Adrenal cortical tumours: 25 years’ experience at the Royal Children’s Hospital, Melbourne
CM McDonnell and MR Zacharin
Department of Endocrinology and Diabetes, Royal Children’s Hospital, Melbourne, Victoria, Australia
Aims: Adrenal cortical tumours remain a rare entity with inconsistent consensus about treatment and follow up. This article reviews 25 years of experience in the Royal Children’s Hospital, Melbourne, Australia.
Methods: All records with the diagnosis of adrenal adenoma or carcinoma between 1976 and 2001 were reviewed, excluding tumours of the adrenal medulla. Details were recorded for age, gender, family history, presentation, biochemistry, imaging, histology, diagnosis, treatment and outcome.
Results: Twelve children (six boys, six girls) were diagnosed with adrenal cortical tumours in this period. Median age at diagnosis was 2.5 years (range 0.5-15.6 years). Six of the 12 children presented with virilization. The remaining six identified by ultrasound performed for hemi-hypertrophy (2), hypertension (2), and fever with abdominal pain (2). Five children had a family history of tumour and two of these five had a p53 mutation demonstrated on molecular genetic analysis. Tumours in five of the 12 children were defined as malignant on histology. Surgery was deemed curative in nine cases. Only one case required further surgery and two required chemotherapy. Time since diagnosis ranged from 1 to 25.8 years. Two children died from complications of the tumour. One other child died following development of a second tumour.
Conclusion: Adrenal cortical tumours should always be considered in the differential for adrenal hormone excess. New information provided by mutational analysis may predict ongoing risks. Lifelong regular follow up is required.
Key words: adrenal; tumour.
Adrenal cortical tumours remain a relatively rare entity when compared to other forms of cancer. Incidence of adrenal cortical tumours is quoted between 0.05% and 0.5% of all childhood cancers.1-3 The term adrenal cortical tumour encom- passes a wide variety of tumours, which can be functioning, or non-functioning. Functioning tumours are commoner in chil- dren and have been noted to secrete a variety of substances including glucocorticoids, mineralocorticoids, androgens or oestrogens. The majority of adrenal tumours occur sporadically but a proportion have been reported in conjunction with various syndromes including Li Fraumeni Syndrome, Beckwith- Wiedemann Syndrome, Multiple Endocrine Neoplasia type 1 and Carney Complex.4 Consensus about treatment and follow up of these tumours has yet to be reached.
In this article we review 25 years of experience within the Royal Children’s hospital in Melbourne. Changing and develop- ing diagnostic tools (e.g. mutational analysis) and pathological diagnostic criteria are reflected in the spectrum of management seen in this cohort. Children diagnosed with adrenal cortical tumours were followed up for limited periods depending on various biochemical markers and histopathological features. Classification of tumours as benign or malignant was based on several criteria including the weight and diameter of the tumour, the mitotic rate, presence of tumour markers (hormonal or genetic), positive family history of tumours and evidence of metastases on presentation.
METHODS
Hospital records of all patients with a diagnosis of adrenal adenoma, carcinoma or adrenal cortical tumour identified from
the local endocrine database and ICD 10 database of the Royal Children’s Hospital, Melbourne between 1976 and 2001 were reviewed. Tumours of the adrenal medulla (phaeochromocytoma, neuroblastoma and ganglioneuroma) were excluded from analysis. Details were recorded for age, gender, family history, presenting complaint, urine and blood biochemistry, imaging, histology, diagnosis, treatment, subsequent follow up and out- come. Family history of tumour was defined as first or second degree relative with any tumour requiring intervention.
RESULTS
Clinical and biochemical features are summarized in Table 1. Treatment and follow up are summarized in Table 2. As the review takes place over a 25-year period, investigations neces- sarily differed based on availability of various methods at time of presentation.
Cases
Twelve children (six girls, six boys) were diagnosed with adrenal cortical tumours in this hospital between 1976 and 2001. Median age at diagnosis was 2.5 years (range 0.5- 15.6 years).
Presentation
Of the 12 children, six presented with features of virilization. The remaining six had an adrenal mass identified by ultrasound
| Patient | Gender | Age at diagnosis (years) | Presentation | 17 OH-P RR: 0.0-0.5 nmol/L | Androstenedione RR: 0.0-1.0 nmol/L | DHEAS RR: 0.0-5.0 umol/L | Testosterone RR: 0.0-0.4 nmol/L |
|---|---|---|---|---|---|---|---|
| 1 | F | 7.5 | Hemi-hypertrophy | – | 0.5 | 0.3 | 0.2 |
| 2 | M | 1 | Virilized | 13.3 | 16.4 | 25 | 21 |
| 3 | M | 0.5 | Hemi-hypertrophy Cushingoid | – | – | 1.7 | 1.8 |
| 4 | F | 1.5 | Virilized | 16.9 | 12.9 | 66.5 | 3.1 |
| 5 | F | 15.6 | Hypertension | - | - | - | - |
| 6 | M | 10.6 | Hypertension | 2.0 | – | – | – |
| 7 | M | 12.8 | Fever, pain | – | 4.4 | 9.9 | 1.4 |
| 8 | M | 4.3 | Fever, pain | - | - | - | - |
| 9 | M | 0.6 | Virilized, Cushingoid | 15.7 | 1.4 | 4.4 | 12.4 |
| 10 | F | 2.2 | Virilized | – | 0.1 | 7.0 | 0.1 |
| 11 | F | 2.6 | Virilized | 14.9 | 4.2 | 44.1 | 7.3 |
| 12 | F | 0.7 | Virilized | 11.0 | 2.7 | 5.6 | – |
17 OHP, 17 hydroxy progesterone; DHEAS, Dihydroepiandrostenedione; RR, reference range.
| Patient | Adrenalectomy | Weight of tumour | Diameter of tumour | Mitoses/high power field | Family history | Metastases | Follow up |
|---|---|---|---|---|---|---|---|
| 1 | Left | 202 g | 8 cm | Moderate | Yes | – | Annual |
| 2 | Right | – | 5 cm | 2-3 | – | – | Discharged |
| 3 | Left | 11 g | 3.5 cm | – | – | – | Every 3 years |
| 4 | Left | 112 g | 7 cm | Low | Yes | Pulmonary | Died |
| 5 | Left | – | 2 cm | None | – | – | Annual |
| 6 | Right | 98 g | 9 cm | 2 | – | – | Annual |
| 7 | Right | 400 g | 11 cm | – | Yes | Hepatic | Died |
| 8 | Left | 405 g | 12 cm | Abundant | – | Pulmonary | Died |
| 9 | Left | – | 6 cm | Moderate | Yes | – | Annual |
| 10 | Left | 10.2 g | 3.5 cm | Moderate | – | – | Discharged |
| 11 | Left | 26 g | 4.2 cm | 3 | – | – | Annual |
| 12 | Left | 13 g | 2.4 cm | Irregular | Yes | – | Annual |
indicated by hemihypertrophy (2/12), hypertension (2/12) or fever with abdominal pain (2/12).
Family history
Five children had a family history of tumour. Two of the five children with a positive family history of tumour and who were tested, demonstrated a p53 mutation on chromosomal analysis.
Genetic analysis
Genetic alterations were noted in four of the six patients who had chromosomal analysis undertaken. Uniparental disomy of 11p15.5 was identified for patient 3 while patients 4, 10 and 12 all had p53 mutations isolated. Patients 4 and 12 had a postive family history of tumour while Patient 10 did not.
Radiology
Ultrasound was used in 11/12 cases. The remaining case was diagnosed in 1976 by laparatomy before ultrasound was avail- able in the hospital. A suprarenal mass was confirmed by
ultrasound in 9/11 cases. The remaining two cases were con- firmed by CT scan.
Treatment
All 12 cases underwent abdominal surgery with either right (n = 3) or left (n = 9) adrenalectomy. This was curative in 9/12 cases.
Tumour pathology
Tumour weight was recorded in 9/12 reports and was above 100 g in four of these nine cases. Mitotic rate was recorded in 9/12 cases and was described as ‘marked’ in 4/9 specimens.
Diagnosis
Breach of capsule, tumour weight >100 g and high mitotic rate were used to determine whether further treatment was required. Only 5/12 could be defined as malignant based on histology. The remaining tumours were termed as adenomas or adrenal cortical tumours.
Recurrence
Three cases recurred after initial surgery. One required further surgery for excision of pulmonary metastases. Two proceeded with chemotherapy with one case also requiring megavoltage radiation therapy.
Follow up
Time since diagnosis ranged from 1 to 25.75 years. Median duration of follow up was 10 years (SD ± 8.02 years). Of the 10 survivors one has had another tumour identified. This patient died following an intracranial haemorrhage secondary to malig- nant glioma. A second patient developed lower limb weakness secondary to an arachnoid cyst in the spinal canal. There was no identified pattern of long-term follow up within the remain- ing eight cases who have been seen by different medical teams at varying intervals.
DISCUSSION
Adrenocortical tumours occur rarely in childhood making up 0.05-0.5% of all paediatric tumours.1-3 The low incidence has contributed to past difficulties standardizing appropriate treat- ment for children with such tumours. A review of the experi- ence in this hospital over the past 25 years shows some changes in management over this period.
All cases of adrenal cortical tumour diagnosed between 1976 and 2001 at the Royal Children’s Hospital, were reviewed for this article. This included functioning and non-functioning, sporadic and syndromic tumours. No difference was noted in gender although other reviews have noted a propensity to females.5,6 Most were left-sided in comparison to studies which have noticed a right-side preponderance2 and no bilateral cases were identified.
Six of our patients presented with features of virilization, confirming other studies of paediatric populations where viri- lizing features were predominant.2,7 Cushingoid features were seen in two patients, one with virilization and one with hemi- hypertrophy. Prognosis of malignant adrenal tumours that produce glucocorticoids in children has been generally viewed as poor in comparison to purely virilizing tumours, which are well encapsulated.8 We did not confirm this observation in this series.
All adrenocortical tumours in children should be resected surgically to offer the best chance of cure.4,9,10 Difficulties remain in the definition of malignant versus benign adrenal cortical tumours in children. A review by Wajchenberg et al. concluded that the pathological features characterized by Weiss and associated with malignant behaviour in adults did not appear to have the same ominous prognosis when present in children. Increasing tumour diameter has been cited as an independent factor predicting malignancy, with the sugges- tion that a tumour greater than 4 cm11 or 5 cm5 in diameter predisposes to malignancy. Seven of our 12 patients had a tumour mass >5 cm dimension, yet only three developed recurrence or metastases. Other criteria have been tumour weight due to the relationship described between increased tumour size and malignancy6,12 and the suggestion that adeno- mas weigh 20-50 g and the majority of cortical tumours >100 g are malignant.13 We found five tumours to be close to or greater than 100 g yet two of the five patients did not develop recurrence or evidence of metastases and are now 12 and 15.7 years, respectively, since tumour diagnosis. Of the
remaining three children, two died as a result of their primary tumours (Patient 7, tumour weight 400 g; Patient 8, tumour weight 405 g). The third patient (Patient 5, tumour weight 112 g) did develop a pulmonary metastasis, which was success- fully resected within 8 months of initial surgery and she remained well to follow up. Death in this case was caused by intracranial haemorrhage, a complication of malignant glioma, which was diagnosed 4 years later.
Retrospective studies of surgical specimens show metastases or invasion to be the only absolute criteria for malignancy.13,14 In adult patients without metastases who underwent curative resection, only mitotic activity was found to be an independent predictor of survival.14 Review of outcome in paediatric series documenting adrenal tumours estimated 5 years survival between 43 and 89%.1
In the past 5 years, increasing emphasis has been laid on identification of tumour markers and examination of genetic predisposition. In our series, two patients had hemihypertrophy. This is an overgrowth condition associated with infradiaphrag- matic tumours of the adrenals, kidneys and liver. Subsequent identification of an adrenal tumour mass was by elective abdominal ultrasound. One of these two patients (Patient 3) also had chromosomal analysis, which showed 11p15.5 parental disomy, confirming Beckwith Wiedemann syndrome (features include macrosomia, visceromegaly, neonatal hypoglycaemia and increased incidence of intra-abdominal cancers).
In total four of the six patients tested were found to have chromosomal abnormalities associated with genetic syndromes. Various p53 mutations and defects in the short arm of chromo- some 11 have been reported from other studies.2 Another hypothesis is that childhood adrenal cortical tumours arise due to persistence of foetal adrenal cells triggered by defective apoptosis, especially in tumours seen in the first 3 years of life.1
These findings imply a high risk for further tumour develop- ment and underline the need for ongoing follow up. The overall role of cell cycle regulatory protein p53 tumour suppressor and Ki67 proliferative index in adrenal carcinoma still remain controversial because the rarity of disease limits an adequate study population size to correlate molecular expression with prognosis. Other potential mechanisms that could be important in adrenocortical tumorigenesis include steroidogenic acute regulatory protein (STAR) and DAX-1.5 New information regarding genetic mutations associated with various forms of cancer raises the concern that children with adrenal cancer may remain at lifelong risk of other cancers.
The incidental finding of adrenal lesions at time of abdomi- nal imaging procedures now commonly performed (‘incidenta- loma’) has led to a search for a reliable marker of possible malignancy. Major histocompatibility complex (MHC) Class II antigen detection in tumour cells has been reported to exclude malignancy.15 This finding has not been replicated in a study of paediatric tumour samples. Serum dihydroepiandrosterone (DHEAS) levels are frequently elevated in patients with clini- cally diagnosed adrenocortical tumours and can be used as a marker of adrenal androgen excess in patients who present with virilization.8,11 Patients with adrenocortical carcinomas have also been noted to have different excretory patterns of urinary steroid metabolites in comparison to adrenal adenomas.16 This may be a useful diagnostic tool.
From our experience and on review of the experience of others we suggest that all adrenal tumours should be classified on weight, mitotic rate, metastases, family history and presence of a tumour marker. Presence of any of these criteria should indicate a need for close surveillance. As non-functioning tumours are rare in children compared to adults, all adrenal masses noted on imaging should be investigated. Hormone
results should be corrected for age and sex reference ranges specific to the assay method used and preferably carried out at a laboratory specializing in steroid assays.
Poor long-term prognosis is seen in many survivors of adrenal cortical tumours due to development of a second primary tumour in later years. In one series, 3/18 paediatric patients studied developed tumours in subsequent years.17 All three tumours were found to be sarcomas and may have been related to the field of radiation. However, 1/12 patients in our study had developed a second significant tumour. It is our belief that all adrenal cortical tumours in children require lifelong follow up despite previous recommendations to the contrary.12
CONCLUSION
Although rare, due to the varied presentation, adrenal cortical tumours should be included in the differential diagnosis for case of adrenal hormone excess. None of the criteria applied to adrenal tumours in adults can safely predict or exclude malig- nancy in children. Mutational analysis should be considered in any case of confirmed tumour but only with appropriate genetic counselling. Finally, ongoing, lifelong follow up is recom- mended in all patients.
ACKNOWLEDGEMENTS
We acknowledge the help and advice of Ronda Greaves and Shankar Kanumakala in the preparation of this manuscript.
REFERENCES
1 Wilkin F, Gagne N, Paquette J, Oligny LL, Deal C. Paediatric adrenocortical tumours: molecular events leading to insulin-like growth factor II gene overexpression. J. Clin. Endocrinol. Metab. 2000; 85: 2048-56.
2 Ciftci AO, Senocak ME, Tanyel FC, Buyukpamukcu N. Adreno- cortical tumours in children J. Paed. Surg. 2001; 36: 549-54.
3 Chudlee RM, Kay R. Adrenocortical carcinoma in children. Urol. Clinics N. Am. 1989; 16: 469-79.
4 Latronico AC, Chrousos GP. Adrenocortical tumours. J. Clin. Endocrinol. Metab. 1997; 82: 1317-24.
5 Ghizzoni L, Mastorakos G, Vottero A. Adrenal hyperandrogenism in children J. Clin. Endocrinol. Metab. 1999; 84: 4431-5.
6 Patil KK, Ransley PG, McCullagh M, Malone M, Spitz L. Functioning adrenal neoplasms in children. BJU Int. 2002; 89: 562-5.
7 Mendonca BB, Lucon AM, Menezes CA. Clinical, hormonal and pathological findings in a comparative study of adrenocortical neoplasms in childhood and adulthood. J. Urol. 1995; 154: 2004-9.
8 Migeon CJ, Lanes RL. Hypo- and hyperfunction of the adrenal cortex. In: Lifshitz F, ed. Paediatric Endocrinology, 3rd edn. New York, USA: Marcel Dekker, 1996; 336-43.
9 Wolthers OD, Cameron FJ, Scheimberg I. Androgen secreting adrenocortical tumours. Arch. Dis. Child 1999; 80: 46-50.
10 Wajchenberg BL, Albergaria-Pereira MA, Mendonca BB. Adrenocortical carcinoma. Clinical and laboratory observations. Cancer 2000; 88: 711-36.
11 Mantero F, Arnaldi G. Investigation protocol: adrenal enlarge- ment. Clin. Endocrinol. 1999; 50: 141-6.
12 Xiao XR, Ye LY, Shi LX, Cheng GF, Li YT, Zhou BM. Diagnosis and treatment of adrenal tumours. a review of 35 years experience. Br. J. Urol. 1998; 82: 199-205.
13 Saeger W. Histopathological classification of adrenal tumours. Eur. J. Clin. Invest. 2000; 30: 58-62.
14 Stojadinovic A, Ghossein RA, Hoos A. Adrenocortical carcinoma: Clinical, morphologic and molecular characterization. J. Clin. Oncol. 2002; 20: 941-50.
15 Marx C, Worlersdorfer GW, Brown JW, Scherbaum WA, Bornstein SR. MHC Class II expression-a new tool to assess dignity in adrenocortical tumours. J. Clin. Endocrinol. Metab. 1996; 81: 4488-91.
16 Kikuchi E, Yanaihara H, Nakashima J et al. Urinary Steroid Profile in Adrenocort Tumours. Biomed. Pharmacother. 2000; 54 (Suppl. 1): 194-7.
17 Driver CP, Birch J, Gough DCS, Bruce J. Adrenal cortical tumours in childhood. Ped. Haem. Onc. 1998; 15: 527-32.