European Journal of Pediatrics Springer-Verlag 1984
Steroid biochemistry of virilising adrenal tumours in childhood
J.W. Honour1*, D.A. Price2, N.F. Taylor3, H. B. Marsden2, and D. B. Grant4
1 Cobbold Laboratories and Dept. of Chemical Pathology, Middlesex Hospital Medical School, Mortimer Street, London WIN 8AA, England
2 Royal Manchester Children’s Hospital, Pendlebury, Manchester M271HA, England
3 Division of Clinical Chemistry, MCR Clinical Research Centre, Watford Road, Harrow, Middlesex HA13UJ, England
4 Hospital for Sick Children, Great Ormond Street, London WC1N3JH, England
Abstract. Steroid excretion in urine of 12 infants with virilising adrenal tumours has been determined using gas chromato- graphy. In six children, (Group A, five female, one male) aged 2.8-5.3 years, very high urinary excretions of 17 oxosteroids (>40 µmol/24 h) were largely accounted for by dehydroepi- androsterone (DHA). In one of the girls, the pattern of steroids excreted in urine was similar to that of newborn infants, with high excretions of 16-oxygenated derivatives of DHA. The histology of this tumour suggested a neoplasia of fetal-type adrenocortical cells. Very large tumours were found in three of the infants, two of whom have died and one has multiple meta- stases. From the other three children, small, well-encapsulated adenomas were successfully removed. Six children (Group B), had moderately elevated 17-oxosteroid excretions (8-17µmol/ 24 h). In five of these cases (four female, one male) aged 0.8-5 years, 118-hydroxyandrosterone was a consistently prominent urinary steroid. In one boy, aged 7.7 years, 17-oxosteroid excretion was 15 umol/24 h and the major steroids in urine were metabolites of pregnenolone. These six children have survived with no clinical evidence of recurrent tumour.
The in vivo functional activities of the tumours can be de- duced from the different profiles of steroids in urine. These have revealed heterogeneous patterns of steroid biosynthesis.
Key words: Adrenal gland neoplasms - Urine steroids - Gas chromatography - Mass spectrometry
Introduction
Adrenal androgen secreting tumours typically cause rapid viril- isation after the neonatal period. Only rarely is virilisation pre- sent at birth due to exposure to androgens from a fetal tumour [22]. In many cases high excretion rates for dehydroepiandro- sterone (DHA) have been reported. This has been considered an hallmark of malignancy [27] and is reflected in high urinary excretion of 17-oxosteroids. Further studies have shown increased excretions of other steroids, e.g. 16a-hydroxy DHA [11], 118-hydroxy DHA [37], 16-androstenes [16] and some hormones which are normally important biosynthetic precur- sors but only minor secretory products, e.g. 118-hydroxy- androstenedione [32]. In some of these tumours urinary 17- oxosteroids may be only moderately elevated [11,28]. Testo- sterone-producing adrenal tumours are very rare [6,8].
We have studied the excretion patterns of steroids in urine from infants prior to the removal of adrenal tumours in order to assess the steroidogenic activities of the tumours in vivo. The urinary steroid profiles are compared with the results of other laboratory investigations, and these data are reviewed in rela- tion to the gross and microscopic appearance of the excised tumours, and the prognosis.
Methods
The detailed methods for profile analysis of steroids in urine have been described previously [34]. Briefly, steroids were ex- tracted from urine with Amberlite XAD-2 resin or Sep-Pak C18 cartridges as described by Shackleton and Whitney [35]. The steroid conjugates were hydrolysed with the sulphatase and ß-glucuronidase of snail (Helix pomatia) digestive juice. After 36 h hydrolysis, steroids were recovered from these reac- tions by again using Amberlite XAD-2 or Sep-Pak cartridges and the extracts were purified during chromatography on Sephadex LH-20 with cyclohexane: ethanol (4:1 v/v) as the eluting solvent. Methyloximetrimethylsilyl ether derivatives were prepared before temperature programmed gas chromato- graphic analysis on a 20 m OV-1 coated capillary column with flame ionisation detector. Each chromatogram displays all the products of the major steroids in urine. The compounds were provisionally identified by the GC retention times of the peaks in the analysis and where necessary mass spectrometry was used for confirmation.
Histology was reviewed by one histopathologist (H.B.M.) and a tumour was judged malignant if the capsule was penetra- ted or if there was invasion of vascular sinuses. Less significant criteria were lack of an orderly arrangement of cells, bizarre hyperchromatic cytology, presence of mitoses and of necrosis.
Patients
The clinical histories of the three boys and nine girls, presenting with virilisation due to adrenal tumours, are summarised in Table 1. Their ages ranged from 0.8 to 7.7 years. In most cases the children presented with virilisation of less than 12 months duration. One child had hemi-hypertrophy (Case 8) and an- other (Case 10) had close family history of malignancy, namely carcinoma of the breast in the mother and rhabdomyosarcoma in a brother. Only two children (Cases 1 and 7) also had mild
* Corresponding author
| Case | Age (years) | Sex | Duration of virilising symptoms (months) | Tumour | site and size | Histopathology | Outcome |
|---|---|---|---|---|---|---|---|
| 1 | 3 | F | 3 | (L) | 12.5 cm | Carcinoma | Multiple metastases |
| 2 | 2.8 | F | 2 | (L) | 4 cm | Adenoma | Alive and well 4 years |
| 3 | 4 | F | 6 | (L) | 4 cm | Adenoma | Alive and well 4 years |
| 4 | 4 | F | 6 | (R) | 13 cm, renal infiltration | Carcinoma | Died 5 months post-operation |
| 5 | 5.3 | M | 4 | (R) | 9 cm, adherent to diaphragm | Carcinoma | Died 4 years post-operation |
| 6 | 4.6 | F | 12 | (R) | 3 × 1.5 cm | Adenoma fetal type cells | Alive and well 3.5 years |
| 7 | 0.7 | F | 2 | (L) | 3 cm | Adenoma | Alive and well 2.5 years |
| 8 | 2 | M | 5 | (L) | Multiple nodules 2 × 1.2 cm and 0.8 cm | Adenoma | Alive and well 7 years |
| 9 | 4.3 | F | 2 | (R) | 4.5 × 2.5xcm | Adenoma | Alive and well 3.5 years |
| 10 | 1.7 | F | 10 | (L) | 3.5 cm | Carcinoma | Alive and well 2 years |
| 11 | 2.5 | F | 7 | (R) | 5× 5×3.5cm | Adenoma | Alive and well 1 year |
| 12 | 7.7 | M | 10 | (L) | 6 × 5 × 3.5cm | Adenoma | Alive and well 1.5 years |
| Case | Serum testosterone (nmol/l) | Total urinary 17 oxosteroids (pmol/24h) | Major steroids in urine | Individual steroid excretion rates (pmol/24h) | |
|---|---|---|---|---|---|
| Group A* | 1 | 11.5 | 175 | Aetiocholanolone | 10 |
| DHA | 42 | ||||
| 16a-Hydroxy DHA | 16 | ||||
| 2 | 33 | 43 | DHA | 12 | |
| 3 | 3.4 | 64 | Androsterone | 24 | |
| Aetiocholanolone | 3 | ||||
| DHA | 28 | ||||
| 4 | ND | 230 | DHA | 50 | |
| 5 | ND | 160 | DHA | 37 | |
| 6 | 11.5 | 58 | Androsterone | 6 | |
| Aetiocholanolone | 2 | ||||
| DHA | 6 | ||||
| 16a-Hydroxy DHA | 33 | ||||
| Group B* | 7 | 2.6 | 8 | Androsterone | 3 |
| Aetiocholanolone | 1 | ||||
| DHA | 2 | ||||
| 118-Hydroxyandrosterone | 7 | ||||
| 16x-Hydroxy DHA | 2 | ||||
| 5-Androstene-33,16x,173-triol | 4 | ||||
| 8 | Raised | 10 | 118-Hydroxyandrosterone | 6 | |
| 5-Androstene-33,16a,178-triol | 3 | ||||
| 9 | ND | 11 | 118-Hydroxyandrosterone | 3 | |
| 10 | 6.2 | 17 | Androsterone | 6 | |
| 118-Hydroxyandrosterone | 7 | ||||
| 16a-Hydroxy DHA | 13 | ||||
| 11 | 1.8 | 14 | 113-Hydroxyandrosterone | 5 | |
| 5-Pregnene-33,16x,20-triol | 20 | ||||
| 12 | 2.1 | 15 | Androsterone | 3 | |
| Pregnene-33,16a,20a-triol | 7 | ||||
| 16a-Hydroxypregnenolone | 6 | ||||
| Pregnene-33,16x,200-triol | 8 |
ND-not determined
* The groups are defined by the levels of excretion of 17-oxosteroids
clinical features of Cushing’s syndrome but in neither was this confirmed biochemically.
Tumours were palpable in only two infants (Cases 1 and 4) but were visualised in all others except Cases 5, 6 and 8 by X-ray or ultrasound. In the latter case the pre-operative diagnosis was made in 1975 by demonstrating a testosterone gradient be- tween the adrenal vein blood on the left side compared with the periphery. At surgery most tumours were less than 5 cm in dia- meter and were encapsulated permitting complete resection (Table 1). Tumours were classified partly by their histological appearances. The sections of one tumour (Case 6) revealed bizarre giant cells with atypical nuclei showing cytoplasmic intranuclear inclusions similar to those seen in fetal adreno- cortical cytomegaly. In Case 5 a large tumour (>9 cm diame- ter) had invaded surrounding tissues and the tumour bed was irradiated following resection. In Cases 1 and 4 o,p’DDD treat- ment [24] afforded temporary remission of symptoms. How- ever, multiple metastases developed in these three children and two died. The third child (Case 1) has had subsequent surgical removal of several abdominal metastases. The other nine chil- dren have been followed for up to 7 years and show no clinical evidence of recurrent tumour. The signs of virilisation have regressed and physical development has been normal except for continued pubertal development in Cases 6 and 12 and this has been treated with cyproterone acetate.
Results
The production of androgens was assessed by determinations of urinary excretion of 17-oxosteroids, plasma testosterone
DHA
3
C
A
S
2
6
9
1
8
11
14
concentrations and from the profiles of urinary steroids obtain- ed from analysis by capillary column gas chromatography. Uri- nary 17-oxosteroids were variable elevated above the normal range (2-6 pmol/24 h) and were not suppressed after admini- stration of dexamethasone. Six children (Group A, Table 2) were distinguished by having very high urinary excretions of 17- oxosteroids. Detailed GC analysis of steroid excretion in urine from Cases 1-5 confirmed that 17-oxosteroid excretion excee- ding 40 umol/24 h was largely attributable to dehydroepiandro- sterone (Fig. 1). In one female child (Case 6) 17-oxosteroid excretion was 58 pmol/24 h. This was not totally attributable to dehydroepiandrosterone and the excretions of androsterone, aetiocholanolone and 16a-hydroxy DHA were also raised. Excretion of this latter steroid is usually only so prominent in the neonatal period.
The other six children (Group B) had only moderately ele- vated excretion of 17-oxosteroids (8-15 umol/24 h). GC profile analysis of steroids in urine from three of these children (Cases
5
IIBhydroxy Androsterone
8
A
S
C
1
3
2
18
16
9
A
17
S
15
C
12
1
8
11
2
3
7,8 and 9) indicated that 118-hydroxyandrosterone was the major androgen metabolite (Fig. 2). In Cases 10 and 11, 118- hydroxyandrosterone was also a prominent steroid in urine but 16a-hydroxy-DHA or pregnene-33,16a,20a-triol respectively, were also quantitatively important. In one boy (Case 12) the abnormal steroids in urine were metabolites of pregnenolone (Fig.3). These included pregnenediol, 16a-hydroxypregneno- lone and 5-pregnene-38,16,20x(and )-triols.
17-Oxosteroid excretion returned to normal in most of the children post-operatively. The profiles post-operatively con- firm removal of the source of excess androgen. The incomplete removal of steroidogenic tumour tissue from Cases 1 and 4 was indicated by the high excretion rates of steroids in urine collect- ed 3 months and 1 week after surgery respectively. Despite ir- radiation, androgen excretion continued to increase in Case 1 until metastases were removed.
Discussion
The GC profiles of urine steroids described above reflect the secretion of a range of steroid hormones and their precursors by adrenal tumours but two distinct patterns have been repeat- edly observed, distinguished by whether DHA (n =6) or 118- hydroxyandrosterone (n=5) are prominent steroids. In nor- mal children under 8 years the excretion rates of these two steroids would each be less than 1 µmol/24 h. The GC patterns of steroids in urine from patients with adrenal tumours secret- ing DHA has been described previously [26,34] but this type of tumour is not confined to children [5,31]. In the present series, conjugated steroids in urine were hydrolysed by using the en- zymes of Helix pomatia digestive juice. If conjugates had been separated from free steroids before separate hydrolysis the pro- duction of DHA-sulphate could have been determined. The pattern of steroids in urine from Case 12 reflects the high secre- tion of pregnenolone and as such resembles the activity of an adrenal tumour in an adult male with no clinical signs of andro- gen excess and in whom the tumour was regarded as non-func- tioning [13].
The observed patterns of steroid metabolites are assumed to reflect the respective excess secretion of either DHA or 118- hydroxyandrostenedione by tumours although both com- pounds are normally secreted by the adrenals. In previous stud- ies of infants with virilising adrenal tumours, arterio-venous concentration differences across the adrenal have been shown for these steroids [4,32]. 118-hydroxyandrostenedione is con- sidered the only androgen uniquely secreted by the adrenals, although the plasma concentrations are usually very low (<1 nmol/l). Both DHA and 118-hydroxyandrosterone react in the Zimmermann colorimetric method for 17-oxosteroids: 16a-hydroxy DHA does not react with this reagent. Of the 12 children studied the excretion rates for 17-oxosteroids were much higher in the cases with the DHA-secreting tumours.
We observed variable excretions of 16-oxygenated-38- hydroxy-5-ene steriods (Table 2) which are normally only pro- minent in urine during the neonatal period and this suggests that certain virilising tumours might be derived from remnants of fetal adrenal cortex [20]. Such origin is further supported by the similarities in histological appearance (as in Case 6 of the present series and in Craig and Landing [9] and Gregory et al. [17]) by the notable absence of 30-hydroxysteroid dehydro- genase activity [15] and also by their increased incidence before the age of 3 years [19]. On the other hand high excretion rates
of 16a-hydroxy-DHA and 5-androstene-38,16x,173-triol have also been reported in adults with adrenal [12] or breast carci- noma [2]. In adults undergoing surgery the metabolism of DHA can change until the excretion of 16x-hydroxylated metabolites becomes the major fate of DHA. The apparent lowering of excretion rates of 17-oxosteroids by the stress of surgery [1] is thus accounted for because 16x-hydroxy-DHA does not react with the Zimmermann reagent. The results of urinary steroid profile analysis alone do not however distinguish secretion of the 16-hydroxylated steroid from possible formation by peri- pheral conversion.
The steroids in urine are the hepatic metabolites of the cir- culating hormones or precursors. In general, the 3-oxo-4-ene groups of the active steroids are chemically reduced and the products are conjugated with glucuronic or sulphuric acid before renal excretion. The steroids secreted by tumours can thus be deduced from the nature and excretion rates of the me- tabolites in urine. The biosynthetic pathways to androgens from pregnenolone involve the integrated activities of four enzymes. Two distinct routes, the 4-ene and 5-ene pathways are generally recognised, depending upon whether 38-hyd- roxysteroid dehydrogenase is the first or last step in the sequence. From our own results high excretion of DHA is pre- sumed to reflect low tumour activity of the 36-hydroxysteroid dehydrogenase for pregnenolone. Goldman et al. [15] establis- hed by histochemical techniques that some virilising tumours which excrete excessive quantities of 30-hydroxy-5-ene stero- ids lack this enzyme activity. These observations have been confirmed in vitro [7,30]. Cells cultured from such a tumour release DHA in response to ACTH stimulation in much greater quantities than androstenedione or testosterone.
Adrenal tumours secreting 118-hydroxyandrostenedione were considered to be present in 5 of the 12 cases described above. The early recognition of such cases may be difficult without profile analysis because the excretion of urinary 17- oxosteriods is only marginally elevated (Table 2). 118-Hyd- roxyandrostenedione is probably synthesised along the 4-ene pathway and homogenates of a similar adrenal tumour have been shown to form 118-hydroxyandrostenedione from proge- sterone [16]. Such tumours may have low activity of 17ß-hydro- xysteroid dehydrogenase but have 17« and 118-hydroxylases [25,33]. However, there are difficulties in determining the true functional activities of tumour tissue using in vitro cell culture of tumour tissue. This is illustrated by the steroidogenic proper- ties of confluent, monolayer cultures of normal adrenal cells which predominantly secrete 118-hydroxylated steroids com- pared with non-confluent cultures of the same cells which pre- dominantly secrete 11-deoxysteroids and mimic the activity of many of the cultures prepared from malignant cells [29]. The excess productions of pregnenolone in Case 12 indicates low activity of both the 36-hydroxysteroid dehydrogenase and of the 17a-hydroxylase in these tumour cells.
The inborn errors of steroid metabolism are usually the con- sequence of a single enzyme defect; our results show that adre- nal tumours exhibit multiple enzyme differences from normal endocrine tissues [30]. Huhtaniemi et al. [21] suggested that the overall secretion of C19 steroids by virilising adrenal tumours is due to absence of the 21-hydroxylase. Our studies of steroid excretion in urine also indicate that single hormone measure- ments, e.g. DHA [3,23] or of the urinary 17-oxosteroids are not always appropriate in view of the observed heterogeneity of functional activity in virilising adrenal tumours. Although the
secretion of steroids can be determined directly, e.g. in the venous blood draining the tumour [14] multiple hormone assays would be required for this distinction. We therefore believe that a urine steroid profile affords the best means of assessing steroidogenesis in vivo, particularly when urinary 17- oxosteroid excretion is only marginally elevated. Using profile analysis the outcome of surgery can be assessed even if there are changes in functional activity with recurrence of tumours [18,36].
Acknowledgements. We are grateful to Drs. G. Snodgrass, C.L. Brown, D. I. Johnston and G.M. Addison also Mr. C. Selby for their co-operation with these studies. Michael Madigan, Mary Daly and Jacqui Kent are thanked for their technical assistance.
References
1. Adams JB, Salasoo S (1973) A major pathway of adrenal C19-5-ene steroid metabolism under conditions of stress: formation of 16- hydroxyandrost-5-enes. Acta Endocrinol (Copenh) 72: 319-329
2. Adams JB, Wong MSF (1968) A correlation between urinary ste- roid metabolites and pathways of steroidogenesis in human breast tumour tissue. Lancet II : 1163-1166
3. Allen WM, Hayward SJ, Pinto A (1950) A colour test for dehydro- epiandrosterone and closely related steroids for use in the diagnosis of adrenocortical tumours. J Clin Endocrinol Metab 10:54-70
4. Baulieu EE (1962) Studies of conjugated 17-ketosteroids in a case of adrenal tumour. J Clin Endocrinol Metab 22: 501-507
5. Bègue R-JB, Brun JM, Desgres J, Padieu P (1976) Steroides urinai- res dans un corticosurrenalome virilisant. J Steroid Biochem 7:583-591
6. Burr IM, Graham T, Sullivan J, Hartman WH, O’Neill J (1973) A testosterone secreting tumour of the adrenal producing virilisation in a female infant. Lancet II : 643-644
7. Cahen LA, Villee DB, Powers ML, Crigler JF (1978) A virilising adrenocortical tumour in a female infant. In vivo and in vitro bio- chemical characteristics. J Clin Endocrinol Metab 47: 300-306
8. Costin G, Goebelsmann U, Kogut MD (1977) Sexual precocity due to a testosterone producing adrenal tumour. J Clin Endocrinol Metab 45: 912-919
9. Craig JM, Landing BH (1951) Anaplastic cells of the fetal adrenal cortex. Am J Clin Pathol 21 : 940-949
10. David RR (1967) Adrenal virilism in childhood. Ann NY Acad Sci 142:787-793
11. David R, August G, Gandy HM (1968) A virilising tumour with borderline elevation of urinary 17-ketosteroids. Pediatrics 42 : 139- 148
12. Fantl V (1973) Isolation of urinary 3x,16a-dihydroxyandrost-5- ene-17-one in adrenal carcinoma. J Endocrinol 56: 615-616
13. Fukushima KD, Gallagher TF (1963) Steroid production in “non- functioning” adrenal cortical tumour. J Clin Endocrinol 23:923- 927
14. Gabrilove JL, Seaman AT, Sabet R, Mitty HA, Nicolis GL (1981) Virilising adrenal adenoma with studies on the steroid content of the adrenal venous effluent and a review of the literature. Endocr Rev 2: 462-470
15. Goldman AS, Bongiovanni AM, Yakovac WC, Prader A (1964) Study of A5,38 hydroxysteroid dehydrogenase activity in normal, hyperplastic and neoplastic adrenal cortical tissue. J Clin Endo- crinol Metab 24:894-909
16. Gower DB, Daly JR, Snodgrass GJAI, Stern MI (1970) Steroid excretion and biosynthesis with special reference to C19-16-ene ste- roids in an infant with a virilising adrenocortical carcinoma. Acta Endocrinol (Copenh) 63: 562-576
17. Gregory T, Gardiner LI, Gower DB, Bicknell DC, Barlow MJ (1979) Studies of 16-androstenes in an infant with virilising adrenal carcinoma. Am J Dis Child 133: 294-297
18. Halmi KA, Lascari AD (1971) Conversion of virilisation to femini- sation in a young girl with adrenal cortical carcinoma. Cancer 27:931-935
19. Hayles AB, Hahn HBS, Sprague RG, Bahn RC, Priestly JT (1966) Hormone secreting tumours of the adrenal cortex in children. Pediatrics 37: 19-25
20. Hillman DA, Roy AD, Gatei DG (1972) A virilising tumour deriv- ed from cells of the foetal cortex in a year old African child. East Afr Med J 49: 386-393
21. Huhtaniemi I, Kahri AI, Pelkonen R, Salmenpaara M, Sivula A, Vihko R (1978) Ultrastructural and steroidogenic characteristics of an androgen producing adrenocortical tumour. Clin Endocrinol 8:305-314
22. Kenny FM, Hashida Y, Askari A, Sieber WH, Fetterman GH (1968) Virilising tumours of the adrenal cortex. Am J Dis Child 115:445-458
23. Korth-Schutz S, Levine LS, New MI (1976) Dehydroepiandro- sterone sulphate levels in a rapid test for abnormal adrenal andro- gen secretion. J Clin Endocrinol 42: 1005-1013
24. Korth-Schutz S, Levine LS, Roth JA, Saenger P, New MI (1977) Virilising adrenal tumour in a child suppressed with dexametha- zone for three years. Effect of o,p”DDD on serum and urinary androgens. J Clin Endocrinol Metab 44: 433-439
25. Lisboa BP, Strassner M, Nocke-Finck L, Breuer H, Bayer M (1978) Studies on the metabolism of steroid hormones on a virili- sing adrenal cortex adenoma. Endokrinologie 72:311-326
26. MacDonald MG, Savin JA, Browning MG (1976) Virilising adrenal tumour in a 9-year-old female. Clin Exp Dermatol 1:331- 335
27. Mills IH (1964) Clinical aspects of adrenal function. Blackwell, Oxford
28. Neto RM, Oliviera JAM, Sa MFS, Cornicelli J (1979) A virilising tumour with borderline elevation of urinary 17-ketosteroids and histochemical demonstration of a deficiency of 5/4 ene isomerase 3ß-hydroxysteroid dehydrogenase enzymatic system. Acta Pae- diatr Scand 68:459-464
29. Neville AM, O’Hare MJ (1982) The distinction between benign and malignant adrenocortical tumours causing Cushings’ and or the adrenogenital syndromes: Structural and functional criteria. In: Neville AM, O’Hare MJ (eds) The human adrenal cortex. Sprin- ger-Verlag, pp 186-201
30. Neville AM, Webb JL, Symington T (1969) The in vitro utilisation of (4-14℃)-dehydroepiandrosterone by human adrenocortical tum- ours associated with virilism. Steroids 13:821-833
31. Pfaffenberger CD, Horning EC (1975) High-resolution biomedical gas chromatography: Determination of human urinary steroid met- abolites using glass open tubular capillary column. J Chromatogr 112: 581-594
32. Saez JM, Rivarola MA, Migeon CJ (1967) Studies of androgens in patients with adrenocortical tumours. J Clin Endocrinol 27:615- 623
33. Saez JM, Loras B, Morera AM, Bertrand J (1971) Studies of andro- gens and their precursors in adrenocortical virulising carcinoma. J Clin Endocrinol Metab 32: 462-469
34. Shackleton CHL, Honour JW (1977) Simultaneous estimation of urinary steroids by semi-automated gas chromatography. Clin Chim Acta 69: 267-283
35. Shackleton CHL, Whitney J (1980) Use of Sep-Pak cartridges for urinary steroid extraction: evaluation of the method for use prior to gas chromatographic analysis. Clin Chim Acta 107:231-243
36. Spaulding SW, Masuda T, Osawa Y (1980) Increased 178-hydroxy- steroid dehydrogenase activity in a masculinising adrenal adenoma in a patient with isolated testosterone overproduction. J Clin Endo- crinol Metab 50: 537-540
37. Wiedenfeld J, Shaeffer JM, Landau H, Schiller M, Zadik Z (1978) 118-hydroxydehydroepiandrosterone in a case of virilising adrenal adenoma. Isolation from urine and mitrochondrial conversation from dehydroepiandrosterone. J Clin Endocrinol Metab 47:102- 104