|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Older Age Is an Adverse Prognostic Factor in Stage I, Favorable Histology Wilms’ Tumor Treated With Vincristine Monochemotherapy: A Study by the United Kingdom Children’s Cancer Study Group, Wilm’s Tumor Working Group
From the Department of Paediatric Oncology, Royal Marsden Hospital/Institute of Cancer Research, Sutton; Department of Histopathology, Royal Manchester Children’s Hospital, Manchester; Department of Histopathology, University Hospital of Wales, Cardiff; United Kingdom Children’s Cancer Study Group Data Centre, Leicester; and John Radcliffe Hospital, Oxford, United Kingdom. Address reprint requests to K. Pritchard-Jones, MD, Department of Paediatric Oncology, Institute of Cancer Research/Royal Marsden Hospital, Brookes Lawley Bldg, Cotswold Rd, Sutton, Surrey, SM2 5NG, United Kingdom; e-mail: kathy.pritchard-jones{at}icr.ac.uk.
Purpose: To identify clinical prognostic factors in children with stage I, favorable histology (FH) Wilms’ tumor treated with vincristine monochemotherapy after immediate nephrectomy to define subgroups for consideration of further reduction in treatment intensity. Patients and Methods: During two consecutive trials of the United Kingdom Children’s Cancer Study Group (UKW2 and UKW3, 1986 to 2001), 242 children with stage I FH Wilms’ tumor were treated with immediate nephrectomy followed by 10 weekly injections of vincristine 1.5 mg/m2. Event-free survival (EFS) and overall survival (OS) were compared by age group. Results: The 4-year EFS rate was 93.2%, 87.2%, and 71.3% for children less than 2 years old, 2 to 4 years old, and 4 years old or older at diagnosis, respectively (log-rank, P = .001); the corresponding 4-year OS rate was 98.1%, 95.0%, and 87.2% (log-rank, P = .01). There were no toxicity- or procedure-related deaths. In multivariate analysis, specimen weight was not of independent prognostic value (P = .66). Among the 186 children younger than 4 years at diagnosis, there were 17 relapses and five deaths, compared with 16 relapses and eight deaths among the 56 children at least 4 years old at diagnosis. OS after relapse was surprisingly poor (61.6% at 4 years). Conclusion: Treatment for stage I FH Wilms’ tumor is generally successful using vincristine monotherapy after immediate nephrectomy, and therefore, the risks of dactinomycin hepatopathy can be avoided. However, age at least 4 years is a significant adverse prognostic factor. This treatment schedule should be considered in any trial of treatment reduction in very young children with stage I FH Wilms’ tumor, regardless of tumor size, and we suggest that the upper age limit for the reduced therapy be set at 4 years.
WILMS’ TUMOR (nephroblastoma) is an embryonal kidney cancer of childhood that reflects the close relationship between malformation and tumor development. Approximately 85% of children with Wilms’ tumor are long-term survivors of their disease with relatively simple chemotherapy and surgical removal of the affected kidney.1–3 The success of treatment for children with Wilms’ tumor is now at a level where increasing consideration is given to the balance of risk versus benefit for each treatment modality. Such considerations are particularly appropriate in the group of Wilms’ tumors with the lowest risk of relapse, especially stage I favorable histology (FH) Wilms’ tumors, which has an expected overall survival (OS) rate of 95%. Both the National Wilms’ Tumor Study Group (NWTSG) 5 trial and the International Society of Pediatric Oncology (SIOP) 93–01 study focused on treatment reduction for this group of patients.4 Indeed, the NWTSG 5 trial included a treatment arm in which no adjuvant treatment was given after complete nephrectomy for small stage I tumors in children less than 2 years of age at diagnosis, based on their excellent survival in the previous study. This arm was closed prematurely in 1998 when the relapse-free survival decreased below the stringently set lower limit of 90%.5,6 However, most patients treated without chemotherapy had no recurrence, indicating that stage I Wilms’ tumor in young children can often be cured with surgery alone. In the SIOP 93–01 study, the investigators asked a randomized question about the duration of postoperative chemotherapy and concluded that a total of eight doses of vincristine and three fractionated doses of dactinomycin were sufficient treatment for stage I Wilms’ tumor without histologic evidence of anaplasia.4 The United Kingdom has pioneered the treatment of stage I, FH Wilms’ tumor using only vincristine chemotherapy after immediate nephrectomy. This approach was first adopted in the Medical Research Council national trials that ran during the 1970s and in which vincristine was combined with surgery and postoperative radiotherapy.7 Radiotherapy was dropped for stage I patients in the United Kingdom Children’s Cancer Study Group (UKCCSG) UKW1 trial, which ran from 1980 to 1986.8 In this study, 104 patients with stage I, FH Wilms’ tumor were treated with weekly intravenous vincristine 1.5 mg/m2 for 10 to 11 consecutive weeks followed by a further five doses in 3-week intervals, for a total treatment duration of 25 to 26 weeks. The event-free survival (EFS) in this group was 90%. In the UKW2 trial, which ran from 1986 to 1991 and recruited 136 patients with stage I, FH tumors, the duration of vincristine chemotherapy after immediate nephrectomy was shortened to just 10 weekly injections at the same dosage.1 This treatment schedule was also adopted as the standard arm for stage I tumors after immediate nephrectomy in the UKW3 trial. This trial was designed as a randomized study to compare the EFS and stage distribution of localized tumors treated with nephrectomy followed by chemotherapy (the NWTSG approach) versus preoperative chemotherapy and delayed nephrectomy, with both arms receiving stage-appropriate postoperative treatment. In this article, we analyze the outcome of 242 patients with stage I, FH Wilms’ tumor treated with nephrectomy followed by vincristine monochemotherapy in the UKW2 and UKW3 trials to identify factors that may aid patient selection for this treatment method and for other treatment modifications that may be contemplated in the future.
Patients Eligible patients for inclusion in this analysis were less than 16 years old at diagnosis, had a stage I, FH Wilms’ tumor based on the local pathologist’s assessment, and had been treated with immediate nephrectomy followed by 10 weekly injections of intravenous vincristine in the successive UKW2 (1986 to 1991) and UKW3 (1991 to 2001) clinical trials. Although central pathology review was performed on 87% of specimens, the results were not generally available to influence treatment decisions. Therefore, this analysis is based on treatment received according to the local pathologist’s assessment of stage and histology. Ten patients with stage I, FH tumors diagnosed during the same period were excluded from this analysis because, contrary to the recommended protocol, six were given no chemotherapy and four had received additional dactinomycin. During the study period, a total of 1,046 children were diagnosed with Wilms’ tumor in the United Kingdom, representing 96% of all patients with this diagnosis recorded in the national population-based childhood cancer registry held at the Childhood Cancer Research Group in Oxford, United Kingdom (C. Stiller, personal communication, November 2002). The UKW2 and UKW3 trials were approved by the local research ethics committee at each participating hospital, and written informed consent was obtained from the parent or guardian of each patient. Absence of lung metastases was determined by chest radiograph in two planes (or in only one plane if the child also had a computed tomography [CT] scan of the thorax). Absence of intra-abdominal metastases was determined by ultrasound or CT scan and by inspection at the time of surgery. Tumor stage and histologic subtype were defined by the local pathologist’s assessment. During both trials, stage I disease was defined as it was in the NWTSG 4 study,3 as tumor limited to the kidney and completely excised, with the surface of the renal capsule intact, with no evidence of rupture before or during removal and no residual tumor apparent beyond the margins of resection. The microsubstaging variables identified by Weeks et al9 and used in the NWTSG 5 study10 were not applied in either trial because their prognostic significance was not established at the time of their conception. Anaplasia was defined by standard morphologic criteria, and both focal and diffuse anaplastic tumors were excluded from this treatment arm in the UKW2 and UKW3 trials. Central pathology review of the nephrectomy specimen was requested on all tumors as part of both trials. There was central review of histology in 86.2% of the tumors and of tumor stage in 83.1% of the tumors treated in the UKW2 trial; the lower percentage of central review of tumor staging reflects the adequacy of material submitted. In the UKW3 trial, central review was only pursued in patients who were either randomized or included in this study. In these two groups, the proportion of patients who had review of histology and tumor stage was 86% and 91%, respectively. Details of all chemotherapy doses given and toxicity (World Health Organization) were documented for each patient on individual flow sheets.
Treatment Schedule
Statistical Methods Confidence intervals (CIs) for EFS and OS were calculated using Rothman’s method.12 The prognostic variables of age and tumor weight were tested separately using the log-rank test. These variables were also entered sequentially into a multivariate Cox proportional hazards model. Variables representing the protocols (ie, UKW2 and UKW3) were added to the final model to allow for any differential effect of the protocol. The proportional hazards assumption was checked by visual inspection of log-log survival plots.
The continuous variables of age (under 2 years v 2 to 4 years v
Two hundred forty-two children diagnosed with stage I, FH Wilms’ tumor between June 1986 and March 2001 were eligible for and enrolled onto this study. Ten patients with stage I, FH tumors who were diagnosed during the same period were not included because six of these patients were given no chemotherapy and four were given dactinomycin in addition to vincristine. The mean and median vincristine doses received per course are shown in Table 1  . Overall, infants or children less than 10 kg received a median vincristine dose of 50 µg/kg compared with 63 to 64 µg/kg for older children or those weighing more than 10 kg. Only 12 patients (5%) required a dose reduction of more than 25% of their calculated protocol dose. The median number of courses requiring modification was 2.5 (range, one to nine courses).
Toxicity Grade 3 or 4 toxicity in any system was reported in 35 patients, as listed in Table 2 . One of the most problematic side effects of vincristine is neurotoxicity, so all grades of neurotoxicity were examined in detail, as listed in Table 3. These latter data represent a minimum because it is possible that not all the clinically anticipated low-grade toxicity was systematically reported. Fifty-nine patients (24%) were reported as experiencing a maximum of grade 1 or 2 neurotoxicity, consisting mainly of pain or paraesthesia, weakness, or constipation. These symptoms generally responded either to simple analgesia or laxatives and did not require dose reductions. Only six patients had grade 3 or 4 neurotoxicity, including seizures in two patients, grade 3 toxicity (stumbling gait, absent reflexes, and jaw pain) in three patients, and severe ptosis in one patient. However, even in these patients, dose reductions or omissions were short lived, and all patients received at least seven doses of vincristine. None of these patients relapsed.
Survival At a median follow-up of 9.21 years (range, 0.4 to 11.9 years; median follow-up, 10.3 years in UKW2 and 4.69 years in UKW3), there have been a total of 34 events, comprising 33 relapses, with 12 patients dying from uncontrollable tumor progression, and one late death from a non–tumor-related cause. EFS and OS for the entire group are shown in Fig 1 . The 2- and 4-year OS rates are excellent at 96.5% (95% CI, 93.3% to 98.2%) and 94.7% (95% CI, 90.9% to 96.9%), respectively. The 2- and 4-year EFS rates are also satisfactory at 87.9% (95% CI, 83.1% to 91.4%) and 86.5% (95% CI, 81.5% to 90.2%), respectively. Most of the tumor-related events (88%, 29 of 33 events) occurred within 2 years of initial diagnosis.
The age at diagnosis of patients with stage I, FH Wilms’ tumor is significantly younger than other Wilms’ tumors, with a median age at diagnosis of 1.95 years for stage I, FH Wilms’ tumors compared with 3.2 years for all other Wilms’ tumor patients diagnosed during the same period (Mann-Whitney, P < .0001). The age distribution was as follows: 122 patients (50%) were less than 2 years old, 64 (26%) were 2 to 4 years old, and 56 (23%) were  4 years old. When EFS was stratified by age at diagnosis, there was a striking difference, with children aged 4 years having a significantly worse 4-year EFS (71.3%; 95% CI, 58.3% to 81.5%) than younger patients (Fig 2A). There was a smaller but still significant adverse impact of older age on OS (Fig 2B).
To allow comparison of our data with the surgery-only approach of the NWTSG 5 trial, we performed a multivariate analysis of the combined impact of age and specimen weight on EFS and OS. The same cutoff weight of 550 g, as in the NWTSG study, was used to define smaller tumors.5,6 There were 113 nephrectomy specimens weighing less than 550 g; among these patients, there were 13 events and two deaths. Ninety-one patients had a recorded specimen weight of 550 g, and in this group, there were 16 events and seven deaths. Thirty-eight patients had no specimen weight recorded, and in this group, there were five events and four deaths. Although tumor weight showed a significant positive correlation with age (Spearman rank correlation, n = 242, P = .02), the distribution of events in relation to specimen weight was not significant. This remains true if we restrict the analysis to patients less than 2 years old at diagnosis, in which the 4-year EFS rates were 89.9% (95% CI, 79.5% to 95.3%) and 95.1% (95% CI, 83.6% to 98.6%) for patients with a nephrectomy specimen weight less than 550 g (n = 64) or 550 g (n = 41), respectively. Corresponding 4-year OS rates were 98.4% (95% CI, 91.8% to 99.7%) and 97.4% (95% CI, 86.8% to 99.5%), with only one death in each group.
In multivariate analysis, age
Sites of Relapse Thirty three patients experienced a tumor recurrence, which was confined to the lungs in 21 patients and to the tumor bed in seven patients, or occurred as a metachronous lesion in the remaining kidney in two patients. One patient relapsed in the lungs and flank, and one patient each developed bone or brain metastases. Median follow-up from time of relapse was 8 years. The OS rate after relapse was disappointing, considering that first-line therapy had been relatively modest, with a 4-year OS of 61.6% (95% CI, 44.1% to 76.5%; Fig 3 ). The site of relapse did not bear any clear relationship to any alterations of abdominal tumor stage by the review panel (see Pathology).
Pathology Central pathology review was performed in 87% of the tumors in the entire group. The initial nephrectomy specimen had been centrally reviewed in 30 of the 33 relapsing patients, without knowledge of relapse status by the reviewing pathologist. Twenty-five tumors were confirmed as stage I tumors and three had a higher tumor stage assigned by the panel (two stage II tumors and one stage III tumor, all three relapsed in the lungs alone); in two patients, there was insufficient material to assign a review stage (one patient relapsed in the lungs, and the other relapsed in the tumor bed). This 12% rate of discrepancy between local and review staging compares with a rate of 4% for the 196 of 209 nonrelapsing tumors that had pathologic stage review (P = .145, Fisher’s exact test). In view of this apparent difference, additional tumor material was received for review in eight of 15 relapsing patients treated in the UKW3 period. Assigned tumor stage was raised by the panel in five of these patients (four stage II and one stage III), and one case of clear-cell sarcoma was diagnosed. However, a similar level of in-depth central review was not performed on the nonrelapsing patients, and it should be noted that the nonrelapsing patients included six patients whose assigned tumor stage was raised to stage II (n = 5) or stage III (n = 1) by the review panel but who had been successfully treated with vincristine alone. Review of histology confirmed 28 tumors to be FH Wilms’ tumor and showed one tumor to be a clear-cell sarcoma of the kidney and one tumor to be an anaplastic Wilms’ tumor. Both of these patients subsequently died of disease progression despite appropriate second-line therapy after relapse.
The results of this analysis show that short-duration, weekly vincristine chemotherapy is a well-tolerated and very effective treatment for stage I, FH Wilms’ tumor after immediate nephrectomy. It is of importance that half these patients are under 2 years old, an age group in which the potentially life-threatening hepatotoxicity of additional dactinomycin is of particular concern.14 The age distribution of patients with stage I, FH tumors is virtually identical in this study to that of the NWTSG 3 study, in which 51% of patients were less than 2 years old, and 21% were greater than 4 years old.15
Age
An alternative or additional explanation for the better prognosis of children less than 4 years old may relate to vincristine dose exposure, for which there are few data in younger children. The original infant dose modifications in UKW2 were conservative, leading to some children receiving only a quarter of the dose received by a child just over 1 year of age. Because of some unexpected relapses in this age group, adjustments were made to the recommended dose modifications. As shown in Table 1
The prognostic significance of age was also noted in the first National Wilms’ Tumor Study, in which children with stage I tumors were randomly assigned to single-agent dactinomycin with or without postoperative radiotherapy.19 The 2-year relapse-free survival was 89% for children less than 2 years old at diagnosis compared with 67% for children Analysis of the relapsing patients in this study underlines the importance of meticulous abdominal staging of Wilms’ tumor, although the discrepancy between local and central review of tumor stage did not reach statistical significance. The incidence of isolated relapse in the lungs (64%) in this study is similar to the rate (71%) reported in the NWTSG 3 study, in which children with stage I tumors received chemotherapy with two drugs.15 This observation indicates that the use of only chest radiographs to exclude lung metastases at diagnosis does not miss clinically significant micrometatases when chemotherapy consists of vincristine alone. However, a related UKCCSG study on the role of CT scanning of the lungs in children with Wilms’ tumor reached a different conclusion.20 Eight of the patients with pulmonary relapse from this study, all treated in the UKW2 trial, were included in the analysis of Owens et al.20 Owens et al demonstrated that, of nine children with stage I, FH Wilms’ tumor with normal chest x-rays but nodules noted on chest CT scan, five relapsed. Based on these data, the current recommendation in the United Kingdom is that no child should be considered for vincristine monotherapy unless they have a negative chest CT scan. However, there is clearly a need to collect further data on this clinically important point. Such prospective data collection forms part of the newly opened Wilms’ tumor 2001 trial of SIOP. It is also apparent from the NWTSG 5 study that a substantial subset of these younger patients with stage I, FH Wilms’ tumor are curable by surgery alone and presumably do not have micrometastases. This is confirmed in our study by the outcome of the six patients excluded from the survival and treatment analyses because they were not given chemotherapy. Five of these children never relapsed, and the sixth patient is also a long-term survivor, having received treatment for pulmonary metastases. The future challenge lies in the selection of patients who do not need adjuvant chemotherapy of any sort.
In conclusion, for children treated for Wilms’ tumor by immediate nephrectomy and shown to have a stage I, FH tumor, single-agent short-duration vincristine chemotherapy is a safe and effective treatment. Our data indicate that children aged
We thank all the staff involved in the management of children with Wilms’ tumor at the following institutions, which participated in the UKW2 and UKW3 clinical trials: United Kingdom Children’s Cancer Study Group (UKCCSG) centers: Addenbrooke’s Hospital, Cambridge; Alder Hey Children’s Hospital, Liverpool; St Bartholomew’s and The Royal London Hospital; The Hospital for Sick Children, Great Ormond Street; and The Middlesex Hospital, London; John Radcliffe Hospital, Oxford; Leicester Royal Infirmary, Leicester; Llandough Hospital, Cardiff; Our Lady’s Hospital for Sick Children, Dublin; Queen’s Medical Centre, Nottingham; Royal Aberdeen Children’s Hospital, Aberdeen; Royal Hospital for Sick Children, Belfast; Royal Hospital for Sick Children, Bristol; Royal Hospital for Sick Children, Edinburgh; Royal Hospital for Sick Children, Glasgow; Royal Manchester Children’s Hospital, Manchester; Royal Marsden Hospital, Sutton; Royal Victoria Infirmary, Newcastle-on-Tyne; Sheffield Children’s Hospital, Sheffield; Southampton General Hospital, Southampton; St James’ University Hospital, Leeds; and The Children’s Hospital, Birmingham, United Kingdom; other centers: Rikshospitalet, Olso, Norway; and Royal Hospital for Sick Children, Melbourne; and Adelaide Women and Children’s Hospital, Adelaide, Australia. We also thank the following members of the UKCCSG Wilms’ Tumor Working Group for helpful comments and discussion: Mr. Peter Gornall, Dr Richard Grundy, Dr Juliet Hale, Dr Gill Levitt, Dr Cathy Owens, Dr Pat Sartori, Dr Rosemary Shannon, Dr Roger Taylor, and Miss Jenny Walker, the UKCCSG Data Centre, especially Helen Middleton and Kath Robinson for data management, and Barbara Smith for secretarial support.
We thank Dr Pat Morris Jones and Dr John Martin for the original idea of using vincristine monotherapy, all clinicians caring for children with Wilms’ tumor at participating hospitals in the UKW2 and UKW3 trials, staff at the United Kingdom Children’s Cancer Study Group Data Centre, Dr Jon Pritchard for critical review of the manuscript, and Cancer Research United Kingdom and the Royal Marsden Hospital Charitable Funds for financial support. (See online appendix for full list of contributors and acknowledgments.)
Supported by Cancer Research United Kingdom and Royal Marsden Hospital Charitable Funds, Sutton, United Kingdom. Presented at the Thirty-Fourth Congress of the International Society of Paediatric Oncology, Porto, Portugal, October 11, 2003.
1. Mitchell C, Jones PM, Kelsey A, et al: The treatment of Wilms’ tumour: Results of the United Kingdom Children’s Cancer Study Group (UKCCSG) second Wilms’ tumour study. Br J Cancer 83:602–608, 2000[CrossRef][Medline]
2. Tournade MF, Com-Nougue C, de Kraker J, et al: Optimal duration of preoperative therapy in unilateral and nonmetastatic Wilms’ tumor in children older than 6 months: Results of the Ninth International Society of Pediatric Oncology Wilms’ Tumor Trial and Study. J Clin Oncol 19:488–500, 2001
3. Green DM, Breslow NE, Beckwith JB, et al: Effect of duration of treatment on treatment outcome and cost of treatment for Wilms’ tumor: A report from the National Wilms’ Tumor Study Group. J Clin Oncol 16:3744–3751, 1998 4. De Kraker J, Graf N, Pein F, et al: The SIOP-Nephroblastoma Study 93-01 two years after closure for randomisation: An update. Med Pediatr Oncol 39:235, 2002 (abstr O060) 5. Green DM, Breslow NE, Beckwith JB, et al: Treatment outcomes in patients less than 2 years of age with small, stage I, favorable-histology Wilms’ tumors: A report from the National Wilms’ Tumor Study. J Clin Oncol 11:91–95, 1993[Abstract]
6. Green DM, Breslow NE, Beckwith JB, et al: Treatment with nephrectomy only for small, stage I/favorable histology Wilms’ tumor: A report from the National Wilms’ Tumor Study Group. J Clin Oncol 19:3719–3724, 2001
7. Medical Research Council’s Working Party on Embryonal Tumors in Childhood: Management of nephroblastoma in childhood: Clinical study of two forms of maintenance chemotherapy. Arch Dis Child 53:112–119, 1978
8. Pritchard J, Imeson J, Barnes J, et al: Results of the United Kingdom Children’s Cancer Study Group first Wilms’ Tumor Study. J Clin Oncol 13:124–133, 1995 9. Weeks DA, Beckwith JB, Luckey DW: Relapse-associated variables in stage I favorable histology Wilms’ tumor: A report of the National Wilms’ Tumor Study. Cancer 60:1204–1212, 1987[CrossRef][Medline] 10. Green DM, Beckwith JB, Weeks DA, et al: The relationship between microsubstaging variable, age at diagnosis, and tumor weight of children with stage I/favorable histology Wilms’ tumor: A report from the National Wilms’ Tumor Study. Cancer 74:1817–1820, 1994[CrossRef][Medline] 11. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457–481, 1958[CrossRef] 12. Parmar MKB, Machin D: Survival Analysis: A Practical Approach. Chichester, England, United Kingdom, J Wiley & Sons, 1995 13. Armitage P, Berry G: Statistical Methods in Medical Research (ed 3). Oxford, England, United Kingdom, Blackwell Science, 1994 14. Bisgno G, de Kraker J, Weirich A, et al: Veno-occlusive disease of the liver in children treated for Wilms’ tumor. Med Pediatr Oncol 29:245–251, 1997[CrossRef][Medline] 15. Breslow N, Sharples K, Beckwith JB, et al: Prognostic factors in nonmetastatic, favorable histology Wilms’ tumor: Results of the Third National Wilms’ Tumor Study. Cancer 68:2345–2353, 1991[CrossRef][Medline] 16. Crom WR, de Graaf SSN, Synold T, et al: Pharmacokinetics of vincristine in children and adolescents with acute lymphocytic leukaemia. J Pediatr 125:642–649, 1994[CrossRef][Medline] 17. Gidding CEM, Meeuwsen-de Boer GJ, Koopmans P, et al: Vincristine pharmacokinetics after repetitive dosing in children. Cancer Chemother Pharmacol 44:203–209, 1999[CrossRef][Medline] 18. Groninger E, Meeuwsen-de Boer T, Koopmans P, et al: Pharmacokinetics of vincristine monotherapy in childhood acute lymphoblastic leukaemia. Pediatr Res 52:113–118, 2002[Medline] 19. D’Angio GJ, Evans AE, Breslow N, et al: The treatment of Wilms’ tumor: Results of the National Wilms’ Tumor Study. Cancer 38:633–646, 1976[CrossRef][Medline]
20. Owens CM, Veys PA Pritchard J, et al: Role of chest computed tomography at diagnosis in the management of Wilms’ tumor: A study by the United Kingdom Children’s Cancer Study Group. J Clin Oncol 20:2768–2773, 2002 21. Shaul DB, Scheer B, Rokhsar S, et al: Risk factors for early infection of central venous catheters in pediatric patients. J Am Coll Surg 186:654–658, 1998[CrossRef][Medline] Submitted January 10, 2003; accepted June 10, 2003.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
