This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Maris, J. M. Articles by Matthay, K. K. Search for Related Content PubMed PubMed Citation Articles by Maris, J. M. Articles by Matthay, K. K.

Molecular Biology of Neuroblastoma

From the Division of Oncology, Children's Hospital of Philadelphia, and Department of Pediatrics, University of Pennsylvania School of Medicine, Philadelphia, PA; and Department of Pediatrics, University of California at San Francisco School of Medicine, San Francisco, CA.

Address reprint requests to John M. Maris, MD, The Children's Hospital of Philadelphia, Division of Oncology, ARC 902A, 324 South 34th St, Philadelphia, PA 19104-4318; email maris{at}email.chop.edu

ABSTRACT

PURPOSE AND RESULTS: Neuroblastoma, the most common solid extracranial neoplasm in children, is remarkable for its clinical heterogeneity. Complex patterns of genetic abnormalities interact to determine the clinical phenotype. The molecular biology of neuroblastoma is characterized by somatically acquired genetic events that lead to gene overexpression (oncogenes), gene inactivation (tumor suppressor genes), or alterations in gene expression. Amplification of the MYCN proto-oncogene occurs in 20% to 25% of neuroblastomas and is a reliable marker of aggressive clinical behavior. No other oncogene has been shown to be consistently mutated or overexpressed in neuroblastoma, although unbalanced translocations resulting in gain of genetic material from chromosome bands 17q23-qter have been identified in more than 50% of primary tumors. Some children have an inherited predisposition to develop neuroblastoma, but a familial neuroblastoma susceptibility gene has not yet been localized. Consistent areas of chromosomal loss, including chromosome band 1p36 in 30% to 35% of primary tumors, 11q23 in 44%, and 14q23-qter in 22%, may identify the location of neuroblastoma suppressor genes. Alterations in the expression of the neurotrophins and their receptors correlate with clinical behavior and may reflect the degree of neuroblastic differentiation before malignant transformation. Alterations in the expression of genes that regulate apoptosis also correlate with neuroblastoma behavior and may help to explain the phenomenon of spontaneous regression observed in a well-defined subset of patients.

CONCLUSION: The molecular biology of neuroblastoma has led to a combined clinical and biologic risk stratification. Future advances may lead to more specific treatment strategies for children with neuroblastoma.

NEUROBLASTOMA IS THE third most common pediatric cancer and is responsible for approximately 15% of all childhood cancer deaths. It is an embryonal cancer of the postganglionic sympathetic nervous system, which most commonly arises in the adrenal gland. Neuroblastoma is the most common malignant disease of infancy,¹ and 96% of cases occur before the age of 10 years.² Despite the fact that neuroblastoma is sometimes diagnosed in the perinatal period,³ no environmental influences or parental exposures that impact on disease occurrence have been identified consistently.⁴ Thus, the etiology of neuroblastoma remains obscure.

The clinical hallmark of neuroblastoma is heterogeneity, with the likelihood of tumor progression varying widely according to anatomic stage and age at diagnosis. In general, children diagnosed before 1 year of age and/or with localized disease are curable with surgery and little or no adjuvant therapy. Some of these tumors undergo spontaneous regression or differentiate into benign ganglioneuromas. D'Angio et al⁵ first recognized a distinct subset of infant patients who present with a unique pattern of extensive disseminated disease but who reliably have spontaneous disease regression (stage 4S). In contrast, older children often have extensive hematogenous metastases at diagnosis, and the majority die from disease progression despite intensive multimodal therapy. This clinical diversity correlates closely with several molecular biologic features of neuroblastoma.

Investigation of the molecular biology of neuroblastoma began with the cytogenetic characterization of tumor-derived cell lines. The majority of neuroblastoma cell lines show double minute chromatin bodies (DMs) or homogeneously staining regions (HSRs), both representing DNA amplification, and deletions of the short arm of chromosome 1.^6,7 Thus, these early observations clearly demonstrated that both gain and loss of genetic material are common during neuroblastoma evolution, which is consistent with our current concepts of tumorigenesis involving both oncogene activation and tumor suppressor gene inactivation. Therefore, this review discusses the molecular biology of neuroblastoma in terms of gain or loss of genetic information, followed by a description of the frequently observed alterations in the expression of genes that affect survival, differentiation, and apoptosis of neuroblastoma cells.

GAIN OF GENETIC MATERIAL: AMPLIFICATION AND OVEREXPRESSION

MYCN Amplification
MYCN was originally cloned in 1983 by identifying an amplified DNA sequence with partial homology to the MYC proto-oncogene (c-myc) in neuroblastoma cell lines with DMs and/or HSRs.^8,9 The MYCN gene (Table 1) is located on the distal short arm of chromosome 2 (2p24). Genetic material from this region is transposed to DMs during the process of amplification, and the DMs may then be linearly integrated into random chromosomal regions as HSRs. The process of amplification usually results in 50 to 400 copies of the gene per cell, with correspondingly high levels of protein expression.¹⁰ Intermediate copy level numbers (ie, three to 10 copies) may reflect either low-level amplification or aneuploidy. Brodeur et al¹¹ were the first to show that MYCN amplification occurs in a substantial subset of primary untreated neuroblastomas and is highly correlated with advanced stage. Seeger et al¹² then demonstrated a strong association with rapid disease progression and a poor prognosis. Subsequent investigations have confirmed that MYCN gene amplification greater than 10 copies per haploid genome is a powerful prognostic variable independent of anatomic stage, age, and multiple other biologic measures. Analysis of MYCN remains an essential component of disease evaluation for newly diagnosed neuroblastoma patients and serves as a paradigm for the utility of molecular biologic information in cancer treatment stratification.

MYCN is a proto-oncogene normally expressed in the developing nervous system and selected other tissues. The MYCN gene product, MycN, is a nuclear phosphoprotein with a short half-life and regions of marked similarity (overall 38% amino acid identity) to c-myc.¹³ Like all Myc family proteins, MycN contains an N-terminal transactivation domain (Myc box) and a C-terminal region containing a basic helix-loop-helix/leucine zipper (bHLH-LZ) motif. The bHLH-LZ region mediates DNA binding as well as interactions with other bHLH-LZ proteins such as Max and Mad (Fig 1). For MycN to activate transcription, it must first dimerize to Max.¹⁴ Max is a ubiquitously expressed nuclear protein with a long half-life that lacks the amino-terminal transactivation domain that characterizes the Myc protein family. At steady state in a quiescent (G₀) cell, Max expression is high and favors the formation of Max/Max homodimers that repress transcription. However, with increased production of MycN, as with entrance into the cell cycle or as a result of genomic amplification, heterodimerization of MycN and Max occurs. This leads to transcriptional activation of an as yet undefined series of growth-promoting genes. In addition, heterodimerization of Max with other nuclear proteins such as Mad and Mxi1 also function to repress transcription by competing with MycN for Max binding. The specificity of Myc transactivation is partly mediated through binding to a DNA sequence termed an E-box that is present in the promoter region of target genes. Several genes thought to be upregulated by Myc proteins have been identified,¹⁵ but the exact target genes affected by MycN overexpression in neuroblastomas have not yet been defined.

View larger version (21K): [in this window] [in a new window]   Fig 1. Model of MycN interactions with other proteins. Genomic amplification of MYCN leads to increased MycN nuclear protein. Heterodimerization of MycN and Max is favored in the presence of excess MycN (arrows). Transcriptional activation of target genes occurs after interaction of the MycN/Max complex with the promoter region through an E-box motif (CACGTG). In the absence of MycN, it is postulated that the transcriptional repressing protein complexes Max/Max, Max/Mad, and Max/Mxi1 predominate and function to repress transcription.

MYCN functions as a classic dominant oncogene. Forced expression of MYCN can transform normal cells, usually in cooperation with oncogenic ras.^16-18 Overexpression of MYCN can rescue embryonic fibroblasts from senescence,¹⁹ and addition of MYCN antisense RNA to MycN-overexpressing neuroblastoma cell lines can decrease proliferation and/or induce differentiation.^20,21 In addition, targeted overexpression of human MYCN in the neuroectoderm of transgenic mice reliably produces tumors that resemble neuroblastoma in a dose-dependent manner.²²

Although it is clear that MYCN amplification identifies a subset of neuroblastomas with highly malignant behavior, the precise role, if any, that the MycN protein plays in nonamplified tumors remains controversial. Some neuroblas-toma cell lines express high levels of MycN protein without gene amplification,^23,24 and this may be because of alterations in normal protein degradative pathways²³ rather than loss of MYCN transcriptional autoregulation.²⁵ In primary tumors, MycN expression can be detected in all disease stages, suggesting no correlation with clinical phenotype.¹⁰ However, in the subset of tumors without MYCN amplification, some studies have suggested that MycN expression correlates inversely with survival probability, whereas others have found no correlation (reviewed in Bordow et al²⁶). Further studies in a larger cohort of consistently treated patients with standardized methods are necessary to determine if quantitative assessment of MycN expression lends prognostic information in tumors that lack MYCN amplification.

Other Genes Coamplified With MYCN
The genomic region amplified with MYCN is large, usually on the order of 500 to 1,000 kb. Therefore, it has been postulated that additional genes located near MYCN may contribute to the ultimate tumor phenotype when coamplified. High-resolution restriction mapping of the MYCN locus has shown that a 130-kb core domain is the consistent target of amplification.²⁷ To date, no other gene besides MYCN has been identified in this core domain.²⁸ However, the RNA helicase gene DDX1 maps within 300 kb 5' of MYCN and is coamplified in approximately 40% to 50% of neuroblastomas with MYCN amplification.²⁹ DDX1 can transform NIH 3T3 cells and allow for establishment of sarcomatous primary tumors in immunodeficient mice.³⁰ Nevertheless, DDX1 amplification has not been identified in the absence of MYCN amplification. Thus, MYCN is primarily responsible for the aggressive nature of the neuroblastomas with amplification at the 2p24.1 locus, but DDX1 may contribute to the highly malignant nature of some MYCN amplified tumors.

Gain of 17q
Recurrent abnormalities of the long arm of chromosome 17 were first identified by Gilbert et al³¹ in an analysis of Giemsa-banded karyotypes derived from neuroblastoma primary tumors and cell lines. It recently has become apparent that gain of 17q genetic material is perhaps the most common genetic abnormality in primary neuroblastomas. Unbalanced 1;17 translocations occur frequently in primary neuroblastomas^32,33 and usually result in loss of distal 1p with concomitant gain of distal 17q material.³⁴ However, the 17q translocation breakpoints are heterogeneous and often involve other partner chromosomes.³⁵ Recently, comparative genomic hybridization (CGH) analyses have shown that 17q21-qter gain occurs in 50% to 75% of primary tumors (Fig 2).^36-39 A 25-megabase region at 17q23.1-17qter was shown to be the smallest region of gain by fluorescence in situ hybridization mapping.⁴⁰ Thus, this genomic region is the likely location of a gene (or genes) that contributes to neuroblastoma tumorigenesis when present in increased copy number. Unbalanced 17q gain is associated with adverse prognostic features and may therefore be useful for treatment stratification.^41,42

View larger version (49K): [in this window] [in a new window]   Fig 2. Comparative genomic hybridization of primary neuroblastomas. Summary of the chromosomal gains (green) and losses (red) present in a representative panel of 29 primary tumors.37 Thick bars represent high level genomic amplification and each thick bar at the MYCN locus (2p24) represents the pattern of gain in two distinct cases.

DNA Content
The majority of neuroblastoma cell lines and advanced primary tumors have either a near-diploid or near-tetraploid DNA content. In contrast, favorable neuroblastomas, especially those from infants, usually have a hyperdiploid or near-triploid DNA index. The diploid/tetraploid tumors are characterized by chromosomal rearrangements, including amplification, deletion, and unbalanced translocations. In contrast, the hyperdiploid tumors typically have whole chromosome gains with few structural rearrangements. It therefore seems that there is a fundamental difference between these types of neuroblastomas, with the diploid/tetraploid tumors demonstrating generalized genomic instability, and the hyperdiploid tumors having a basic defect in the machinery of mitosis and chromosomal segregation.⁴³ Look et al⁴⁴ first noted the association of hyperdiploidy with favorable prognostic features, especially in children younger than 2 years of age.⁴⁵ Flow cytometric analysis of cell nuclei remains a relatively simple method for measuring DNA content that can lend prognostic information in this age group, although multivariate analyses will be required to show whether it is independent of more specific genetic prognostic variables.

Other Oncogenes
No other human oncogene has been shown to be consistently mutated, overexpressed, or amplified in neuroblastoma. NRAS was originally identified as a transforming gene in a neuroblastoma cell line, and ras cooperates with MycN to transform embryonic fibroblasts.¹⁶ In addition, targeting of HRAS overexpression to the neuroectoderm of mice causes ganglioneuromas and occasional neuroblastomas.⁴⁶ However, activating ras mutations in neuroblastoma primary tumors are only rarely observed.^47,48 Coamplification of MYCL or MDM2 with MYCN has been observed in a few neuroblastoma cell lines,^49-51 and recent CGH studies have identified novel regions of amplification at 2p13-14, 2p23, 3q24-26, 4q33-35, and 6p11-22 in occasional tumors.^36-38 However, the biologic significance of these observations has not yet been defined.

LOSS OF GENETIC MATERIAL: TUMOR SUPPRESSOR LOCI

Neuroblastoma Predisposition
Constitutional chromosomal abnormalities and associated conditions. Germline chromosomal abnormalities, which can predispose to cancer and point to a causative gene, have only rarely been identified in children with neuroblastoma. Two neuroblastoma patients with germline interstitial deletions within chromosome band 1p36 have been identified.^52,53 Both patients were diagnosed during infancy, and one child had multifocal primary tumors. The constitutional deletions overlap the location of a putative 1p36 tumor suppressor gene (see Chromosome 1 Alterations in Neuroblastoma, below), suggesting that germline absence of a gene within this region may predispose to the development of neuroblastoma. In addition, Laureys et al^54,55 described an infant with neuroblastoma and a constitutional balanced translocation involving 1p36 (t(1;17)(p36;q12-21)). At least 14 other cases of constitutional chromosomal rearrangements in neuroblastoma patients have been identified,⁵⁶ but the lack of a consistent pattern indicates that many of these rearrangements may be coincidental rather than causal.

Although coincident congenital anomalies occur only rarely, some neuroblastoma patients have been described with synchronous disorders of other neural crest-derived tissues. Hirschsprung disease, central hypoventilation (Ondine's curse), and/or neurofibromatosis type 1^57,58 have been described in both sporadic and familial neuroblastoma patients, suggesting the existence of a global disorder of neural crest-derived cells (neurocristopathy). Thus, genes mutated in each of the aforementioned conditions may be involved in neuroblastoma tumorigenesis.

RET encodes a tyrosine kinase receptor normally involved in neuronal differentiation. Inactivating mutations occur in a subset of patients with familial Hirschsprung disease, but activating mutations lead to the multiple endocrine neoplasia type 2 syndrome. A potential involvement of RET in neuroblastoma tumorigenesis was suggested by mice carrying a metallothionein/Ret fusion transgene developing retroperitoneal small round blue cell tumors suggestive of neuroblastoma.⁵⁹ However, no mutations were found in a panel of 16 neuroblastoma cell lines,⁶⁰ nor was familial neuroblastoma found to be linked to the RET locus at 10q11.2.^61,62 The other currently identified genes that are mutated in hereditary Hirschsprung disease have not been analyzed yet for somatic mutations in neuroblastomas.

There have been several reports of an association of neurofibromatosis type 1 and neuroblastoma,⁶³ but epidemiologic analyses have suggested that this is coincidental.⁵⁸ Two studies have documented NF1 gene mutations in some neuroblastoma cell lines,^64,65 and a patient with neurofibromatosis type 1 and neuroblastoma whose tumor had a homozygous deletion of the NF1 gene was reported.⁶⁶ However, familial neuroblastoma is not linked to the NF1 locus, even in a family in which the proband had both conditions.⁵⁷ Thus, germline inactivation of NF1 does not seem to predispose to neuroblastoma, but somatically acquired inactivation may occur as a later event in tumor evolution in some cases.

Hereditary neuroblastoma. Although neuroblastoma usually occurs sporadically, 1% to 2% of patients have a family history of the disease.^3,61,62,67 This is similar to the other embryonal cancers of childhood in which familial predisposition is observed. Familial neuroblastoma is inherited in an autosomal dominant Mendelian fashion with incomplete penetrance. Affected children from these families differ from those with sporadic disease in that they are often diagnosed at an earlier age (usually infancy) and/or they have multiple primary tumors.^3,62,67 These clinical characteristics are hallmarks of the "two-mutation" cancer predisposition model first proposed for retinoblastoma.⁶⁸ It therefore seems likely that familial neuroblastoma occurs as a result of a germline mutation in one allele of tumor suppressor gene or genes. In addition, Knudson and Strong³ proposed that a new germinal mutation in a predisposition gene may account for the initiation of tumorigenesis in up to 22% of nonfamilial neuroblastomas as well.

There is remarkable heterogeneity among patients with familial neuroblastoma. Within individual families, the disease can vary from asymptomatic and spontaneously regressing neuroblastoma to rapidly progressive and fatal disease.⁶² Thus, the timing of inactivation of the second tumor suppressor gene allele and additional mutations are postulated to confer the ultimate clinical phenotype.⁶² The clinical heterogeneity of familial neuroblastoma may partially explain its rarity because some tumors remain occult or regress and are never detected, whereas others result in death before reproductive age.

Many candidates for the familial neuroblastoma gene locus have been proposed, including several regions that contain putative tumor suppressor loci (see below). Linkage analysis at candidate loci has excluded each of these regions, including the distal short arm of chromosome 1.^61,62,69 A genome-wide search for linkage may be necessary to localize a familial neuroblastoma predisposition gene.

Chromosome 1 Alterations in Neuroblastoma
Brodeur et al^7,70 first recognized that deletions of the short arm of chromosome 1 (1p) were a common karyotypic feature of neuroblastoma cell lines and advanced tumors. Molecular genetic studies have confirmed these observations by documenting loss of heterozygosity (LOH) in 19% to 36% of primary tumors.^71-78 Addition of an intact human chromosome 1p to a 1p-deleted neuroblastoma cell line induced cellular differentiation and/or death.⁷⁹ It therefore seems likely that a neuroblastoma suppressor gene is located on the short arm of chromosome 1, and that this gene is inactivated in at least one third of primary neuroblastomas.

The majority of 1p deletions are large, encompassing most of the short arm. Virtually all neuroblastomas with 1p deletions identified in the literature delete a common region that includes distal 1p36.2 and all of 1p36.3.⁵⁶ In addition, this defined consensus region is also deleted in all neuroblastoma cell lines examined with 1p allelic loss^73,78,80,81 and in the germline of two neuroblastoma patients with coincident congenital anomalies.^52,53 Recent high-resolution mapping studies have confirmed the existence of a common region of deletion within chromosome subbands 1p36.2-36.3 and refined the critical region to approximately one million bases of DNA (Fig 3).^73,78

View larger version (27K): [in this window] [in a new window]   Fig 3. Localization of a putative 1p36 neuroblastoma suppressor gene. At left is an ideogram of the short arm of chromosome 1 with cytogenetic bands indicated. A genetic linkage map of 1p35-p36 with commonly used polymorphic markers at centiMorgan distances is also shown. Candidate genes are indicated in red at relative map locations. The blue rectangles indicate the smallest region of overlap of 1p deletions as reported by various investigators in the years indicated.7,53,71,78 The current smallest region of overlap of all deletions at 1p36.3 is approximately 5 cM, which correlates to 1 megabase of DNA and excludes all candidate genes except HKR3 and DR3.

Several studies have indicated that there might be additional 1p neuroblastoma suppressor genes. Schleiermacher et al⁸² reported three cases with interstitial deletions of 1p32 only, suggesting the possibility of an additional suppressor gene in this proximal region. Takeda et al⁸³ divided their cohort of tumors with 1p LOH into those with large terminal deletions and those with smaller interstitial deletions. They reported that the tumors with larger deletions also generally had MYCN amplification and poor survival probability, whereas those with smaller deletions were more likely to have a single copy of MYCN and usually had a favorable clinical outcome. They postulated that the more proximal region of 1p deleted in the larger deletion cases harbored a second suppressor gene associated with biologically aggressive tumors. Caron et al⁸⁴ also found that tumors with MYCN amplification generally had 1p deletions extending proximal to 1p36, but single-copy tumors more often had small terminal deletions of 1p36 only. Furthermore, the MYCN-amplified tumors in the latter study showed no preference for the parental origin of the deleted chromosome, but the MYCN single-copy tumors preferentially deleted the maternal-derived allele. They hypothesized that at least two tumor suppressor genes are located within 1p35-36: an imprinted distal 1p36 locus defined by the smallest region of overlap common to all distal 1p-deleted neuroblastomas, and a nonimprinted 1p35-36.1 locus that is deleted in MYCN-amplified aggressive tumors.

Several genes have been analyzed as possible candidates for the 1p36 neuroblastoma suppressor gene. These include the TP53 homolog TP73⁸⁵; the CDK2 homolog CDC2L1 (p58)⁸⁶; the transcription factors HKR3,⁸⁷ DAN,⁸⁸ PAX7,⁸⁹ ID3,⁹⁰ and E2F2⁹¹; the transcription elongation factor TCEB3 (Elongin A)⁹²; and two members of the tumor necrosis factor receptor family, TNFR2⁹³ and DR3.⁹⁴ However, each of these genes except HKR3 and DR3 are located outside the current consensus region, and no mutations have been found in the nondeleted allele of any candidate.⁸⁰

There is a strong correlation between 1p LOH and high-risk features such as age greater than 1 year at diagnosis, metastatic disease, and MYCN amplification.^75-77,95 Thus, 1p LOH occurs frequently in the more malignant subset of neuroblastomas. However, there have been contrasting opinions concerning the independent prognostic significance of 1p LOH.^75-77,95,96 A recently completed retrospective study of uniformly treated Children's Cancer Group patients indicated that 1p36 LOH independently predicts for a decreased event-free survival but not overall survival.⁹⁷ This apparent discrepancy may be explained by the fact that many relapses in patients with locoregional neuroblastoma are salvageable. Thus, 1p LOH analysis may identify patients with low- or intermediate-risk features who are more likely to have disease relapse.

Chromosome 11 Alterations in Neuroblastoma
Several lines of evidence suggest that a neuroblastoma suppressor gene is located on the long arm of chromosome 11 (11q). Chromosome 11q deletions have been noted in approximately 15% to 20% of reported neuroblastoma karyotypes,⁹⁸ and transfer of an intact chromosome 11 into the neuroblastoma cell line NGP induced differentiation.⁷⁹ Constitutional rearrangements of chromosome 11q also have been observed in some neuroblastoma patients.⁵⁶ Molecular genetic studies have demonstrated loss of 11q in approximately one third of primary tumors.^36-38,74,99 A recently completed survey of 295 primary neuroblastomas has found evidence for allelic deletion in 44% of cases.^100,101 The common region of LOH mapped to 11q23, indicating that this is the most likely location of an 11q neuroblastoma suppressor gene. In contrast to 1p LOH, there was a striking inverse correlation of 11q LOH and MYCN amplification. Although there was no survival disadvantage for patients whose tumors had 11q LOH in univariate analyses, there was a significant decrease in overall survival probability when the MYCN unamplified subset of patients was analyzed.^100,101 Thus, 11q is the most common region of allelic deletion currently identified in primary neuroblastomas and may be a marker of an unfavorable phenotype independent of MYCN amplification.

Chromosome 14 Alterations in Neuroblastoma
Deletion of the long arm of chromosome 14 is also a common abnormality in neuroblastomas. LOH for 14q was first identified in 1989 in six of 12 neuroblastomas studied by Suzuki et al¹⁰² using a single restriction fragment length polymorphism marker within 14q32. Subsequent studies have shown collectively that 49 of 180 informative tumors (27%) have allelic loss at 14q32.^72,74,99,103 Unlike chromosome 1, there are only a few reports of cytogenetically visible deletions or rearrangements that involve 14q.^70,104 This may indicate that deletions of chromosome 14 are too small to be identified easily by conventional cytogenetic analysis or that the mechanism of LOH involves homologous mitotic recombination. A recent study of 372 primary neuroblastomas with markers evenly spaced along 14q showed LOH in 22%, with a common region of deletion within 14q23-qter.¹⁰⁵ LOH for 14q was highly correlated with 11q LOH, inversely related to MYCN amplification and present in all clinical risk groups, indicating that this abnormality may occur early in tumor development.

Deletions of Other Chromosomal Regions and Alterations in Known Tumor Suppressor/DNA Repair Genes
There are other regions of the genome that are frequently deleted in neuroblastoma tissues, suggesting the existence of additional tumor suppressor genes. There have been reports of LOH and/or allelic imbalance at chromosome arms 3p,¹⁰⁶ 4p,¹⁰⁷ 5q,¹⁰⁸ 9p,^109,110 and 18q,¹¹¹ but these seem to occur at lower frequency than 1p LOH. Recent CGH studies have confirmed much of this polymerase chain reaction–based data.^36-39

In addition to the fact that no neuroblastoma-specific suppressor genes have been cloned, there is currently no evidence for consistent mutation in known tumor suppressor genes. TP53 is the most frequently mutated gene in human cancer, but it is only rarely inactivated by deletion or mutation in neuroblastomas.^112-115 Moll et al¹¹⁶ have reported aberrant cytoplasmic localization of the p53 protein in neuroblastoma and found evidence for dysregulated G1/S checkpoint control. However, other investigators have shown that DNA damage to neuroblastoma cells causes normal translocation of wild-type p53 to the nucleus and induction of p21.¹¹⁷ CDKN2 encodes p16, another cell cycle control protein commonly inactivated in human cancers, but no mutations or deletions have been found in neuroblastomas.¹¹⁸ The DCC and DPC4 genes are located at 18q, a region that is frequently deleted in primary neuroblastomas. Although expression of DCC may be altered in some cell lines and tumors, no inactivating mutations of DCC or DPC4 have been identified.^111,119

Loss of function of a wide array of DNA repair proteins can lead to a "mutator" phenotype and DNA instability in human cancer. However, microsatellite instability is only observed rarely in primary neuroblastomas and cell lines.^73,78,120 Therefore, it seems unlikely that alterations in DNA repair genes play a major role in neuroblastoma tumorigenesis.

ALTERATIONS IN GENE EXPRESSION

Neurotrophin Signaling Pathways
Because of the central role of neurotrophin signaling in normal neuronal development, there has been considerable interest in alterations in these pathways during neuroblastoma neoplastic transformation. Neurotrophins critical to the development of the sympathetic nervous system include nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), neurotrophin-3, and neurotrophin-4. Neurotrophin signaling is mainly mediated through the Trk family of tyrosine kinases. To date, three high-affinity neurotrophin receptor genes relevant to neuroblastoma have been cloned (TRKA, TRKB, and TRKC), and their expression is developmentally regulated.¹²¹ TrkA serves as the high-affinity receptor for NGF, and developing neurons differentiate in response to this ligand or enter an apoptotic pathway if NGF is withdrawn. TrkB binds both BDNF and neurotrophin-4 with high-affinity, and TrkC is the high affinity receptor for neurotrophin-3.

Differential expression of the neurotrophin receptors is strongly correlated with the biologic and clinical features of neuroblastomas. Some primary neuroblastomas differentiate in vitro in the presence of NGF but die in its absence. Transfection of TRKA into a non–TrkA-expressing neuroblastoma cell line restores the ability to differentiate in response to NGF.^122,123 TRKA expression is inversely related to disease stage and MYCN amplification status.¹²⁴ Thus, high TRKA expression is a marker of "favorable" neuroblastomas and is correlated with an increased probability of disease survival.^125-130 In contrast, full-length TrkB (there is also a truncated isoform lacking the tyrosine kinase) is expressed preferentially in advanced stage, MYCN amplified neuroblastomas.¹³¹ Many of these tumors also express BDNF, establishing an autocrine pathway promoting cell growth and survival.^131,132 TRKB is either expressed in low amounts, or as the truncated isoform, in biologically favorable tumors. Lastly, TRKC is expressed in favorable neuroblastomas, essentially all of which also express TRKA.^129,133,134

It therefore seems that the expression pattern of neurotrophin receptors at the time of malignant transformation, which is related to the degree of neuroblastic differentiation, has a critical influence on tumor behavior. An evolving model of neurotrophin ligand/receptor interactions in neuroblastomas speculates that low-stage tumors, particularly those in infants, usually express high levels of TrkA.⁴³ If NGF is present, TrkA-expressing cells will terminally differentiate. In contrast, if NGF is limiting, TrkA-expressing cells will enter a programmed cell death pathway.¹³⁵ MYCN-amplified higher-stage tumors usually express TrkB and rely on autocrine production of BDNF to maintain a growth-promoting signal, even in the absence of exogenous neurotrophins.

Overexpression of Multidrug Resistance Genes
Many neuroblastomas respond well to initial chemotherapy but eventually progress either on treatment or shortly after it has stopped. In addition, neuroblastoma cell lines derived from tumors at relapse show significantly increased resistance to standard chemotherapeutic agents compared with those established at diagnosis.^136,137 Thus, it seems that acquired drug resistance is an important cause of neuroblastoma treatment failures.

The most well-characterized mechanism of multidrug resistance involves increased expression of the adenosine triphosphate–dependent efflux pump P-glycoprotein.¹³⁸ The PGY1 gene (previously named MDR1) at chromosome subband 7q21.1 encodes the ubiquitously expressed P-glycoprotein. In cancer cells, increased expression of P-glycoprotein through transcriptional activation, increased messenger RNA stability, or genomic amplification has been postulated to be responsible for acquired resistance to natural product drugs such as vinca alkaloids, anthracyclines, and epipodphyllotoxins.¹³⁹ PGY1 expression in primary neuroblastomas increases after exposure to chemotherapy.^140-142 However, the contribution of de novo PGY1 overexpression to neuroblastoma behavior has been controversial. Chan et al¹⁴³ reported that P-glycoprotein expression was restricted to advanced-stage tumors and correlated with a poor response to chemotherapy in 67 untreated neuroblastomas, but other groups were unable to confirm these findings using either immunohistochemical¹⁴⁴ or RNA-based methodologies.^145,146 In addition, PGY1 expression is inversely correlated with MYCN expression¹⁴⁷ and may be restricted to the normal stromal cells in neuroblastoma primary tumor biopsy specimens.¹⁴⁸ Lastly, PGY1 may be transcriptionally activated in response to differentiation rather than in response to chemotherapy.¹⁴⁹ Taken together, it seems that enhanced PGY1 expression is rarely a cause of de novo drug resistance in untreated neuroblastomas.

MRP encodes the multidrug resistance-associated protein, which is also an efflux pump that can render a cell resistant to natural product drugs when overexpressed.¹³⁸ In a study of 60 primary untreated neuroblastomas, Norris et al¹⁴⁶ showed that MRP expression was strongly correlated with MYCN expression and patient survival. These data and the identification of E-box motif consensus sequences in the MRP promoter region¹⁵⁰ suggested a potential interaction between the MycN protein and MRP expression. Transfection of MYCN antisense RNA into a neuroblastoma cell line that expressed both MRP and MYCN caused downregulation of MRP messenger RNA to undetectable levels.¹⁵¹ Therefore, MycN overexpression may transcriptionally activate MRP and lead to the drug resistance phenotype, even in untreated primary tumors. However, a direct demonstration of MycN mediating increased MRP transcription has not yet been provided.

Apoptotic Signaling Pathways
Neuroblastoma has the highest rate of spontaneous regression observed in human cancers. Children with stage 4S neuroblastoma often have initial progression of multifocal disease followed by rapid tumor involution. Delayed implementation of normal apoptotic pathways has been proposed as an explanation for this phenomenon.¹⁵²

Entrance into a programmed cell death pathway can originate with exogenous (presence or absence of ligand) or endogenous (DNA damage) signals. NGF withdrawal is a major signal for apoptosis in the developing nervous system and mediates the elimination of redundant cells. Thus, signal transduction through the neurotrophin receptors can initiate the apoptotic pathway. Other cell surface proteins are involved with initiation of apoptosis in neuronal cells and neuroblastomas (reviewed by Brodeur and Castle¹⁵³). Members of the tumor necrosis factor receptor family, such as p75 (binds NGF with low affinity) and CD95/Fas (binds Fas ligand), as well as members of the retinoic acid receptor family, can mediate the induction of apoptosis in some neuroblastoma cell lines.^154,155 In addition, increased CD95 expression seems to be an essential component of chemotherapy-induced apoptosis in neuroblastomas.¹⁵⁵

Intracellular molecules responsible for relaying the apoptotic signal include the Bcl-2 family of proteins. Apoptosis-suppressing genes such as BCL2 and BCLX are highly expressed early in neuronal ontogeny. BCL2 is highly expressed in most neuroblastoma cell lines^156-158 and primary tumors,^159-163 and the level of expression is inversely related to the proportion of cells undergoing apoptosis and the degree of cellular differentiation.^152,163,164 There have been conflicting reports regarding the correlation between the level of expression of Bcl-2 in primary tumors and prognostic variables,^{160-162,165,166} but overall, the evidence suggests that there is no significant correlation. However, the Bcl-2 family of proteins may play an important role in acquired resistance to chemotherapy. Transfection of cDNA encoding either Bcl-2 or Bcl-X_L into neuroblastoma cells caused resistance to alkylator agent–induced apoptosis in a dose-dependent manner.^157,167

Caspases are proteolytic enzymes responsible for the execution of the apoptotic signal. Recently, two studies have shown that increased expression of interleukin 1ß–converting enzyme (caspase-1) and other caspases in primary neuroblastoma is associated with favorable biologic features and improved disease outcome.^135,168,169 These observations are consistent with the hypothesis that neuroblastomas prone to undergoing apoptosis are more likely to spontaneously regress and/or respond well to cytotoxic agents.

Metastasis-Suppressing Genes
Despite the fact that approximately 50% of patients have disseminated disease at diagnosis, little is known about the biology of neuroblastoma invasion and metastasis. The cell surface glycoprotein CD44 has been postulated to influence tumor cell adhesion and has been shown to be overexpressed in several murine and human cancers.^170,171 However, the converse is true for neuroblastoma, with CD44 expression undetectable in the vast majority of neuroblastoma cell lines¹⁷² and primary tumors with dissemination at diagnosis.^173,174 CD44 expression is inversely correlated with MYCN amplification and has been suggested to be independently prognostic for decreased survival probability.^173,175 CD44 is located at 11p13, a genomic region that is deleted in approximately one third of MYCN single-copy neuroblastomas but only a very small percentage of MYCN amplified tumors (J.M.M., unpublished observations, May 1999).

The nucleoside diphosphate kinase gene family may play a role in metastasis suppression and provides another example of altered expression of a putative oncoprotein in neuroblastoma, but again in a pattern opposite that observed in other human malignancies. NME1 (NM23-H1) encodes the nucleoside diphosphate kinase A protein (nm23A), which was originally identified on the basis of its reduced expression in a murine melanoma metastasis model.¹⁷⁶ Subsequently, reduced expression of nm23A was correlated with increased likelihood of metastases in human carcinomas.^177,178 However, increased nm23A expression was noted in advanced (stages III and IV) primary neuroblastomas.^179,180 In addition, missense mutations in a highly conserved areas of the coding region near the catalytic domain (serine 120glycine) were identified in a subset of these advanced tumors with increased expression.^180,181 Overexpression of NME1 has been correlated with a decreased probability of survival.^180,182 The mutated nm23A protein has both decreased enzymatic activity and alterations in its normal protein-binding capabilities,¹⁸³ suggesting that overexpression of mutated nm23A may act in a dominant-negative fashion. NME1 maps to 17q22, a genomic region that is frequently present in increased copy number in advanced neuroblastomas. Clarification of whether there is an association between 17q gain and NME1 overexpression requires additional studies.

It is assumed that activation of matrix degrading proteolytic enzymes, such as the matrix metaloproteinases, is required for local invasion and metastasis. MMP9 (gelatinase B), one of best-studied matrix metaloproteinases, seems to be selectively overexpressed in tissue extracts from stage 4 primary neuroblastomas as opposed to localized tumors.¹⁸⁴ Like other human cancers, it also seems that MMP9 upregulation occurs primarily in the surrounding stromal cells rather than in the tumor cells.¹⁸⁴

Telomerase
The integrity of chromosomal ends is maintained by the ribonucleoprotein telomerase. Increased telomerase activity is detectable in most cancer cells and seems to be a prerequisite for malignant transformation.¹⁸⁵ Hiyama et al¹⁸⁶ were the first to show that telomerase expression was detectable in the vast majority of neuroblastomas (96%) but not in ganglioneuromas or normal adrenal tissue. In addition, very high levels of telomerase activity may correlate with adverse prognostic features and poorer survival probability.^187-189 Therefore, although elevated telomerase expression may simply be a marker of escape from cellular senescence, markedly increased levels may be associated with genomic instability and an increased likelihood of additional mutational events. Semiquantitative detection of telomerase RNA and/or protein subunit expression should be analyzed prospectively to determine if expression independently predicts for aggressive clinical behavior.

CONCLUSION

Neuroblastoma is a heterogeneous childhood malignancy with a complex biology. As originally proposed by Brodeur et al,^43,56 the evidence suggests that there are at least three distinct types of neuroblastoma, and a working conceptual model for the genetic evolution of these tumors is emerging (Fig 4).¹⁹⁰ Initiation of tumorigenesis is thought to occur in an uncommitted neuroblastic cell. Initiation events are unknown but may involve inactivation of a tumor suppressor gene that also may predispose to neuroblastoma when one allele is mutated in the germline. The degree of differentiation at the time of neoplastic transformation determines the neurotrophin receptor expression pattern. Approximately one third of tumors will be characterized by significant TRKA expression. These tumors (type 1) seem to have a fundamental defect in mitotic disjunction because they are hyperdiploid due to whole chromosome gains, and structural rearrangements are rarely observed. Patients with type 1 tumors are usually cured with surgery alone. On the other hand, tumors that predominantly express TRKB are characterized by genomic instability. Structural rearrangements occur and unbalanced 17q gain occurs in most. Tumors that also acquire loss of genetic information on the long arms of chromosomes 11 and/or 14 rarely have 1p loss or MYCN amplification (type 2). However, tumors that acquire 1p loss often also acquire MYCN amplification and a highly malignant clinical behavior (type 3). Current survival probability for patients with a type 3 tumor is less than 25% and at an undefined intermediate value for patients with a type 2 tumor. More precise definition of the molecular changes in neuroblastomas may allow for more specific therapies with subsequent improvements in overall rates and quality of cure.

View larger version (27K): [in this window] [in a new window]   Fig 4. Hypothetical model of genetic origin of neuroblastoma. Undifferentiated neuroblasts throughout the developing sympathetic nervous system are targets for mutational events leading to tumorigenesis. The degree of neuroblastic differentiation at the time of malignant transformation determines neurotrophin expression pattern. TrkA-expressing tumors are more likely to have errors in mitotic recombination, resulting in whole chromosome gains. These tumors (type 1 tumors) are often near triploid (3n) and can differentiate in response to NGF or undergo programmed cell death in the absence of NGF. TrkB-expressing tumors more often have generalized genomic instability (although they may have mitotic instability also) with structural chromosomal alterations. Tumors with genomic instability often have unbalanced gain of distal 17q material. Loss of 11q and/or 14q material usually occurs with 17q gain and often without 1p deletion and MYCN amplification (type 2 tumors). However, 17q gain may be the result of an unbalanced translocation with 1p, with concomitant loss of 1p material. Tumors with 1p loss (by unbalanced translocation or other mechanisms) often contain MYCN amplification (type 3 tumors). Thus, inactivation of a 1p gene may be a prerequisite for MYCN amplification.

ACKNOWLEDGMENTS

We thank Garrett M. Brodeur for helpful discussions, Peter S. White for careful review of the manuscript, and Burt G. Feuerstein for providing the CGH figure.

NOTES

J.M.M. is supported in part by the American Society of Clinical Oncology, the American Cancer Society Career Development Awards, and grants no. R01 CA78545 and U10 CA78966 from the National Institutes of Health, Bethesda, MD. K.K.M. Matthay is supported in part by grant no. U10 CA17829 from the National Institutes of Health, Bethesda, MD.

REFERENCES

1. Gurney JG, Ross JA, Wall DA, et al: Infant cancer in the U.S.: Histology-specific incidence and trends, 1973 to 1992. J Pediatr Hematol Oncol 19:428-432, 1997[Medline]

2. Grovas A, Fremgen A, Rauck A, et al: The National Cancer Data Base report on patterns of childhood cancers in the United States. Cancer 80:2321-2332, 1997[Medline]

3. Knudson AG Jr, Strong LC: Mutation and cancer: Neuroblastoma and pheochromocytoma. Am J Hum Genet 24:514-532, 1972[Medline]

4. Bunin GR, Ward E, Kramer S, et al: Neuroblastoma and parental occupation. Am J Epidemiol 131:776-780, 1990[Abstract/Free Full Text]

5. D'Angio GJ, Evans AE, Koop CE: Special pattern of widespread neuroblastoma with a favourable prognosis. Lancet 1:1046-1049, 1971[Medline]

6. Biedler JL, Helson L, Spengler BA: Morphology and growth, tumorigenicity, and cytogenetics of human neuroblastoma cells in continuous culture. Cancer Res 33:2643-2652, 1973[Abstract/Free Full Text]

7. Brodeur GM, Sekhon GS, Goldstein MN: Chromosomal aberrations in human neuroblastomas. Cancer 40:2256-2263, 1977[Medline]

8. 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[Medline]

9. Kohl NE, Kanda N, Schreck RR, et al: Transposition and amplification of oncogene-related sequences in human neuroblastomas. Cell 35:359-367, 1983[Medline]

10. Seeger RC, Wada R, Brodeur GM, et al: Expression of N-myc by neuroblastomas with one or multiple copies of the oncogene. Prog Clin Biol Res 271:41-49, 1988[Medline]

11. Brodeur G, Seeger RC, Schwab M, et al: Amplification of N-myc in untreated human neuroblastomas correlates with advanced disease stage. Science 224:1121-1124, 1984[Abstract/Free Full Text]

12. 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[Abstract]

13. Slamon DJ, Boone TC, Seeger RC, et al: Identification and characterization of the protein encoded by the human N-Myc oncogene. Science 232:768-772, 1986[Abstract/Free Full Text]

14. Wenzel A, Cziepluch C, Hamann U, et al: The N-Myc oncoprotein is associated in vivo with the phosphoprotein Max (P20/22) in human neuroblastoma cells. EMBO J 10:3703-3712, 1991[Medline]

15. Grandori C, Eisenman RN: Myc target genes. Trends Biochem Sci 22:177-181, 1997[Medline]

16. Yancopoulos GD, Nisen PD, Tesfaye A, et al: N-myc can cooperate with ras to transform normal cells in culture. Proc Natl Acad Sci U S A 82:5455-5459, 1985[Abstract/Free Full Text]

17. Schwab M, Varmus HE, Bishop JM: Human N-myc gene contributes to neoplastic transformation of mammalian cells in culture. Nature 316:160-162, 1985[Medline]

18. Small MB, Hay N, Schwab M, et al: Neoplastic transformation by the human gene N-myc. Mol Cell Biol 7:1638-1645, 1987[Abstract/Free Full Text]

19. Schwab M, Bishop JM: Sustained expression of the human protooncogene Mycn rescues rat embryo cells from senescence. Proc Natl Acad Sci U S A 85:9585-9589, 1988[Abstract/Free Full Text]

20. Negroni A, Scarpa S, Romeo A, et al: Decrease of proliferation rate and induction of differentiation by a mycn antisense DNA oligomer in a human neuroblastoma cell line. Cell Growth Differ 2:511-518, 1991[Abstract]

21. Schmidt ML, Salwen HR, Manohar CF, et al: The biological effects of antisense N-myc expression in human neuroblastoma. Cell Growth Differ 5:171-178, 1994[Abstract]

22. Weiss WA, Aldape K, Mohapatra G, et al: Targeted expression of MYCN causes neuroblastoma in transgenic mice. EMBO J 16:2985-2995, 1997[Medline]

23. Cohn SL, Salwen H, Quasney MW, et al: Prolonged N-Myc protein half-life in a neuroblastoma cell line lacking N-Myc amplification. Oncogene 5:1821-1827, 1990[Medline]

24. Wada RK, Seeger RC, Brodeur GM, et al: Human neuroblastoma cell lines that express N-myc without gene amplification. Cancer 72:3346-3354, 1993[Medline]

25. Sivak LE, Tai KF, Smith RS, et al: Autoregulation of the human N-myc oncogene is disrupted in amplified but not single-copy neuroblastoma cell lines. Oncogene 15:1937-1946, 1997[Medline]

26. Bordow SB, Norris MD, Haber PS, et al: Prognostic significance of MYCN oncogene expression in childhood neuroblastoma. J Clin Oncol 16:3286-3294, 1998[Abstract]

27. Reiter JL, Brodeur GM: High-resolution mapping of a 130-kb core region of the MYCN amplicon in neuroblastomas. Genomics 32:97-103, 1996[Medline]

28. Reiter JL, Brodeur GM: MYCN is the only highly expressed gene from the core amplified domain in human neuroblastomas. Genes Chromosom Cancer 23:134-140, 1998[Medline]

29. George RE, Kenyon RM, McGuckin AG, et al: Investigation of co-amplification of the candidate genes ornithine decarboxylase, ribonucleotide reductase, syndecan-1 and a DEAD box gene, DDX1, with N-myc in neuroblastoma—United Kingdom Children's Cancer Study Group. Oncogene 12:1583-1587, 1996[Medline]

30. George RE, Thomas H, McGuckin AG, et al: The DDX1 gene which is frequently co-amplified with MYCN in neuroblastoma is itself oncogenic. Eur J Cancer (in press)

31. Gilbert F, Feder M, Balaban G, et al: Human neuroblastomas and abnormalities of chromosomes 1 and 17. Cancer Res 44:5444-5449, 1984[Abstract/Free Full Text]

32. Savelyeva L, Corvi R, Schwab M: Translocation involving 1p and 17q is a recurrent genetic alteration of human neuroblastoma cells. Am J Hum Genet 55:334-340, 1994[Medline]

33. Van Roy N, Laureys G, Cheng NC, et al: 1;17 translocations and other chromosome 17 rearrangements in human primary neuroblastoma tumors and cell lines. Genes Chromosom Cancer 10:103-114, 1994[Medline]

34. Van Roy N, Laureys G, Van Gele M, et al: Analysis of 1;17 translocation breakpoints in neuroblastoma: Implications for mapping of neuroblastoma genes. Eur J Cancer 33:1974-1978, 1997

35. Lastowska M, Roberts P, Pearson AD, et al: Promiscuous translocations of chromosome arm 17q in human neuroblastomas. Genes Chromosom Cancer 19:143-149, 1997[Medline]

36. Brinkschmidt C, Christiansen H, Terpe HJ, et al: Comparative genomic hybridization (CGH) analysis of neuroblastomas: An important methodological approach in paediatric tumour pathology. J Pathol 181:394-400, 1997[Medline]

37. Lastowska M, Nacheva E, McGuckin A, et al: Comparative genomic hybridization study of primary neuroblastoma tumors: United Kingdom Children's Cancer Study Group. Genes Chromosom Cancer 18:162-169, 1997[Medline]

38. Plantaz D, Mohapatra G, Matthay KK, et al: Gain of chromosome 17 is the most frequent abnormality detected in neuroblastoma by comparative genomic hybridization. Am J Pathol 150:81-89, 1997[Abstract]

39. Altura RA, Maris JM, Li H, et al: Novel regions of chromosomal loss in familial neuroblastoma by comparative genomic hybridization. Genes Chromosom Cancer 19:176-184, 1997[Medline]

40. Meddeb M, Danglot G, Chudoba I, et al: Additional copies of a 25 Mb chromosomal region originating from 17q23.1-17qter are present in 90% of high-grade neuroblastomas. Genes Chromosom Cancer 17:156-165, 1996[Medline]

41. Caron H: Allelic loss of chromosome 1 and additional chromosome 17 material are both unfavourable prognostic markers in neuroblastoma. Med Pediatr Oncol 24:215-221, 1995[Medline]

42. Lastowska M, Cotterill S, Pearson AD, et al: Gain of chromosome arm 17q predicts unfavourable outcome in neuroblastoma patients: U.K. Children's Cancer Study Group and the U.K. Cancer Cytogenetics Group. Eur J Cancer 33:1627-1633, 1997

43. Brodeur GM, Maris JM, Yamashiro DJ, et al: Biology and genetics of human neuroblastomas. J Pediatr Hematol Oncol 19:93-101, 1997[Medline]

44. 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[Abstract]

45. 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[Abstract]

46. Sweetser DA, Kapur RP, Froelick GJ, et al: Oncogenesis and altered differentiation induced by activated Ras in neuroblasts of transgenic mice. Oncogene 15:2783-2794, 1997[Medline]

47. Moley JF, Brother MB, Wells SA, et al: Low frequency of ras gene mutations in neuroblastomas, pheochromocytomas, and medullary thyroid cancers. Cancer Res 51:1596-1599, 1991[Abstract/Free Full Text]

48. Ireland CM: Activated N-Ras oncogenes in human neuroblastoma. Cancer Res 49:5530-5533, 1989[Abstract/Free Full Text]

49. Jinbo T, Iwamura Y, Kaneko M, et al: Coamplification of the LMYC and NMYC oncogenes in a neuroblastoma cell line. Cancer Res 80:299-303, 1989

50. Corvi R, Savelyeva L, Breit S, et al: Non-syntenic amplification of MDM2 and MYCN in human neuroblastoma. Oncogene 10:1081-1086, 1995[Medline]

51. Van Roy N, Forus A, Myklebost O, et al: Identification of two distinct chromosome 12-derived amplification units in neuroblastoma cell line NGP. Cancer Genet Cytogenet 82:151-154, 1995[Medline]

52. Biegel JA, White PS, Marshall HN, et al: Constitutional 1p36 deletion in a child with neuroblastoma. Am J Hum Genet 52:176-182, 1993[Medline]

53. White PS, Thompson PM, Seifried BA, et al: Detailed molecular analysis of 1p36 in neuroblastoma. Eur J Cancer (in press)

54. Laureys G, Speleman F, Opdenakker G, et al: Constitutional translocation t(1;17)(p36;q12-21) in a patient with neuroblastoma. Genes Chromosom Cancer 2:252-254, 1990[Medline]

55. Laureys G, Speleman F, Versteeg R, et al: Constitutional translocation t(1;17)(p36.31-p36.13;q11.2-q12.1) in a neuroblastoma patient: Establishment of somatic cell hybrids and identification of PND/A12M2 on chromosome 1 and NF1/SCYA7 on chromosome 17 as breakpoint flanking single copy markers. Oncogene 10:1087-1093, 1995[Medline]

56. Brodeur GM: Clinical and biological aspects of neuroblastoma, in Vogelstein B, Kinzler KW (eds): The Genetic Basis of Human Cancer. New York, NY, McGraw-Hill, 1998, pp 691-711

57. Maris JM, Chatten J, Meadows AT, et al: Familial neuroblastoma: A three-generation pedigree and a further association with Hirschsprung disease. Med Pediatr Oncol 28:1-5, 1997[Medline]

58. Kushner BH, Hajdu SI, Helson L: Synchronous neuroblastoma and von Recklinghausen's disease: A review of the literature. J Clin Oncol 3:117-120, 1985[Abstract]

59. Iwamoto T, Taniguchi M, Wajjwalku W, et al: Neuroblastoma in a transgenic mouse carrying a metallothionein/ret fusion gene. Br J Cancer 67:504-507, 1993[Medline]

60. Hofstra RM, Cheng NC, Hansen C, et al: No mutations found by RET mutation scanning in sporadic and hereditary neuroblastoma. Hum Genet 97:362-364, 1996[Medline]

61. Maris JM, Kyemba SM, Rebbeck TR, et al: Familial predisposition to neuroblastoma does not map to chromosome band 1p36. Cancer Res 56:3421-3425, 1996[Abstract/Free Full Text]

62. Maris JM, Kyemba SM, Rebbeck TR, et al: Molecular genetic analysis of familial neuroblastoma. Eur J Cancer 33:1923-1928, 1997

63. Knudson AG Jr, Amromin GD: Neuroblastoma and ganglioneuroma in a child with multiple neurofibromatosis: Implications for the mutational origin of neuroblastoma. Cancer 19:1032-1037, 1966[Medline]

64. The I, Murthy A, Hannigan G, et al: Neurofibromatosis type 1 gene mutations in neuroblastoma. Nat Genet 3:62-66, 1993[Medline]

65. Johnson M, Look A, DeClue J, et al: Inactivation of the NF1 gene in human melanoma and neuroblastoma cell lines without impaired regulation of GTP-Ras. Proc Natl Acad Sci U S A 90:5539-5543, 1993[Abstract/Free Full Text]

66. Martinsson T, Sjoberg RM, Hedborg F, et al: Homozygous deletion of the neurofibromatosis-1 gene in the tumor of a patient with neuroblastoma. Cancer Genet Cytogenet 95:183-189, 1997[Medline]

67. Kushner BH, Gilbert F, Helson L: Familial neuroblastoma: Case reports, literature review, and etiologic considerations. Cancer 57:1887-1893, 1986[Medline]

68. Knudson AG Jr: Mutation and cancer: Statistical study of retinoblastoma. Proc Natl Acad Sci USA 68:820-823, 1971[Abstract/Free Full Text]

69. Tonini GP, Lo Cunsolo C, Cusano R, et al: Loss of heterozygosity for chromosome 1p in familial neuroblastoma. Eur J Cancer 33:1953-1956, 1997

70. Brodeur GM, Green AA, Hayes FA, et al: Cytogenetic features of human neuroblastomas and cell lines. Cancer Res 41:4678-4686, 1981[Medline]

71. Fong CT, Dracopoli NC, White PS, et al: Loss of heterozygosity for the short arm of chromosome 1 in human neuroblastomas: Correlation with N-myc amplification. Proc Natl Acad Sci U S A 86:3753-3757, 1989[Abstract/Free Full Text]

72. Fong CT, White PS, Peterson K, et al: Loss of heterozygosity for chromosomes 1 or 14 defines subsets of advanced neuroblastomas. Cancer Res 52:1780-1785, 1992[Abstract/Free Full Text]

73. White PS, Maris JM, Beltinger C, et al: A region of consistent deletion in neuroblastoma maps within human chromosome 1p36.2-36.3. Proc Natl Acad Sci U S A 92:5520-5524, 1995[Abstract/Free Full Text]

74. Takita J, Hayashi Y, Kohno T, et al: Allelotype of neuroblastoma. Oncogene 11:1829-1834, 1995[Medline]

75. Maris JM, White PS, Beltinger CP, et al: Significance of chromosome 1p loss of heterozygosity in neuroblastoma. Cancer Res 55:4664-4669, 1995[Abstract/Free Full Text]

76. Gehring M, Berthold F, Edler L, et al: The 1p deletion is not a reliable marker for the prognosis of patients with neuroblastoma. Cancer Res 55:5366-5369, 1995[Abstract/Free Full Text]

77. Caron H, van Sluis P, de Kraker J, et al: Allelic loss of chromosome 1p as a predictor of unfavorable outcome in patients with neuroblastoma. N Engl J Med 334:225-230, 1996[Abstract/Free Full Text]

78. Martinsson T, Sjoberg RM, Hedborg F, et al: Deletion of chromosome 1p loci and microsatellite instability in neuroblastomas analyzed with short-tandem repeat polymorphisms. Cancer Res 55:5681-5686, 1995[Abstract/Free Full Text]

79. Bader SA, Fasching C, Brodeur GM, et al: Dissociation of suppression of tumorigenicity and differentiation in vitro effected by transfer of single human chromosomes into human neuroblastoma cells. Cell Growth Differ 2:245-255, 1991[Abstract]

80. White PS, Maris JM, Sulman EP, et al: Molecular analysis of the region of distal 1p commonly deleted in neuroblastoma. Eur J Cancer 33:1957-1961, 1997

81. Cheng NC, Van Roy N, Chan A, et al: Deletion mapping in neuroblastoma cell lines suggests two distinct tumor suppressor genes in the 1p35-36 region, only one of which is associated with N-myc amplification. Oncogene 10:291-297, 1995[Medline]

82. Schleiermacher G, Peter M, Michon J, et al: Two distinct deleted regions on the short arm of chromosome 1 in neuroblastoma. Genes Chromosom Cancer 10:275-281, 1994[Medline]

83. Takeda O, Homma C, Maseki N, et al: There may be two tumor suppressor genes on chromosome arm 1p closely associated with biologically distinct subtypes of neuroblastoma. Genes Chromosom Cancer 10:30-39, 1994[Medline]

84. Caron H, van Sluis P, van Hoeve M, et al: Allelic loss of chromosome 1p36 in neuroblastoma is of preferential maternal origin and correlates with N-myc amplification. Nat Genet 4:187-190, 1993[Medline]

85. Kaghad M, Bonnet H, Yang A, et al: Monoallelically expressed gene related to p53 at 1p36, a region frequently deleted in neuroblastoma and other human cancers. Cell 90:809-819, 1997[Medline]

86. Lahti JM, Valentine M, Xiang J, et al: Alterations in the PITSLRE protein kinase gene complex on chromosome 1p36 in childhood neuroblastoma. Nat Genet 7:370-375, 1994[Medline]

87. Maris JM, Jensen SJ, Sulman EP, et al: Cloning, chromosomal localization, physical mapping, and genomic characterization of HKR3. Genomics 35:289-298, 1996[Medline]

88. Enomoto H, Ozaki T, Takahashi E, et al: Identification of human DAN gene, mapping to the putative neuroblastoma tumor suppressor locus. Oncogene 9:2785-2791, 1994[Medline]

89. Shapiro DN, Sublett JE, Li B, et al: The gene for PAX7, a member of the paired-box-containing genes, is localized on human chromosome arm 1p36. Genomics 17:767-769, 1993[Medline]

90. Deed RW, Hirose T, Mitchell EL, et al: Structural organisation and chromosomal mapping of the human ID-3 gene. Gene 151:309-314, 1994[Medline]

91. Saito M, Helin K, Valentine MB, et al: Amplification of the E2F1 transcription factor gene in the HEL erythroleukemia cell line. Genomics 25:130-138, 1995[Medline]

92. Aso T, Lane WS, Conaway JW, et al: Elongin (SIII): A multisubunit regulator of elongation by RNA polymerase II. Science 269:1439-1443, 1995[Abstract/Free Full Text]

93. Beltinger CP, White PS, Maris JM, et al: Physical mapping and genomic structure of the human TNFR2 gene. Genomics 35:94-100, 1996[Medline]

94. Marsters SA, Sheridan JP, Donahue CJ, et al: Apo-3, a new member of the tumor necrosis factor receptor family, contains a death domain and activates apoptosis and NF-kappa B. Curr Biol 6:1669-1676, 1996[Medline]

95. Schleiermacher G, Delattre O, Peter M, et al: Clinical relevance of loss of heterozygosity of the short arm of chromosome 1 in neuroblastoma: A single-institution study. Int J Cancer 69:73-78, 1996[Medline]

96. Rubie H, Delattre O, Hartmann O, et al: Loss of chromosome 1p may have a prognostic value in localised neuroblastoma: Results of the French NBL 90 study—Neuroblastoma Study Group of the Societe Francaise d'Oncologie Pediatrique (SFOP). Eur J Cancer 33:1917-1922, 1997

97. Maris JM, Guo C, Blake D, et al: A comprehensive analysis of chromosome 1p deletions in neuroblastoma. Eur J Cancer (in press)

98. Mertens F, Johansson B, Mitelman F: Chromosomal imbalance maps of malignant solid tumors: A cytogenetic survey of 3185 neoplasms. Cancer Res 57:2765-2780, 1997[Abstract/Free Full Text]

99. Srivatsan ES, Ying KL, Seeger RC: Deletion of chromosome 11 and of 14q sequences in neuroblastoma. Genes Chromosom Cancer 7:32-37, 1993[Medline]

100. Maris JM, Guo C, White PS, et al: Allelic deletion at chromosome band 11q14-23 is common in neuroblastoma. Eur J Cancer (in press)

101. Guo C, White PS, Blake D, et al: Allelic deletion at 11q23 is common in MYCN single copy neuroblastomas. Oncogene (in press)

102. Suzuki T, Yokota J, Mugishima H, et al: Frequent loss of heterozygosity on chromosome 14q in neuroblastoma. Cancer Res 49:1095-1098, 1989[Abstract/Free Full Text]

103. Takayama H, Suzuki T, Mugishima H, et al: Deletion mapping of chromosomes 14q and 1p in human neuroblastoma. Oncogene 7:1185-1189, 1992[Medline]

104. Petkovic I, Cepulic M: Cytogenetic analysis of primary neuroblastoma with del(1), del(14), hsr, and dmin chromosomes. Cancer Genet Cytogenet 55:231-234, 1991[Medline]

105. Thompson PM, Seifried BA, Kyemba SK, et al: Loss of heterozygosity for chromosome 14q in neuroblastoma. Eur J Cancer (in press)

106. Ejeskar K, Aburatani H, Abrahamsson J, et al: Loss of heterozygosity of 3p markers in neuroblastoma tumours implicate a tumour-suppressor locus distal to the FHIT gene. Br J Cancer 77:1787-1791, 1998[Medline]