Interpretation of results and summary of profile findings in steroid disorders

URINARY STEROID PROFILING SERVICE: Guidelines
INTRODUCTION
This national service originated at the MRC Clinical Research Centre and grew out of research by Dr. C.H.L. Shackleton from the Centre’s opening in 1970. Many steroid disorders were investigated by capil ary gas chromatography (GC) and gas chromatography-mass spectrometry (GC-MS) for the first time and numerous new steroid metabolites were identified. Dr. N.F. Taylor joined the laboratory in 1973 and took over control of the service in 1983. It was transferred to the Department of Clinical Biochemistry at King’s in 1989, gaining recognition as an SAAS service soon after. It continues to be operated and directed by Dr. Taylor. There are close clinical links with paediatric (Dr. C. Buchanan) and adult (Prof. A. McGregor) endocrinology. The department has ful CPA accreditation. Research and development related to the service are continuous. In–house and col aborative projects are always in progress and contract analyses are also carried out. Records of every patient ever investigated are maintained, filed under patient name. These can general y be rapidly accessed if a sending laboratory or clinician wishes to cal to discuss results and interpretation.
Urinary steroid profiling by GC and GC-MS provides qualitative and quantitative data on excretion of steroid metabolites. Most of these originate in the adrenal cortex, while the gonads after puberty contribute to the metabolites of androgens and 17α- hydroxyprogesterone. A profile provides a composite picture of the quantitatively major biosynthetic and catabolic pathways. This is not synonymous with biological importance: metabolites of cortisol, progesterone, corticosterone and testosterone are readily detected whereas those of oestradiol and aldosterone are not, unless analysed by GC-MS in selected ion monitoring mode. We formerly used GC for the primary analysis but now use GC-MS to generate quantitative data. There are internal quality control procedures in place and we participate in the UKNEQUAS scheme run by University Col ege London in conjunction with ERNDIM.
APPLICATIONS
Steroid profiling is ideal for detection of altered steroid metabolism, whether due to defects in steroidogenic enzymes, associated with certain clinical states (eg: anorexia, porphyrias) or as a result of drug treatment. When 24 hour col ections are made, profiling can also be used to detect changes in steroid excretion rates. This approach offers much greater discrimination of smal changes, eg: when checking for adrenal suppression resulting from treatment with inhaled steroids for asthma, but is clearly not appropriate in many other circumstances. Potential errors due to incomplete col ection can be minimised by expression of steroid levels relative to creatine or completeness can be checked with an ingested marker, such as para-aminobenzoic acid (PABA). For infants and young children, we have developed a novel method for the quantitative extraction of 24 hour col ections of disposable nappies. While we do not encourage this for regular use because the method is labour intensive, it offers an entirely comparable quality of analysis with the advantage that col ections can be made at home and avoid the expense and stress of hospital admission.
We recommend that inborn errors of steroid metabolism always be confirmed by profile analysis. Several disorders require lifelong treatment so that a definitive identification is essential in terms of both patient welfare and cost. We have encountered many examples of confusion being caused by sole reliance on plasma steroid assays. Profiling enables the functional degree of enzyme deficit to be estimated, by comparing levels of the enzyme’s precursor and product. We have encountered partial forms of al the major steroidogenic enzyme deficiencies. Profiling is not routinely necessary for regular monitoring of patients on replacement treatment but is occasional y useful for investigating problems that arise.
Profiling has special relevance to the perinatal period, when gender ambiguity or salt wasting may indicate a possible inborn error of steroid metabolism. These need to be identified rapidly post partum to anticipate life threatening steroid deficiency and in the case of sexual ambiguity, to establish the sex of rearing as soon as possible. There are marked differences in steroid metabolism at this stage of life and standard steroid immunoassays frequently give discrepant results owing to the presence of cross-reactants. Another possible source of confusion is the occurrence of transient enzyme deficiencies. The identities of many of the major metabolites in infancy were first established by us and we continue to investigate the changes in metabolism that take place over time in both normal infants and those with inborn errors or other clinical disorders. For example, we have recently used col ections on nappies to derive longitudinal data on a cohort of severely preterm infants fol owed at defined intervals until 3 months post normal term.
NORMAL RANGES
Our report forms compare the profile findings with age-appropriate normal range data for neonates, children within the age ranges 2-6, 7-10 and 11-17 years, adult males and adult females. As required, we list less usual metabolites such as in screening for steroid sulphatase deficiency in pregnancy, and investigating adults treated with prednisolone. We hold normal data for children listed for each half year of life, so that deviations from normal values can be quoted in more detail where relevant. SUMMARY OF PROFILE FINDINGS IN STEROID METABOLIC
DISORDERS
Almost al known inborn errors of steroid metabolism can be readily identified in a single steroid profile. The exceptions are lipoid adrenal hyperplasia (StAR protein defect), in which absence of steroid metabolites cannot provide distinction from hypoadrenalism, and 17-hydroxysteroid dehydrogenase deficiency, which requires analysis of serum androstenedione and testosterone. The distinguishing features of each disorder are summarised below.
3β -Hydroxysteroid dehydrogenase (3β-HSD) deficiency
Increased metabolites of 17a-hydroxypregnenolone and DHA. Low/absent androgen, corticosterone and cortisol metabolites.
5α-Reductase (SRD5α2) deficiency
Increased ratio 5β/5α reduced metabolites of androgens, corticosterone and cortisol
metabolites. In newborns with ambiguous genitalia, this disorder cannot be looked for by profiling until 1-3 months of age, because the diagnostic pairs of 5-reduced metabolites are not detectable until this stage.
11β-Hydroylase (CYP 11B1) deficiency
Increased metabolites of 11-deoxycortisol and androgens. DOC metabolites are much increased but difficult to quantitate. Low/absent corticosterone and cortisol metabolites.
Corticosteroid 11-dehydrogenase (11β-HSD 2) deficiency & Corticosteroid-11-
oxoreductase (11β-HSD 1) deficiency
Increased ratio of cortisol/cortisone metabolites, decreased adrenal steroid metabolites and decreased ratio of cortisol/cortisone metabolites, increased adrenal steroid metabolites respectively.
17α-Hydroxylase (CYP 17) deficiency
Increased metabolites of progesterone, DOC and corticosterone. Absent androgen and cortisol metabolites.
17,20-Lyase (CYP 17) deficiency
Absent androgen metabolites.
21-Hydroxylase (CYP 21) deficiency
Increased 17α-hydroxyprogesterone and 21-deoxycortisol metabolites. Low cortisol and
corticosterone metabolites. The salt-wasting and simple virilising forms cannot be distinguished.
Placental sulphatase (STS) deficiency & X-linked ichthyosis
Increased steroid sulphates in maternal urine. Low oestriol. No differences from normal are
seen in the urinary steroids from individuals with steroid sulphatase deficiency but serum
cholesterol sulphate is elevated.
Aldosterone synthase (CYP 11B2) deficiency, (hypoaldosteronism)
Increased corticosterone metabolites. GC-MS analysis shows that tetrahydroaldosterone, the
major metabolite of aldosterone, is low or normal.
Cytochrome P450 oxoreductase (POR) deficiency (associated with Antley-Bixler
syndrome)
Increased 17α-hydroxyprogesterone, 21-deoxycortisol and corticosterone metabolites.
Relatively low cortisol metabolites.
7-Dehydrocholesterol reductase (DHCR7) deficiency (associated with Smith Lemli
Opitz syndrome, SLOS)
Increased 7- and 8-dehydro versions of the usual steroid metabolites.
Aromatase (P450arom) deficiency
Oestrogen deficiency in pregnancy
OTHER CLINICAL DISORDERS AMENABLE TO INVESTIGATION BY
URINARY STEROID PROFILING
Pseudohypoaldosteronism (apical sodium channel defect)
Increased corticosterone metabolites. GC-MS analysis shows that tetrahydroaldosterone is
increased.
Congential adrenal hypoplasia
In the 'anencephalic' form in infancy, DHA and pregnenolone metabolites are absent. In
other forms, al steroids are absent.
Steroid producing tumours
Adrencortical tumours usual y produce very distinctive profiles. We have more than 100
examples and have classified 7 major patterns of steroid excretion. Steroid metabolite levels
offer some indication of likelihood of recurrence. Some adrencortical tumours do not secrete
steroids, while occasional y the profile mimics a partial enzyme deficiency. Tomographic
methods have reduced the importance of profiling for initial diagnosis, but we recommend
submission of urine for analysis because the steroids in excess can be defined and used as
markers of recurrence during fol ow up. Gonadal tumours are sometimes strongly indicated
by increases in androgen and/or 17α-hydroxyprogesterone metabolites. More often, they are
suggested by the finding of a normal profile in conjunction with increased serum sex steroids.
Precocious puberty/virilisation in children
Specific adrenal causes of androgen excess, ie: virilising forms of congenital adrenal
hyperplasia (CAH) or a tumour are readily identifed. Precocious increases in adrenal
androgen synthesis ('premature adrenarche') are frequently indicated by increases in
androgen sulphate excretion and increases of the other major androgen metabolites, with or without accompanying increases of cortisol metabolites. Thus, while profiling per se cannot distinguish precocious adrenarche from precocious puberty, it can indicate whether an increased contribution of androgen from the adrenals is likely, and provide reassurance when, as is usual y the case, there is no serious adrenal pathology.
Hirsutism in women
This has been an area of research for us for a number of years and close links are
maintained with Prof. McGregor’s team. Excess of adrenal steroid metabolites is a common
finding, and we believe that there are a number of common causes, including mild
glucocorticoid resistance and increased cortisol metabolic clearance as a result of increase in
the oxidative activity of 11β-hydroxysteroid dehydrogenase. Profiling readily distinguishes
partial forms of CAH. Polycystic ovary syndrome is not distinguishable by profiling. Many clinicians consider that, once important pathological causes of androgen excess have been eliminated, further investigation is fruitless since use of the antiandrogen cyproterone acetate is the most effective treatment, irrespective of the source of the androgen excess. If the clinician has a high “index of suspicion” that a hirsute patient is unusual this would be justification for profiling: by this means we have identified three of the six families known to us in the World with corticosteroid 11-oxoreductase deficiency.
Cushing's syndrome
The most appropriate first line test is urinary free cortisol. If Cushing’s is considered
probable, profiling can help to distinguish adrenocortical tumour from other causes, and give
an indication of the duration of the cortisol excess. If Cushing’s is due to ACTH excess, al
steroids are increased. Adrenocortical tumours usual y produce a similar pattern but with
additional unusual metabolites. In both types, the relative proportions of the cortisol
metabolites are abnormal, the degree of abnormality depending on the duration and severity
of the excess cortisol production. The more acute the increase, the more that cortisol
metabolites are increased relative to cortisone metabolites and the more free cortisol is
increased relative to cortisol metabolites. Similarly, cortisol metabolites are increased relative
to androgen metabolites, but less so if the excess stimulation has been long term. As in the
discussion of adrenocortical tumour above, profiling can be used to provide early warning of
recurrence.
Hypertension
Profiling is only appropriate after aldosterone and renin measurements have been obtained.
The causes due to inborn errors of metabolism are readily distinguishable. Conn's syndrome
may be distinguishable from bilateral adrenal hyperplasia on the basis of an increase or
decrease of corticosterone metabolites.
Glucocorticoid treatment
We have found on a number of occasions that over-the-counter remedies obtained abroad
that claim to be herbal in origin contain glucocorticoids. We can identify some constituents,
such as dexamethasone, by direct analysis of the product, but a reliable way to establish if
glucocorticoid activity is present is to compare profiles on and off treatment in order to look for
evidence of adrenal suppression. Some synthetic glucocorticoids used in conventional
treatment, such as prednisolone, give rise to urinary metabolites. A single profile can provide
information on steroid uptake and metabolism and degree of adrenal suppression caused.
Drug treatment
Changes in steroid metabolism due to some drugs (eg metyrapone, ketoconazole) can be
defined.
Anabolic steroid abuse
This results in testicular atrophy in the male, with consequent reduction of androgen
metabolites. Since the normal range for these is wide, profiling cannot provide proof of
abuse. For specific identification of anabolic agents, we recommend the GC-MS screen
provided by Prof. D.A. Cowan at The Drug Control Centre, King’s Col ege, Waterloo Campus
(020 7848 4779).
SAMPLE COLLECTION CONSIDERATIONS
Few drugs interfere with steroid profile analysis, but it is important to be aware that glucocorticoid treatment suppresses adrenocortical activity, so that steroids diagnostic of an adrenal disorder may no longer be detectable during their use. Treatment with cortisol (hydrocortisone, cortisone acetate) also makes it impossible to assess the ability to synthesise cortisol, since the urinary metabolites from exogenous and endogenous sources are identical. If CAH is suspected and glucocorticoid replacement has already been started, the dose should be tailed off before col ection if treatment has been given only briefly. Otherwise, it is necessary to change the glucocorticoid to dexamethasone and stimulate with depot Synacthen for up to 5 days before sample col ection.
For investigation of salt wasting states in infants, diagnostic increases of corticosterone metabolites may be abolished if electrolyte and/or mineralocorticoid treatment has normalised the plasma electrolytes. It is therefore advisable to tail off treatment as far as possible before sample col ection.
When it is appropriate that steroid profiling be carried out on two or more samples, for example in a stimulation or suppression test, we regard this as one investigation and bil accordingly. If a disorder in a newborn cannot be diagnosed on the first specimen, the repeat is not charged for. Col ection on two different days rather than one without such a reason never enhances the result, and we therefore only analyse one sample if a pair is submitted.
STORAGE AND SENDING OF SAMPLES
For most purposes, a 24 hour col ection is preferable to a random sample. No preservative is needed. Samples can be stored at 4° for short periods or frozen if longer. A portion (20ml in a plastic universal container) should be sent by first class post. Smal er samples can usual y be accommodated. If freezing in a universal container, the vessel must be kept upright: even if plenty of room for expansion is al owed. Freezing horizontal y results in the cap being displaced, leading to leakage of a concentrate of the sample.
The request form should include brief clinical details and drug treatments and any specific questions that are being addressed: without this it is difficult to provide any focussed comment in the report.
Please indicate if an urgent response is required. Written reports are usual y returned in 2-3 weeks, but telephoned reports can be given within one week or less. We normal y expect to telephone or fax results on young infants as soon as we obtain them. Comments are provided with results and further explanation, advice and repeat copies of reports are always available. While it is not necessary to phone before submitting samples, we welcome discussion of patients and are happy to suggest alternatives to profiling if warranted.
Address samples to: Dr. N.F. Taylor, Department of Clinical Biochemistry, King’s Col ege School of Medicine, Bessemer Road, London SE5 9PJPhone (direct line): 020 3299 4131 Fax: 020 7737 7434 e-mail:

Source: http://www.kingspath.info/resources/pdf/test/2/kingspath%20usp%20guidelines%202009.pdf