THE IMPACT OF TUMOR STAGE ON PROGNOSIS IN CHILDREN WITH ADRENOCORTICAL CARCINOMA
SILVIO TUCCI, JR., ANTONIO C. P. MARTINS,* HAYLTON J. SUAID, ADAUTO J. COLOGNA AND RODOLFO B. REIS
From the Division of Urology, Department of Surgery, Medical School of Ribeirao Preto, University of Sao Paulo, Sao Paulo, Brazil
ABSTRACT
Purpose: We evaluated treatment outcomes in children with adrenocortical carcinoma.
Materials and Methods: We studied 34 children with a median age of 3 years. In 27 of 28 patients without intracaval extension complete surgical excision was accomplished, while tumor resection combined with thrombectomy was carried out in 5 of 6 children with vascular invasion. In 2 children with cavoatrial thrombus the thrombectomy required cardiopulmonary bypass with deep hypothermia and circulatory arrest. Children with incomplete excision of the tumor and/or stage IV disease received adjuvant chemotherapy.
Results: Ultrasonography, computerized tomography and magnetic resonance imaging exhib- ited specificity of 100% in the diagnosis of vascular invasion, and sensitivity of 50%, 66% and 100%, respectively. Patient age, tumor stage or size and vascular invasion were associated with survival in univariate analysis. Tumor stage was the only independent factor associated with survival in multivariate analysis. The overall 5-year survival rates according to tumor stage were 100% in stage I, 85% in stage II, 40% in stage III and 0% in stage IV. Of 11 children with local recurrence only 2 were alive without disease at 96 and 204 months after reoperation with complete tumor excision. Only 2 of 6 patients with vascular invasion were disease-free at 17 and 50 months. A total of 10 children with stage IV disease treated with chemotherapy died within a median of 6 months.
Conclusions: Tumor stage was the most relevant prognostic factor for children with adreno- cortical carcinoma. Reoperation for local tumor recurrence and thrombectomy for inferior vena caval tumor invasion should be attempted whenever possible.
KEY WORDS: adrenal gland neoplasms; adrenocortical carcinoma; thrombosis; vena cava, inferior
Adrenocortical carcinoma (ACC) is rare, especially in chil- dren. The prevalence of inferior vena cava (IVC) invasion in adults varies from 6% to 30%, while in children the rate is 10%.1-3 Complete surgical excision of the tumor is the only potentially curative treatment but data on outcomes are scarce.1,2,4 IVC tumor extension is reportedly the most rele- vant prognostic factor in children.3 In contrast, tumor ex- tending into IVC has also been described as a feature of no significance with regard to outcome.5 The aim of this study was to analyze outcomes in children with ACC with and without extension into the IVC.
MATERIALS AND METHODS
We retrospectively reviewed the records of 34 children and adolescents (13 males and 21 females) with ACC who were treated surgically at our institution between January 1975 and December 2003. Median patient age was 3 years (range 1 to 17, mean + SD 4.8 + 4.3). The tumor arose from the left gland in 14 patients and the right in 20. Inclusion criteria were adequate pathological data and a followup of 1 year or more, except for those who died of the disease. Only 1 patient was excluded from the study because he was lost to followup 2 months postoperatively with no evidence of disease.
The diagnosis of ACC was based on the histological criteria of Weiss et al (Appendix 1).6 The disease was staged accord- ing to the system of Henley et al (Appendix 2).7 Tumors were also classified into categories according to the level of ceph-
alad extension of tumor thrombus, ie level I-tumor throm- bus extended from the renal vein into the infrahepatic seg- ment of the IVC less than 2 cm from the renal vein, level II-tumor thrombus extended to below the hepatic vein, level III-tumor thrombus extended into the suprahepatic seg- ment of the IVC but not into the atrium and level IV-tumor thrombus extended into the right atrium.8-10
All patients underwent appropriate clinical and hormonal investigation. An 11-year-old female had the only nonfunc- tional ACC (tumor weight 2,350 gm, diameter 25 cm). Chil- dren with functional tumors exhibited virilization (19), Cushing’s (4) and mixed syndromes (10).
Ultrasonography and computerized tomography (CT) were performed in all patients. Magnetic resonance imaging (MRI) was indicated to differentiate ACC from other pathology in the upper abdomen (2 cases), to depict more precisely vascu- lar invasion detected by ultrasound and/or CT (4) and to demonstrate extension into surrounding structures (8). MRI revealed vascular extension in 1 case (level I) that was not diagnosed by ultrasound or CT. One additional child with a small tumor (3 cm) did not undergo MRI, and the diagnosis of tumor thrombus (level I) was established intraoperatively.
All 28 patients without IVC tumor extension underwent laparotomy culminating in total adrenalectomy in 12, adre- nalectomy plus lymphadenectomy in 9 and en bloc adrenal- ectomy, lymphadenectomy and nephrectomy in 6. In 1 pa- tient the tumor was unresectable.
In 2 of the 6 patients with IVC invasion adrenalectomy and longitudinal cavotomy plus thrombectomy were performed via laparotomy. In the child in whom IVC invasion was only detected intraoperatively the surgery was performed by video
* Correspondence: Hospital das Clínicas de Ribeirão Preto, Univer- sity of São Paulo, Av. Bandeirantes # 3900, Ribeirão Preto, São Paulo, Brazil 14048-900 (e-mail: arpmartins@convex.com.br).
laparoscopy because he had been misdiagnosed with adrenal adenoma. The remaining 3 patients with cavoatrial throm- bus underwent exploration via laparotomy to verify the pos- sibility of surgical resection of the primary tumor. Two of these patients subsequently underwent thoracotomy and car- diopulmonary bypass with deep hypothermic (15C) circula- tory arrest (durations 40 and 45 minutes) with exsanguina- tion for complete thrombectomy. In 1 patient the tumor was unresectable.
Patients with primary tumor unable to undergo com- plete resection, or stage IV disease, IVC extension or local relapse (19) received adjuvant chemotherapy consisting of repeated cycles of a combination of several drugs, includ- ing mitotane, doxorubicin, cyclophosphamide, vincristine and 5-fluorouracil. External beam radiotherapy at doses of 25 to 40 Gy delivered for 2 to 4 weeks was indicated for pain palliation in 3 children with bone metastases.
Continuous variables were analyzed using the unpaired t test. Survival curves were calculated using the Kaplan-Meier method. Comparisons between groups of patients were made using log rank univariate analysis and Cox regression mul- tivariate analysis. All analyses were conducted using Stata® 6.0 statistical software. P values less than 0.05 were consid- ered statistically significant.
RESULTS
Detection rates for adrenal mass were 82% for ultrasound (this method overlooked 6 small ACCs) and 100% for CT. Ultrasonography, CT and MRI exhibited specificity of 100% for the diagnosis of IVC invasion, and sensitivity of 50%, 66% and 100%, respectively.
Patient age, tumor stage and size, and vascular invasion were associated with survival in univariate analysis (see figure). The results of multivariate analysis are shown in table 1. Survival rates according to several parameters are listed in table 2.
Repeat CT revealed a partial response in the liver metas- tasis in 2 of 10 children with stage IV disease treated with chemotherapy. However, all of these patients died within a median of 6 months. Radiotherapy was effective for pain palliation in 3 of 3 children.
Of 32 children who underwent complete excision of the primary lesion 11 (34%) had locoregional recurrence (table 3).
Three of these patients underwent reoperation. One died during surgery due to anesthetic complications, and 2 were alive without disease at 96 and 204 months of followup. Eight children with local relapse deemed unsuitable for surgery received chemotherapy and died of disease within a median of 7 months.
Intracaval tumor invasion was independent of tumor size
| TABLE 1. Influence of risk factors on survival | |||
|---|---|---|---|
| Factors | p Value | Hazard Ratio | 95% CI |
| Tumor stage | 0.02 | 8.8 | 1.6-47.1 |
| Pt age | 0.06 | 5.5 | 1.3-23.8 |
| Tumor IVC invasion | 0.07 | 5.1 | 0.9-27.1 |
| Tumor size | 0.43 | 1.8 | 0.4-9.5 |
| TABLE 2. Survival according to tumor stage, local recurrence and vascular invasion | ||
|---|---|---|
| Parameters | No. Pts/Total No. (%) | Followup (mos) |
| Stage: | ||
| I | 5/5 (100) | 60 |
| II | 12/14 (86) | 60 |
| III | 2/5 (40) | 60 |
| IV | 0/10 (0) | 60 |
| Local relapse | 2/11 (18) | 96 |
| IVC invasion | 2/6 (33) | 40 |
| All pts | 20/34 (59) | 60 |
(table 4). The only major complication after thrombectomy occurred in 1 patient with cardiopulmonary bypass, in whom a large retroperitoneal collection developed that required open surgical drainage. Only 2 of 6 patients with IVC tumor thrombus were alive and disease-free at 17 and 50 months of followup (table 5).
DISCUSSION
Ultrasonography is a generally accepted screening tool for adrenal neoplasms. However, it lacks sufficient sensitivity to
AGE
SIZE
100
100
Percent Survival
N=20
<5 years
Percent Survival
N=21
<9cm
N=13
50
50
N=14
≥5 years
≥9cm
Logranktest p=0.0002; HR=0.13
Logrank test p=0.0001; HR=0.12
0
0
0
100
200
300
0
100
200
300
Months
Months
STAGE
100
1 - N = 5
100
TUMOR THROMBUS
Percent Survival
II - N = 14
Percent Survival
Without IVC Thro mbus
50
III - N = 5
50-
N = 28
With IVC Thro mbus
IV - N = 10
N = 6
Logranktest: p<0.0001
Logranktest p = 0.06; HR = 2.7
0
100
200
300
0
0
0
100
200
300
Months
Months
| Tumor Recurrence | No. Pts | Tumor Size (cm) | ||
|---|---|---|---|---|
| Mean + SD | Median (range) | |||
| Yes | 11 | 13.7 ±5.3 | 14.0 (5.5-25.0) | |
| No p Value (95% CI of mean difference)* | 21 | 6.5 ± 3.0 | 6.3 (2.0-12.0) | |
| 0.001 (3.4-10.9) | ||||
* Unpaired t test with Welch correction.
| TABLE 4. IVC tumor involvement versus primary tumor weight and size | |||||
|---|---|---|---|---|---|
| IVC Tumor Involvement | No. Pts | Primary Tumor Size (cm) | Primary Tumor Wt (gm) | ||
| Mean ± SD | Median (range) | Mean ± SD | Median (range) | ||
| Yes | 6 | 8.6 ± 4.6 | 9.4 (3-14.5) | 298 ± 300 | 290 (11-828) |
| No | 28 | 9.1 ± 5.1 | 8.0 (2-25) | 343 ± 489 | 215 (5-2,350) |
| 2-Sided p values (95% CI)* | 0.79 (-4.07-5.26) | 0.92 (-381.1-471.2) | |||
* Unpaired t test.
| TABLE 5. Characteristics of patients with intracaval tumor thrombus, type of surgery performed and outcomes | ||||||
|---|---|---|---|---|---|---|
| Pt No .- Sex-Age (yrs) | Side | Tumor Size (cm) | Stage | Surgery | Outcome | |
| Followup (mos) | Status | |||||
| 1-F-5 | Rt | 8.6 | II | Adrenalectomy via laparotomy + thrombectomy via cavotomy | 7 | Dead |
| 2-M-1 | Rt | 3.0 | I | Video laparoscopic adrenalectomy + thrombectomy via cavotomy | 50 | Alive |
| 3-M-2 | Rt | 14.5 | IV | Adrenalectomy via laparotomy + ipsilateral nephrectomy + cavoatrial thrombectomy/cardiopulmonary bypass/deep hypothermic circulatory arrest | 7 | Dead |
| 4-F-1 | Lt | 3.3 | I | Adrenalectomy via laparotomy + ipsilateral nephrectomy + thrombectomy via cavotomy | 17 | Alive |
| 5-F-7 | Rt | 10.2 | IV | Unresectable | 5 | Dead |
| 6-F-5 | Rt | 12.0 | IV | Adrenalectomy via laparotomy + cavoatrial thrombectomy/ cardiopulmonary bypass/deep hypothermic circulatory arrest | 12 | Dead |
detect small adrenal masses.11 The ultrasound detection rate of 85% of ACCs compared to the CT detection rate of 100% as ascertained in our study confirms this observation. Among high resolution imaging techniques CT is a first line modality for evaluation of adrenal masses, although the origin of the mass sometimes is difficult to discern.1,11 The most sensitive method of diagnostic imaging is MRI because it allows better assessment of tumor thrombosis in blood vessels and can visualize primary ACC, secondary metastatic tumors, adeno- mas and pheochromocytomas.1,11, 12 Our results also under- line the observation that MRI has better sensitivity in the diagnosis of vascular invasion in children.
In a review of 602 primarily adult cases the 5-year survival rates for stages I to IV ACC were 30% to 45%, 12.5% to 57%, 5% to 18% and 0%, respectively.1 The survival rate for each stage observed in our sample seems higher than that re- ported for adults, except for stage IV disease. Such a differ- ence has been noted previously, and some investigators sug- gest that ACCs in children are clinically and histologically different from those in adults.1,3 However, it is not clear whether we need distinct histological criteria to distinguish benign from malignant lesions in children and adults.
The system of Weiss et al6 is 1 of at least 3 sets of histo- logical criteria considered useful in the diagnosis of malig- nancy. Each system has unique features, and in some cases one system may predict benign behavior, whereas another system may suggest a malignant outcome.1,6 Wieneke et al proposed a list of macroscopic and microscopic criteria of malignancy for adrenal tumors in children but they were not able to establish a clean break point for distinction between benign and malignant disease.3 They analyzed 74 children
with ACC and concluded that IVC invasion and increased mitotic activity independently increase the malignant out- come, while other factors such as patient age, tumor size and extension into periadrenal fat and/or organs are less rele- vant. However, in their study tumor stage was not assessed. Our data suggest that tumor stage is the most relevant factor in predicting outcome of ACC in children.
Reoperation for recurrent disease should be taken into account whenever possible because without further surgical treatment the prognosis is poor, as observed in our study and others.13 The locoregional recurrence rate of 34% found in our setting is in accordance with findings published by oth- ers.13 The high rate of recurrent disease emphasizes the need for careful followup.
Mitotane has been used as the mainstay of chemotherapy for ACC in many series. In some studies mitotane adminis- tration has yielded a transient benefit in controlling endo- crine symptoms but no improvement in survival,1 while in others the results have been more encouraging.14 Several alternative drugs, including etoposide, 5-fluorouracil, cispla- tin, vincristine, suramin and doxorubicin, have been used with varying results.1 The proportion of partial objective response and the median survival time observed in our pa- tients treated with chemotherapy confirm previous reports and indicate that the efficacy of adjuvant chemotherapy de- serves further evaluation.15,16 The efficacy of radiotherapy for pain palliation in our patients is in agreement with other reports, and this method should be used whenever neces- sary.1,15
The prevalence of vascular invasion in 17% of our patients is within the range published for adults.1,2 Since no associ-
| TABLE 6. Outcomes of patients with ACC extending into IVC | ||||
|---|---|---|---|---|
| References | Tumor Stage | No. Pts | Followup (mos) | No. Surviving |
| Present series | I | 2 | 17, 12 | 2 |
| Hedican and Marshall4 | I/II | 12 | 0,* 3,1 5, 5, 8,1 10,1 18,1 21,1 22, 24, 60, 80 | 6 |
| Ortiz Gorraiz et al17 | I/II | 1 | 2+ | 0 |
| Present series | II | 1 | 2+ | 0 |
| Figueroa et al10 | II | 2 | 60, 72 | 2 |
| Alaoui et al18 | II | 1 | 48 | 1 |
| Nakanoma et al19 | II | 1 | 4+ | 0 |
| Mingoli et al2 | II | 1 | 7+ | 0 |
| Lucon et al20 | III | 5 | 5,1 12,1 15,1 15,+ 180 | 1 |
| Stewart et al9 | III | 1 | 15+ | 0 |
| Present series | IV | 2 | 6,1 7+ | 0 |
| Ekici and Ciancio8 | IV | 1 | 18 | 1 |
| Figueroa et al10 | IV | 2 | 12, 60+ | 1 |
| Hedican and Marshall4 | III/IV | 15± | 2,1 5,1 10, 12, 12, 12, 14,+ 15, 17, 36,+ 36 | 7 |
| Stewart et al9 | III/IV | 2 | 17,1 29 | 1 |
In some reports it was not possible to distinguish precisely thrombus level I from II, or III from IV.
* Patient died perioperatively.
7 Patient died of disease.
# Four patients in this subpopulation died perioperatively.
ation was found between IVC invasion and tumor size, as reported elsewhere,5,10 perhaps even in small tumors vascu- lar invasion should be investigated preoperatively.
The actual survival rate and the influence of ACC throm- bus level on outcome are unknown due to the paucity of data. A review of the literature demonstrates disease-free survival rates of 39% (11 of 28 patients with stage III/IV, median followup 15 months) and 48% (9 of 21 patients with stage I/II, median followup 24 months) in 49 patients undergoing re- section of primary ACC plus thrombectomy (table 6).
CONCLUSIONS
Tumor stage is the most relevant prognostic factor for ACC in children. Reoperation for local recurrence and thrombec- tomy for IVC tumor invasion should be tried whenever sur- gical excision is deemed suitable.
| Criteria | Features |
|---|---|
| 1 | Nuclear grade III or IV |
| 2 | Mitotic rate greater than 5 per 50 high power field |
| 3 | Atypical mitotic figures |
| 4 | Clear cells constituting 25% of tumor or less |
| 5 | Diffuse architecture |
| 6 | Tumor cell necrosis |
| 7 | Invasion of venous structures |
| 8 | Invasion of sinusoidal structures |
| 9 | Invasion of tumor capsule |
At least 3 features had to be present for a definitive diagnosis of adreno- cortical carcinoma, based on the criteria of Weiss et al.6
| Criteria: | |
|---|---|
| T1 | Primary tumor smaller than 5 cm, no local invasion |
| T2 | Primary tumor 5 cm or greater, no local invasion |
| T3 | Primary tumor of any size, locally invading to but not involving adjacent organs |
| T4 | Primary tumor of any size, locally invading adjacent organs |
| N0 | No regional positive nodes |
| N1 | Positive regional nodes |
| M0 | No distant metastasis |
| M1 | Distant metastasis present |
| Stages: | |
| I | T1N0M0 |
| II | T2N0M0 |
| III | T1-2N1M0 or T3N0M0 |
| IV | T4N0M0 or T3N1M0 or T1-4N0-1M1 |
REFERENCES
1. Ng, L. and Libertino, J. M .: Adrenocortical carcinoma: diagnosis, evaluation and treatment. J Urol, 169: 5, 2003
2. Mingoli, A., Nardacchione, F., Sgarzini, G., Marzano, M., Ciccarone, F. and Modini, C .: Inferior vena cava involvement by a left side adrenocortical carcinoma: operative and prognos- tic considerations. Anticancer Res, 16: 3197, 1996
3. Wieneke, J. A., Thompson, L. D. and Heffess, C. S .: Adrenal cortical neoplasms in the pediatric population: a clinicopatho- logic and immunophenotypic analysis of 83 patients. Am J Surg Pathol, 27: 867, 2003
4. Hedican, S. P. and Marshall, F. F .: Adrenocortical carcinoma with intracaval invasion. J Urol, 158: 2056, 1997
5. Lee, J. E., Berger, D. H., el-Naggar, A. K., Hickey, R. C., Vassilopoulou-Sellin, R., Gagel, R. F. et al: Surgical manage- ment, DNA content, and patient survival in adrenal cortical carcinoma. Surgery, 118: 1090, 1995
6. Weiss, L. M., Medeiros, L. J. and Vickery, A. L., Jr .: Pathologic features of prognostic significance in adrenocortical carci- noma. Am J Surg Pathol, 13: 202, 1989
7. Henley, D. J., Van Heerden, J. A., Grant, C. S., Carney, J. A. and Carpenter, P. C .: Adrenal cortical carcinoma-a continuing challenge. Surgery, 94: 926, 1983
8. Ekici, S. and Ciancio, G .: Surgical management of large adrenal masses with or without thrombus extending into the inferior vena cava. J Urol, 172: 2340, 2004
9. Stewart, J. R., Carey, J. A., McDougal, W. S., Merrill, W. H., Koch, M. O. and Bender, H. W., Jr .: Cavoatrial tumor throm- bectomy using cardiopulmonary bypass without circulatory arrest. Ann Thorac Surg, 51: 717, 1991
10. Figueroa, A. J., Stein, J. P., Lieskovsky, G. and Skinner, D. G .: Adrenal cortical carcinoma associated with venous tumor thrombus extension. Br J Urol, 80: 397, 1997
11. Frilling, A., Tecklenborg, K., Weber, F., Kuhl, H., Muller, S., Stamatis, G. et al: Importance of adrenal incidentaloma in patients with a history of malignancy. Surgery, 136: 1289, 2004
12. Bonfig, W., Bittmann, I., Bechtold, S., Kammer, B., Noelle, V., Arleth, S. et al: Virilizing adrenocortical tumours in children. Eur J Pediatr, 162: 623, 2003
13. Meyer, A., Niemann, U. and Behrend, M .: Experience with the surgical treatment of adrenal cortical carcinoma. Eur J Surg Oncol, 30: 444, 2004
14. Nussey, F., Seckl, J. and Smyth, J .: Adrenal carcinoma treated with chemotherapy can result in prolonged survival. Clin On- col, 16: 157, 2004
15. Schulick, R. D. and Brenman, M. F .: Adrenocortical carcinoma. World J Urol, 17: 26, 1999
16. Williamson, S. K., Lew, D., Miller, G. J., Balcerzack, S. P., Baker, L. H. and Crawford, E. D .: Phase II evaluation of cisplatin and etoposide followed by mitotane at disease progression in pa- tients with locally advanced or metastatic adrenocortical car- cinoma. Cancer, 88: 1159, 2000
17. Ortiz Gorraiz, M., Tallada Bunuel, M., Vicente Prados, F. J., Rodriguez Herrera, F., Rosales Leal, J. L., Honrubia Vilchez, B. et al: Left adrenal carcinoma with caval thrombosis. Arch Esp Urol, 56: 485, 2003
18. Alaoui, O. A., Ehirchiou, A., Ahallat, M., Sabbah, F., Mahassini, N. and Tounsi, A .: Malignant adrenocortical tumour with in- ferior vena cava invasion. Ann Chir, 128: 262, 2003
19. Nakanoma, T., Ueno, M., Nonaka, S., Tsukamoto, T. and Deguchi, N .: Left adrenocortical cancer with inferior vena cava tumor thrombus-a case report. Nippon Hinyokika Gakkai Zasshi, 92: 34, 2001
20. Lucon, A. M., Machado, M. C., Pereira, M. A., Mendonca, B. B., Praxedes, J. N. and Arap, S .: Surgery for adrenal tumours with thrombus in the supra-diaphragmatic infra-atrial infe- rior vena cava, with no cardiopulmonary bypass. BJU Int, 94: 70, 2004
EDITORIAL COMMENT
This is an important and useful review of a large series of a rare pediatric tumor. It is worthwhile reading for pediatric urologistsbe- cause it summarizes a great deal of timely and useful information.
There are a number of points made that bear emphasis. Unlike adults, most children with adrenocortical carcinomas have evidence of a hormonally active tumor. In this series there was only 1 inactive tumor. Virilization was the most common presentation.
With respect to imaging, when this type of tumor is suspected it is clear from this review that computerized tomography is the best way to detect the tumor. CT is much more likely to demonstrate a lesion than ultrasound. It is also important to emphasize that if a tumor is identified, magnetic resonance imaging is best to detect evidence of vascular invasion. This point is particularly important because there is no association between tumor size and vascular invasion.
Surgery in this type of cancer is critical because it is complete excision that gives the best prognosis for these children. Knowledge of vascular invasion is clearly important in planning surgery. It is also noteworthy that almost 25% of the patients in this series re- quired nephrectomy as part of the en bloc tumor resection. While reoperation for recurrence is appropriate, it is noteworthy that in this series long-term survival was achieved in only 2 of 11 patients undergoing reoperation. Finally, the overall survival in children with adrenocortical carcinoma is better than in adults.
In conclusion, the authors are to be congratulated for putting together an excellent review of a rare but important type of cancer in children.
Howard Snyder Division of Pediatric Urology Children’s Hospital of Philadelphia Philadelphia, Pennsylvania