HORMONE RESEARCH IN PÆDIATRICS
Horm Res Paediatr DOI: 10.1159/000488855
Received: December 21, 2017 Accepted: March 26, 2018 Published online: May 25, 2018
Adrenocortical Carcinoma in Children: A Clinicopathological Analysis of 41 Patients at the Mayo Clinic from 1950 to 2017
Nidhi Guptaª Michael Riverab Paul Novotny” Vilmarie Rodriguezd Irina Bancose Aida Lteifa
ªDivision of Pediatric Endocrinology and Metabolism, Mayo Clinic, Rochester, MN, USA; bDepartment of Anatomic Pathology, Mayo Clinic, Rochester, MN, USA; “Department of Health Sciences Research, Mayo Clinic, Rochester, MN, USA; dDivision of Pediatric Hematology-Oncology, Mayo Clinic, Rochester, MN, USA; eDivision of Endocrinology, Diabetes, Metabolism and Nutrition, Mayo Clinic, Rochester, MN, USA
Keywords
Adrenal tumor . Child . Adrenalectomy . Mitotane . Prognosis
Abstract
Background/Aims: Adrenocortical carcinoma (ACC) is an ag- gressive childhood cancer. Limited evidence exists on a def- inite histopathological criterion to differentiate ACC from adrenocortical adenoma. The aim of this study was to inves- tigate the clinicopathological data of children with ACC, identify prognostic factors, and validate a histopathological criterion to differentiate ACC from adrenocortical adenoma. Methods: This retrospective cohort included 41 children, fol- lowed at the Mayo Clinic from 1950 to 2017 (onset of symp- toms ≤21 years). Outcomes of interest were: alive with no evidence of disease, alive with evidence of disease, and dead of disease. Results: Median age at onset of symptoms was 15.7 years (n = 41; range, 0.2-21 years). Female:male ratio was 3.6:1. Mixed symptomatology (>1 hormone abnormali-
ty) was the most common presentation (54%, n = 22). Sixty- six percent of patients (n = 27 out of 41) underwent total adrenalectomy. Metastatic disease was more common in children aged >12 years (p = 0.002 compared to <4 years). The most common sites of metastases were the liver and lungs. Overall 2-year and 5-year survival rates were 61% (95% CI 45-77) and 46% (95% CI 30-62), respectively. Metas- tasis at the time of diagnosis was independently associated with poor prognosis (risk ratio 13.7%; 95% CI 3.9-87.7). Weiss criteria (29%) and modified Weiss criteria (33%) were less ac- curate in younger patients (<12 years), compared to the Wieneke index (100%). Conclusion: The presence of metas- tases was an independent prognostic factor. The Wieneke index was the most accurate in predicting clinical outcomes in younger children. @ 2018 S. Karger AG, Basel
This study was presented as an oral presentation at the 2017 annual meeting of the Endocrine Society (April 2017, Orlando, FL, USA) and received an Endocrine Society Outstanding Abstract Award.
KARGER
@ 2018 S. Karger AG, Basel
E-Mail Lteif.Aida@mayo.edu or gupta.nidhi@mayo.edu
Introduction
Adrenocortical carcinoma (ACC) is a rare but aggres- sive childhood cancer with a reported incidence of 0.2- 0.3 new cases per 1 million children per year [1-4]. There is remarkable geographical variation with the in- cidence in Southern Brazil being 12-18 times higher than in the United States, likely due to the high preva- lence of a founder TP53 mutation in Brazil [2, 5]. ACC commonly presents with virilization and Cushing syn- drome and rarely with feminization and hyperaldoste- ronism [1, 6-8]. ACC can also be nonfunctional and di- agnosed incidentally during the evaluation of abdomi- nal pain, fatigue or other nonspecific symptoms [9]. The majority of ACC are sporadic, but some are associated with a genetic defect. Individuals with the Li-Fraumeni syndrome [10], Beckwith-Wiedemann syndrome [11], multiple endocrine neoplasia type 1 [12], and familial adenomatous polyposis [13] have an increased risk of ACC.
Surgery is the mainstay of treatment. Even after com- plete resection, a high risk of recurrence of ACC re- mains. Age <4 years, tumor size <10 cm [14], virilization alone [15], tumor volume <200 cm3 [16, 17], tumor weight and stage [18], metastatic disease [14], and qual- ity of surgery [14, 19] have been suggested as important prognostic markers. Despite multimodality approaches including mitotane [an adrenolytic drug: (o,p’-DDD: 1,1-dichloro-2-o-chlorophenyl)-2-(p-chlorophenyl) ethane], chemotherapy with cisplatin, etoposide and doxorubicin (CED), and radiotherapy, prognosis of pe- diatric ACC remains poor with an estimated 5-year sur- vival rate ranging from 30 to 90% [2, 4, 15, 18, 20].
There is considerable overlap in clinical, radiological and histological features of ACC and adrenocortical ad- enoma, which makes decision-making challenging. Due to the rarity and heterogeneity of pediatric ACC, limited evidence exists on a definite histopathological criterion to differentiate ACC from adrenocortical adenomas. The Weiss score is widely used to assess ACC in adults [21]. The modified Weiss score [9, 22] and the Wieneke index [18] have been proposed for assessing ACC in pe- diatrics.
We reviewed the clinicopathological features of 41 pediatric patients with ACC that have been evaluated at the Mayo Clinic over the last 67 years (1950-2017). We also evaluated the validity of the Weiss score, modified Weiss score, and the Wieneke index, all of which have so far been validated in limited pediatric studies [1, 16, 23].
Materials and Methods
Mayo Clinic Adrenal Disorders Database
The Mayo Clinic Adrenal Disorders Database is a retrospective registry of patients (n = 3,348) of any age with adrenal disorders, evaluated at the Mayo Clinic since January 1, 2000. Data on pa- tients from 1950 to 1999 were obtained through the Division of Biomedical Statistics and Informatics at the Mayo Clinic.
Eligibility and Exclusion Criteria
We searched for patients with clinically and/or pathologically confirmed (based on the Weiss score) diagnosis of ACC (Interna- tional Classification of Diseases Code ICD-9-CM 194.0 [before September 30, 2015] and ICD-10-CM C74.00, C74.01, C74.02 [af- ter October 1, 2015]) with age at onset of symptoms ≤21 years. Patients evaluated at the Mayo Clinic from January 1, 1950 to Oc- tober 31, 2017 were included. Patients with adrenocortical adeno- ma, neuroblastoma, pheochromocytoma, and other tumors of the adrenal medulla were excluded.
Search Methods
Electronic medical records were available for patients evalu- ated from 1997 to 2017. Paper medical records were obtained for patients evaluated from 1950 to 1996. Demographics, clinical fea- tures, family history, laboratory analysis, radiological findings, surgical data, pathology reports, treatment description, and long- term outcomes were extracted for each patient. Available ar- chived pathology slides (n = 25) were reviewed by an expert pa- thologist (M.R.) and the Weiss score, modified Weiss score, and Wieneke index were determined (see online suppl. Table 1; see www.karger.com/doi/10.1159/000488855 for all online suppl. material) [9, 18, 21]. The pathologist was blinded to patient out- come information. Complete follow-up data were available for 32 patients. Vital statistics of 28 patients were obtained from elec- tronic/paper medical records, of 2 patients from the National Death Index [24], and of 2 patients from direct patient contact. Of the 9 patients who were lost to follow-up, 3 patients were from outside the United States. This study was conducted in accor- dance with the guidelines of the Code of Federal Regulations (CFR), Title 45, Part 46 and the Institutional Review Board at the Mayo Clinic, Rochester, MN, USA. The requirement to obtain informed consent and assent was waived in accordance with 45 CFR 46.116.
Tumor Staging and Outcome Definitions
Disease staging was defined according to the European Net- work for the Study of Adrenal Tumors (ENSAT) classification, i.e., stages I and II were defined as localized tumors ≤5 or >5 cm, respectively (stage I, T1NoMo; stage II, T2NOM0); stage III was de- fined as tumors that infiltrated surrounding tissue or displayed positive regional lymph nodes or tumor thrombus in the vena cava /renal vein (T1-2N1M0 or T3-4N0-1Mo); stage IV consisted of tumors with distant metastases (T1-4N0-1M1) (see online suppl. Table 2) [3, 25].
Three outcomes of interest were: alive with no evidence of dis- ease, alive with evidence of disease, and dead of disease. The pri- mary endpoints of this study were overall 2-year survival and over- all 5-year survival. Time to event was defined as the interval be- tween date of onset of symptoms and date of death resulting from any cause or date of last follow-up.
Gupta/Rivera/Novotny/Rodriguez/ Bancos/Lteif
| Clinical feature | Age <4 years (n = 10) | Age ≥4 to ≤12 years (n = 4) | Age >12 years (n = 27) | All patients (n = 41) |
|---|---|---|---|---|
| Age at onset of symptoms, years | ||||
| Median | 0.5 | 7.9 | 16.5 | 15.7 |
| Range | 0.2-2.2 | 6.5-10.0 | 12.5-21.0 | 0.2-21.0 |
| Sex, n | ||||
| Female | 7 | 3 | 22 | 32 |
| Male | 3 | 1 | 5 | 9 |
| Female:male ratio | 2.3:1 | 3:1 | 4.4:1 | 3.6:1 |
| Type of presentation, n | ||||
| Virilization only | 3 | 2 | 2 | 7 |
| Feminization onlyª | 0 | 0 | 0 | 0 |
| Cushing syndrome only | 1 | 0 | 1 | 2 |
| Hypertension only | 1 | 0 | 2 | 3 |
| Mixed tumorb | 4 | 1 | 17 | 22 |
| Nonfunctional tumor | 1 | 1 | 4 | 6 |
| Unknown“ | 0 | 0 | 1 | 1 |
| Duration of symptoms, months | ||||
| Median | 3.3 | 5.4 | 2.4 | 3.0 |
| Range | 1.2-12 | 1.2-14.4 | 1.2-14.4 | 1.2-14.4 |
a Indicated by the presence of gynecomastia in males and premature thelarche in females. b Indicated by clin- ical and/or laboratory evidence of abnormal production of more than one hormone, including aldosterone or estrogen. “The patient was diagnosed at another institution and the initial medical records could not be accessed.
Statistical Analysis
Descriptive analyses were used to present clinical and macro- scopic features. Two-year survival and 5-year survival rates were estimated using the Kaplan-Meier method. Heterogeneity in sur- vival among strata of selected variables was assessed using the log- rank test. A multivariate Cox proportional hazards model was used to identify factors associated with the risk of mortality. Multivariate hazard ratios, with 95% CI, were computed considering factors that turned out to be statistically significant at univariate analysis, apply- ing a forward stepwise regression method. Age, disease stage, and metastatic disease were the variables utilized in multivariate regres- sion analysis because complete data on 41 patients were available for these three variables only. Sensitivity, specificity, and accuracy of the Weiss score, modified Weiss score, and Wieneke index were esti- mated. Clinically benign versus malignant outcomes were com- pared with pathologically benign versus malignant outcomes. Sta- tistical analysis was performed using JMP Pro version 10.0.0.
Results
Clinical Features
Forty-one patients met the prespecified inclusion cri- teria (Table 1). Median age of all patients at onset of symptoms was 15.7 years (range, 0.2-21 years). High fe- male preponderance was noted in all age categories with
an overall female:male ratio of 3.6:1. Mixed symptom- atology representing more than one hormonal abnormal- ity was the most common presentation (54%, n = 22) fol- lowed by virilization alone (17%, n = 7). Evidence of iso- lated feminization in males or premature thelarche in females was not noted, except in 1 male patient who pre- sented at 18 years of age with gynecomastia and hyperten- sion. Isolated hyperaldosteronism was reported in 2 pa- tients with onset of symptoms at 16.5 and 19.8 years, re- spectively. Precocious pubarche was noted in 8 patients, of which 4 were less than 1 year of age, 3 were 1.5-2.2 years and 1 was 7 years old.
Median duration of symptoms before diagnosis of ACC was made was 3.0 months (range, 1.2-14.4 months) (p = 0.609 between the three age groups). None of the 41 patients had a known family history of ACC. One patient tested positive for Beckwith Wiedemann syndrome and 1 tested negative for any associated genetic syndromes (Fig. 1).
Macroscopic Features
Sixty-six percent of the patients (n = 27 of 41) under- went total adrenalectomy (Table 2). ACC was described
Color version available online
2
a
b
0
C
RIC
2|
3
4|
5
6
7
8
9
1
METRIC
d
1
1
2|
1
3|
1.
4|
5|
6
7
8
as inoperable in 4 patients, all of whom were >12 years old. Patients aged <4 years had a smaller median tumor diameter and lower median tumor weight as compared to those aged ≥12 years (tumor diameter: 6.4 vs. 10.8 cm; tumor weight: 80 vs. 435 g, p < 0.005 for both). The tumor was localized to the adrenal gland in 15 out of the 36 pa- tients (42%) for whom these data were available.
Metastatic disease was less frequent in patients aged <4 years (2 out of 10; 20%), compared to those aged >12 years (22 out of 27; 81%) (p = 0.002). Overall, the most common sites of metastases were the liver (73%), lungs (69%), and retroperitoneal lymph nodes (27%). A major- ity of patients (n = 26, 63%) were classified by the ENSAT system as stage IV (T1-4N0-1M1).
Treatment Outcomes
Stage I (n = 2)
After resection of the primary spinal adrenocortical tumor at 5 months and recurrent spinal lesion at 11 months, 1 patient was treated with mitotane and 8 cycles of CED. She was in complete remission at 6.3 years. The second patient underwent surgery at 1.8 years, manifest- ed high-grade chondroblastic osteosarcoma of the right sacroiliac area at 16 years, and was treated with hemipel- vectomy, cisplatin, doxorubicin, and methotrexate.
Stage II (n = 11)
All patients were treated with surgery only. Six pa- tients are alive with no evidence of disease (outcome un- known for 4 patients). One patient developed local recur- rence and died a year after resection of recurrent lesion. One patient was diagnosed with familial adenomatous
| Macroscopic feature | Age <4 years (n = 10) | Age ≥4 to ≤12 years (n = 4) | Age >12 years (n = 27) | All patients (n = 41) |
|---|---|---|---|---|
| Surgery, n | ||||
| Partial adrenalectomy | 1 | 0 | 2 | 3 |
| Total adrenalectomy | 7 | 3 | 17 | 27 |
| Radical resection | 1 | 1 | 4 | 6 |
| Inoperable | 0 | 0 | 4 | 4 |
| Other | 1ª | 0 | 0 | 1 |
| Primary site, n | ||||
| Right | 5 | 1 | 11 | 17 |
| Left | 4 | 3 | 16 | 23 |
| Other | 1ª | 0 | 0 | 1 |
| Maximum tumor diameterc, cm | ||||
| Median | 6.4 | 9.8 | 10.8 | 9.5 |
| Range | 2.0-11.0 | 6.0-14.0 | 4.0-20.0 | 2.0-20.0 |
| Tumor weightd, g | ||||
| Median | 80 | 515 | 435 | 276 |
| Range | 20-250 | 180-712 | 25-1,046 | 20-1,046 |
| Tumor extension, n | ||||
| Localized to adrenal gland | 5 | 2 | 8 | 15 |
| Extending into adjacent tissue | 3 | 2 | 11 | 16 |
| Extending into adjacent organ | 0 | 0 | 5 | 5 |
| Other | 2ª, b | 0 | 3b | 5 |
| Vena cava invasion, n | ||||
| Present | 0 | 0 | 5 | 5 |
| Absent | 7 | 3 | 10 | 20 |
| Other | 3ª, b | 1b | 12b | 16 |
| Metastatic disease, n | ||||
| No metastases | 8 | 2 | 5 | 15 |
| Lung only | 0 | 0 | 3 | 3 |
| Liver only | 1 | 0 | 4 | 5 |
| Lung and liver only | 0 | 1 | 3 | 4 |
| Multiple sites | 1e | 1f | 128 | 14 |
| Recurrent diseaseh, n | ||||
| Present | 1 | 1 | 8 | 10 |
| Absent | 9 | 3 | 19 | 31 |
| Tumor stagei, n | ||||
| Stage I | 2 | 0 | 0 | 2 |
| Stage II | 5 | 2 | 4 | 11 |
| Stage III | 1 | 0 | 1 | 2 |
| Stage IV | 2 | 2 | 22 | 26 |
| Type of treatment, n | ||||
| Surgery only | 8 | 2 | 10 | 20 |
| Surgery + chemotherapy | 2 | 2 | 13 | 17 |
| Inoperable (chemotherapy + radiotherapy) | 0 | 0 | 4 | 4 |
a Resection of primary spinal adrenocortical carcinoma in 1 patient. b Data not available. “Based on operative or pathology report; estimated from imaging if tumor was inoperable. ªData was available for: age group <4 years, n = 9; for age group ≥4 to ≤12 years, n = 3, for age group >12 years, n = 14, for all patients, n = 26. e Brain; cervical and thoracic spinal cord. f Liver, lumbar spine and parietal bone. ” Included combinations of ≥2 of these sites: lung (n = 11), liver (n =10), thoracic vertebrae (n = 2), bones (n = 2), kidney (n = 2), spleen (n = 2), brain (n = 1), ovary (n = 1), retroperitoneum (n = 4), retroperitoneal lymph nodes (n = 7), mediastinal lymph nodes (n = 3), supraclavicular lymph nodes (n = 1), left pulmonary artery (n = 1). h At the site of the primary tumor. ’ ENSAT, European Network for the Study of Adrenal Tumors. Stage I, T1NoMo; stage II, T2NoMo; stage III, T1-2N1Mo or T3-4No-1Mo; stage IV, T1-4No-1M1.
| Feature | Numberª | 2-year survival, % (95% CI) | 5-year survival, % (95% CI) | log-rank p value |
|---|---|---|---|---|
| All patients | 41 | 61 (45-77) | 46 (30-62) | |
| Age | ||||
| <4 years | 10 | 100 (-) | 100 (-) | 0.002* |
| ≥4 to ≤12 years | 4 | 75 (32-100) | 75 (32-100) | |
| >12 years | 27 | 43 (23-63) | 21 (5-37) | |
| Sex | ||||
| Female | 32 | 63 (45-81) | 48 (30-66) | 0.654 |
| Male | 9 | 56 (23-89) | 42 (9-75) | |
| Type of presentation | ||||
| Virilization only | 7 | 86 (61-100) | 86 (61-100) | 0.352 |
| Mixed tumor | 22 | 56 (34-78) | 28 (8-48) | |
| Nonfunctional | 6 | 50 (11-89) | 50 (11-89) | |
| Maximum tumor diameter | ||||
| <10 cm | 18 | 87 (69-100) | 80 (60-100) | 0.001* |
| ≥10 cm | 21 | 46 (24-68) | 26 (6-46) | |
| Tumor weight | ||||
| <400 g | 16 | 93 (81-100) | 85 (67-100) | 0.002* |
| ≥400 g | 10 | 50 (19-81) | 20 (0-45) | |
| Tumor extension | ||||
| Localized to adrenal gland | 15 | 86 (68-100) | 78 (56-100) | 0.001* |
| Extending into adjacent tissue | 16 | 48 (23-73) | 27 (5-49) | |
| Extending into adjacent organ | 5 | 40 (0-83) | 0 (-) | |
| Metastatic disease | ||||
| Yes | 26 | 39 (19-59) | 18 (0.4-36) | <0.0001* |
| No | 15 | 93 (81-100) | 86 (68-100) | |
| Tumor stage | ||||
| I-II | 13 | 100 (-) | 92 (76-100) | <0.001* |
| III-IV | 28 | 40 (20-60) | 21 (3-39) | |
| a Missing values were excluded. * p < 0.05. | ||||
polyposis, which was treated with subtotal colectomy and chemotherapy.
Stage III (n = 2)
One patient with the Beckwith-Wiedemann syndrome was treated with surgery at 6 months and is alive with no evidence of disease. The second patient received surgery and mitotane and died 5 months after diagnosis, at 21 years.
Stage IV (n = 26)
Four patients had inoperable, metastatic ACC and died of disease 1-8 months after diagnosis (adjuvant ther- apy: radiotherapy in 2, mitotane in 1, mitotane, cisplatin and etoposide in 1). Three patients were treated with sur- gery and died at 10-17 years (0.6-9.4 years after diagno- sis). After resection of the primary tumor, the remaining
19 patients received chemotherapy (mitotane in 6; vary- ing combinations of CED, mitotane, paclitaxel, mitoxan- trone, 5-fluorouracil, gemcitabine, bleomycin, docetaxel, cyclophosphamide, carboplatin, and ifosfamide in 13). Of these 19 patients, 14 died at a median age of 1.35 years from diagnosis (range, 0.2-7.8 years), 1 is alive without evidence of disease, and outcome is unknown for 4.
Survival Analysis
Overall 2-year survival rate was 61% (95% CI 45-77). Overall 5-year survival rate was 46% (95% CI 30-62) (Ta- ble 3). In univariate analysis, age <4 years, maximum tu- mor diameter <10 cm, tumor weight <400 g, tumor local- ization to adrenal gland, absence of metastatic disease at the time of diagnosis, and low stage (I/II) were associated with better 2-year survival and 5-year survival. Kaplan- Meier survival curves are shown in Figure 2. Applying a
| Criteria | Age ≤12 years | Age >12 years | All patients | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| n AND | AWD | DOD | n | AND | AWD | DOD | n | AND | AWD | DOD | |
| Patients considered to have malignant pathology | |||||||||||
| Weiss | 6 5 | – | 1 | 14 | 1 | 1 | 12 | 20 | 6 | 1 | 13 |
| Modified Weiss | 5 4 | – | 1 | 11 | 1 | 2 | 8 | 16 | 5 | 2 | 9 |
| Wieneke index | 1 – | – | 1 | 10 | – | 1 | 9 | 11 | – | 1 | 10 |
| Patients considered to have benign pathology | |||||||||||
| Weiss | 1 1 | – | – | 0 | – | – | – | 1 | 1 | – | – |
| Modified Weiss | 1 1 | – | – | 0 | – | – | – | 1 | 1 | – | – |
| Wieneke index | 5 5 | – | – | 2 | 2 | – | – | 7 | 7 | – | – |
AND, alive with no evidence of disease; AWD, alive with evidence of disease; DOD, dead of disease. ª Malignant pathology was de- fined by Weiss score ≥3, modified Weiss score ≥3, Wieneke index score ≥4. b Data are presented only for patients on whom complete histological and clinical outcomes were known. ” Age groups “<4 years” and “24 to ≤12 years” were combined due to limited sample size in the latter age group.
1.0
1.0
Survival probability
0.8
Survival probability
0.8
<4 years (n = 10)
0.6
≥4 years to ≤ 12 years
n = 41
0.6
(n = 4)
0.4
0.4
0.2
0.2
>12 years (n = 27)
0
0
0
2
4
6
8
10
0
2
4
6
8
10
a
Survival time, years
b
Survival time, years
1.0
1.0
Survival probability
0.8
Survival probability
Metastases absent (n = 15)
0.8
Stage I-II (n = 13)
0.6
0.6
0.4
0.4
0.2
Metastases present (n = 26)
0.2
Stage III-VI (n = 28)
0
0
0
2
4
6
8
10
0
2
4
6
8
10
c
Survival time, years
d
Survival time, years
forward stepwise regression method using age, disease stage, and metastatic disease, presence of metastases at the time of diagnosis of ACC was the only independent factor associated with poor prognosis (risk ratio 13.7%; 95% CI 3.9-87.7).
Accuracy of Histopathological Criteria
Outcomes that were clinically benign versus malig- nant were analyzed based on the pathological features of
the tumor (Table 4). A malignant pathology was defined by a Weiss score ≥3, modified Weiss score ≥3, and Wieneke index score ≥4. Data on clinical outcomes as well as pathology were available for a smaller number of patients (n = 21 for Weiss criteria, n = 17 for modified Weiss criteria, n = 18 for Wieneke index). Therefore, a clear assessment of performance of each criterion could not be made. However, certain trends were noted (Table 5). Weiss criteria were less accurate in the younger pa-
| Criteria | Age ≤12 years | Age >12 years | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| sensitivity, % | specificity, % (95% CI) | PPV, % | NPV, % | accuracy, % | sensitivity, % | specificity, % (95% CI) | PPV, % | NPV, % | accuracy, % | |
| Weiss | 100 | 17 (1-64) | 17 | 100 | 29 | 100 | 0 (0-97) | 93 | – | 93 |
| Modified Weiss | 100 | 20 (1-72) | 20 | 100 | 33 | 100 | 0 (0-98) | 91 | – | 91 |
| Wieneke index | 100 | 100 (48-100) | 100 | 100 | 100 | 100 | 100 (16-100) | 100 | 100 | 100 |
PPV, positive predictive value; NPV, negative predictive value; - , insufficient n.
tients (29% in ≤12 years vs. 93% in >12 years). Modified Weiss criteria were also less accurate in the younger pa- tients (33% in ≤12 years vs. 91% in >12 years). The Wieneke index was more accurate than Weiss and mod- ified Weiss criteria in both age groups. A significant overlap in 95% CI for specificity was noted for all three criteria, thereby limiting the interpretation of these re- sults.
Discussion
The findings from this study provide strong evidence that the natural course of ACC is significantly different and favorable in children younger than 4 years of age compared to those older than 12 years of age. Presence of metastases at the time of diagnosis of ACC is an indepen- dent prognostic factor for survival. Histopathological classification of ACC in pediatrics remains controversial. The Wieneke index was the most accurate in predicting clinical outcomes in younger children in our study.
Biphasic Age Distribution
Adrenocortical tumors (ACT) have been reported to have either early age of onset (<4 years) [3, 9, 14-16] or biphasic age distribution (<5 years and >10 years) [18, 26]. These studies included patients from Brazil [9, 15], the USA [14, 18], the Netherlands [3], Turkey [26], and other countries [27], thereby disputing any geographical pattern. A majority of our patients, which was primarily an American cohort, had onset of symptoms after 12 years of age.
Limited data are available on the onset and course of ACT during early infancy [28-33]. In the International Pediatric Adrenocortical Tumor Registry (IPACTR), 12% of cases (adenomas and carcinomas combined) were diagnosed during the first year of life [15]. Seven of our
patients (17%) had onset of symptoms within the first year of life. Three of these patients are alive with no evi- dence of disease (at 3 years, 18 years and 38 years of age), 1 died at 10 years, and outcome was not known on the remaining 3 patients. Previous reports have suggested fa- vorable outcomes after treatment in infantile ACC [30, 31, 33]. A newborn girl had spontaneous regression of skin metastases and cerebral lesion by 4 months of age following surgical resection of right adrenal primary tu- mor [29]. On the other hand, poor clinical outcomes were described by Satge et al. [28] in 14 children with congen- ital ACT.
Variability in Clinical Presentation
Consistent with previous studies [3, 9, 15, 16, 26, 34], mixed symptomatology (virilization, Cushing syndrome, and hyperaldosteronism) was the most common presen- tation of ACC across all age groups in our cohort. One patient presented at 5 months of age with lower extrem- ity weakness and was diagnosed with ectopic extramedul- lary intradural ACC with 2 intact adrenal glands. Ectopic ACT are rare and have been described previously in the spinal canal [35] and intrathoracic cavity [36].
Nonfunctional tumors comprised 12% of our patients. Fifteen percent of patients older than 12 years had non- functional presentation, compared to only 0.1% nonfunc- tional tumors in less than 4-year-olds. About 20-30% of adult ACC present as nonfunctional tumors [37], sug- gesting likely resemblance of the natural course of ACC in adults with that in older children, compared to young- er children. This further supports the hypothesis pro- posed by Michalkiewicz et al. [15] that “ACTs comprise at least two distinct subtypes based on age at onset.” In IPACTR, nonfunctional tumors were observed in 10.2% of patients (median age 3.2 years; range, 0-19 years); however, data included adrenal adenomas and carcino- mas [15].
Treatment Strategies and Outcomes
Our patients were treated with adjuvant therapy in set- tings of metastatic, recurrent, inoperable or stage IV dis- ease. In IPACTR [15], for localized disease, surgery alone was recommended. However, 12% of patients with stage I or II received adjuvant chemotherapy including CED and mitotane. Stage III or IV tumors were treated with intensive chemotherapy. Recurrent disease was treated with surgical resection, if feasible, always followed by che- motherapy. A similar treatment approach was employed by Dall’Igna et al. [16] in 58 Italian children enrolled in a prospective national study (TREP; Rare Tumors in Pedi- atric Age). In a cohort of 111 children with ACC from the National Cancer Database (NCDB), adjuvant systemic therapy was utilized in 30% of children who underwent surgical resection [14].
Redlich et al. [38] described treatment outcomes of 60 German patients (age 0.24-17.8 years) treated according to the nonrandomized, single-arm study GPOH-MET-97. Chemotherapy with vincristine, ifosfamide, carboplatin, etoposide, and doxorubicin was provided to 34 patients (56.6%) and mitotane therapy to 32 patients (53.3%). Mi- totane treatment longer than 6 months and mitotane lev- els greater than 14 mg/L were found to be associated with significantly better survival. Radiation therapy alone or in combination with chemotherapy has rarely been used in previous studies [3, 15, 18].
Mitotane has been associated with significant gastro- intestinal, neuropsychiatric and hepatic toxicity [38, 39]. Twenty patients in our study were treated with mitotane in different settings (neoadjuvant, adjuvant, and salvage), with doses ranging from 0.5 to 10 g per day. Of these 20 patients, 15 died of disease, 1 is alive with no evidence of disease and outcome is unknown for 4 patients. Adverse effects noted in our patients included sedation, depres- sion, blurred vision, gastrointestinal symptoms, extrem- ity weakness, ototoxicity, mild renal insufficiency, bone marrow suppression, and drug reaction. Mitotane had to be discontinued within 6 weeks to 6 months in 7 patients due to intolerable side effects. Godil et al. [33] reported multiple additional complications including seizures, gy- necomastia, and abnormal thyroid function tests in a ne- onate with metastasizing congenital ACC. These issues resolved after discontinuation of mitotane.
Survival and Prognosis
Overall 2-year and 5-year survival rates (61 and 46%, respectively) of our patients were comparable to those re- ported previously [4, 14, 15, 17]. In IPACTR [15], at a median follow-up of 2 years and 5 months, 61.8% patients
remained alive. The 5-year overall survival estimate was 54.7% (95% CI 48.7-60.7). The age at which survival de- clines remains unclear. Results from the Surveillance, Ep- idemiology, and End Results (SEER) database indicated that overall 5-year survival for patients younger than 4 years was 91.1% (95% CI 74.8-97.1) and that it reduced significantly for patients aged 5-19 years to 29.8% (95% CI 16.8-44.1) [4]. Age less than 4 years was an important predictor of survival in our patients as well as in several other cohorts [3, 14-16]. This was likely related to a smaller tumor size (<10 cm) [18], lower tumor weight (<400 g), lower tumor volume (<200 cm3) [16], disease localized to adrenal gland [14, 18], less advanced disease (stage I/II) [15], and lower incidence of metastatic disease in younger children compared to older ones [14].
Histopathological Classification
Rarity of ACC and lack of data on association between histopathological features and clinical outcomes contrib- ute to this puzzle. We found that the majority of our pa- tients younger than 12 years of age had a clinically benign outcome despite malignant histological features based on Weiss criteria [21] and modified Weiss criteria [8, 22]. Investigators have reported similar discrepancy between correlation of histological classification of ACC based on adult criteria (specifically Weiss criteria) and patient out- comes [16, 18, 40]. Correlation between high Weiss score and worse clinical outcomes was not observed in 12 Dutch children (median age 4.1 years); however, stronger correlation was noted with application of the Wieneke index instead, though their sample size was small [3, 18].
Validity of the Wieneke index was further demonstrat- ed by Chatterjee et al. [1, 41], in a group of 13 children with ACT (mean age 2.9 years). Applying the Wieneke index, 6 patients were assigned to benign, 1 to the inter- mediate, and 6 to the malignant pathology group, which was eventually consistent with the clinical behavior of these tumors. However, when Weiss criteria were ap- plied, 3 tumors with benign clinical behavior were as- signed to malignant pathology. The Wieneke index was applied in 24 Italian children (median age 5 years) with good significance [16]. Our findings corroborated that the Wieneke index was more accurate in predicting clin- ical outcomes at least in younger children. However, con- sidering that adrenocortical adenomas were excluded from the study, false-positive cases may have been under- estimated.
Limitations and Strengths
Our study had certain limitations including the retro- spective nature of the data. Genetic testing was not per- formed on the majority of the patients (except 2), primar- ily due to lack of availability of genetic testing over the study period. Therefore, no conclusion regarding asso- ciation of TP53 mutations with ACC could be made. In- formation on immunohistochemical staining was lack- ing, though validity of these stains in pediatrics has not been established [18]. As mentioned previously, histo- pathological and mortality data were available for a lim- ited number of patients, which made definite validation of histopathological criteria challenging. Data on growth abnormalities and endocrine laboratory evaluation were available for a limited number of patients as well.
The Mayo Clinic is a tertiary care referral center. Our cohort included local, regional and international patients over the last 6 decades, some of whom had undergone initial diagnostic evaluation and management elsewhere. These patients might have received treatment that differs from current practice at the Mayo Clinic. In pediatric ACC, various registries have utilized either the IPACTR or ENSAT staging system [3, 15, 16]. We did not use the IPACTR staging system because that requires data on postoperative hormonal levels, tumor margin status, and presence of microscopic residual tumor [2]. These data were not available for a majority of our patients.
The strengths of our study relate to the measures un- dertaken to reduce the effect of bias in histological inter- pretation of archived pathological slides. We acknowl-
edge that larger studies on pediatric ACT (adenomas plus carcinomas) have been published previously. These stud- ies have reported cohorts of patients from multicenter registries over a period of 10-35 years [1, 3, 4, 14-17, 26, 42]. Our study is the largest single institution experience with longest duration of data on pediatric ACC, reported so far.
Conclusions
ACC in younger children has a distinct natural history compared to older children. Though age <4 years was an important predictor of survival in our study, metastatic disease at the time of diagnosis of ACC was an indepen- dent prognostic factor. Based on our findings, we advo- cate utilization of the Wieneke index (in conjunction with clinical features) in predicting outcomes at least in young- er children, though validation studies with a larger sam- ple size are required. Efficacy of adjuvant systemic thera- py remains unclear. We recommend observation after complete surgical resection in patients with localized dis- ease and a low Wieneke index score (<4). Collaborative, international prospective trials with various adjuvant therapy options would be the ideal next step.
Disclosure Statement
The authors declare that there are no conflict of interests.
References
1 Chatterjee G, DasGupta S, Mukherjee G, Sen- gupta M, Roy P, Arun I et al. Usefulness of Wieneke criteria in assessing morphologic characteristics of adrenocortical tumors in children. Pediatr Surg Int. 2015 Jun;31(6): 563-71.
2 Sandrini R, Ribeiro RC, DeLacerda L. Child- hood adrenocortical tumors. J Clin Endocri- nol Metab. 1997 Jul;82(7):2027-31.
3 Kerkhofs TM, Ettaieb MH, Verhoeven RH, Kaspers GJ, Tissing WJ, Loeffen J et al. Adre- nocortical carcinoma in children: first popu- lation-based clinicopathological study with long-term follow-up. Oncol Rep. 2014 Dec; 32(6):2836-44.
4 McAteer JP, Huaco JA, Gow KW. Predictors of survival in pediatric adrenocortical carci- noma: a Surveillance, Epidemiology, and End Results (SEER) program study. J Pediatr Surg. 2013 May;48(5):1025-31.
5 Rodriguez-Galindo C, Figueiredo BC, Zam- betti GP, Ribeiro RC. Biology, clinical charac- teristics, and management of adrenocortical tumors in children. Pediatr Blood Cancer. 2005 Sep;45(3):265-73.
6 Wajchenberg BL, Albergaria Pereira MA, Me- donca BB, Latronico AC, Campos Carneiro P, Alves VA et al. Adrenocortical carcinoma: clinical and laboratory observations. Cancer. 2000 Feb;88(4):711-36.
7 Frasch W, Gnekow A, Bolkenius M, Wagner T, Dhom G, Dörr HG et al. Adrenal cortex carcinoma. A rare cause of a Conn syndrome in childhood. Monatsschr Kinderheilkd. 1992 Feb; 140(2):95-101.
8 Ribeiro RC, Figueiredo B. Childhood adreno- cortical tumours. Eur J Cancer. 2004 May; 40(8):1117-26.
9 Ribeiro RC, Michalkiewicz EL, Figueiredo BC, DeLacerda L, Sandrini F, Pianovsky MD et al. Adrenocortical tumors in children. Braz J Med Biol Res. 2000 Oct;33(10):1225-34.
10 Kleihues P, Schäuble B, zur Hausen A, Estève J, Ohgaki H. Tumors associated with p53 germline mutations: a synopsis of 91 families. Am J Pathol. 1997 Jan; 150(1):1-13.
11 Lapunzina P. Risk of tumorigenesis in over- growth syndromes: a comprehensive review. Am J Med Genet C Semin Med Genet. 2005 Aug; 137C(1):53-71.
12 Gatta-Cherifi B, Chabre O, Murat A, Niccoli P, Cardot-Bauters C, Rohmer V et al. Adrenal involvement in MEN1. Analysis of 715 cases from the Groupe d’etude des Tumeurs Endo- crines database. Eur J Endocrinol. 2012 Feb; 166(2):269-79.
13 Gaujoux S, Pinson S, Gimenez-Roqueplo AP, Amar L, Ragazzon B, Launay P et al. Inactiva- tion of the APC gene is constant in adrenocor- tical tumors from patients with familial ade- nomatous polyposis but not frequent in spo- radic adrenocortical cancers. Clin Cancer Res. 2010 Nov; 16(21):5133-41.
14 Gulack BC, Rialon KL, Englum BR, Kim J, Talbot LJ, Adibe OO et al. Factors associated with survival in pediatric adrenocortical car- cinoma: An analysis of the National Cancer Data Base (NCDB). J Pediatr Surg. 2016 Jan; 51(1):172-7.
15 Michalkiewicz E, Sandrini R, Figueiredo B, Miranda EC, Caran E, Oliveira-Filho AG et al. Clinical and outcome characteristics of chil- dren with adrenocortical tumors: a report from the International Pediatric Adrenocor- tical Tumor Registry. J Clin Oncol. 2004 Mar; 22(5):838-45.
16 Dall’Igna P, Virgone C, De Salvo GL, Ber- torelle R, Indolfi P, De Paoli A et al. Adreno- cortical tumors in Italian children: analysis of clinical characteristics and P53 status. Data from the national registries. J Pediatr Surg. 2014 Sep;49(9):1367-71.
17 Cecchetto G, Ganarin A, Bien E, Vorwerk P, Bisogno G, Godzinski J et al. Outcome and prognostic factors in high-risk childhood ad- renocortical carcinomas: A report from the European Cooperative Study Group on Pedi- atric Rare Tumors (EXPERT). Pediatr Blood Cancer. 2017 Jun;64(6):64.
18 Wieneke JA, Thompson LD, Heffess CS. Ad- renal cortical neoplasms in the pediatric pop- ulation: a clinicopathologic and immunophe- notypic analysis of 83 patients. Am J Surg Pathol. 2003 Jul;27(7):867-81.
19 Teinturier C, Pauchard MS, Brugières L, Lan- dais P, Chaussain JL, Bougnères PF. Clinical and prognostic aspects of adrenocortical neo- plasms in childhood. Med Pediatr Oncol. 1999 Feb;32(2):106-11.
20 Redlich A, Boxberger N, Strugala D, Früh- wald MC, Leuschner I, Kropf S et al. Systemic treatment of adrenocortical carcinoma in children: data from the German GPOH-MET 97 trial. Klin Padiatr. 2012 Oct;224(6):366- 71.
21 Weiss LM. Comparative histologic study of 43 metastasizing and nonmetastasizing adreno- cortical tumors. Am J Surg Pathol. 1984 Mar; 8(3):163-9.
22 Bugg MF, Ribeiro RC, Roberson PK, Lloyd RV, Sandrini R, Silva JB et al; Brazilian Group for Treatment of Childhood Adrenocortical Tumors. Correlation of pathologic features with clinical outcome in pediatric adrenocor- tical neoplasia. A study of a Brazilian popula- tion. Am J Clin Pathol. 1994 May; 101(5): 625-9.
23 Magro G, Esposito G, Cecchetto G, Dall’Igna P, Marcato R, Gambini C et al. Pediatric adre- nocortical tumors: morphological diagnostic criteria and immunohistochemical expres- sion of matrix metalloproteinase type 2 and human leucocyte-associated antigen (HLA) class II antigens. Results from the Italian Pe- diatric Rare Tumor (TREP) Study project. Hum Pathol. 2012 Jan;43(1):31-9.
24 National Death Index. National center for health statistics 2015 [cited 2016 June 17]. Available from: http://www.cdc.gov/nchs/ ndi/index.html.
25 Fassnacht M, Johanssen S, Quinkler M, Buc- sky P, Willenberg HS, Beuschlein F et al; Ger- man Adrenocortical Carcinoma Registry Group; European Network for the Study of Adrenal Tumors. Limited prognostic value of the 2004 International Union Against Cancer staging classification for adrenocortical carci- noma: proposal for a Revised TNM Classifica- tion. Cancer. 2009 Jan; 115(2):243-50.
26 Ciftci AO, Senocak ME, Tanyel FC, Büyük- pamukçu N. Adrenocortical tumors in chil- dren. J Pediatr Surg. 2001 Apr;36(4):549-54.
27 Ribeiro RC, Pinto EM, Zambetti GP, Rodri- guez-Galindo C. The International Pediatric Adrenocortical Tumor Registry initiative: contributions to clinical, biological, and treat- ment advances in pediatric adrenocortical tu- mors. Mol Cell Endocrinol. 2012 Mar;351(1): 37-43.
28 Satge D, Philippe E, Ruppe M, Levy JM, Lutz P, Walter P. Neonatal carcinoma. Review of the literature apropos of a case. Bull Cancer. 1988;75(4):373-84.
29 Saracco S, Abramowsky C, Taylor S, Silver- man RA, Berman BW. Spontaneously re- gressing adrenocortical carcinoma in a new- born. A case report with DNA ploidy analysis. Cancer. 1988 Aug;62(3):507-11.
30 García E, Cordero Gallardo G, Lugo-Vicente H. Adrenocortical carcinoma in a female in- fant: a case report. Bol Asoc Med P R. 2013; 105(4):52-5.
31 Hishiki T, Kazukawa I, Saito T, Terui K, Mit- sunaga T, Nakata M et al. Diagnosis of adre- nocortical tumor in a neonate by detection of elevated blood 17-hydroxyprogesterone mea- sured as a routine neonatal screening for congenital adrenal hyperplasia: a case report. J Pediatr Surg. 2008 Oct;43(10):e19-22.
32 Kakkar N, Vasishta RK, Lamba A, Trehan A, Marwaha RK. Special feature: pathological case of the month. Denouement and discus- sion: congenital adrenocortical carcinoma. Arch Pediatr Adolesc Med. 2000 Dec;154(12): 1267-8.
33 Godil MA, Atlas MP, Parker RI, Priebe CJ, Zerah MM, Kane P et al. Metastatic congeni- tal adrenocortical carcinoma: a case report with tumor remission at 3 1/2 years. J Clin Endocrinol Metab. 2000 Nov;85(11):3964-7.
34 Hubertus J, Boxberger N, Redlich A, von Sch- weinitz D, Vorwerk P. Surgical aspects in the treatment of adrenocortical carcinomas in children: data of the GPOH-MET 97 trial. Klin Padiatr. 2012 Apr;224(3):143-7.
35 Kepes JJ, O’Boynick P, Jones S, Baum D, Mc- Millan J, Adams ME. Adrenal cortical adeno- ma in the spinal canal of an 8-year-old girl. Am J Surg Pathol. 1990 May; 14(5):481-4.
36 Medeiros LJ, Anasti J, Gardner KL, Pass HI, Nieman LK. Virilizing adrenal cortical neo- plasm arising ectopically in the thorax. J Clin Endocrinol Metab. 1992 Dec;75(6):1522-5.
37 Else T, Kim AC, Sabolch A, Raymond VM, Kandathil A, Caoili EM et al. Adrenocortical carcinoma. Endocr Rev. 2014 Apr;35(2):282- 326.
38 Redlich A, Boxberger N, Strugala D, Früh- wald MC, Leuschner I, Kropf S et al. Systemic treatment of adrenocortical carcinoma in children: data from the German GPOH-MET 97 trial. Klin Padiatr. 2012 Oct;224(6):366- 71.
39 Luton JP, Cerdas S, Billaud L, Thomas G, Guilhaume B, Bertagna X et al. Clinical fea- tures of adrenocortical carcinoma, prognostic factors, and the effect of mitotane therapy. N Engl J Med. 1990 Apr;322(17):1195-201.
40 Faria AM, Almeida MQ. Differences in the molecular mechanisms of adrenocortical tu- morigenesis between children and adults. Mol Cell Endocrinol. 2012 Mar;351(1):52-7.
41 Das S, Sengupta M, Islam N, Roy P, Datta C, Mishra PK et al. Weineke criteria, Ki-67 index and p53 status to study pediatric adrenocorti- cal tumors: is there a correlation? J Pediatr Surg. 2016 Nov;51(11):1795-800.
42 Liou LS, Kay R. Adrenocortical carcinoma in children. Review and recent innovations. Urol Clin North Am. 2000 Aug;27(3):403-21.