described is an adaptation of that published previously (18, 20). Conjugated and unconjugated steroids in a random urine sample were absorbed onto a Varian Bond Elut C18/OH solid phase extraction column previously primed with methanol and water. The volume of urine extracted was dependent on the urine creatinine concentration. A volume of urine containing 75 µg creatinine was determined to be acceptable for analysis by GC/MS. Thus, the formula for determining the volume of urine (in microliters) to be extracted was 75,000/creatinine value in mg/liter. Following water washing, the steroids were eluted with methanol (4 ml) and dried under nitrogen. The steroid extract was subjected to enzyme hydrolysis using 36 µl per sample ß-glucuronidase/arylsulfatase (Roche Diagnostics, Indianapolis, IN) and 10 mg per sample sulfatase (type H-1, from Helix pomatia, Sigma, St. Louis, MO) in 3 ml acetate buffer (pH 4.7) to release the steroids from conjugation. The samples were then incubated for 3 h at 55 C. The unconjugated steroids were again extracted on a C18/OH column. After the addition of internal standards (cholesteryl butyrate and stigmasterol) to the dried extract, steroids were initially derivatized using 2% methoxamine HCl in pyridine (1 h at 55 C), which protects carbonyl groups (except those at C-11). The samples were dried and subjected to a secondary derivatization with trimethylsilylimidazole (4 h at 120 C), which resulted in a virtually complete silylation of all hydroxyl groups. The derivatizing reagents were removed by chromatography on hydroxyalkoxypropyl dextran type IX (formerly Lipidex, Sigma) columns using cyclohexane as the mobile phase (18). The cyclohexane eluates were dried under nitrogen and reconstituted in 200 µl cyclohexane in preparation for GC/MS analysis.
GC/MS analysis was carried out on an HP 5973 instrument (Hewlett-Packard, Wilmington, DE) housing a 15 m DB1 column (J and W Scientific, Folsom, CA). The samples were injected using an HP 7683 series autosampler and were eluted using cold trapping, with the column temperature ramping from 50 C to 285 C over 41 min. The flow of helium through the column was 1 ml/min. The eluting steroids were detected by selected ion monitoring (SIM). The specific ions used for each steroid are listed in Table 1
.
|
-hydroxy-tetrahydrocompound A (6-hydroxy-THA) or 6-hydroxytetrahydro substance S (6-hydroxy-THS), which are important neonatal markers for 17-hydroxylase and 11ß-hydroxylase deficiency syndromes, respectively (15, 17). Calibration for these was achieved using urine extract from affected patients as follows: The steroid extracts with internal standards were analyzed by GC/MS in scanning mode, and the total ion current (TIC) peak areas of components identified as 6-hydroxy-THA and 6-hydroxy-THS were integrated. These steroids elute between the first and second internal standards, and we related the peak areas to the mean TIC peak areas of the two internal standards. We assumed for the purpose of this study that peak area responses in TIC of different compounds were identical for components of similar retention time (an approximate but practical assumption). Based on this assumption, we determined the approximate concentration of 6-hydroxy-THA and 6-hydroxy-THS in the patient samples so they could be used for instrument calibration. The samples were analyzed again in SIM mode to obtain calibration response curves to be used in quantitating these compounds in the patient samples. After response curves had been obtained for each steroid analyte to be measured, quantitation of the steroids was performed using the HP ChemStation software. Peak areas for the steroids were obtained and compared with calibration response curves for quantitation, using the internal standard stigmasterol to correct for procedural losses. The raw results obtained were multiplied by the ChemStation software by 13,333 to obtain the desired microgram per gram creatinine units.
The quantitated steroid data were exported to an Excel spreadsheet (Microsoft, Seattle, WA) in which the diagnostic ratios were calculated (described in Results and Discussion).
Statistics
The Mann-Whitney U test was used to statistically compare the steroid ratio data for 21-hydroxylase deficiency with data from the normal and diseased groups at the different ages. These statistics should be interpreted with caution because of the relatively small study population. Additional studies are required to extend this to the general population. Statistical analysis was not performed on the steroid ratio data for the 3ß-HSD, 11ß-hydroxylase, and 17-hydroxylase deficiencies because of the low number of disease state samples available.
Results
Quantitative analysis
Normals.
Thirty-four adrenal steroid hormone metabolites were measured. Reference ranges were developed for the 3ß-hydroxy-5-ene steroids as well as for the DHEA and pregnenolone metabolites produced in the adrenal fetal zone. Table 1
lists the mean and observed range for the 15 metabolites. Figure 1 shows the adrenal steroid biosynthetic pathway and the steroid precursor for each of the 15 referenced metabolites.
|
lists the quantitative excretions for 28 patients with CAH. Although the mean excretion of all 3ß-hydroxy-5-ene steroids is higher than normal in affected infants, there is considerable overlap in the ranges. Specific analytes for 21-hydroxylase, 17-hydroxylase, and 3ßHSD defects are also notably elevated. Individual steroid values for 11ß-hydroxylase deficiency were not listed because creatinine values were unavailable. The excretion of cortisol metabolites is not markedly different between normals and CAH patients, probably reflecting remaining maternal contribution.
|
Table 3 reports the range of precursor and product metabolite ratios for the four types of CAH. The rationale for choice of the numerator in the ratio is addressed in the discussion. The product (cortisol) metabolites used as denominator are the summation of tetrahydrocortisone, -cortolone, and ß-cortolone, referred to as cortisol metabolites (CM). Two ratios are not based on the CM denominator: pregnanetriol/pregnanetriolone (5-pregnene-3ß, 17, 20-triol/5ß-pregnane-3, 17, 20-triol-11-one) for discriminating the 21-hydroxylase defect from 3ßHSD deficiency and 16-hydroxypregnenolone/16-hydroxy DHEA for diagnosis of 17-hydroxylase deficiency.
|
).
|
|
|
|
We describe the development of a robust method independent of 24-h excretion or values relative to creatinine and based on the measurement of steroid precursor/product ratios. These ratios represent the ratio of the urinary metabolite of an adrenal precursor (e.g. 17-OHP) to the urinary metabolites of the product (cortisol). When testing both children and adults, we used the summation of three important CM for the denominator, namely: tetrahydrocortisone, tetrahydrocortisol, and 5-tetrahydrocortisol. The latter two steroids are virtually immeasurable in neonates because of the dominance of the 11-keto-containing steroids in the perinatal period. In neonates, we would have preferred to use the three dominant neonatal metabolites (tetrahydrocortisone, 6-hydroxytetrahydrocortisone, and 6-hydroxycortolone) (21, 22); however, authentic steroids are not available for the two latter compounds, rendering instrument calibration difficult. Therefore, we decided to use the sum of tetrahydrocortisone, -cortolone, and ß-cortolone (abbreviated as CM) as the denominator. Typically, this group represents 20–50% of excreted cortisol metabolites (21, 22). Any use of a CM denominator is potentially compromised in the first few days of life because some cortisol precursor presumably originated in the mother. However, the maternal contribution to neonatal cortisol metabolites is probably negligible after 4 d of life. This, of course, also applies to the numerator analyte because maternal derived pregnanetriol, for example, will contribute to neonatal excretion.
21-Hydroxylase deficiency
Three diagnostic ratios use well-known analytes, i.e. 17-hydroxypregnenolone x 100/CM (ratio 1), pregnanetriol/CM (ratio 2), and pregnanetriolone x 100/CM (ratio 3) (18). We also measured the 15ß,17-dihydroxypregnanolone x 100/CM ratio (ratio 4) based on the numerator being a specific neonatal diagnostic analyte for CAH (16). It is clear from Fig. 2 that in all cases the patients with 21-hydroxylase deficiency exhibited no overlap with the normal neonates. We had a sufficiently large sample size to break down the result into three age groups (i.e. first day of life, d 2–4, and 5 d and above. It is evident that, although discrimination between normal and affected individuals is readily achievable on the first day of life, the separation between them increases with age of the infants owing to both the ratio increasing for affecteds and decreasing for normals. The widening gap between normal and affected ranges can be attributed to reduction in the maternal contribution to steroid excretion and increasing neonatal adrenal stimulation in response to falling cortisol availability.
3ßHSD deficiency
The two samples from a single patient with diagnosed 3ßHSD II deficiency showed a profile distinctive for the disorder. The excretion of 3ß-hydroxy-5-ene steroids (relative to creatinine) were extremely elevated (Table 2). Except for their high excretion, the quantitatively major neonatal 3ß-hydroxy-5-ene steroids do not provide diagnostic specificity for the condition. The classical analyte of the condition, pregnenetriol (5-pregnene-3ß,17,20-triol), first identified by Bongiovanni (23, 24) in urine from older patients with the disorder, is a relatively minor component of neonatal 3ßHSD urine but nonetheless remains the discriminating analyte. Pregnenetriol is the major urinary metabolite of 17-hydroxypregnenolone. The pregnenetriol x 100/CM ratio (ratio 5) was 356 on d 7 of life and 149 on d 14 of life, and the highest recorded ratio for 2- to 4-d-old normal infants was 17 (Fig. 3).
Distinguishing 3ßHSD deficiency from 21-hydroxylase has always been challenging. Bongiovanni (23) noted early that the 21-hydroxylase deficiency analytes pregnanetriol and 17-hydroxypregnanolone were also elevated in 3ßHSD deficiency, and this has been verified by other authors (14). In addition, serum analysis of 17-hydroxyprogesterone results in elevated values that may lead to false diagnosis (25). In our 3ßHSD-deficient patient, ratios 1, 2, and 4 (Fig. 2) all were elevated, possibly indicating the presence of 21-hydroxylase deficiency. In addition, two 21-hydroxylase-deficient patients had ratio 5 values within the range of the 3ßHSD patient (Fig. 3). One feature of 3ßHSD noted by Bongiovanni (23, 24) is the virtual absence of pregnanetriolone, and this proves to be a necessary discriminant. Ratio 3, alone among the 21-hydroxylase deficiency ratios, is normal. Thus, the two primary diagnostic ratios, pregnanetriol x 100/CM (ratio 5), which should be elevated, and pregnanetriolone x 100/CM (ratio 3), which should be normal, could be combined to give the secondary ratio pregnenetriol/pregnanetriolone (ratio 6). Plots of this ratio against normals and 21-hydroxylase patients are shown in Fig. 3. This ratio clearly distinguishes between the 21-hydroxylase and 3ß-HSD deficiencies.
In summary, for 3ßHSD the following criteria must be met: elevated pregnenetriol x 100/CM, normal pregnanetriolone/CM, and elevated pregnenetriol/pregnanetriolone.
11ß-Hydroxylase deficiency
Reichstein’s substance S (11-deoxycortisol) is the primary steroid overproduced in 11ß-hydroxylase deficiency, and its tetrahydro derivative has for many years been recognized as the urinary marker of the disorder when excreted in excess. In children and adults, the absence of 11-oxygenated androgens and highly elevated androsterone and etiocholanolone levels are also notable features that are less evident in neonates. In newborns, the active 6-hydroxylase results in 6-hydroxy-THS being the second most important urinary analyte (15). Both analytes are normally excreted in extremely low amounts and could not be quantified in all normal neonates. Plots of the ratios THS/CM (ratio 7) and 6-hydroxy-THS/CM (ratio 8) are shown in Fig. 4. The two patients with 11ß-hydroxylase defect had extremely high values for these diagnostic ratios.
17-Hydroxylase deficiency
The major hormonal steroid overproduced in 17-hydroxylase deficiency is corticosterone, and in previous studies we have shown that Kendall’s compound A (11-dehydrocorticosterone) is dominant in the neonate. Both 5-tetrahydro compound A (5THA) and tetrahydro compound A are quantitatively important, but the major corticosterone metabolite is 6-hydroxy-THA (17, 26). This is in contrast to children and adults in whom metabolites of corticosterone (Kendall’s compound B) dominate, e.g. tetrahydrocorticosterone or 5-tetrahydrocorticosterone. Thus, the primary analytes we measure for diagnosis of 17-hydroxylase deficiency are tetrahydro compound A, 5THA, and 6-hydroxy-THA. Deficiency of 17-hydroxylase prevents the formation of C19 steroids; therefore, the excretion of all such steroids is minimal. The ratio of 16-hydroxypregnenolone/16-hydroxy-DHEA is useful as a discriminant. In a previous study of a different patient (17), this ratio was reported to be more than 100 rather than approximately unity for normals. The elevated ratio has now been verified in the patient studied here (Fig. 5).
We have adopted the following ratios for 17-hydroxylase deficiency: 5THA x 100/CM (ratio 9), 6-hydroxy-THA x 100/CM (ratio 10), and 16-hydroxypregnenolone/16-hydroxy-DHEA (ratio 11). Values for ratios 10 and 11 are plotted in Fig. 5 and clearly demonstrate the deficiency of 17-hydroxylase. Ratio 9 is equally discriminating but is not illustrated.
In summary, methodology that can be used for routine analysis of neonatal urine for four causes of CAH has been described. This methodology is noninvasive; requires no ACTH stimulation; and provides discrimination of 21-hydroxylase, 17-hydroxylase, 3ßHSD, and 11ß-hydroxylase deficiencies in the first days of life. For 21-hydroxylase deficiency, the predominant enzyme deficiency for CAH, we have shown that reliable diagnosis can be made on samples collected within 24 h of birth. The method can be used independently or in conjunction with serum steroid assays. It allows for diagnosis of CAH within 36 h upon receipt of the urine in the laboratory. Specificity of analysis is achieved by requiring at least two, and sometimes more, specific analytes for each condition to be measured.
There is one additional form of CAH that is not addressed in this manuscript, namely congenital lipoid adrenal hyperplasia. This is a rare variant of the disorder caused by deficiency of steroid acute regulatory protein (StAR). This is a protein associated with cholesterol transport from outer to inner mitochondrial membrane (6, 27), a process necessary for introducing the precursor to the steroidogenic cascade. This is a disorder causing severe adrenal insufficiency and male pseudohermaphroditism. Neonates with lipoid adrenal hyperplasia excrete almost no steroids; those that are excreted in the first days of life are of maternal origin. A combination of adrenal hyperplasia and negligible steroid excretion is diagnostic of the condition.
Current trends in diagnosis are moving toward molecular rather than biochemical analysis. Direct mutation analysis for diagnosis of CAH is possible, particularly for 21-hydroxylase deficiency. It is used for prenatal diagnosis of the condition and as a complement to neonatal screening because the serum 17-hydroxyprogesterone immunoassay, in the absence of preextraction and chromatography, can produce false-positive results (28, 29, 30, 31, 32). However, routine molecular diagnosis for the 21-hydroxylase defect remains difficult because of the presence of the CYP21 pseudogene and the wide range of mutations reported in affected patients (1, 2, 28). These include gene deletions, large gene conversions, point mutations, and small deletions. Screening for common mutations has been proposed, but the best way to avoid missing uncommon mutations is to sequence the entire gene (2, 32, 33). Although automated sequencing methods are not yet widely available, DNA chip technology is the most attractive option for routine high throughput diagnosis but remains a research technique (2). Mutation analysis for the 11ß-hydroxylase, 17-hydroxylase, and 3ß-HSD defects also remain research methods.
Thus, it is our view that mass spectrometry provides the best bridging technology between multiple serum immunoassay methods and routine DNA analysis for CAH diagnosis. The random urine steroid metabolite profile provides a rapid, simplified, comprehensive, differential diagnosis for CAH.
Acknowledgments
We are indebted to Drs. Maria New, Stefan Wudy, and Janos Homoki for supplying urine samples from several CAH patients and to Dr. George Joannou for providing important reference compounds. Cedric Shackleton, Ph.D., is particularly indebted to Esther Roitman who for many years served in the development of the reported methodologies.
Footnotes
Abbreviations: 6-Hydroxy-THA, 6-Hydroxy-tetrahydrocompound A; 6-hydroxy-THS, 6-hydroxytetrahydro substance S; 3ß-HSD, 3ß-hydroxysteroid dehydrogenase; CAH, congenital adrenal hyperplasia; CM, cortisol metabolite; GC/MS, gas chromatography/mass spectrometry; 17OHP, 17-hydroxyprogesterone; SIM, selected ion monitoring; TIC, total ion current.
Received November 14, 2001.
Accepted April 13, 2002.
References
-hydroxylase/17, 20 lyase deficiency: from clinical investigation to molecular definition. Endocr Rev 12:91–108-hydroxyprogesterone radioimmunoassay kit when used with samples from neonates. Clin Chem 34:2070–2075-hydroxyprogesterone radioimmunoassay. Clin Chem 38:1830–1837, 15ß, 17-trihydroxy-5ß-pregnan-20-one in the perinatal diagnosis of congenital adrenal hyperplasia (CAH) due to a 21-hydroxylase deficiency. J Steroid Biochem 14:901–912[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  V. Dhir, H. E. Ivison, N. Krone, C. H. L. Shackleton, A. J. Doherty, P. M. Stewart, and W. Arlt Differential Inhibition of CYP17A1 and CYP21A2 Activities by the P450 Oxidoreductase Mutant A287P Mol. Endocrinol., August 1, 2007; 21(8): 1958 - 1968. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. M. Kushnir, A. L. Rockwood, W. L. Roberts, E. G. Pattison, W. E. Owen, A. M. Bunker, and A. W. Meikle Development and Performance Evaluation of a Tandem Mass Spectrometry Assay for 4 Adrenal Steroids Clin. Chem., August 1, 2006; 52(8): 1559 - 1567. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Keen-Kim, J. B. Redman, R. U. Alanes, M. M. Eachus, R. C. Wilson, M. I. New, J. M. Nakamoto, and R. G. Fenwick Validation and Clinical Application of a Locus-Specific Polymerase Chain Reaction- and Minisequencing-Based Assay for Congenital Adrenal Hyperplasia (21-Hydroxylase Deficiency) J. Mol. Diagn., May 1, 2005; 7(2): 236 - 246. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Homma, T. Hasegawa, E. Takeshita, K. Watanabe, M. Anzo, T. Toyoura, K. Jinno, T. Ohashi, T. Hamajima, Y. Takahashi, et al. Elevated Urine Pregnanetriolone Definitively Establishes the Diagnosis of Classical 21-Hydroxylase Deficiency in Term and Preterm Neonates J. Clin. Endocrinol. Metab., December 1, 2004; 89(12): 6087 - 6091. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Olgemoller, A. A. Roscher, B. Liebl, and R. Fingerhut Screening for Congenital Adrenal Hyperplasia: Adjustment of 17-Hydroxyprogesterone Cut-Off Values to Both Age and Birth Weight Markedly Improves the Predictive Value J. Clin. Endocrinol. Metab., December 1, 2003; 88(12): 5790 - 5794. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
