Adrenocortical Tumors in Children
By Arbay O. Ciftci, Mehmet Emin Şenocak, F. Cahit Tanyel, and Nebil Büyükpamukçu Ankara, Turkey
Background/Purpose: Etiopathogenesis and management of pediatric adrenocortical tumors (ACTs) is still obscure because of the limited number of cases. The aim of this study is to present a clear picture of the entire spectrum of pediatric ACTs by reviewing one of the largest noncollected pediatric series treated in a single medical center.
Methods: Records of children treated for ACTs in our unit between 1970 and 1999, inclusive, were reviewed. Informa- tion recorded for each patient included age, sex, clinical characteristics, diagnostic methods, stage of disease, treat- ment, pathologic findings, and outcome. The patients were subdivided into 2 groups: group I, patients with adrenocor- tical carcinoma (ACC) and group II, patients with adrenocor- tical adenoma (ACA). These groups were analyzed with re- gard to parameters mentioned above.
Results: There were 30 children treated for ACTs in the study period with a mean age of 6.7 + 4.2 years (range, 2.5 to 13 years). Of these, 20 had ACC, and 10 had ACA. The tumors were right sided in 22 patients, left sided in 6 and bilateral in 2. Analysis of each group with regard to age and site of tumor showed no significant difference. Endocrine dysfunc- tion was noted in 83% of the patients and virilization was the most common presentation followed by Cushing’s syn- drome. The most striking difference between 2 groups was the prepondarance of virilization in group Il and Cushing’s syndrome in group I. In the latter, 14 patients presented with palpable abdominal mass and 3 patients with distant metas- tases. The mean time from initial symptoms to diagnosis was 8.1 ± 0.2 months, and this interval was similar in 2 groups, in functional and nonfunctional tumors, and in both sexes. Ultrasound scan, computerized tomography, mag- netic resonance imaging, intravenous pyelography, and an- giography were used for the diagnosis. All patients with ACA had localized disease, whereas 80% of the patients with ACC had regional or metastatic disease. Total excision was done
in all patients with ACA, but only in 13 patients with ACCs. Of the latter, 2 patients underwent ipsilateral nephrectomy, and 1 patient had right hepatic lobectomy plus nephrectomy. Adjuvant chemotherapy consisting of mitotane (n = 12), mitotane plus cisplatin and etoposide (n = 2) was com- menced. Seven patients with ACC had distant metastases postoperatively. The presence of regional disease at presen- tation was associated with a significantly shorter disease- free interval. All patients presenting with nonfunctional ACC (n = 4), functional ACC that have been totally resected (n = 4), and partially resected (n = 3) died of disease within the first 2.5 years after diagnosis. There was no significant dif- ference between the functional and nonfunctional ACCs with regard to survival rate. All patients who had distant metas- tases postoperatively and who had partial excision died. Of the surviving 9 patients with ACC, there are 6 known long- term survivors who are still alive.
Conclusions: ACAs are treated by total excision satisfactorily without any complication. For the time being, the most important aspect of therapy for ACCs is early diagnosis and total excision. Partial excision and advanced-stage disease are the major determinants of poor outcome. None of the clinical, laboratory, or pathologic features are reliable predic- tors for recurrence and discrimination of malignancy in ACTs. Because of the steadily increasing incidence of pre- cancerous genetic syndromes of adrenal glands and poor prognosis of ACCs, childhood patients of endocrine disor- ders should receive a detailed and vigorous diagnostic eval- uation and appropriate treatment as given to adults. Patients with ACTs should be entered into multi-institutional trials to adequately assess effective chemotherapy and radiotherapy protocols and molecular mechanisms of oncogenesis.
J Pediatr Surg 36:549-554. Copyright @ 2001 by W.B. Saunders Company.
INDEX WORDS: Adrenocortical, carcinoma, adenoma.
A LTHOUGH they are different entities, both adreno- cortical adenoma (ACA) and adrenocortical carci- noma (ACC) are evaluated together under the title of adrenocortical tumors (ACTs) in the literature.1-5 ACTs constitute less than 0.2% of all pediatric neoplasms and account for 6% of all adrenal tumors in children with an estimated incidence of 0.3 per 1 million population.1 They may occur in all age groups with 2 peaks occurring before 10 years of age and the fourth to fifth decade. ACTs are more common in girls and bilaterality has been reported in 2% to 10% of the cases.2,3 Contrary to the adrenal neuroblastomas, little is known about the etio- pathogenesis, clinical characteristics, and management
of pediatric ACTs because of the limited number of cases reported in the literature.3,4 Because of the rarity, many large series consist mostly of adult and few pediatric cases collected from the literature or gathered from
From the Department of Pediatric Surgery, Hacettepe University Medical Faculty, Ankara, Turkey.
Address reprint requests to Arbay O. Ciftci, MD, Associate Professor of Pediatric Surgery, Hacettepe University Medical Faculty, 06100, Ankara, Turkey.
Copyright @ 2001 by W.B. Saunders Company 0022-3468/01/3604-0003$35.00/0 doi:10.1053/jpsu.2001.22280
| Group I (ACC) | Group II (ACA) | Total | |
|---|---|---|---|
| Mean age (yr) | 6.3 ± 4.3 | 7.5 ± 4.9 | 6.7 ± 4.2 |
| Age range (yr) | 2.5-12.5 | 5-13 | 2.5-13 |
| Sex | |||
| Female | 11 | 8 | 19 |
| Male | 9 | 2 | 11 |
| Site of tumor | |||
| Right | 14 | 8 | 22 |
| Left | 4 | 2 | 6 |
| Bilateral | 2 | — | 2 |
several centers, and very few of these consist of both adenomas and carcinomas.5 Therefore, a retrospective clinical trial was performed to present a clear picture of the entire spectrum of pediatric ACTs based on our experience and a brief literature review. The current study consists of one of the largest noncollected pediatric ACT series treated in a single medical center.
MATERIALS AND METHODS
Records of children treated for ACTs in our unit between 1970 and 1999, inclusive, were reviewed retrospectively. Information recorded for each patient included age, sex, past medical and family history, clinical characteristics, diagnostic methods, stage of the disease, treat- ment, pathologic findings, and outcome. Extent of disease at presenta- tion was determined by the results of radiologic studies (ultrasonogra- phy [US], computerized tomography [CT], magnetic resonance imaging [MRI], and angiography) and was corroborated by the findings at surgery and on pathologic examination. Localized disease was defined as disease confined to one of the adrenal glands. Regional disease was defined as disease extending beyond the limits of the adrenal gland to the surrounding organs and tissues, including the regional lymph nodes. Metastatic disease was defined as disease in which there were distant metastases, whether regional disease was present or absent. In the patients presenting without metastases ini-
tially, the disease-free interval was defined as the time between curative surgery and a local recurrence or the occurrence of distant metastases. While analyzing the results, the patients were subdivided into 2 groups according to the diagnosis: group I, patients with ACC and group II, patients with ACA. These groups were analyzed with regard to param- eters mentioned above. Mann-Whitney U Wilcoxon Rank Sum and the Kruskal Wallis tests were used for statistical analysis. P value of less than .05 was considered significant. When appropriate, data were recorded as mean ± SD.
RESULTS
There were 30 children treated for ACTs in the study period. Of these, 20 had ACC and 10 had ACA. The mean age of the patients at presentation was 6.7 ± 4.2 years (range, 2.5 to 13 years) with a female to male ratio of 19:11. The tumors were right sided in 22 patients, left sided in 6, and bilateral in 2 patients. Analysis of each group with regard to age and site of tumor found no significant difference apart from sex ratio (Table 1). In group II, girls clearly predominated (8 girls, 2 boys), whereas group I consisted of 11 girls and 9 boys.
Predisposing factors were noted in group I. There was 1 patient who had congenital adrenal hyperplasia and 1 patient with horse-shoe kidney. A cousin of another patient had died of ACC. Clinical characteristics of our series are shown in Table 2. Endocrin dysfunction was noted in 83% (n = 25) of the patients. Virilization was the most common presentation, and the symptoms were precocious development of public and axillary hair, deepening of the voice, accelerated height, acne, and enlargement of the genitals. Hypertension, central obe- sity, buffalo hump, and moon face were the most com- mon symptoms of Cushing’s syndrome. One patient presented with the symptoms of hyperaldosteronism in- cluding hypertension, polyuria polydypsia, and hypoka-
| Feature | Group I | Group II | ||||||
|---|---|---|---|---|---|---|---|---|
| No | Stage | Treatment | Result | No | Stage | Treatment | Result | |
| Functional tumors | ||||||||
| Virilization | 6 | Localized (2) | TE (4) | Exitus:3 | ||||
| Regional (4) | TE + nephrectomy (1) | Cured:3 | 6 | Localized | TE | Cured | ||
| TE + nephrectomy + | ||||||||
| hepatic lobectomy (1) | ||||||||
| Cushing's syndrome | 6 | Localized (2) | TE (4) | Exitus:2 | 2 | Localized | TE | Cured |
| Regional (4) | NTE (1) | Cured:4 | ||||||
| TE + nephrectomy (1) | ||||||||
| Giant mass + Cushing's | ||||||||
| syndrome + distant metastases | 1 | Metastatic | PE (1) | Exitus: 1 | ||||
| Virilization + Cushing's syndrome | 3 | Regional | TE (2) | Exitus:1 | ||||
| NTE (1) | Cured:2 | |||||||
| Hyperaldosteronism | 1 | Localized | TE | Cured | ||||
| Nonfunctional tumors | ||||||||
| Giant mass | 2 | Regional | IB (1) | Exitus:2 | ||||
| PE (1) | ||||||||
| Giant mass + distant metastases | 2 | Metastatic | IB (2) | Exitus:2 | ||||
| Pain, anorexia, weight loss | 1 | Localized | TE | Cured | ||||
Abbreviations: TE, total excision; PE, partial excision; NTE, near total excision; IB, incisional biopsy.
lemia. The most striking difference between the 2 groups was the preponderance of virilization in group II and Cushing’s syndrome in group I. At presentation, abdom- inal mass was palpable in 14 patients, and distant me- tastases to the lung (n = 2) and lung and liver (n = 1) were recorded in 3 patients of group I.
The mean time from initial symptoms to diagnosis was 8.1 ± 0.2 months (range, 3 to 18 months). This interval was similar in both groups (7.9 months in group I and 8.3 months in group II), in functional and nonfunctional ACCs, and in the female and male patients. Elevated levels of urinary-free cortisol and several plasma and urinary steroids (17-ketosteroids) were found in patients with functional tumors. A positive correlation was found between the clinical stage and urinary levels of steroid metabolites. Advanced-stage tumors had the highest lev- els of urinary metabolites. Diagnostic procedures were ultrasound scan, computerized tomography, magnetic resonance imaging, intravenous pyelography, and an- giography (Figs 1 and 2).
Preoperative steroid replacement was done, and trans- abdominal approach was used in all patients. Laparot- omy findings showed that all patients with ACA had localized disease, whereas 80% (n = 16) of the patients with ACC had regional or metastatic disease. An unre- sectable giant tumor associated with widespread intraab- dominal disease was noted in 5 patients, and only inci- sional biopsies could be taken in 3 of them. Partial excision was done in the remaining 2 patients. Total excision was done in all patients with ACA, but only in 13 patients with ACC. Of the latter, 2 patients underwent ipsilateral nephrectomy, and 1 patient had right hepatic lobectomy plus nephrectomy. Partial (n = 2) and near- total (n = 2) excision were done in 4 patients (Table 2).
L
11200
+50
Adjacent invasion and intraperitoneal tumor seeding were found in all patients with metastatic disease or giant abdominal mass in whom total excision cannot be per- formed. Operative mortality was not seen in our series.
Macroscopically, a wet weight of more than 300 g (range, 300 to 1,500 g), a grossly lobulated cut surface, presence of necrotic areas, calcifications, and hemor- rhages were noticed in 16 malignant tumors. Only 4 malignant tumors mimicked adenoma macroscopically. Adenomas were less than 250 g (range, 70 to 250 g), encapsulated, and well circumscribed with homogeneous texture and color. Two adenomas were thought to be malignant with regard to macroscopic appearance be- cause of necrotic areas and lobulated cut surface with heterogeneous texture. Vascular and capsular invasion, frequent mitoses, nuclear atypia, and hyperchromasia were the most common suggestive features of malig- nancy. Nuclear pleomorphism and broad fibrous bands, although limited to malignant tumors also were found in 2 adenomas, which were thought initially to be malig- nant.
Patients with ACA were treated successfully by total excision without any concomitant therapy. Adjuvant che- motherapy consisting of mitotane (n = 12), mitotane plus cisplatin, and etoposide (n = 2) was commenced postoperatively in group ACC. In addition to the meta- static disease at presentation (n = 3), 7 patients (35%) had metastases consisting of lung and liver (n = 3), liver and lymph nodes (n = 2), and lung and bone (n = 2) during mitotane therapy within 2 years’ time. The initial surgical procedures performed in these patients were total excision (n = 3), partial excision (n = 2), near total excision (n = 1), and incisional biopsy (n = 1). Second- look surgery for resection of the liver and lung metasta- ses were done in 3 patients successfully. Disease-free interval was 9 + 1.3 months for ACC patients. The presence of regional disease at presentation was associ-
ated with a significantly shorter disease-free interval (3.4 ± 0.3 months, P < . 05). In group I, all patients presenting with nonfunctional tumors (n = 4), functional tumors that have been totally resected (n = 4), and partially resected tumors (n = 3) died of disease within the first 2.5 years after diagnosis. There was no signifi- cant difference between the functional and nonfunctional ACCs with regard to survival figures. All patients who had distant metastases postoperatively and who had par- tial excision died (55%). The surviving 9 patients (45%) had functional ACCs either as localized (n = 4) or regional (n = 5) disease, but all of these had total excision. Of these, there are 6 known long-term survi- vors (30%) who are still alive and had been on mitotane therapy. The remaining 3 patients were lost to follow-up 4 years after the operation.
DISCUSSION
The current study confirms literature findings with regard to epidemiologic data such as ACTs are more common in girls and most occur before the age of 10. Although some reports have indicated predominant site to be the left adrenal gland,1,6 we have found a right- sided preponderance. Various p53 gene mutations and p-arm defects of chromosome 11 have been identified in ACTs. Li Fraumeni and Beckwith-Wiedeman syndromes are the common anomalies associated with ACTs.7,8 Congenital anomalies of the kidney and congenital ad- renal hyperplasia are known to increase the risk of ACTs as noted in our 2 patients. Activation of various proto- oncogenes, inactivation of tumor suppressor genes, inhi- bition of senescence or apoptosis, and changes in adre- nocortical tissue specific factors have been proposed as potential mechanisms for adrenocortical tumorigenesis.8
ACTs can be hormonally silent (nonfunctional) or hormone secreting (functional). Functional tumors pro- duce symptoms as a result of overproduction of cortical hormones, and most of the children present with endo- crine symptoms in contrast to adults, who usually present with nonfunctional tumors.9 Virilization was the most common presentation in our series followed by Cush- ing’s syndrome. Although this was in agreement with the literature findings,10 when we analyzed the ACC patients alone, we found that number of patients presenting with Cushing’s syndrome was equal to patients with viriliza- tion, which was in contrary to the literature findings.5,14 This fact showed that patients with benign ACTs were less likely to present with Cushing’s syndrome. It is extremely rare to encounter a patient with adrenal tumor who shows primary aldosteronism based solely on ex- cessive aldosterone production or feminization caused by estrogen. However, in our series, a 13 year-old-boy presenting with the signs and symptoms of hyperaldo- steronism was found to have adrenal adenoma. Our experience showed that diagnostic and therapeutic strat-
egies did not differ in these cases. The percentage of nonfunctional tumors was 20% in ACC group, 10% in ACA group, and 17% in all series, which has been one of the lowest published figures.5 In the literature, the mean time from initial symptoms to diagnosis ranges from 6 to 14 months, which covers our interval of 8.1 ± 0.2 months.11-13 This interval did not differ between group I and II or between functional and nonfunctional tumors, although literature findings emphasize that functional tumors are diagnosed earlier12,14 than nonfunctional tu- mors, which usually present with distant metastases, inoperable abdominal masses, and pain as noted in our series. This conflict is most probably caused by misin- terpretation of the clinical picture of nonfunctional tu- mors by the parents of our patients. There is no evidence that nonfunctional tumors have a higher growth rate that results in advanced-stage tumors in a shorter period.
Intravenous pyelography and angiography have been replaced by the widespread use of US, CT, and MRI in the last decade. New techniques provided detailed infor- mation about size, homogenity, presence of calcifica- tions, areas of necrosis, extent of local invasion, and status of vascular structures.15 Achieving all this detailed preoperative information has diminished the need to explore the abdomen thoroughly by big incisions but has not improved the survival rates in our series. We have not used MRI characteristic to distinguish among pri- mary ACC, nonfunctional ACA and pheochromocyto- mas by comparing the ratio of signal intensity of each mass to that of liver.16 In spite of advantages, we noted that the requirement for sedation or anesthesia made MRI and CT imaging of children difficult. Iodocholes- terol scanning, venography, and arteriography currently are reserved for the circumstances in which CT or MRI cannot supply the information needed.
Certain aspects of the treatment must be emphasized to prevent morbidity and enhance the chances for satis- factory results. First of all, steroid replacement in ac- cordance with adrenocortical function monitorization should be done preoperatively and postoperatively be- cause previous low survival figures of ACTs in part reflects inaccurate management of steroid balance.5,6,8,13 Preoperative assessment of renal function also should be obtained in case nephrectomy is required as noted in our 2 patients. Radical excision with en bloc resection of any local invasion is the best treatment approach in ACTs. We have used the transabdominal approach rather than retroperitoneal or posterior approaches not only for bet- ter exposure, but this offers the best opportunity to evaluate the contralateral adrenal gland, as well as to assess the extent of metastatic disease within the abdo- men. Complete resection of the tumor totally cures ACA patients; however, 30% to 50% of patients with ACC have local or distant recurrences in spite of an apparently curative surgery as observed in our series.17 We believe
that patients with solitary metastases or a locally recur- rent tumor with a long disease-free interval also should be considered for surgical resection. Lung, liver, lymph nodes and gastrointestinal tract were the major sites of metastases in our series, in descending order of fre- quency as mentioned in the literature.1,3,4,8,10 Although we have not performed in every case, retroperitoneal node dissection is recommended for therapeutic and prognostic considerations.13 Our experience showed that frequent imaging studies and endocrine evaluation was of utmost importance to detect recurrence in the long- term follow-up of these patients.
Although we have not noticed a significant effect of mitotane on survival rate in our series, in view of the literature findings, which shows an objective response to mitotane varying from 34% to 61%,8,13,18,19 we think that mitotane can be used as adjuvant therapy. Many other drugs with an adrenolytic effect, a combination of mito- tane or different cytotoxic agents, and radiotherapy also have not improved survival rates, although some studies of complete remission were reported.17-20
Malignant ACTs are defined clinically as producing metastases or resulting in death from direct effects of tumor.9 Although weight more than 150 g, gross and microscopic necrosis, calcifications, broad fibrous bands, vascular and capsular invasion, mitoses, and aneuploidy are more commonly associated with malignant ACTs, none is widely accepted and universally applied.10,13 Pediatric benign ACTs were found to have more mitoses, necrosis, broad fibrous bands, and moderate to severe pleomorphism than were adult benign tumors, and tumor weight more than 500 g was proposed to be the most important predictor of malignancy.9 However, several fatal pediatric cases with smaller tumor weight also had been reported.21 In another pediatric series, histologic type and tumor weight were found to be statistically significant indicators of outcome. Ploidy and prolifera- tive indexes were found to have no effect on prognosis of pediatric ACTs.22 Our experience and literature findings showed that there is no sufficient data to say that the criteria for malignancy should be different between chil- dren and adults.
Various investigators such as Hough, Weiss, and Sloo- ten developed histologic and nonhistologic indexes of malignancy in ACTs, but none of these parameters had a statistically significant relationship to survival rate or tumor behavior when presence or absence of pediatric
metastases at presentation were taken into account.23 According to our results, atypical mitoses, necrosis, cap- sular and vascular invasion, and big size tumors had a greater propensity to exhibit malignancy than stated in previous reports,9-13 but the clinical behavior was the best determinant of malignancy. We have not encoun- tered any second malignant tumor associated with ACTs in our series, although a high incidence of 24% is reported in the literature.13
Statistically significant prolonged survival rate was reported in girls, localized tumors, patients who under- went extirpative procedure, and those who had a disease- free interval longer than 12 months.13 In their extensive review, Latranico and Chrousos8 reported that patients with regional and metastatic disease had a very poor prognosis, and in spite of aggressive surgical therapy, the 5-year survival rate was less than 30%. In his excellent review, Sawin24 reported the long-term survival rate of children with ACCs between 10% and 46%. Complete resection was associated with an improved survival rate (67%) despite a 38% local recurrence rate. There were no survivors after partial resection, even when adjuvant chemotherapy or radiotherapy was used. Mortality and distant metastases rates of current series are within the limits mentioned in the literature.
Our experience showed that ACAs are treated by total excision satisfactorily without any complication. For the time being, the most important aspect of therapy for ACC is early diagnosis and total excision. The duration of symptoms before diagnosis and resection was in- versely related to survival rate in ACCs. A statistically significant difference in survival rate was not found between functional and nonfunctional ACCs in keeping with the experience of others.1,2,4,8,19 Younger age was not found to be correlated with good prognosis contrary to previous reports.3 None of the patients with partial excision recovered as reported in the literature.8,24 Partial excision and advanced-stage disease were the most de- terminants of poor outcome. Because of the steadily increasing incidence of precancerous genetic syndromes of adrenal glands and poor prognosis of ACCs, child- hood cases of endocrine disorders should receive a de- tailed and vigorous diagnostic evaluation and appropriate treatment as given to adults. Patients with ACTs should be entered into multi-institutional trials to assess ade- quately effective chemotherapy and radiotherapy proto- cols and molecular mechanisms of oncogenesis.17
REFERENCES
1. Chudler RM, Kay R: Adrenocortical carcinoma in children. Urol Clin North Am 16:469-479, 1989
2. Prommier RF, Brennan MF: An eleven-year experience with adrenocortical carcinoma. Surgery 112:963-971, 1992
3. Sabbage CC, Avilla SG, Schultz C, et al: Adrenocortical carci-
noma in children: Clinical aspects and prognosis. J Pediatr Surg 28:841-843, 1993
4. Neblett WW, Frexes-Steed M, Scott HW: Experience with adre- nocortical neoplasm in childhood. Am Surg 53:117-125, 1987
5. Bertagna C, Orth DN: Clinical and laboratory findings and results
of therapy in 58 patients with adrenocortical tumors admitted to a single medical center (1951 to 1978). Am J Med 71:855-875, 1981
6. Luton JP, Cerdas S, Billaud L, et al: Clinical features of adreno- cortical carcinoma, prognostic factors, and the effect of mitotane therapy. N Engl J Med 322:1195-1201, 1990
7. Ciftci AO, Salman AB, Tanyel FC, et al: Bilateral multiple adrenal pseudocysts associated with incomplete Beckwith-Wiedemann Syndrome. J Pediatr Surg 32:1388-1390, 1997
8. Latranico AC, Chrousos GP: Extensive experience: Adrenocorti- cal tumors. J Clin Endocrinol Metabol 82:1317-1324, 1997
9. Cagle PJ, Hough AJ, Pysher TJ, et al: Comparison of adrenal cortical tumors in children and adults. Cancer 57:2235-2237, 1986
10. Bergada I, Venara M, Maglio S, et al: Functional adrenal cortical tumors in pediatric patients. A clinicopathologic and immuno- histochemical study of a long term follow-up series. Cancer 77:771- 777, 1996
11. Mendonca BB, Lucon AM, Menezes CAV, et al: Clinical, hormonal and pathological findings in a comparative study of adrenal cortical neoplasms in childhood and adulthood. J Urol 154:2004-2009, 1995
12. Nader S, Hickey RC, Sellin RV, et al: Adrenal cortical carci- noma: A study of 77 cases. Cancer 52:707-711, 1983
13. Didolkar MS, Bescher RA, Elias EG, et al: Natural history of adrenal cortical carcinoma: A clinicopathologic study of 42 patients. Cancer 47:2153-2161, 1981
14. Lipsett MB, Herta R, Ross GT: Clinical and pathophysiologic aspects of adrenocortical carcinoma. Am J Med 35:374-383, 1963
15. Petrus LV, Hall TR, Boachat MI, et al: The pediatric patient with suspected adrenal neoplasm: Which radiological test to use? Med Pediatr Oncol 20:53-57, 1992
16. Shapiro B, Fig LM, Gross MD, et al: Contributions of nuclear endocrinology to the diagnosis of adrenal tumors. Recent Results Cancer Res 118:113-138, 1990
17. Johnson DH, Greco FA: Treatment of metastatic adrenal cortical carcinoma with cisplatin and etoposide. Cancer 58:2198-2202, 1986
18. Percarpio B, Knowlton AH: Radiation therapy of adrenal corti- cal carcinoma. Acta Radiol 15:288-292, 1976
19. Karakousis CP, Uribe J, Moore R: Adrenal adenocarcinomas: Diagnosis and management. J Surg Oncol 16:385-389, 1981
20. Flack MR, Pyle RG, Mullen NM, et al: Oral gossypol in the treatment of mesenteric adrenal cancer. J Clin Endocrinol Metabol 76:1019-1024, 1993
21. Ribeiro RC, Neto RS, Schell MJ, et al: Adrenocortical carci- noma in children: A study of 40 cases. J Clin Oncol 8:67-74, 1990
22. Bugg MF, Ribeiro RC, Roberson PK, et al: Correlation of pathologic features with clinical outcome in pediatric adrenocortical neoplasia. A study of Brazilian population. Am J Clin Pathol 101:625- 629, 1994
23. Evans HL, Vassilopoulou-Sellin R: Adrenal cortical neoplasms. A study of 56 cases. Anat Pathol 105:76-86, 1996
24. Sawin RS: Functioning adrenal neoplasms. Semin Pediatr Surg 6:156-163, 1997