当前位置: 首页 > 期刊 > 《临床肿瘤学》 > 2005年第9期 > 正文
编号:11332312
Hyperdiploidy Plus Nonamplified MYCN Confers a Favorable Prognosis in Children 12 to 18 Months Old With Disseminated Neuroblastoma: A Pediat
http://www.100md.com 《临床肿瘤学》
     the Dana-Farber Cancer Institute, Harvard Medical School

    Pediatric Hematology/Oncology, Boston Floating Hospital for Infants & Children, Boston, MA

    Children's Oncology Group, University of Florida, Gainesville, FL

    Northwestern University, Feinberg School of Medicine, Chicago, IL

    Children's Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia, PA

    University of Alabama at Birmingham, Birmingham, AL

    ABSTRACT

    PURPOSE: To determine predictive strength of tumor cell ploidy and MYCN gene amplification on survival of children older than 12 months with disseminated neuroblastoma (NB).

    PATIENTS AND METHODS: Of 648 children with stage D NB enrolled onto the Pediatric Oncology Group NB Biology Study 9047 (1990-2000), 560 children were assessable for ploidy and MYCN amplification. Treatment of patients older than 12 months varied; most receiving high-dose chemotherapy with stem-cell rescue. Infants received standard chemotherapy, depending on MYCN status and ploidy.

    RESULTS: Among stage D MYCN-amplified patients, 4-year event-free survival (EFS) ± SE had no prognostic significance for tumor cell ploidy for patients either younger than 12 months or 12 months old. However, among stage D nonamplified-MYCN patients, 4-year EFS for those with tumor hyperdiploidy (DNA index [DI] > 1) was clearly superior to those with diploidy (DI 1): younger than 12 months, 83.7% ± 4.4% (n = 87) versus 46.2% ± 13.8% (n = 13; P = .0003); and for 12- to 24-month-old children, 72.7% ± 10.2% (n = 22) versus 26.7% ± 13.2% (n = 16; P = .0092). Further analysis suggested better prognoses in the 12- to 18-month-old subgroup with hyperdiploid tumors (4-year EFS, 92.9% ± 7.2%) compared with the 19- to 24-month-old subgroup (4-year EFS, 37.5% ± 21.0%; P = .0037). In children older than 24 months, outcome was dire (< 20% long-term survival), regardless of ploidy or MYCN status.

    CONCLUSION: Children 12 to 18 months old with metastatic NB had favorable outcomes with high-dose therapy if their tumors were hyperdiploid and lacked MYCN amplification. This subgroup may respond well to contemporary chemotherapy, and could be spared intensive myeloablative therapy with stem-cell rescue.

    INTRODUCTION

    Neuroblastoma, the most common extracranial solid tumor in children, is derived from primordial neural crest cells that ultimately populate the sympathetic ganglia and adrenal medulla. The heterogeneity of the genetic and histopathologic features of these neuroblasts, together with their diverse anatomic locations, results in a broad spectrum of clinical manifestations, from spontaneous regression of tumors in some infants younger than 12 months to widespread metastasis of drug-resistant tumors in older children. Although overall survival rates for neuroblastoma have improved only modestly over the past several decades, there has been marked progress in the ability to recognize prognostic subgroups based on clinical and biologic risk factors, including age,1 disease stage,2 tumor cell DNA index (ploidy),3 MYCN oncogene copy number,4,5 and histology.6,7 In general, these studies have demonstrated a high cure rate for patients with resectable tumors, regardless of age, primary site, or biologic risk factors,8 whereas the responses of patients with metastatic neuroblastoma have been poor, with some notable exceptions such as infants.9-12 For example, we have shown (1) that hyperdiploidy in combination with a nonamplified MYCN copy number identifies a subgroup of infants with metastatic disease who respond well to standard therapy with cyclophosphamide and doxorubicin, (2) that diploidy is associated with poorer outcome compared with hyperdiploidy in infants with advanced disease and a nonamplified MYCN copy number treated on Pediatric Oncology Group (POG) protocols, and (3) that MYCN amplification is a statistically significant marker of higher-risk disease within both the diploid and hyperdiploid subgroups.3,13-15

    Despite the utility of tumor cell ploidy and other biologic variables in assigning infants to risk-based therapies, comparable progress has not been achieved in older children with metastatic tumors, whose 3-year survival rates still range from 15% to 46%.11,16,17 Our previous observations on a limited number of patients suggested that the prognostic importance of neuroblast genetic markers might not be confined to infants, but might extend into the 12- to 24-month-old subgroup, which constitutes approximately 20% of children with neuroblastoma.13 To test this prediction, we reviewed the medical records of 1,667 patients enrolled onto a single POG study from 1990 to 1999 (Neuroblastoma Biology Study 9047). The event-free survival (EFS) of patients with amplified or nonamplified MYCN stage D tumors was then analyzed by age and ploidy classification. The results indicate that children 12 to 18 months old with disseminated hyperdiploid neuroblastoma without MYCN amplification have a very favorable prognosis, similar to previous findings in infants,3,13-15 and may benefit from less aggressive therapy.

    PATIENTS AND METHODS

    Study Population and Eligibility Criteria

    All patients in this analysis were enrolled onto POG Neuroblastoma Biology Study 9047. To be eligible for registration, patients had to be younger than 21 years with newly diagnosed neuroblastoma and to have tumor tissue available for biologic studies, which included determinations of DNA ploidy and MYCN gene status at the POG Reference Laboratories (Memphis, TN; St Louis, MO; Boston, MA) under the supervision of Drs G.M. Brodeur and A.T. Look. Informed consent for POG 9047 was obtained for all patients. Patients registered early during the study period were staged according to POG criteria,8 as follows: stage A, localized resectable tumors; stage B, localized but unresectable tumors with negative noncontiguous lymph nodes; stage C, metastasis to noncontiguous lymph nodes; stage D, metastasis beyond the lymph nodes; and stage Ds, metastasis limited to liver, skin, and bone marrow. Most patients registered during the latter part of the study period were staged according to International Neuroblastoma Staging System (INSS) criteria.18 Although the two staging systems differ in some respects, their major provisions are similar. Age was defined as the age at diagnosis.

    There were 2,429 patients enrolled onto study 9047 from February 7, 1990, to April 5, 2001. In order to have sufficient follow-up observations for the survival analyses, we included only patients registered before December 31, 1999 (n = 1,667), 648 of whom had stage D tumors.

    Treatment

    Treatments administered to these children were not uniform, consisting of regimens specified by individual institutional protocols as well as different POG protocols. In general, patients younger than 12 months were treated on or according to a phase 3 study of "intermediate-risk" neuroblastoma (POG 9243), open from 1992 to 1996. The treatment of infants with stage D neuroblastoma was stratified according to MYCN status and tumor cell ploidy. Patients with hyperdiploid nonamplified MYCN tumors were treated with five to seven cycles of cyclophosphamide and doxorubicin with additional platinum and etoposide, depending on response. Patients with diploid or MYCN amplified tumors were treated with cisplatin and teniposide15 or carboplatin and etoposide, alternating with ifosfamide and etoposide for eight to 12 cycles (manuscript submitted for publication). Children older than 12 months were treated primarily with intensive platinum- and alkylator-based chemotherapy. Many, but not all, patients received consolidation therapy with bone marrow or stem-cell support, regardless of their ploidy status. Informed consent was obtained for all patients enrolled onto the treatment protocols, and all studies were approved by the institutional review boards of the participating institutions.

    Analysis of DNA Index

    DNA index (DI) or tumor cell ploidy was determined as described previously.13 A tumor stem line was considered to have a DNA content indistinguishable from that of normal diploid cells (DI of 1.0) if the percentage of cells in the diploid G0/G1 peak of the DNA histogram was at least 30% greater than the percentage of normal blood leukocytes determined from the morphologic analysis. Hyperdiploid stem lines were those with a DNA index of more than 1.0. For patients with multiple tumor stem lines, analysis of the impact of ploidy on prognosis was based on the DNA index of the lowest ploidy stem line.

    Analysis of MYCN Amplification

    Before July 1993, MYCN copy number was determined by Southern blot analysis using previously described standard methods.4,19 All samples were studied at least in duplicate, and the MYCN copy number of amplified samples (defined as > 3 copies of the MYCN gene per haploid genome or > 6 copies per cell) was determined by serial dilution and laser densitometry. After July 1993, fluorescence in situ hybridization (FISH) was used to determine the presence of MYCN amplification as previously described.20,21 MYCN amplification status was determined with a cosmid clone from the MYCN genomic locus used as a probe20 or with a labeled MYCN probe purchased from Vysis Inc (Downers Grove, IL). Tumors with cells containing 10 copies of MYCN or homogenously staining regions that hybridized to the MYCN probe were considered amplified.

    The results of Southern blotting and FISH analysis were prospectively compared and a MYCN copy number of 10 was determined to be the optimal cutoff by FISH,21 as the vast majority of amplified tumors have very large numbers of double minutes in each tumor cell. By Southern blotting, any normal cells in the tissue were included in the measurement, whereas by FISH, each tumor nucleus was visualized directly. Thus, in our estimation, the two cutoff points are as equivalent as possible in terms of reliably designating MYCN gene amplification.

    Statistical Analysis

    Tests of association were performed with the 2 or Fisher's exact test (for categorical variables), with a logistic regression model (for dependent binary variables), and with an ordinal logistic regression model (for ordinal dependent variables). Survival curves were constructed by the methods of Kaplan and Meier,22 with the log-rank test used to determine the prognostic significance of selected variables. Failures, or "events," are defined as relapse, progressive disease, secondary malignancy, or death. Analysis of EFS was based on the time to occurrence of the first event or the time to last contact with the patient, if no event occurred. The end point for analysis of overall survival (OS) was the time of death or the time of last contact (for surviving patients). EFS and OS rates were calculated as the rate ± SE. P values of less than .05 were considered statistically significant. Sample sizes varied from analysis to analysis due to missing values for the factors under study; however, all patients with available measurements for a particular factor were included in a given analysis.

    RESULTS

    Relationship of Tumor Cell Ploidy to Age and MYCN Copy Number

    Of the 1,667 patients enrolled onto POG Study 9047, 648 patients (39.3%) had stage D tumors, 528 of which were assessable for both tumor cell ploidy and MYCN amplification. Table 1 lists significant associations of ploidy with age at diagnosis and MYCN copy number. The frequency of hyperdiploid tumors was highest, 75.2% (97 of 129), in infants decreasing to 50.5% (54 of 107) in patients 12 to 24 months old and 54.8% (177 of 323) in patients older than 24 months (P < .0001). In the 0- to 24-month-old subgroup, the proportion of patients with MYCN-amplified tumors increased with increasing age and then remained fairly constant in those children older than 24 months (P < .0001; data not shown). MYCN amplification was most often associated with tumor cell diploidy. Of the 216 patients with diploid tumors, 97 patients (44.9%) had MYCN amplification, compared with 82 (26.3%) of 312 patients with hyperdiploid tumors (P < .0001; Table 1). The significant association between MYCN amplification and a diploid DNA content was clearest in infants and persisted in patients 12 to 24 months old, but was not evident in patients older than 24 months; of 135 older children with diploid tumors, 46 children (34.1%) had MYCN amplification compared with 46 (27.7%) of 166 children with hyperdiploid tumors (P = .2585).

    Prognostic Significance of Hyperdiploidy and Nonamplified MYCN

    In both univariate and multivariate analyses of EFS rates among infants (0 to 365 days old) and older children with MYCN amplified stage D tumors, hyperdiploidy failed to identify patients with a better prognosis (data not shown). This underscored the overriding predictive strength of an amplified MYCN copy number,4,5 prompting us to focus on the subgroup with nonamplified MYCN stage D tumors. The 4-year EFS rates for infants with these features clearly depended on tumor cell ploidy: 83.7% ± 4.4% (n = 87) for the subgroup with hyperdiploid tumors versus 46.2% ± 13.8% (n = 13) for patients with diploid tumors (P = .0003; Fig 1A, Table 2). The prognostic benefit of hyperdiploidy extended to patients 12 to 24 months old (366 to 730 days old), with a 4-year EFS rate of 72.7% ± 10.2% (n = 22) for the subgroup with hyperdiploid tumors, compared with only 26.7% ± 13.2% (n = 16) for patients with diploid tumors (P = .0092; Fig 1B, Table 2). The OS results for these groups were similar to those given in Table 2 based on EFS (data not shown).

    A preliminary examination of the data revealed that most of the failures in the 12- to 24-month-old subgroup had occurred in patients older than 18 months, leading us to repeat the Kaplan-Meier analysis in the 12- to 18-month-old subgroup. As shown in Figure 2A, the 4-year EFS rate for the 12- to 18-month-old subgroup (366 to 550 days old) with hyperdiploid nonamplified MYCN tumors (n = 14) was 92.9% ± 7.2%, similar to the results for the corresponding subgroup of infants (Fig 1A), but significantly different from that for the corresponding subgroup of 19- to 24-month-old patients (551 to 730 days old; Fig 2B), whose 4-year EFS rate was only 37.5% ± 21.0% (P = .0037). In the combined 12- to 24-month-old group, diploidy significantly predicted a worse prognosis, and in both the 12- to 18-month-old and the 19- to 24-month-old subgroups, tumor cell diploidy appeared to negatively affect prognosis, although these differences did not attain statistical significance (Figs 2A and 2B; Table 2). In children older than 24 months (> 730 days old), the outlook was dire, with a long-term EFS rate of less than 20% regardless of ploidy classification or MYCN copy number (Fig 1C; Table 2).

    Comparison of Individual Treatment Responses in the 12- to 18-Month-Old and the 19- to 24-Month-Old Subgroups

    Although a comparison of the Kaplan-Meier curves (Fig 1B) statistically supports recognition of 12- to 24-month-old children with hyperdiploid nonamplified MYCN stage D tumors as a good prognosis group, we would suggest that this designation might not extend beyond 18 months of age. Table 3 compares the specific clinical responses of all patients whose EFS rates were analyzed in Figures 2A and 2B. Hyperdiploidy was almost uniformly associated with good response to therapy in patients 12 to 18 months old. In this subgroup (n = 14), there were two complete responses to intensive chemotherapy alone and six complete responses to chemotherapy followed by myeloablative treatment. Two additional patients attained complete remission following standard chemotherapy for intermediate-risk disease, as did two others treated with surgery only. Although the latter two patients were centrally classified as having stage D disease, the treating physician made a clinical decision to treat them on a protocol for stage Ds neuroblastoma. Only two children with hyperdiploid tumors were not long-term survivors, one as a result of relapse and the other as a result of drug toxicity. This favorable profile was not duplicated in the 19- to 24-month-old subgroup, where half of the patients either relapsed or had progressive disease in response to high-dose chemotherapy alone or with stem-cell rescue. Diploidy appeared to confer a worse prognosis in the 12- to 18-month-old non–MYCN-amplified tumor subgroup. In the 19- to 24-month-old age group, diploidy conferred a dismal prognosis, with seven of 10 patients showing early resistance to high-dose chemotherapy. Thus, although the number of patients in each subgroup was small, hyperdiploid tumors without MYCN amplification seem more responsive to treatment in children younger than 19 months.

    DISCUSSION

    We have previously shown that a diploid DNA index and MYCN gene amplification confer a very high risk of treatment failure in infants with disseminated neuroblastoma.3,13,15 In those studies, the combination of tumor cell hyperdiploidy with nonamplified MYCN was associated with a very favorable outcome in infants younger than 12 months who had received well-tolerated chemotherapy, even if the disease was widely disseminated at diagnosis. This favorable prognosis was not duplicated in children older than 24 months with stage D disease, all of whom had a very low probability of long-term survival. What remained unclear was the age at which hyperdiploidy and nonamplified MYCN cease to be predictive of a favorable response in toddlers with stage D disease. Currently, children 12 to 24 months old with disseminated neuroblastoma are considered to be at high risk for treatment failure, similar to patients older than 24 months, and most therefore receive myeloablative chemotherapy with peripheral blood stem-cell rescue. Although producing some encouraging results in these high-risk patients,11,12 this intensified treatment increases the risk of serious complications, including transplantation-related death, sepsis, irreversible ototoxicity, nephrotoxicity, secondary myelodysplasia, and secondary tumors.

    To address this issue, we undertook a study of infants and young children with stage D tumors who were registered onto POG Neuroblastoma Biology Protocol 9047. This patient population was selected on the basis of tumor specimens with interpretable DNA ploidy and MYCN measurements; however, the distribution of patients by age was comparable with results reported for unselected series of patients with stage D neuroblastoma. As before,15 a hyperdiploid DNA content was highly predictive of a favorable outcome in infants younger than 12 months with stage D tumors. We also found that in children 12 to 24 months with nonamplified MYCN stage D tumors, hyperdiploidy offered a better outcome than diploidy (Fig 1B; Table 2), confirming our previous data.13 DNA ploidy measurements had no predictive value for children older than 24 months.

    The prognostic contribution of age in neuroblastoma outcome is continuous in its nature, raising the important issue of whether 12- to 24-month-old patients should be regarded as a homogenous risk group. The 4-year EFS rate in children 12 to 18 months old with tumor cell hyperdiploidy and nonamplified MYCN was 92.9% ± 7.2%, decreasing to 37.5% ± 21.0% in the 19- to 24-month-old subgroup and to 24.9% ± 4.5% in older patients. Conceivably, the better outcome in children aged 12 to 18 months with hyperdiploid tumors could be due to the treatment they received; however, as indicated in Table 3, these patients were managed in an essentially similar manner (ie, standard v high-dose chemotherapy with stem-cell support) to the 19- to 24-month-old subgroup. Similarly, an EFS rate of 66.7% ± 22.2% and 11.1% ± 10.5% was seen in those with diploid nonamplified MYCN tumors in the 12- to 18-month-old and 19- to 24-month-old subgroups, respectively. Although our statistical analysis (Table 2) clearly supports a significant prognostic effect of ploidy in patients up to 24 months of age, with stage D disease and nonamplified MYCN, we conservatively recommend that an age of 18 months be used as the cutoff for the assignment of patients to less intensive treatment plans. More extensive modeling on an even larger cohort will be needed before we can definitively support any age cutoff greater than 365 days. Unfortunately, due to the continuous nature of the predictive strength of age, virtually any cutoff in age between 12 and 24 months that results in a sufficient sample size in the two age cohorts would result in a statistically significant difference in outcome between the two age groups. We therefore propose the 18-month cutoff as a beginning guideline based on existing data that can then be tested and refined in subsequent studies. An attractive possibility for the treatment of younger patients with hyperdiploid nonamplified MYCN tumors has been described by Schmidt et al.23 in their analysis of the CCG-3881 study of infants (< 12 months old) with nonamplified stage IV (POG stage D) neuroblastoma. Consisting of cisplatin, cyclophosphamide, doxorubicin, and etoposide (with surgery and local irradiation for residual disease), this regimen is highly effective and has relatively low toxicity. Indeed, the 4-year EFS rate of infants with stage IV disease in CCG-3881 was 93% ± 4% (nonamplified MYCN, irrespective of ploidy), compared with 78.1% ± 4.4% (nonamplified MYCN, irrespective of ploidy) and 83.7% ± 4.4% (nonamplified MYCN, tumor cell hyperdiploidy) in our cohort of infants with stage D disease.

    A major question in any combined analysis of ploidy and MYCN status is whether these variables add clinically significant prognostic information to each other. In a recent study of 44 patients with locoregional neuroblastoma tumors (INSS stages 1, 2, and 3), both MYCN amplification and diploidy had a significant impact on OS, but only diploidy was associated with a reduced EFS rate and progression to stage 4 disease.24 The authors reported that DNA index was the most sensitive predictive marker of progression-free survival, OS, and progression to stage 4 disease, in both multivariate and univariate analyses. Ladenstein et al25 performed a multivariate analysis of age, stage, MYCN status, DNA ploidy, serum ferritin, and lactate dehydrogenase levels in 179 patients with neuroblastoma, showing that DNA diploidy or tetraploidy was the most powerful biologic factor for predicting OS and EFS, followed by MYCN gene copy number. Finally, in an analysis of 134 infants with stage 4 neuroblastoma, Schmidt et al23 for the Children's Cancer Group, showed that Shimada histopathologic classification, serum ferritin, and bone marrow immunocytology lost prognostic significance after adjustment for MYCN copy number. This report did not assess the relationship between MYCN status and ploidy classification. Our analysis showed independent prognostic significance for tumor cell ploidy in the subgroup of infants with nonamplified MYCN stage D tumors (n = 100; Fig 1A).

    Our conservative choice of 18 months as a cutoff point for the favorable impact of tumor cell hyperdiploidy is supported by findings in two earlier studies with limited numbers of patients.26,27 In an analysis of 38 cases of neuroblastoma, Gansler et al26 found that nine of 12 patients with DNA aneuploid tumors who remained disease free were 18 months old or younger, whereas three of four patients who died with such tumors were older than 18 months. In another study,27 DNA index provided important prognostic information on patients who were younger than 19 months: 11 such patients with aneuploid tumors all had a good prognosis, compared with only five of nine in the same group with diploid tumors. However, neither of these studies analyzed the effect of MYCN, precluding any assessment of the independent prognostic contribution of DNA index. In a study of 29 children younger than 19 months, in which the effect of MYCN was taken into account, the difference in disease-free survival for patients with hyperdiploid versus diploid tumors attained statistical significance.14 This study included patients of all stages and did not specifically address stage D. Our study is the first to demonstrate the prognostic impact of hyperdiploidy in patients 12 to 24 months in the context of a nonamplified MYCN copy number and stage D disease.

    It is possible that hyperdiploid neuroblastomas in infants are more sensitive to chemotherapy because of their increased propensity to undergo apoptosis in the face of treatment with antineoplastic agents. Moreover, high levels of the TRKA (NTRK1) gene are associated with a younger age, a favorable disease stage, and a better outcome,28 as well as a hyperdiploid DNA content.29 Given the findings of our study, it is reasonable to assume that these properties may also be true of tumors with hyperdiploid cell content in patients aged 12 to 24 months. Thus, one might profitably examine the tumor cells of this cohort for expression of apoptosis-inducing genes or for TRKA expression, with a view to identifying new therapeutic targets. Similarly, DNA microarrays could potentially be used to delineate the molecular differences underlying the clinical behaviors of hyperdiploid tumors in younger children versus those older than 18 months.

    In conclusion, our data show that the favorable influence of hyperdiploidy in infants with nonamplified MYCN stage D neuroblastoma extends to children as old as 18 months, suggesting that it may be possible to treat these patients effectively with less intensive therapy. We therefore suggest the use of an 18-month age cutoff for the initial trials of reduced chemotherapy for young children with hyperdiploid nonamplified MYCN stage D neuroblastoma.

    Authors' Disclosures of Potential Conflicts of Interest

    Although all authors completed the disclosure declaration, the following author or immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.

    Acknowledgment

    We acknowledge the Pediatric Oncology Group for supplying the tumor samples.

    NOTES

    Supported in part by National Institutes of Health grant No. CA-39771 (GMB) and by the POG Statistics and Data Center grants No. U10 CA29139, CA25408, and the POG Chair's grant No. U10 CA30969. R.E.G. is supported by the National Institute for Neurological Diseases and Stroke grant No. K08NS047983, Young Investigator Awards from the American Society of Clinical Oncology, the Children's Oncology Group, and the Fred Lovejoy Resident Research Award.

    Presented in part at the 39th Annual Meeting of the American Society of Clinical Oncology, May 31-June 3, 2003, Chicago, IL.

    Authors' disclosures of potential conflicts of interest are found at the end of this article.

    REFERENCES

    Breslow N, McCann B: Statistical estimation of prognosis for children with neuroblastoma. Cancer Res 31:2098-2103, 1971

    Evans AE, D'Angio GJ, Randolph J: A proposed staging for children with neuroblastoma: Children's cancer study group A. Cancer 27:374-378, 1971

    Look AT, Hayes FA, Nitschke R, et al: Cellular DNA content as a predictor of response to chemotherapy in infants with unresectable neuroblastoma. N Engl J Med 311:231-235, 1984

    Brodeur GM, Seeger RC, Schwab M, et al: Amplification of N-myc in untreated human neuroblastomas correlates with advanced disease stage. Science 224:1121-1124, 1984

    Seeger RC, Brodeur GM, Sather H, et al: Association of multiple copies of the N-myc oncogene with rapid progression of neuroblastomas. N Engl J Med 313:1111-1116, 1985

    Shimada H, Chatten J, Newton WA Jr, et al: Histopathologic prognostic factors in neuroblastic tumors: Definition of subtypes of ganglioneuroblastoma and an age-linked classification of neuroblastomas. J Natl Cancer Inst 73:405-416, 1984

    Joshi VV, Cantor AB, Altshuler G, et al: Age-linked prognostic categorization based on a new histologic grading system of neuroblastomas: A clinicopathologic study of 211 cases from the Pediatric Oncology Group. Cancer 69:2197-2211, 1992

    Nitschke R, Smith EI, Shochat S, et al: Localized neuroblastoma treated by surgery: A Pediatric Oncology Group Study. J Clin Oncol 6:1271-1279, 1988

    Philip T, Ladenstein R, Zucker JM, et al: Double megatherapy and autologous bone marrow transplantation for advanced neuroblastoma: The LMCE2 study. Br J Cancer 67:119-127, 1993

    Ladenstein R, Philip T, Lasset C, et al: Multivariate analysis of risk factors in stage 4 neuroblastoma patients over the age of one year treated with megatherapy and stem-cell transplantation: A report from the European Bone Marrow Transplantation Solid Tumor Registry. J Clin Oncol 16:953-965, 1998

    Matthay KK, Villablanca JG, Seeger RC, et al: Treatment of high-risk neuroblastoma with intensive chemotherapy, radiotherapy, autologous bone marrow transplantation, and 13-cis- retinoic acid: Children's Cancer Group. N Engl J Med 341:1165-1173, 1999

    Grupp SA, Stern JW, Bunin N, et al: Tandem high-dose therapy in rapid sequence for children with high-risk neuroblastoma. J Clin Oncol 18:2567-2575, 2000

    Look AT, Hayes FA, Shuster JJ, et al: Clinical relevance of tumor cell ploidy and N-myc gene amplification in childhood neuroblastoma: A Pediatric Oncology Group study. J Clin Oncol 9:581-591, 1991

    Cohn SL, Rademaker AW, Salwen HR, et al: Analysis of DNA ploidy and proliferative activity in relation to histology and N-myc amplification in neuroblastoma. Am J Pathol 136:1043-1052, 1990

    Bowman LC, Castleberry RP, Cantor A, et al: Genetic staging of unresectable or metastatic neuroblastoma in infants: A Pediatric Oncology Group study. J Natl Cancer Inst 89:373-380, 1997

    Hartmann O, Valteau-Couanet D, Vassal G, et al: Prognostic factors in metastatic neuroblastoma in patients over 1 year of age treated with high-dose chemotherapy and stem cell transplantation: A multivariate analysis in 218 patients treated in a single institution. Bone Marrow Transplant 23:789-795, 1999

    Frappaz D, Michon J, Coze C, et al: LMCE3 treatment strategy: Results in 99 consecutively diagnosed stage 4 neuroblastomas in children older than 1 year at diagnosis. J Clin Oncol 18:468-476, 2000

    Brodeur GM, Pritchard J, Berthold F, et al: Revisions of the international criteria for neuroblastoma diagnosis, staging, and response to treatment. J Clin Oncol 11:1466-1477, 1993

    Schwab M, Alitalo K, Klempnauer KH, et al: Amplified DNA with limited homology to myc cellular oncogene is shared by human neuroblastoma cell lines and a neuroblastoma tumour. Nature 305:245-248, 1983

    Shapiro DN, Valentine MB, Rowe ST, et al: Detection of N-myc gene amplification by fluorescence in situ hybridization: Diagnostic utility for neuroblastoma. Am J Pathol 142:1339-1346, 1993

    Mathew P, Valentine MB, Bowman LC, et al: Detection of MYCN gene amplification in neuroblastoma by fluorescence in situ hybridization: A Pediatric Oncology Group study. Neoplasia 3:105-109, 2001

    Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958

    Schmidt ML, Lukens JN, Seeger RC, et al: Biologic factors determine prognosis in infants with stage IV neuroblastoma: A prospective Children's Cancer Group study. J Clin Oncol 18:1260-1268, 2000

    Mora J, Cheung NK, Chen L, et al: Survival analysis of clinical, pathologic, and genetic features in neuroblastoma presenting as locoregional disease. Cancer 91:435-442, 2001

    Ladenstein R, Ambros IM, Potschger U, et al: Prognostic significance of DNA di-tetraploidy in neuroblastoma. Med Pediatr Oncol 36:83-92, 2001

    Gansler T, Chatten J, Varello M, et al: Flow cytometric DNA analysis of neuroblastoma: Correlation with histology and clinical outcome. Cancer 58:2453-2458, 1986

    Oppedal BR, Storm-Mathisen I, Lie SO, et al: Prognostic factors in neuroblastoma: Clinical, histopathologic, and immunohistochemical features and DNA ploidy in relation to prognosis. Cancer 62:772-780, 1988

    Nakagawara A, Arima-Nakagawara M, Scavarda NJ, et al: Association between high levels of expression of the TRK gene and favorable outcome in human neuroblastoma. N Engl J Med 328:847-854, 1993

    Kogner P, Barbany G, Bjork O, et al: Trk mRNA and low affinity nerve growth factor receptor mRNA expression and triploid DNA content in favorable neuroblastoma tumors. Prog Clin Biol Res 385:137-145, 1994(Rani E. George, Wendy B. )