SEARCH

SEARCH BY CITATION

Keywords:

  • adult;
  • Wilms;
  • SEER;
  • pediatric;
  • outcome

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

PURPOSE:

To compare the characteristics and outcome of adults and children diagnosed with Wilms' tumor.

METHODS:

The Surveillance, Epidemiology and End Results (SEER) database was analyzed for patients diagnosed with Wilms' tumor between 1973 and 2007. Patients were stratified into pediatric (<16 years) or adult (≥16 years) groups. Overall survival was the primary endpoint.

RESULTS:

A total of 2342 patients (2190 pediatric and 152 adult) with Wilms' tumor were identified. Adult patients were statistically more likely to be staged as localized than pediatric patients (62.5% vs 44.7%), to not receive any lymph node sampling (57.9% vs 16.2%), and to not receive any radiation treatment (74.3% vs 53.9%). Adults had a statistically worse overall survival (OS) than pediatric patients (5-year OS, 69% vs 88%, P<.001) despite the earlier tumor stage. When stratified by treatment era, the OS of all patients treated after 1981 was statistically higher than those treated before (5-year OS, 75% vs 89%, P<.001). Significant predictors of OS on univariate analysis for adults included treatment era, SEER stage, surgery, and radiation treatment. Significant predictors of OS on multivariate analysis of all patients included adult status (hazard ratio, 4.14; P<.001), treatment era, SEER stage, and surgery.

CONCLUSION:

Adults in the SEER database had statistically worse OS than pediatric patients despite previous studies showing comparable outcome when treated on protocol. The worse outcome of SEER adults likely stems from incorrect diagnosis, inadequate staging and undertreatment. We recommend lymph node samplings for all adult Wilms' tumor patients and collaboration with pediatric oncologists. Cancer 2012. © 2011 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

Wilms' tumor is the most common renal tumor in children and represents approximately 95% of all pediatric renal malignancies. The diagnosis of Wilms' tumor in adults, however, is exceedingly rare, with an estimated incidence of only 0.2 cases per million.1, 2 Kilton et al3 created a strict definition for the diagnosis of adult Wilms' tumor in 1980 that included several histological criteria as well as an age requirement of >15 years.

Currently, there have been only case reports and a limited number of published series with relatively small populations of registered patients studying the treatment and outcome of this rare adult tumor.4-18 The 2 primary multi-institutional Wilms' tumor cooperative groups, the National Wilms Tumor Study Group (NWTS) and the International Society of Paediatric Oncology (SIOP), have each published analyses of adult Wilms' tumor cases registered on their respective trials.19, 20 The publications from these groups were based on adult (≥16 years) subsets that were treated identically to the children on protocol with overall and disease-free survival results comparable to those of the children. These outstanding results, however, are in contrast to those of general population-based studies and mono-institutional series that have observed significantly worse outcome for adults compared with children.2, 14, 16 It has been hypothesized that this difference in outcome stems from the relative unfamiliarity of adult oncologists with the specifics of the Wilms' tumor pediatric treatment protocols as well as difficulties and delays associated with diagnosis in adults.2, 20

Although the NWTS and SIOP adult subsets have the distinct advantage of having received a highly regimented diagnostic evaluation and treatment program with standardized and detailed outcome measures, these trials accrued few adult patients and thus had relatively weak statistical power. The recent NWTS report on adult Wilms' tumor had only 23 patients from the NWTS 4 and 5 trials, and the recent SIOP retrospective review included only 30 patients.19, 20 A recent report on adult Wilms' tumor from the EUROCARE study, which consists of data from 67 cancer registries and 22 European countries, included 143 adult patients. Unfortunately, there was no inclusion of any pediatric patient data from that database within the analysis. Additionally, the study was quite limited in associated database parameters and did not have staging information for over two-thirds of the included patients.2 Another report on adult Wilms' tumor patients by Izawa et al21 combined previously published single case reports and small case series (including the previously mentioned SIOP adult Wilms' series) to create a study group of 128 patients. Though they were unable to perform a direct statistical comparison against a correlated pediatric population, they did conclude that the overall survival for adult patients appeared to be worse than historical pediatric patient overall survival rates. Additionally, they postulated that one of the reasons for the worse outcome may be lower rates of lymph node dissections in adult patients compared with pediatric patients, although they were unable to report any lymph node dissection data for the adult patients included in their publication.

Given the relative lack of knowledge and conflicting analyses regarding the clinical behavior and treatment outcome of Wilms' tumor in adults compared with children, we pursued the Surveillance, Epidemiology and End Results (SEER) database to provide a singular source of data from one of the nation's most extensive tumor registries to provide long-term follow-up, strong statistical power, and uniform pediatric and adult patient populations with which to study this rare disease.22

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

Database

Data used in the analysis were acquired from the SEER 17 Registry, which was released in April 2010 and includes tumors diagnosed between 1973 and 2007.22 This database was accessed using SEER*Stat version 6.6.2. The inclusion criteria for the case listing session required that all cases have a known age and an adolescent and young adult site recode diagnosis of 9.1.1 Wilms tumor. The SEER adolescent and young adult classification scheme is based on that proposed by Barr et al22, 23 and includes, by definition, the International Classification of Diseases for Oncology (ICD-O-3) histological diagnostic codes of 8959-8960 with a malignant ICD-O-3 behavior recode. Cases included only a patient's first diagnosis of Wilms' tumor, and those registered as autopsy or death certificate only were excluded from the analysis.

It should be noted that although the SEER database does provide strong statistical power with a large uniform population of both adult and pediatric patients tracked across nearly 4 decades in addition to supplying a sizeable array of prognostic variables, there are rather significant limitations of the SEER database compared with a database from a clinical trial group such as the NWTS or SIOP. Specifically, the SEER database, by design, does not have a central pathology review of tissue specimens. Additionally, although it does track the type of surgery and radiation treatment as well as surgical margin status, number of lymph nodes dissected, and number of positive lymph nodes, it does not provide more extensive surgical details such as the type of surgical approach and operative notes or radiation doses. Perhaps the most significant limitation of the SEER database, however, is the exclusion of chemotherapy details and lack of long-term toxicity parameters.

Statistical Analysis

All statistical analyses were performed with SPSS version 17 (SPSS Inc., Chicago, IL). Patients were classified as either “pediatric” (<16 years) or “adult” (≥16 years) depending on age at the time of diagnosis.3, 19, 20 The SEER staging system was used for staging purposes because this categorization had standardized definitions and consistent data dating back to 1973 for primary kidney tumors. SEER staging defines “localized” cancer as that which is “limited to the organ in which it began, without evidence of spread.” “Regional” cancer is defined as that which has “spread beyond the primary site to nearby lymph nodes or organs and tissues.” “Distant” cancer is defined as that which has “spread from the primary site to distant organs or distant lymph nodes.”22 The listed year of diagnosis was used to categorize the patients into 1 of 4 treatment eras between 1973 and 2007 defined by the publication years of the NWTS 2, 3, and 4 trials.24-26

Patient parameters were statistically compared between age groups to assess for variable heterogeneity. The continuous variables were statistically compared with the Student t test, and categorical variables were compared with Pearson's chi-square test except when the data set was relatively small, in which case the Fisher exact test was used.

The outcomes of the pediatric and adult patient groups were primarily compared on the basis of overall survival (OS) and disease-specific survival (DSS) endpoints. The latter was obtained using the SEER cause-specific death classification as defined by Howlader et al.27 Relative survival was also calculated across the spectrum of ages of patients with Wilms' tumor.22 The Kaplan-Meier method was used to generate survival curves. Survival estimates are presented followed by the standard error. The pediatric and adult patient groups were compared through a univariate analysis on the basis of the log-rank test and a multivariate analysis using Cox proportional hazards regression. For the purposes of the multivariate analysis, the laterality parameter was consolidated into unilateral versus bilateral, and the surgery parameter was consolidated into no primary site surgery or biopsy only versus surgery.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

Patient Characteristics

The SEER 17 registry contained a total of 2352 patients with a diagnosis of Wilms' tumor. Ten of these patients were registered on the basis of autopsy or death certificate only and were excluded, leaving a total of 2342 patients with 2190 (93.5%) pediatric patients (<16 years) and 152 (6.5%) adult patients (≥16 years). As seen in Table 1, the pediatric and adult patients were fairly well balanced with respect to sex and race. There was, however, a statistically significant difference in tumor laterality, because approximately 6.4% of the pediatric patients were bilateral compared with only 0.7% of the adult patients. In analyzing the association of the Wilms' tumor patients with other primary malignancies in the same patient, it is observed that statistically more of the adult patients had additional primary malignancies (both before and after the Wilms' diagnosis) than the pediatric patients.

Table 1. Patient Characteristics
ParameterTotal (N = 2342)Pediatric (n = 2190)Adult (n = 152)P
  • NOS, not otherwise specified; RT, radiation treatment; SD, standard deviation; SEER, Surveillance, Epidemiology and End Results. All data are presented as no. (%) unless indicated otherwise.

  • a

    Student t test.

  • b

    Pearson chi-square test.

  • c

    Fisher exact test.

  • d

    There were 109 patients who had unknown “regional lymph nodes examined” status but reported positive lymph nodes, and there were 3 patients who had “regional lymph nodes examined” status but had an unknown positive lymph node status.

Age, y    
 Median3336 
 Mean (SD)5.6 (10.8)3.2 (2.8)40.3 (19.7)<.001a
 Range0-930-1516-93 
Sex   .089b
 Male1080 (46.1)1020 (46.6)60 (39.5) 
 Female1262 (53.9)1170 (53.4)92 (60.5) 
Race   .777b
 White1860 (79.4)1740 (79.5)120 (78.9) 
 Black342 (14.6)321 (14.7)21 (13.8) 
 Other/unknown140 (6.0)129 (5.9)11 (7.2) 
Laterality   <.001b
 Right1093 (46.7)1019 (46.5)74 (48.7) 
 Left1076 (45.9)1007 (46.0)69 (45.4) 
 Bilateral142 (6.1)141 (6.4)1 (0.7) 
 Unknown31 (1.3)23 (1.1)8 (5.3) 
Primary malignancies    
 First primary2320 (99.1)2184 (99.7)136 (89.5)<.001c
 1 total2295 (98.0)2158 (98.5)137 (90.1)<.001c
 2 total41 (1.8)29 (1.3)12 (7.9) 
 >2 total6 (0.3)3 (0.1)3 (2.0) 
 Later primaries38 (1.6)33 (1.5)5 (3.3).096c
SEER stage   <.001b
 Localized1073 (45.8)978 (44.7)95 (62.5) 
 Regional643 (27.5)623 (28.4)20 (13.2) 
 Distant534 (22.8)510 (23.3)24 (15.8) 
 Unstaged/blank92 (3.9)79 (3.6)13 (8.6) 
Regional lymph nodes examinedd   <.001b
 0442 (18.9)354 (16.2)88 (57.9) 
 1-3551 (23.5)540 (24.7)11 (7.2) 
 4-6284 (12.1)279 (12.7)5 (3.3) 
 7-9110 (4.7)110 (5.0)0 (0.0) 
 10-1264 (2.7)64 (2.9)0 (0.0) 
 >12109 (4.7)100 (4.6)9 (5.9) 
 Unknown782 (33.4)743 (33.9)39 (25.7) 
Regional lymph nodes positived   .314c
 0996 (81.4)975 (81.5)21 (75.0) 
 1-3185 (15.1)180 (15.1)5 (17.9) 
 >343 (3.5)41 (3.4)2 (7.1) 
Tumor size   .017a
 Measured1526 (65.2)1433 (65.4)93 (61.2) 
 Unknown816 (34.8)757 (34.6)59 (38.8) 
 Mean size (SD), cm10.6 (4.4)10.6 (4.2)9.5 (6.3) 
Treatment era   .320b
 1973-1981349 (14.9)328 (15.0)21 (13.8) 
 1982-1989357 (15.2)341 (15.6)16 (10.5) 
 1990-1998594 (25.4)550 (25.1)44 (28.9) 
 1999-20071042 (44.5)971 (44.3)71 (46.7) 
Surgery   <.001b
 No primary site surgery or biopsy only94 (4.0)82 (3.7)12 (7.9) 
 Partial nephrectomy96 (4.1)86 (3.9)10 (6.6) 
 Simple nephrectomy378 (16.1)368 (16.8)10 (6.6) 
 Radical nephrectomy1206 (51.5)1121 (51.2)85 (55.9) 
 Nephrectomy, NOS136 (5.8)129 (5.9)7 (4.6) 
 Surgery, NOS388 (16.6)367 (16.8)21 (13.8) 
 Unknown44 (1.9)37 (1.7)7 (4.6) 
Radiation treatment   <.001c
 None1294 (55.3)1181 (53.9)113 (74.3) 
 External beam1009 (43.1)972 (44.4)37 (24.3) 
 Other/NOS9 (0.4)9 (0.4)0 (0.0) 
 Unknown30 (1.3)28 (1.3)2 (1.3) 
Radiation sequence   <.001c
 RT before surgery87 (3.7)84 (3.8)3 (2.0) 
 RT after surgery879 (37.5)849 (38.8)30 (19.7) 
 RT before and after surgery14 (0.6)14 (0.6)0 (0.0) 
 Unknown/no RT or surgery1362 (58.2)1243 (56.8)119 (78.3) 
Vital status   <.001b
 Alive2013 (86.0)1913 (87.4)100 (65.8) 
 Dead329 (14.0)277 (12.6)52 (34.2) 

When the patients were stratified by treatment era according to NWTS 2, 3, and 4 trial publication dates, the pediatric and adult groups were well balanced (Table 1). There is an increase in the number of cases in later treatment eras due to the expanding geographical coverage of the SEER program.22

Comparison of the SEER stage at presentation between children and adults revealed that the adult patients were statistically lower stage than the pediatric patients, with fewer metastatic presentations (15.8% vs 23.3%), fewer regional cases (13.2% vs 28.4%), and more localized cases (62.5% vs 44.7%). In addition, the adult group had statistically more cases without lymph node samplings than the pediatric group (57.9% vs 16.2%).

Regarding treatment, more adult patients had no surgery or biopsy only compared with pediatric patients (7.9% vs 3.7%). Additionally, adult patients were less likely to receive radiation treatment than pediatric patients (24.3% vs 44.4%). When radiation was administered, it was after surgery in most pediatric and adult cases (Table 1).

Survival Analysis

The survival for all patients was analyzed by treatment era defined by NWTS 2, 3, and 4 trial publication dates (1973-1981, 1982-1989, 1990-1998, and 1999-2007) in Figure 1. There was a statistically significant improvement in the OS of all patients treated after 1981 (5-year OS, 75% ± 2.3%, 89% ± 1.6%, 89% ± 1.3%, and 89% ± 1.1%; P<.001), which would correspond to the introduction of multiagent chemotherapy with doxorubicin. A similar statistically significant improvement in the OS of the adult patients treated after 1981 was also seen (Figure 2A) when the analysis was restricted to the adult group (5-year OS, 38% ± 10.6%, 75% ± 10.8%, 72% ± 6.8%, and 74% ± 6.8%; P = .005). Figure 3A shows that there is a statistically significant worse OS for the adult group compared with the pediatric group (P<.001). After a median follow-up of 7.1 years (range, 0-34.9 years), 87.4% of pediatric patients were alive compared with only 65.8% of adult patients. The 5-, 10-, and 15-year OS estimates for the pediatric group were 88% ± 0.7%, 87% ± 0.8%, and 86% ± 0.9%, respectively. The corresponding OS rates for the adult group were 69% ± 4.1%, 61% ± 4.7%, and 57% ± 5.3%, respectively. Similarly, the DSS rate for the adult group was significantly worse than the pediatric group (Figure 3B), with DSS estimates of 89% ± 0.7%, 88% ± 0.7%, and 88% ± 0.8%, respectively, for the pediatric group and 73% ± 3.9%, 68% ± 4.5%, and 68% ± 4.5%, respectively, for the adult group (P<.001).

thumbnail image

Figure 1. Overall survival is shown for all patients, stratified by treatment era corresponding to the publication dates of National Wilms Tumor Study Group 2 (1981), 3 (1989), and 4 (1998).

Download figure to PowerPoint

thumbnail image

Figure 2. Overall survival is shown for adult patients, stratified by (A) treatment era, (B) Surveillance, Epidemiology and End Results stage, (C) surgery, and (D) radiation in all stages.

Download figure to PowerPoint

thumbnail image

Figure 3. (A) Overall survival and (B) disease-specific survival of the 2190 pediatric Wilms' tumor patients compared with the 152 adult Wilms' tumor patients in the Surveillance, Epidemiology and End Results database are shown.

Download figure to PowerPoint

In Figure 4, the expected survival, relative survival, and number of Wilms' tumor patients in each age category are displayed across the entire spectrum of ages. It is seen that the youngest patients (<1 year) have the highest relative survival and increasing age results in a decrease in relative survival until a nadir is reached around the second to third decade.

thumbnail image

Figure 4. Estimated 5-year expected survival (based on general population data), 5-year relative survival of the Surveillance, Epidemiology and End Results Wilms' tumor patients, and number of patients in each age category are shown, stratified by age. Error bars represent the standard error.

Download figure to PowerPoint

Univariate Analysis

Univariate analysis was performed using the log-rank test to compare treatment era, sex, race, first primary, SEER stage, lymph node sampling, surgery, and radiation treatment within the adult patients on the basis of overall survival. As shown in Figure 2, only treatment era, SEER stage, surgery, and radiation treatment were significant. The 5-year OS for localized Wilms' tumor was 90% ± 3.6%, regional was 32% ± 13.3%, and distant was 23% ± 9.4% (P<.001). Adult patients who underwent surgery had a significantly better OS than those who had biopsy only or received no surgery (5-year OS, 76% ± 4.0% vs 19.4% ± 12.2%; P<.001). Adult patients who did not receive radiation treatment had significantly better OS than those who did (5-year OS, 78% ± 4.3% vs 47% ± 8.8%; P = .004). It should be noted, however, that when radiation administration is analyzed within all distant stage patients (pediatric and adult), there is a statistically significant OS benefit to radiation treatment (5-year OS, 68% ± 4.6% vs 75% ± 2.3%; P = .025) (Figure 5). In Figure 6, the OS of the adult and pediatric groups are compared by stage. The adult patients had a statistically significant worse OS than pediatric patients for all stages including localized (5-year OS, 90% ± 3.6% vs 94% ± 0.8%; P = .001), regional (32% ± 13.3% vs 89% ± 1.4%; P<.001), and distant (23% ± 9.4% vs 76% ± 2.0%; P<.001). Additionally, the adult and pediatric groups were compared via lymph node sampling status on the basis of OS. The adult patients fared significantly worse than the pediatric patients whether they received a lymph node sampling (5-year OS, 74% ± 10.6% vs 92% ± 0.9%; P = .001) or did not receive a lymph node sampling (5-year OS, 75% ± 5.2% vs 85% ± 2.1%; P = .010).

thumbnail image

Figure 5. Overall survival in all distant stage patients (pediatric and adults) is shown, stratified by radiation administration.

Download figure to PowerPoint

thumbnail image

Figure 6. (A) Localized stage, (B) regional stage, and (C) distant stage Wilms' tumor are shown, stratified by age group.

Download figure to PowerPoint

Multivariate Analysis

Multivariate analysis was performed using a Cox proportional hazard model with OS as the endpoint. Only those patients with complete information available for all variables were included (complete case analysis technique, n = 2173 total [2045 pediatric, 128 adult]).28, 29 Lymph node sampling status was not included in the model because it is correlated with stage, and there was a high proportion of unknowns. As seen in Table 2, adult status (hazard ratio, 4.14; P<.001), SEER stage, and treatment era were significant prognostic factors across all patient groups. Adult patients diagnosed with Wilms' tumor after another primary malignancy had worse OS than those in which the Wilms' tumor diagnosis was their first primary malignancy (hazard ratio, 3.64; P = .025). Surgery was of significant benefit in pediatric patients (hazard ratio, 0.27; P<.001), but was not statistically significant for adult patients, though this is likely due to the limited number and reduced statistical power of the patients that did not undergo resection (n = 12). Radiation treatment was not a statistically significant predictor for either the pediatric or adult group.

Table 2. Multivariate Analysis
ParameterAll PatientsPediatric PatientsAdult Patients
HR95% CIPHR95% CIPHR95% CIP
  • CI, confidence interval; HR, hazard ratio; NA, not available; SEER, Surveillance, Epidemiology and End Results.

  • a

    There was only 1 patient with bilateral Wilms' tumor in the adult group.

Age category         
 Pediatric1.00Reference       
 Adult4.142.81-6.10<.001      
Sex         
 Male1.00Reference 1.00Reference 1.00Reference 
 Female0.910.72-1.15.4290.950.73-1.22.6670.710.34-1.49.369
Race         
 White1.00Reference 1.00Reference 1.00Reference 
 Black1.130.81-1.57.4681.000.70-1.44.9832.470.90-6.78.080
 Other/unknown1.540.97-2.44.0681.530.93-2.52.0942.260.60-8.52.228
Laterality         
 Unilateral1.00Reference 1.00Reference 1.00Reference 
 Bilateral0.830.51-1.38.4770.870.52-1.45.588<0.001NAa.987a
Primary malignancies         
 First primary1.00Reference 1.00Reference 1.00Reference 
 Not first primary1.860.77-4.46.1653.330.46-24.20.2353.641.17-11.28.025
SEER stage         
 Localized1.00Reference 1.00Reference 1.00Reference 
 Regional2.051.45-2.90<.0011.641.13-2.37.00910.253.91-26.87<.001
 Distant4.162.92-5.94<.0013.352.31-4.87<.00143.4011.72-160.6<.001
Treatment era         
 1973-19811.00Reference 1.00Reference 1.00Reference 
 1982-19890.400.27-0.59<.0010.400.26-0.60<.0010.380.10-1.38.142
 1990-19980.350.25-0.50<.0010.410.28-0.59<.0010.080.02-0.30<.001
 1999-20070.400.29-0.56<.0010.430.30-0.61<.0010.300.12-0.79.015
Surgery         
 No surgery or biopsy only1.00Reference 1.00Reference 1.00Reference 
 Surgery0.270.17-0.44<.0010.270.16-0.47<.0010.540.16-1.81.315
Radiation         
 None1.00Reference 1.00Reference 1.00Reference 
 External beam1.160.86-1.58.3361.210.87-1.67.2620.980.40-2.38.958

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

The analysis presented in this manuscript compares the outcome of 2142 pediatric Wilms' tumor patients with 152 adult Wilms' tumor patients in the SEER database. Adult patients were found to have a consistently worse outcome than pediatric patients. The etiology of this observed difference is likely multifactorial. It should be noted, though, that the histology of adult Wilms' tumors are presumed to be identical to pediatric tumors and, according to currently available literature, no biological differences are apparent.14, 20, 30 In fact, the 2 major Wilms' tumor trial groups, NWTS and SIOP, concluded from their retrospective review of patients ≥16 years in the NWTS 4-5 (n = 23 + 22) and SIOP 93-01 (n = 30) databases that adult patients, when treated on protocol in a similarly risk-adapted strategy as pediatric patients, may achieve outcomes comparable with pediatric patients (adult 5-year OS, 82.6% and 83%, respectively).19, 20

A major difference between this report and the NWTS and SIOP adult Wilms' tumor data is the percentage of adult Wilms' tumor patients within the database. The SEER database contains 152 adult patients out of a total 2342 Wilms' tumor patients (6.5%). In contrast, the NTWS report had 22 adult patients combined between the NWTS 4 and 5 databases compared with 1638 and 2387 pediatric patients, representing approximately 0.5%. The SIOP report included only 30 adult patients compared with 847 pediatric patients in the SIOP 93-01 trial publication, representing approximately 3%.26, 31 This difference in adult proportions between the SEER and trial group databases likely stems from the reluctance of adult oncologists, who may not have established institutional ties with either NWTS or SIOP, to register adult patients on a specific protocol. The current study is based on the SEER database, which includes a cohort of patients from geographically diverse rural and urban areas treated in academic and community practice settings and likely not treated on any specific trial. As a result, the SEER database appears to have captured a significantly higher percentage of adult Wilms' tumor patients than the pediatric trial databases and is likely more representative of community-based treatment practice trends for this rare tumor.

In analyzing the SEER patient characteristics, adult patients received far fewer lymph node samplings and were significantly more likely to be staged lower than pediatric patients, yet adults had a worse outcome on a stage-for-stage basis. This is in contrast to the SIOP report in which the SIOP adult patients generally presented at higher stages than the SIOP pediatric patients and had nearly twice the rate of irradiation as the SEER adult patients (46% vs 24.3%).20 As discussed, 74% of SEER adult patients did not receive any radiation treatment compared with only 54% of the SEER pediatric patients. It should be noted that both pediatric trial groups, NWTS and SIOP, require lymph node sampling in most of their pediatric protocols.32-36 If lymph node samplings are not performed, it is likely that the number of patients with regional disease would be underestimated, whereas the localized patients would have a worse outcome. The SEER data suggest that adult Wilms' tumor patients, when treated in the community and off protocol, tend to be inadequately staged and are therefore undertreated compared with their pediatric counterparts.

The inadequate staging of adult patients would also be expected to affect chemotherapy administration as well as radiation treatment. Although the SEER database does not have chemotherapy data, when the patients are stratified by treatment era, it is seen that those adult patients treated after 1981 (NWTS 2 trial publication year, which found a DSS benefit to the addition of doxorubicin) had a significantly improved OS. Thus, it is apparent that adult patients, like pediatric patients, benefit from modern chemotherapy regimens. This finding is confirmed by another study by Arrigo et al37 that found an improvement in the 3-year survival rate from 24% to 67% for a series of 24 adults treated with modern 3-drug chemotherapy, including doxorubicin. Current NWTS treatment recommendations include a 2-drug chemotherapy regimen for early stages and favorable histology and a ≥3-drug chemotherapy regimen plus radiation for advanced stages and unfavorable histology.19, 38 Thus, if adult patients were understaged, they might also be expected to be undertreated with chemotherapy.

Another factor likely contributing to the worse outcome of the SEER adult patients compared with protocol adult and pediatric patients may be treatment delays and inappropriate treatment regimens associated with the relative unfamiliarity of adult oncologists with established pediatric Wilms' tumor protocols. A recent mono-institutional retrospective study on 17 adult Wilms' tumor patients found a “very low compliance with specific diagnostic and therapeutic protocols” with a corresponding worse outcome for the adult patients compared with historical pediatric Wilms' tumor results.14 Additionally, the minimization of radiation treatment delay has been shown to be an important factor, because several studies on pediatric patients have concluded that delays in the initiation of radiation beyond 10 days after surgery results in a higher rate of recurrence.39, 40 It might also be expected that there would be significant delay associated with the pathological review of a rare Wilms' tumor diagnosis in an adult.

In addition to delays associated with a pathological diagnosis, a misdiagnosis of more aggressive tumors as Wilms' tumor might also be expected to result in worse outcomes for adults. Common tumors often pathologically confused with Wilms' tumor include renal adenocarcinoma, clear cell sarcoma of the kidney, renal teratoma, and malignant rhabdoid tumor of the kidney.41 In fact, in NWTS 4, it was found that 12 out of 1439 (0.8%) patients originally diagnosed with favorable histology Wilms' tumor by their institution actually had clear cell sarcoma of the kidney on central pathology review.42 A significant portion of the pediatric patients in the SEER database likely have had central pathology review due to concurrent registration on NWTS protocols; however, it is probable that only a small portion of registered adult cases would have received such a thorough examination. A higher rate of pathological misdiagnosis might also result in undertreatment and worse outcome for adult patients. Additionally, a higher percentage of anaplastic tumors in the adult patients would certainly account for the worse outcome of the adult patient group. However, the report by the SIOP group documented only 2 out of 30 adult Wilms' tumors (6.7%) that were anaplastic and only 3 tumors (10%) that were clear cell sarcoma of the kidney.20 Notably, these rates were comparable to those of the pediatric patients in that same SIOP 93-01 trial (7.0% high risk histology and 5.0% clear cell sarcoma of the kidney).43

In conclusion, although previous studies have shown comparable outcomes between adult and pediatric patient populations when adults are treated on protocol, our analysis of the SEER database finds that adults have significantly worse outcome than pediatric patients. This worse outcome likely stems from inadequate staging and incomplete diagnostic evaluations in the general community resulting in treatment delays and undertreatment with radiation as well as possibly chemotherapy. Thus, this body of work serves to support the recommendation for lymph node sampling in all adult Wilms' tumor patients, close collaboration with pediatric surgeons and oncologists, and protocol-based treatment regimens. Additionally, it should be noted that the Children's Oncology Group (COG) has recently raised the enrollment age for their trials to 30 years, hence institutions should consider enrolling all eligible adult patients on COG trials when possible.

FUNDING SOURCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

No specific funding was disclosed.

CONFLICT OF INTEREST

The authors made no disclosures.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES