Histological profile of tumours from MYCN transgenic mice

Histologial profile of tumours from MYCN transgeni mie H C Moore, 1 K M Wood, 2 M S Jakson, 3 M A Lastowska, 3 D Hall, 3 H Imrie, 1 C P F Redfern, 1 P E Lovat, 1 F Ponthan, 1 K O Toole, 1 J Lune, 1 D A Tweddle 1 1 Northern Institute for Caner Researh, Newastle University, Newastle upon Tyne, UK; 2 Department of Cellular Pathology, Royal Vitoria Infirmary, Newastle upon Tyne, UK; 3 Institute of Human Genetis, Newastle University, Newastle upon Tyne, UK Correspondene to: Dr D A Tweddle, Northern Institute for Caner Researh, Paul O Gorman Building, Newastle University, Framlington Plae, Newastle upon Tyne NE2 4HH, UK; D.A. Tweddle@newastle.a.uk Aepted 18 July 2008 Published Online First 4 August 2008 ABSTRACT Bakground: MYCN is the most ommonly amplified gene in human neuroblastomas. This proto-onogene has been overexpressed in a mouse model of the disease in order to explore the role of MYCN in this tumour. Aims: To report the histopathologial features of neuroblastomas from MYCN transgeni mie. Methods: 27 neuroblastomas from hemizygous transgeni mie and four tumours from homozygous mie were examined histologially; Ki67 and MYCN immunoytohemistry was performed in 24 tumours. Results: Tumours obtained from MYCN transgeni mie resembled human neuroblastomas, displaying many of the features assoiated with stroma-poor neuroblastoma, inluding heterogeneity of differentiation (but no overt ganglioni differentiation was seen), low levels of Shwannian stroma and a high mitosis karyorrhexis index. The tumours had a median Ki67 labelling index of 70%; all tumours expressed MYCN with a median labelling index of 68%. The most striking differene between the murine and human neuroblastomas was the presene of tingible body marophages in the transgeni mouse tumours refleting high levels of apoptosis. This has not previously been desribed in human or other murine neuroblastoma models. Conlusions: These studies highlight the histologial similarities between tumours from MYCN transgeni mie and human neuroblastomas, and reaffirm their role as a valuable model to study the biology of aggressive human neuroblastoma. A neuroblasti tumour, the most ommon extraranial solid tumour of hildhood, 1 is an embryonal malignany of neural rest origin with the primary tumour arising at a site within the sympatheti nervous system. 1 Heterogeneity is the hallmark of human neuroblastomas, with spontaneous regression and maturation being ommonly desribed. 2 The different biologial and morphologial harateristis of the tumours an be used to predit their linial behaviour; histologially they an be plaed into one of three ategories: neuroblastoma, ganglioneuroblastoma (nodular or intermixed) or ganglioneuroma. 2 Amplifiation of the proto-onogene MYCN, a member of the MYC family of onogenes, is a well haraterised geneti abnormality of some neuroblastomas. 1 Amplifiation is found in approximately 25% of tumours and is assoiated with advaned stages of the disease and a poor patient prognosis. 3 8 Until reently, the most widely used animal models to study neuroblastoma have been orthotopi or heterotopi rodent xenograft models, and they have been generally used to study drug effiay. To test the hypothesis that amplifiation of MYCN ontributes to tumourigenesis, and to explore the role of MYCN in neuroblastoma, a transgeni mouse model was reated in whih the MYCN onogene was targeted to neuroetodermal ells of developing mie under the influene of a rat tyrosine hydroxylase promoter reating mie that overexpress MYCN in neural rest derived ells. 9 10 Mie develop neuroblastoma tumours resembling those seen in humans, and omparative genomi hybridisation (CGH) has identified onsistent regions of hromosomal gain and loss synteni to those that are found in human neuroblastoma, indiating that this murine model is representative of the human disease. 9 Furthermore, tumours from hemizygous mie have also been reported to undergo amplifiation of the MYCN transgene. 11 These transgeni mie tumours will hopefully be useful in the identifiation of additional genes involved in the tumourigeni proess, and in the development of therapies to treat patients suffering from neuroblastoma. The aim of this study was to report the histopathologial features in 27 neuroblastomas from mie hemizygous for the MYCN transgene and four from mie homozygous for the transgene, to determine how losely they resemble the pathology of human neuroblastoma. METHODS MYCN transgeni mouse tumours Twenty-seven formalin-fixed, paraffin-embedded neuroblastomas from mie hemizygous for the MYCN transgene from strain 129X1/SvJ were studied. Ethial approval for this study was obtained from the loal researh ethis ommittee. Mie were examined every other day after weaning for palpable tumours or evidene of hindleg paralysis, and when present were sarified and the tumour removed for study. For omparison, four tumours from mie homozygous for the MYCN transgene were also studied (129X1/SvJ strain and CBA strain). Table 1 shows harateristis of the tumours in terms of lateny and site of primary tumour. H&E staining of at least one setion from eah tumour was examined by light mirosopy and the histopathologial features in terms of differentiation, presene of Shwannian stroma, estimated mitosis karyorrhexis index (MKI) and presene of alifiation were reorded using the International Neuroblastoma Pathology Classifiation (INPC). 2 The MKI is defined as the perentage of ells in mitosis and karyorrhexis (apoptosis)/5000 neuroblasti ells sored. It is lassified as low if,2%, intermediate if 2 4% and high if.4%. 2 1098 J Clin Pathol 2008;61:1098 1103. doi:10.1136/jp.2007.054627

Table 1 Charateristis of MYCN transgeni mouse tumours Tumour no Sex Lateny Tumour loation Sample Histology{ Neu 4* Male 10 weeks Left abdominal Fr Neu 5* Male 12 weeks Paravertebral Fr, FF Poorly differentiated Neu 7* Female 13 weeks Right abdominal Fr Neu 8* Female 9 weeks Left abdominal Fr, FF Undifferentiated Neu 9{ Male 17 weeks Left abdominal Fr, FF Poorly differentiated Neu 10{ Female 15 weeks Left abdominal Fr, FF Poorly differentiated Neu 11{ Female 6 weeks Left abdominal Fr, FF Poorly differentiated Neu 12{ Male 13 weeks Thorax FF Undifferentiated Neu 13*{ Male 11 weeks Abdominal bilateral Fr, FF Poorly differentiated Neu 14{ Female 13 weeks Left abdominal FF Undifferentiated Neu 15{ Female 6 weeks Abdominal Fr, FF Undifferentiated Neu 16{ Female 9 weeks Abdominal small Fr, FF Undifferentiated Neu 17{ Female 10 weeks Abdominal FF Poorly differentiated Neu 18{ 5 weeks Abdominal FF Undifferentiated Neu 19{ Male 8 weeks Abdominal Fr, FF Undifferentiated Neu 20{ Male 6 weeks Abdominal Fr, FF Undifferentiated Neu 21{ Female 8 weeks Abdominal Fr, FF Poorly differentiated Neu 22*{ Male 8 weeks Abdominal Fr, FF Undifferentiated Neu 23 Female 8 weeks Abdominal Fr, FF Undifferentiated Neu 24{ Female 7 weeks Fr, FF Undifferentiated Neu 25{ Female 9 weeks Abdominal Fr, FF Undifferentiated Neu 27{ Male 6 weeks Abdominal Fr, FF Undifferentiated Neu 28{ Male 8 weeks Abdominal Fr, FF Undifferentiated Neu 29{ Female 11 weeks Abdominal Fr, FF Undifferentiated Neu 30{ Male 13 weeks Abdominal Fr, FF Poorly differentiated Neu 31{ Male 13 weeks Abdominal Fr, FF Undifferentiated Neu 33{ Male 17 weeks Right abdominal Fr, FF Undifferentiated Neu 34{ Female 10 weeks Abdominal Fr, FF Undifferentiated Neu 35{ Female 10 weeks Fr, FF Undifferentiated Neu 36{ Male 10 weeks Fr, FF Undifferentiated Neu 581 Male 6 weeks Abdominal Fr Undifferentiated Neu 591 Male 7 weeks Abdominal FF Undifferentiated Neu 601 Male 8 weeks Abdominal Fr Undifferentiated Neu 611 Female 8 weeks Abdominal Fr Undifferentiated *Used for western blotting. {Used for immunoytohemistry. {International Neuroblastoma Pathology Classifiation. 1Homozygous tumours. Fr, frozen tumour sample; FF, formalin-fixed tumour samples. Immunohistohemistry On one representative tumour, immunohistohemistry for the neural differentiation markers PGP9.5, synaptophysin, neurofilament and NB84 was performed as previously desribed, 12 using normal goat serum to blok bakground staining. PGP9.5 is a neuron-speifi enolase of the ubiquitin C-terminal hydroxylase family found in metabolially dynami regions of the ell. For the detetion of human MYCN transgene expression, a mouse monolonal antibody (MYCN100; gift from Dr Nao Igekaki), at a dilution of 1 in 16 was used, along with the LINK and LABEL detetion system (BioGenex Laboratories, San Remon, California, USA). This involved pressure-ooking antigen retrieval and pre-diluted working solutions of a biotinylated goat anti-mouse seondary antibody (LINK) and a streptavidin horseradish peroxidase solution (LABEL). For Ki67 a rabbit polylonal primary antibody (Abam, Cambridge, UK) was used at a dilution of 1 in 50 with a goat anti-rabbit seondary antibody (Dako, Glostrup, Denmark). The Ki67 antigen is a prototypi ell yle related nulear protein and is expressed by proliferating ells in all phases of the ative ell yle (G 1,S,G 2 and M phase) but is absent in resting (G 0 ) ells. Antibodies against Ki67 are useful to establish the ell growing fration in neoplasms, being quantified immunohistohemially by determining the number of Ki67 positive ells among the total number of ells (Ki67 index). Soring of slides Tumour setions were sored as being either positively or negatively immunostained based on the labelling index (LI). This was the perentage of positive ells out of 1000 ells sored by seleting one or two fields with the highest density of positively stained tumour ells at a magnifiation of 6400. Soring was performed by two independent observers (KW and HM). Conordane was ahieved in over 90% of ases; disrepanies were resolved by resoring of relevant ases by both observers together to reah a onsensus. Western blotting Cell lysates were prepared from six frozen tumours (Neu 4, 5, 7, 8, 13 and 22). Western blotting was performed aording to previously published methods, 13 using the MYCN100 antibody at 1:50, a monolonal atin (Dako) antibody at 1 in 500, and a goat anti-mouse seondary antibody for both. A seletion of ell line lysates were also inluded as positive and negative ontrols: J Clin Pathol 2008;61:1098 1103. doi:10.1136/jp.2007.054627 1099

Figure 1 H&E staining of neuroblastomas from MYCN transgeni mie. (A) Islands of undifferentiated neuroblastoma separated by fibrous septae (Neu 23) (6200). (B) Cells with a high-mitosis karyorrhexis index with numerous intensely stained apoptoti and mitoti ells in the entre (Neu 10) (6400). (C) Islands of larger differentiating neuroblastoma ells with single nuleoli (arrowed) (Neu 10) (6400). (D) Poorly differentiated neuroblastoma with neuropil among islands of undifferentiated tumour (arrowed) (Neu 13) (6400). (E) (Neu 23) and (F) (Neu 13) The presene of tingible body marophages (long arrow) reflets the high levels of apoptosis (short arrow) shown in F (6400). MYCN-amplified IMR-32, 13 MYCN non-amplified SHSY5Y, 13 LAN-6 14 and a ell line established from a transgeni mouse tumour homozygous for the MYCN transgene (3402A). 15 RESULTS Histologial features of MYCN transgeni tumours Twenty-seven formalin-fixed, paraffin-embedded MYCN transgeni hemizygous murine neuroblastomas were studied and the histology reorded using the INPC lassifiation used for human neuroblastoma. 2 H&E staining revealed that the MYCN transgeni mie tumours were predominantly stroma-poor, undifferentiated, small round blue ell tumours laking neuropil. Tumour ells were arranged in islands, separated by fibrous septa (fig 1A). A high MKI was present in all of them and was distinguishable by the fragmentation of nulear material and ondensation of hromatin within abundant apoptoti ells, and the presene of frequent mitoti figures (fig 1B). 2 The majority of tumours showed no evidene of neuropil or differentiation towards ganglion ells, although small areas in a minority of tumours ontained small lusters of differentiating neuroblasts (fig 1C), and neuropil was present (fig 1D). No Homer Wright pseudorosettes were seen in any tumours, and no immature or mature ganglion ells assoiated with Shwannian stroma were seen. A small amount of alifiation was present in a minority of samples, as is often seen in human neuroblastomas, although the biologial signifiane of alifiation has yet to be established. 16 Overall the appearanes of the MYCN murine tumours resembled human undifferentiated and poorly differentiated neuroblastoma (table 1). 2 Compared with the hemizygous tumours, the four homozygous tumours were all undifferentiated with a high MKI (table 1). The most striking differene between the MYCN transgeni mie neuroblastoma and human neuroblastoma was the presene of tingible body marophages in the majority of hemizygous and all homozygous tumours, a feature not previously reported in human neuroblastoma (fig 1E,F). The polylonal PGP9.5 antibody immunostained the murine tumour tested, onfirming the neuroblasti origin (fig 2A,B). The adrenal medulla in fig 2B ated as an internal positive 1100 J Clin Pathol 2008;61:1098 1103. doi:10.1136/jp.2007.054627

Original artile Figure 2 (A) PGP9.5 immunostaining showing strong positivity in neuroblasts and adjaent normal adrenal medulla (Neu 23) (6100). (B) Strong PGP9.5 immunostaining in adrenal medulla and negative adrenal ortex (Neu 23) (6200). (C) Ki67 immunostaining is strongly positive in mitoti ells and negative in apoptoti ells (arrowed) (6400). (D) Ki67 immunostaining showing positive neuroblasts adjaent to negative lymph node ortex (6200). (E) MYCN immunostaining showing heterogeneous nulear expression with negative stromal ells (Neu 25)6400. (F) MYCN immunostaining showing strong positivity in neuroblasts ompared with negative adjaent lymphoytes (6400). ontrol. NB84, synaptophysin and neurofilament, all mouse monolonal antibodies, were negative. Ki67 expression Ki67 expression was examined in 22 hemizygous tumours and all showed positive nulear immunostaining with nuleolar enhanement, although some were stronger than others (fig 2C,D). The median Ki67 labelling index was high at 70% (range 5 95), as expeted for highly proliferative tumours. MYCN expression MYCN was expressed by all hemizygous tumours studied (n = 24), with a median labelling index of 70% (range 10 95) for strongly immunostained ells. Nulear, heterogeneous immunostaining was seen in a variable perentage of neuroblasti ells (fig 2E,F). The appearane was onsistent with a stars in the sky pattern previously reported using immunohistohemistry on frozen MYCN amplified human neuroblastomas3 and some formalin-fixed, paraffin embedded MYCN amplified tumours.5 Speifiity of MYCN immunostaining was onfirmed by J Clin Pathol 2008;61:1098 1103. doi:10.1136/jp.2007.054627 western blotting, whih showed a 64 kd human MYCN protein in all tumours studied, together with a lower moleular weight band at 60 kd. This was possibly due to the seondary goat anti-mouse antibody reating with other mouse immunoglobulins as it was not present in any of the ell line ontrols. The MYCN transgeni mouse tumours expressed omparable levels of MYCN protein to those in the human MYCN amplified IMR32 ell line and the murine ell line 3402A derived from a mouse homozygous for the MYCN transgene (fig 3). In this series of murine transgeni tumours, there was no orrelation between Ki67 and MYCN labelling indies (Pearson orrelation oeffiient: p = 0.487, data not shown). DISCUSSION Human neuroblasti tumours show a spetrum of histology ranging from undifferentiated stroma-poor neuroblastoma through stroma-rih ganglioneuroblastomas to mature ganglioneuromas. Neuroblasti tumours are lassified aording to the INPC as having unfavourable or favourable histology on the basis of patient age and histology.2 Unfavourable tumours 1101

Figure 3 Western blots showing high levels of MYCN expression in a seletion of MYCN transgeni mouse tumours ompared with the MYCN-amplified human neuroblastoma ell line IMR-32, MYCN nonamplified human neuroblastoma ell lines SHSY5Y and LAN-6, and a MYCN transgeni mouse tumour ell line (3402A). The presene of a lower moleular weight 60 kd band in the mouse tumours may be from binding of the seondary antibody to mouse immunoglobulins. inlude stroma-poor, undifferentiated tumours with a high MKI in the,18 month of age group, and undifferentiated or poorly differentiated tumours with a moderate or high MKI in the 18 month to 5 year age group. These are the types most often seen 2 17 18 in MYCN amplified tumours. Histologially the MYCN transgeni mouse tumours are analogous to those seen in MYCN amplified human disease, omprising stroma-poor, undifferentiated and poorly differentiated neuroblastoma with a high MKI (unfavourable histology). This makes the MYCN transgeni mouse tumours an exellent model for studying aggressive MYCN amplified neuroblastoma. All tumours were stroma-poor and no immature or mature ganglion ells assoiated with Shwannian stroma were seen. This is onsistent with the observation that MYCN amplifiation is rarely, if ever, found in ganglion ells of Shwannian stroma-rih ganglioneuroblastoma either intermixed, or the Shwannian stroma-rih areas of nodular ganglioneuroblastoma, or ganglioneuroma. 17 19 The ganglion-like ells desribed in the original report by Weiss et al were also in the ontext of a Shwannian stroma-poor bakground. Neural differentiation was onfirmed by PGP9.5 positivity. The presene of undifferentiated and poorly differentiated neuroblastoma and the absene of differentiating neuroblastoma in the MYCN transgeni tumours is onsistent with the histology found in human MYCN amplified neuroblastomas, in whih the differentiating subtype is rarely found, regardless of age. 17 18 Homer Wright pseudorosettes are typially rings of neuroblasts surrounding a entral ore or eosinophili neuropil. They are found in poorly-differentiated neuroblastoma regardless of MYCN status, and their absene in the MYCN transgeni mouse tumours most likely reflets the low levels of neuropil found in these tumours. The presene of a high MKI in the MYCN transgeni mouse tumours is onsistent with the presene of a high MKI in 45/51 (88%) MYCN amplified human neuroblastomas in one series. 17 An intermediate MKI has been reported in MYCN amplified human tumours from patients aged 18 months to 5 years. In MYCN amplified tumours from patients over the age of 5 years, whih are very rarely found, a low MKI has been reported. 18 Thus the presene of a high MKI in all of the MYCN transgeni mie tumours studied in this series is most representative of the biology of the MYCN amplified human tumours in the,18 month age group. The most notable differene between these murine-derived tumours and human neuroblastoma was the presene of tingible body marophages, a feature seen in a variety of aggressive tumour types and harateristially in Burkitt lymphoma, where C-MYC is ativated and overexpressed via a t(8;14) or a t(2;8) transloation of C-MYC alongside the immunoglobulin heavy or light hain promoter respetively, leading to C-MYC overexpression. 20 These marophages are generally found in the germinal entres of lymph nodes where their presene reflets high rates of proliferation and apoptosis, as the marophages ingest apoptoti lymphoytes. This is the first report of these marophages being present in neuroblastomas; their presene reflets the high rate of ell death and proliferation ourring within the tumours. Sine most human MYCN amplified neuroblastomas have a high MKI, as noted above, it is intriguing that these tingible body marophages have never, to our knowledge, been reported in human neuroblastoma. This begs the question of whether marophage infiltration is a peuliarity of mouse tumour tissue; the neuroblastoma mouse tissue ould be sereting some kind of hemoattrating marophage fator not sereted in human neuroblastoma. However, if this was the ase, one would expet an immune response with lymphoyte infiltration of the tumours whih was not observed in any of the MYCN transgeni tumours studied. Comparison with other mouse models of neuroblastoma is not possible as they are either human xenograft tumours in immunodefiient mie, 10 or syngenei murine neuroblastoma models whih do not eliit an immune response. Lymphoyti infiltration of human neuroblastomas has been observed to our most often and extensively in neuroblastomas from patients with the opsolonus myolonus syndrome (OMS) whih are usually INPC differentiating or stroma-rih tumours, very rarely MYCN amplified and usually assoiated with a good prognosis. 21 22 In neuroblastomas from patients without OMS, foal lymphoyte infiltration has been reported but not from patients with metastati or MYCN amplified neuroblastoma. Thus the apparent lak of lymphoyte infiltration in the MYCN transgeni tumours is onsistent with the pauity of lymphoyti infiltration in MYCN amplified human neuroblastoma. It is most likely that the presene of tingible body marophages in the MYCN transgeni tumours an be explained by the very high levels of mitosis and apoptosis ourring in these tumours, as ours in Burkitt lymphoma. Indeed the very high Ki67 proliferative index (median of 70%) found in transgeni mouse tumours is omparable to that found in high grade lymphomas. In human MYCN amplified neuroblastoma, although the Ki67 index is signifiantly higher than in nonamplified tumours, a median of 36% was reported in MYCN amplified tumours in one study of 19 MYCN amplified tumours 23 and a median of 31% in another study of 38 MYCN amplified tumours. 24 The observation of undifferentiated and poorly differentiated neuroblastoma in the MYCN transgeni tumours is also onsistent with in vitro data reporting high levels of MYCN in undifferentiated, MYCN amplified neuroblastoma ell lines, whih are redued after retinoi aid indued differentiation, 25 and also onsistent with the differentiation indued by short interfering RNA mediated MYCN knokdown. 26 27 It has been reported that the presene of My Max heterodimers in the nulei of neuroblasti tumours prevents the ell from differentiating in the absene of exogenous differentiating agents. 28 This histologial study of MYCN transgeni mouse tumours has onfirmed that the MYCN transgeni mouse represents an 1102 J Clin Pathol 2008;61:1098 1103. doi:10.1136/jp.2007.054627

Take-home messages MYCN transgeni mouse tumours are stroma-poor, undifferentiated or poorly differentiated with a high mitosis karyorrhexis index. This histologial appearane most losely resembles the histology of MYCN amplified tumours in the,18 month age group. Tingible body marophages refleting high levels of apoptosis were found in MYCN transgeni mouse tumours, but have not been previously reported in human neuroblastoma. MYCN transgeni mouse tumours provide an exellent model to study the biology of aggressive MYCN amplified human neuroblastoma. exellent model for studying the biology of human MYCN amplified neuroblastoma, and for the future investigation of potential MYCN-direted targeted therapies. Aknowledgements: We thank Dr W Weiss, University of California (San Franiso) for the MYCN transgeni mie, Dr Nao Igekaki (urrent address: University of Illinois) for the MYCN 100 antibody, and Dr Mihelle Haber (Children s Caner Researh Institute, Sydney, Australia) for the 3402A MYCN transgeni mouse ell line. Funding: Supplied by the North of England Children s Caner Researh Fund, the Newastle Healthare Charity and the Neuroblastoma Soiety. DAT is a Department of Health Cliniian Sientist. Competing interests: None delared. REFERENCES 1. Brodeur GM. Neuroblastoma: biologial insights into a linial enigma. Nat Rev Caner 2003;3:203 16. 2. Shimada H, Ambros IM, Dehner LP, et al. Terminology and morphologi riteria of neuroblasti tumours. Caner 1999;86:349 63. 3. Seeger RC, Wada R, Brodeur GM, et al. Expression of N-my by neuroblastomas with one or more multiple opies of the onogene. Prog Clin Biol Res 1988;271:41 9. 4. Seeger RC, Brodeur GM, Sather H, et al. Assoiation of multiple opies of the N-my onogene with rapid progression of neuroblastomas. N Engl J Med 1985;313:1111 6. 5. Chan HSL, Gallie BL, DeBoer PS, et al. MYCN protein expression as a preditor or neuroblastoma prognosis. Clin Can Res 1997;3:1699 706. 6. Kohl NE, Kanda N, Shreh RR. Transposition and amplifiation of onogene-related sequenes in human neuroblastoma. Cell 1983;35:359 67. 7. Shwab M, Varmus HE, Bishop JM. Human gene N-my ontributes to neoplasti transformation of mammalian ells in ulture. Nature 1985;316:160 2. 8. Cohn SL, Tweddle DA. MYCN amplifiation remains prognostially strong 20 years after its linial debut. Eur J Caner 2004;40:2639 42. 9. Weiss WA, Aldape K, Mohapatra G, et al. Targeted expression of MYCN auses neuroblastoma in transgeni mie. EMBO J 1997;16:2985 95. 10. Dyer MA. Mouse models of hildhood aner of the nervous system. J Clin Pathol 2004;57:561 76. 11. Hansford LM, Thomas WD, Keating JM, et al. Mehanisms of embryonal tumor initiation: distint roles for MyN expression and MYCN amplifiation. Pro Natl Aad Si USA 2004;101:12664 9. 12. Bomken SN, Redfern K, Wood KM, et al. Limitations in the ability of NB84 to detet metastati neuroblastoma ells in bone marrow. J Clin Pathol 2006;59:927 9. 13. Tweddle DA, Malolm AJ, Cole M, et al. p53 ellular loalisation and funtion in neuroblastoma: evidene for defetive G1 arrest despite WAF1 indution in MYCNamplified ells. Am J Pathol 2001;158:2067 77. 14. Wada RK, Seeger RC, Brodeur GM. Charaterisation of human neuroblastoma ell lines that lak N-my gene amplifiation. Prog Clin Biol Res 1988;271:57 69. 15. Cheng AJ, Ching Cheng N, Ford J, et al. Cell lines from MYCN transgeni murine tumours reflet the moleular and biologial harateristis of human neuroblastoma. Eur J Caner 2007;43:1467 75. 16. Joshi VV, Rao RV, Cantor AB, et al. Modified histologi grading of neuroblastoma by replaement of mitoti rate with mitosis karyorrhexis index: a liniopathologial study of 223 ases from the Paediatri Onology Group. Caner 1996;77:1582 8. 17. Shimada H, Stram DO, Chatten J, et al. Identifiation of subsets of neuroblastomas by ombined histopathologi and N-my analysis. J Natl Caner Inst 1995;87:1470 6. 18. Goto S, Umehara S, Gerbing RB, et al. Histopathology (International Neuroblastoma Pathology Classifiation) and MYCN status in patients with peripheral neuroblasti tumours: a report from the Children s Caner Group. Caner 2001;92:2699 708. 19. Umehara S, Nakagawa A, Matthay KK, et al. Histopathology defines prognosti subsets of ganglioneuroblastoma, nodular. Caner 2000;89:1150 61. 20. Ferry JA. Burkitt s lymphoma: linopathologial features and differential diagnosis. Onologist 2006;11:375 83. 21. Cooper R, Khakoo Y, Matthay KK, et al. Opsolonus-myolonus-ataxia syndrome in neuroblastoma: histopathologi features a report from the Children s Caner Group. Med Pediatr Onol 2001;36:623 9. 22. Gambini C, Conte M, Bemini G, et al. Neuroblasti tumours assoiated with opsolonus-myolonus syndrome: histologial, immunohistohemial and moleular features of 15 Italian ases. Virhows Arh 2003;442:555 62. 23. Rudolph P, Lappe T, Hero B, et al. Prognosti signifiane of the proliferative ativity in neuroblastoma. Am J Pathol 1997;150:133 45. 24. Krams M, Hero B, Berthold F, et al. Proliferation marker KI-S5 disriminates between favorable and adverse prognosis in advaned stages of neuroblastoma with and without MYCN amplifiation. Caner 2002:94;854 61. 25. Thiele CJ, Reynolds CP, Israel MA. Dereased expression of N-my preedes retinoi aid-indued morphologial differentiation of human neuroblastoma. Nature 1985;313:404 6. 26. Nara K, Kusafuka T, Yoneda A, et al. Silening of MYCN by RNA interferene indues growth inhibition, apoptoti ativity and ell differentiation in a neuroblastoma ell line with MYCN amplifiation. Int J Onol 2007;30:1189 96. 27. Kang JH, Ryhahou PG, Ishola TA, et al. MYCN silening indues differentiation and apoptosis in human neuroblastoma ells. Biohem Biophys Res Commun 2006;351:192 7. 28. Wenzel A, Cziepluh C, Hamann U, et al. The N-My onoprotein is assoiated in vivo with the phosphoprotein Max (p20/22) in human neuroblastoma ells. EMBO J 1991;10:3703 12. J Clin Pathol 2008;61:1098 1103. doi:10.1136/jp.2007.054627 1103