nature template - pc word 97 · web viewinterpretation amplification of the mycn oncogene may...
TRANSCRIPT
Discovery of MYCN-amplified retinoblastoma with functional
retinoblastoma protein in very young childrenNot Knudson’s
Retinoblastoma:
One-Hit Cancer Initiated by the MYCN Oncogene?
Diane E Rushlow, Berber M Mol,* Jennifer Y Kennett,* Berber M Mol,* Stephanie Yee,*
Sanja Pajovic, Brigitte L Thériault, Nadia L Prigoda-Lee, Clarellen Spencer, Helen
Dimaras, Timothy W Corson, Renee Pang, Christine Massey, Roseline Godbout, Zhe
Jiang, Eldad Zacksenhaus, Katherine Paton, Annette C Moll, Claude Houdayer, Anthony
Raizis, William Halliday, Wan L Lam, Paul C Boutros, Dietmar Lohmann, Josephine C
Dorsman, Brenda L Gallie
*These authors share second authorship
Retinoblastoma Solutions and the Toronto Western Hospital Research Institute, Campbell
Family Cancer Research Institute, Princess Margaret Cancer Centre, University Health
Network; Informatics and Biocomputing Platform, Ontario Institute for Cancer Research;
Departments of Hematology/Oncology, Ophthalmology and Visual Science and of
Pathology, Hospital for Sick Children; and Departments of Molecular Genetics,
Ophthalmology, Medical Biophysics, Pathobiology and Lab Medicine, University of Toronto,
Toronto, ON, Canada (D E Rushlow, BSc, S Yee, MSc, S Pajovic, PhD, B L Thériault, PhD, N L
Prigoda-Lee, MSc, C Spencer, BSc, Z Jiang, BSc, E Zacksenhaus, PhD, R Pang, MA, C Massey,
MSc, H Dimaras, PhD, P C Boutros, PhD, W Halliday, MD, Prof B L Gallie, MD); Departments of
Clinical Genetics, Ophthalmology and Pediatric Oncology/Hematology, VU University
Medical Center Amsterdam, Amsterdam, The Netherlands (B M Mol, MSc, A C Moll, MD, Prof J
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
C Dorsman, PhD); British Columbia Cancer Research Centre and Departments of
Ophthalmology and Pathology & Laboratory Medicine, University of British Columbia,
Vancouver, BC, Canada (J Y Kennett, MSc, K Paton, MD, Prof W L Lam, PhD); Departments of
Clinical Genetics, Ophthalmology and Pediatric Oncology/Hematology, VU University
Medical Center Amsterdam, Amsterdam, The Netherlands (B M Mol, MSc, A C Moll, MD, Prof J
C Dorsman, PhD); Eugene and Marilyn Glick Eye Institute, Departments of Ophthalmology,
Biochemistry and Molecular Biology, Indiana University School of Medicine Indianapolis,
Indiana, USA (Timothy W Corson, PhD); University of Alberta, Cross Cancer Institute,
Edmonton, Alberta (Prof Roseline Godbout, PhD); Service de Génétique Oncologique, Institut
Curie and Université Paris Descartes Paris, France (C Houdayer, PhD); Department of
Molecular Pathology, Canterbury Health Laboratories Christchurch, New Zealand (A Raizis,
PhD); and Institut für Humangenetik, Universitätsklinikum, Essen, Germany (Prof D Lohmann,
MD)
Correspondence to Dr Brenda L. Gallie, Campbell Family Cancer Research Institute, Princess Margaret
Cancer Centre, University Health Network, Rm 8-415, 610 University Ave, Toronto, ON, M5G 1M9, Canada,
2
1
2
3
4
5
6
7
8
9
10
11
12
13
15
16
17
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Word count 297/300
Summary
Background Retinoblastoma is the childhood retinal cancer that defined tumour suppressor genes.
By analysing age of diagnosis, Knudson proposed that two “hits” initiate retinoblastoma, later
attributed to mutation of both alleles of the retinoblastoma suppressor gene, RB1, in tumours.
Persons with hereditary retinoblastoma carry a heterozygous constitutional RB1 mutation; one
additional hit initiates retinoblastoma and/or other cancers. Non-hereditary retinoblastoma arises
when both RB1 alleles are damaged in developing retina (RB1-/- ).
Methods We analysed clinical data, genomic copy-number, histology, RB1 gene expression and
protein function, and retinal gene expression, in unilateral non-familial retinoblastomas with no
evidence of RB1 mutations (RB1+/+ ), compared to RB1-/- tumours.
Findings We found that 2·7% (29/1068) of unilateral non-familial retinoblastomas had no evidence
of RB1 mutations (RB1+/+ ). Surprisingly, half of the RB1+/+ tumours had high-level MYCN oncogene
amplification (28 to 121 copies), while no RB1-/- primary tumours showed MYCN amplification
(p<0·0001). RB1+/+ MYCNA cell lines expressed functional RB1 protein. RB1+/+ MYCNA tumours had
fewer overall genomic copy-number changes and distinct, aggressive histology. MYCN
amplification was the sole copy-number change detected in one RB1+/+ MYCNA retinoblastoma.
Median age at diagnosis of RB1+/+ MYCNA tumours was 4·5 months, compared to 24 months for
non-familial unilateral RB1-/- retinoblastoma. Children diagnosed with unilateral non-familial large
retinoblastoma at six months of age or less have >18% chance of RB1+/+ MYCNA tumour.
Interpretation Amplification of the MYCN oncogene may initiate RB1+/+ MYCNA retinoblastoma
despite normal RB1 genes. Although they are young, it is unlikely that these children carry a
hereditary risk for other retinoblastoma or cancers. Since these tumours may rapidly become
3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
extraocular, removal of the eye of young children with potential large RB1+/+ MYCNA retinoblastoma
is recommended.
Funding NCI-NIH; CIHR; Canadian Retinoblastoma Society; Hyland Foundation; Ontario
Ministry of Health and Long Term Care; Toronto Netralya and Doctors Lions Clubs; and
Foundations Avanti-STR and KiKa.
Word count 298/300
Summary
Background Retinoblastoma is the childhood retinal cancer that defined tumour suppressor genes.
By analysing age of diagnosis, Knudson proposed that two “hits” initiate retinoblastoma, later
attributed to mutation of both alleles of the retinoblastoma suppressor gene, RB1, in tumours.
Persons with hereditary retinoblastoma carry a heterozygous constitutional RB1 mutation; one
additional hit initiates retinoblastoma or other cancers. Non-hereditary retinoblastoma arises when
both RB1 alleles are damaged in developing retina (RB1-/-).
Methods Our international collaboration determined the proportion of 1068 unilateral non-familial
retinoblastomas with no evidence of RB1 mutations despite high-sensitivity assays. We analysed
clinical data, genomic copy-number changes, histology, RB1 gene expression and protein function,
comparing RB1+/+and RB1-/- tumours.
Findings No evidence of RB1 mutation (RB1+/+) was found In 2·7% (29/1068) of unilateral non-
familial retinoblastomas. Surprisingly, half of these had high-level MYCN oncogene amplification
(28 to 121 copies), while no RB1-/- primary tumours showed MYCN amplification. . Analysis of three
revealed (EZ – otherwise sounds like only three RB+/+ tumor show functional RB)RB1+/+MYCNA
tumours had fewer overall genomic copy-number changes and distinct, aggressive histology. MYCN
amplification was the sole copy-number change detected in one RB1+/+MYCNA retinoblastoma.
Median age at diagnosis of RB1+/+MYCNA tumours was 4·5 months, compared to 24 months for
non-familial unilateral RB1-/- retinoblastoma. We calculate an 18·4% chance that A child diagnosed
4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
with unilateral non-familial retinoblastoma at six months of age or less will have an RB1+/+MYCNA
tumour.
Interpretation Amplification of the MYCN oncogene may initiate RB1+/+MYCNA retinoblastoma
despite normal RB1 genes. Despite their young age at diagnosis, it is very unlikely that these
children carry a hereditary risk for more other cancers. But since these aggressive tumours may
rapidly disseminate become extra-ocular, removal of the eye of these young children with unilateral
non-familial RB1+/+ MYCNA retinoblastoma is important/ - or recommended.
Funding NCI-NIH; CIHR; Canadian Retinoblastoma Society; Hyland Foundation; Ontario
Ministry of Health and Long Term Care; Toronto Netralya and Doctors Lions Clubs; and
Foundations Avanti-STR and KiKa.
5
1
2
3
4
5
6
7
8
9
10
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Word count 2998/3000
Background
Retinoblastoma set the paradigm for tumour suppressor genes, with Knudson’s classic hypothesis
predicting that two rate-limiting hits initiate this childhood eye cancer.1 The two hits were later
attributed to the retinoblastoma gene (RB1).2 Approximately 40% of retinoblastoma is bilateral.
AThe 40% of children who have with bilateral retinoblastoma carries carry a heterozygous,
constitutional RB1 mutation and is are predisposed to retinoblastoma; one additional hit, in
whichdamages the second RB1 allele is damaged and initiates retinoblastoma, and/or other cancers
later in life. In 70% of tumours, the second hit involves loss of the normal allele with duplication of
the mutant allele (loss of heterozygosity, LOH). Approximately The 60% of children have with
unilateral non-familial retinoblastoma usually have. Most non-familial unilateral retinoblastomas
arise when both RB1 alleles are damaged only in developing retina, but 15% of them carry a
heritable, constitutional RB1 mutation. Accepted dogma is that damage todisruption of both RB1
alleles (RB1-/-) is required forinitiates all retinoblastoma development.2-4
The heterozygous mutant RB1 allele is identifiable in blood of 95% of bilaterally affected
persons. RB1 mutations or promoter methylation, are detected on both alleles (RB1-/-) in 95% of
unilateral retinoblastomas.5, 6 The possibility that some unilateral retinoblastomas with no detectable
RB1 mutations occur by an independent mechanism has not been previously explored.
We report the first identification of unilateral retinoblastomas with normal RB1 alleles and high-
level MYCN gene amplification (MYCNA). These unilateral RB1+/+MYCNA retinoblastomas are
characterized by distinct histology, few of the genomic copy-number changes characteristic of
retinoblastoma, and very early age of diagnosis. This newly recognized sub-type of retinoblastoma
has immediate diagnostic, genetic counselling, and therapeutic implications.
6
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Methods (details in webappendix)
Clinical samples
Tissues and clinical data were provided for identification of RB1 mutations for in clinical care of
children and their families. Research Ethics Board approvals for theto use of de-identified data and
tissues after clinical analyses, are on file at each participating site. Although not required, the
Toronto and Essen patients provided additional informed consent.
Mutation analyses
Standard of care analyses that identify 95% of RB1 (Gen bank accession #L11910) mutant alleles5-7
were applied at each site, including DNA sequencing, quantitative multiplex PCR (QM-PCR) or
Multiplex Ligation-Dependent Probe Amplification (MLPA), and RB1 promoter methylation
testing. Intragenic and closely-linked RB1 microsatellite markers were used to determine zygosity
of the RB1+/+ tumours.
Genome copy-number analyses
Genomic copy-numbers of five genes were determined by QM-PCR (Toronto) or MLPA and single
nucleotide polymorphism (SNP) analyses (Amsterdam). Sub-megabase resolution array
comparative genomic hybridization (aCGH) or SNP array were used to assess overall genomic
copy-numbers and normal cell contamination.
Protein expression studies
Paraffin-embedded sections of retinoblastomas and adjacent normal retinas were stained for full-
length pRb protein (antibodies targeting N- and C-terminal pRb) and N-Myc.8 Western blots were
performed on RB1+/+MYCNA, RB1-/- and control cell lines. Function of pRB was assessed by
7
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
immunostainingblotting with phosphor- specific for phosphorylated pRb antibody and co-
immunoprecipitation of pRb and its major effector molecule, E2F1.
RNA gene expression studies
Expression of RB1, MYCN, and genes reflecting the proliferative status and retinal derivation of
tumours were assessed using End-Point Reverse Transcriptase PCR (RT-PCR) and/or Quantitative
Real-Time PCR.
Age of Diagnosis analysis
Ages at diagnosis vs. proportion not yet diagnosed were analysed to assess minimal number of
events for tumour initiation. Likelihood of children having RB1+/+MYCNA tumours at different ages
of diagnosis was estimated.
Role of the funding sources
The sponsors of the study had no roles in study design, data collection, data analysis, data or
interpretation, or writing of the manuscript. No author was paid to write this article. All authors had
full access to all data in the study; the corresponding author (BLG) had final responsibility for the
decision to submit for publication. NCI-NIH grant 5R01CA118830-05 supported the early
discovery at the Canadian site. Canadian Institutes for Health Research grants (MOP-86731, MOP-
77903, MOP-110949) supported the aCGH studies, and MOP-77710 supported the pRb
phosphorylation studies. The Canadian Retinoblastoma Society, Hyland Foundation and Toronto
Netralya and Doctors Lions Clubs provided critical funding for additional experiments. The Ontario
Ministry of Health and Long Term Care provided infrastructure. Solutions by Sequence supported
the overall project, data analysis and manuscript preparation. The Dutch study was made possible
by Avanti-STR and KiKa, while VUmc provided infrastructure.
8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Results
The Toronto lab identifies both tumour mutations (or promoter methylation) in >95% of tumours
(as of October 10, 2012, 616/642, 96.0%) from unilateral probands with no known family history.
In 3-4% of tumours an only one RB1 mutation is identified on only one allele, and in 1·6% of
tumours, no RB1 mutation is found. Low-level mosaicism may account for the 5% "no mutation
found" blood samples from bilaterally affected patients.5 However, undetected RB1 mutations in
fully tested tumours (RB1+/+) were are believed to include translocations, deep intronic mutations,
or mutations in unknown RB1 regulatory regions. Unlike most cancers, normal cell contamination
of retinoblastoma tumours is rare, and detectable by RB1 sequence, sensitive allele-specific PCR,
and microsatellite analysis (EZ – polish sentence – undetectable/detactable…). Rarely, significant
normal cell contamination is associated with very small tumours.
In our clinical work in Toronto, we (DER, BLG) found seven unilateral retinoblastoma tumours
with no RB1 mutations, no promoter methylation, and no LOH at RB1 LOH (RB1+/+). We used QM-
PCR to measure in RB1+/+ tumours, copy-numbers of genes at 6p, 1q, 16q and 2p that are
characteristically gained or lost in RB1-/- retinoblastoma tumours.9 To our surprise, 5/7 RB1+/+
tumours showed dramatic MYCN oncogene amplification (MYCNA). To validate this observation,
we collaborated with RB1 clinical laboratories in Germany, France, and New Zealand, to study
similar RB1+/+ tumours in which they had found no detectable RB1 mutations or promoter
methylation, no RB1 LOH, and low normal cell contamination (tTable 1). After analysis was
completeinitial studies, we learned that the Amsterdam lab group had independently discovered
three RB1+/+MYCNA tumours. We report our combined data after statistical analysis showed that
frequencies were not different (p = 0·.08) among the five clinical labs. Retinoblastoma T101,
predicted clinically and pathologically to be an RB1+/+MYCNA tumour (figure 4C), was later shown
to havehad 40 copies of MYCN, and is included in some analyses.
9
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Since our standard detection sensitivity >95% to find an the RB1 mutation in blood of bilaterally
affected persons is >95%,5-7 the probability of finding no RB1 mutations in a tumour with no RB1
LOH is equivalent to the probability of missing two independent RB1 mutations in one sample
(0·05 x 0·05) or 0·25%. By combining our data on 1068 unilateral non-familial tumours, we
identified 29 RB1+/+tumours (2·7%), 10-fold more than expected (p = 6 x 10 -45) (ttable 1). This
suggested that some RB1+/+ tumours might originate by a mechanism other than Knudson’s two RB1
mutations.
To characterize copy numbers of genes commonly gained or lost in retinoblastomas,9 we used
QM-PCR (Toronto) or MLPA/SNP (Amsterdam) analyses. MYCN copy-number was elevated in
27/30 (90%) RB1+/+ vs. 60/93 (65%) RB1-/- retinoblastomas (p = 3·4 x 10-4) (table S65). No RB1-/-
primary tumours showed over >10 MYCN copies. In comparison, 16/30 (53%) RB1+/+ tumours
showed high-level MYCN amplification (28 to 121 copies of MYCN), (RB1+/+ MYCNA
retinoblastoma) MYCN copy-number in RB1+/+ tumours showed bimodal distribution, with 16/30
(53%) RB1+/+ tumours showing high-level MYCN amplification (28 to 121 copies of MYCN),
henceforth referred to as RB1+/+MYCNA retinoblastoma ((p <0·0001) (tables 1, S65, figure 1). The
remaining 14 RB1+/+ tumours showed between 2 and 10 MYCN copies. For ten children with
RB1+/+MYCNA tumours, normal DNA from available blood showed two MYCN copies.
Compared to RB1-/- retinoblastoma, the 16 RB1+/+MYCNA tumours showed lower frequency of
copy-number changes in four other genes characteristic of retinoblastoma: gain of oncogenes KIF14
(62% vs. 19%; p = 0·002), DEK and E2F3 (57% vs. 6%; p = 0·0002) and loss of tumour suppressor
gene CDH11 (56% vs. 13%; p = 0·002) (tables S65, S76). SNP analysis indicated that the level of
normal cell contamination in the RB1+/+ MYCNA tumours was low and similar to RB1-/- tumours
(table S2).
10
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
To probe for other genomic gains/losses, we studied DNA from 48 unilateral retinoblastomas by
aCGH10 and 3 by SNP analysis (Amsterdam) (14 RB1+/+MYCNA, 12 RB1+/+, 25 RB1-/-(+)) (tables S65,
S87, figure 2A). None of the RB1+/+MYCNA retinoblastomas showed any evidence of copy-number
changes or translocations11 at the RB1 locus. Except for MYCN copy-number, aCGH (figure 2A)
confirmed a reduced frequency in RB1+/+MYCNA retinoblastomas of the specific genomic copy-
number changes characteristic of RB1-/- retinoblastomas (figure 1, table S76).9 The RB1+/+MYCNA
retinoblastomas also showed overall significantly fewer altered bp and aCGH clones than the RB1-/-
retinoblastomas (p = 0·033) (figure 2C, DS3, table S87).
MYCN amplicons of 14 (11 by aCGH, 3 by SNP analysis) RB1+/+MYCNA retinoblastomas, as
well as one RB1+/- tumour (T33) with 73 copies of MYCN, and one RB1-/- primary tumour (RB381)
with 9·2 copies of MYCN, were narrow, spanning 1·1 to 6·3 Mbp encompassing MYCN (figures
2BS4, 5, S2, table S87). Importantly, the sole copy-number change detected in one RB1+/+MYCNA
retinoblastoma (E4) was 48 copies of 2p24·.2-24·.3 encompassing MYCN. The minimal common
amplicon defined by RB1+/+(-) MYCNA retinoblastomas T33 and P2 spanned 513 kbp containing only
MYCN. RB1+/+MYCNA retinoblastomas T5 and P2 also defined a minimal common amplicon
including only MYCN. Of the remaining 35 tumours tested by aCGH, 24 unilateral tumours showed
no gain or loss at MYCN, and 11 had moderate gain spanning a broad region of at least 28 Mbp of
chromosome 2p, too large to meet the definition of amplification.12
Three (23%) RB1+/+MYCNA tumours showed 17q21·.3-qter or 17q24·.3-qter gain; two
RB1+/+MYCNA tumours showed 11q loss. Both regions are commonly altered in neuroblastoma,13, 14
but rare in RB1-/- retinoblastoma (present and published data15, 16). Other changes in RB1+/+MYCNA
retinoblastomas not often seen in RB1-/- retinoblastoma were gains at 14q and 18q, and losses at 11p
(figures 2A, S53).
Retinoblastoma T33 (RB1+/-) showed high-level MYCN amplification and loss of one copy of
most of 13q, including RB1; we suspect that amplification of MYCN initiated proliferation, followed
11
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
by 13q deletion. Since T33 showed numerous characteristics of RB1+/+MYCNA tumours, we included
T33 with MYCNA retinoblastomas in many analyses (figures 2A & B, 4A, S2, S4S4, S6).
RB1+/+MYCNA retinoblastomas expressed functional RB1 protein (pRB). pRB. Primary pRB. The
RB1+/+MYCNA retinoblastomas and the expected usual normal retinal cell types expressing pRb
(ganglion cells, specific inner nuclear cells, and cone photoreceptors)17 stained showed nuclear
staining for C-terminal (figure 3A, S6) and N-terminal (data not shown) epitopes of the RB1 protein
(pRB), pRB, while RB1-/- and RB1+/- tumours were negative or stained weakly for pRB, depending
on their RB1 mutations. Western blots and immunoprecipitation on of in 3 RB1+/+ MYCNA cell lines
(A3, and RB522, T101) derived from RB1+/+MYCNA primary retinoblastomas, showed demonstrated
functional nuclear pRb that was both phosphorylated and un-phosphorylated, full-length (figures
S6, S7), normally hypo- and hyperphosphorylated (figure 3B, S7), and bound to pRBE2F1 (figure
3C), the major target of pRb in control of the cell cycle.18{Sahin, 2010 #16727;Templeton, 1991
#18635} {Buchkovich, 1989 #17829} (figures 3B, S5).Full-length 2·8 kbp RB1 transcripts were
detected by end-point and real-time RT-PCR in the 3 Three RB1+/+MYCNA primary retinoblastomas
for which mRNA was available, end-point and real-time RT-PCR showed full-length 2·8 kbp RB1
transcripts at levels comparable to fetal retina (figures 3C, 3E3D, S8,, table S98). In contrast, most
RB1-/- retinoblastomas expressed low RB1 transcript levels.
RB1+/+MYCNA primary retinoblastomas (figure 3A) and three derived cell lines (data not shown)
stained strongly for N-Myc protein. RB1+/+MYCNA retinoblastomas showed increased N-Myc
protein and transcripts compared to RB1-/- primary retinoblastomas and fetal retina (figure 3B, C, E,
table S8). MYCN and MKI67 transcripts (indicative of proliferation) were at low levels in adult
retina, at and were detectedhigher levels in fetal retina and RB1-/- retinoblastomas without MYCN
amplification, and at very high levels in primary RB1+/+MYCNA retinoblastoma and RB1-/- cell lines
with MYCN amplification (figure 3AC, E, F, table S8). , and RB1-/- retinoblastomas. RB1+/+MYCNA
tumours showed reduced expression of the oncogene KIF14,9 in contrast to normal fetal retina, and
12
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
RB1-/- primary retinoblastomas and cell lines with which over-express high KIF14 expression (figure
3FE, table S8).
RB1+/+MYCNA tumours expressed embryonic retinal cell markers consistent with a retinal origin.
The mRNAs of cone cell marker X-arrestin19 and CRX, a marker of retinal and pineal lineage
tumours, (strongly expressed in RB1-/- retinoblastoma but not in neuroblastoma,20) were expressed in
fetal retina, human adult retina, and three RB1+/+MYCNA, and four RB1-/- primary retinoblastomas
(figure 3DS6).
Children with RB1+/+MYCNA retinoblastomas were much younger at diagnosis than children with
unilateral RB1-/- retinoblastomas. The median age at diagnosis of 17 children (with 16 RB1+/+MYCNA
and one RB1+/-MYCNA [(T33]) tumours) was 4·5 months, significantly younger than children with
unilateral sporadic RB1-/- (24·0 months, p<10-4) or RB1+/+ (21·5 months, p<10-4) retinoblastomas
(figure 4A, tables S4A & S5). We estimate that 18·.4% of children diagnosed with non-familial,
unilateral retinoblastoma at age six months or younger will have RB1+/+MYCNA retinoblastoma
(table S4B).
Analysis of age at diagnosis vs. proportion not yet diagnosed led Knudson to propose that two
hits initiate retinoblastoma.1 Our data from 79 unilaterally affected RB1-/- patients was consistent
with Knudson’s model and fit a two-hit curve, representative of two independent mutational events
in a tumour suppressor gene (figure 4B). Similar analysis for RB1+/+MYCNA tumours was
inconclusive: while the data points for twelve children diagnosed before 10 months approximated
the calculated one-hit curve, ages for the older children deviated (figure 4B).
On histological examination, RB1+/+MYCNA tumours were distinctive, displaying undifferentiated
cells with large, prominent, multiple nucleoli, and necrosis, apoptosis, and little calcification,
similar to other MYCNA embryonic tumours, such as neuroblastoma21 (figures 4C, S45). The
Flexner-Wintersteiner rosettes22 and nuclear molding characteristic of prototypic RB1-/-
retinoblastoma (figure 4D) were absent.
13
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Clinically, the RB1+/+MYCNA retinoblastomas were large and invasive, considering the very
young age of these children (figure 4C, E, and S76). Three RB1+/+MYCNA retinoblastomas (RB522,
T101, and A3) from enucleated eyes grew rapidly into cell lines, unlike the RB1-/- retinoblastomas
that grow poorly in tissue culture. One RB1+/+MYCNA retinoblastoma had already invaded the optic
nerve past the cribriform plate at age 11 months, a feature of aggressive disease23, 24 (figure 4E).
However, all the children in this study with RB1+/+MYCNA tumours were cured by removal of their
affected eye with no adverse outcomes; none developed retinoblastomas in their other eye.
14
1
2
3
4
5
6
7
8
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Discussion
Knudson’s analysis of retinoblastoma was fundamental to the concept that normal genes suppress
cancer,1 leading to the identification of the RB1 tumour suppressor gene,2 widely assumed to initiate
all retinoblastoma. We now describe a previously unrecognized type of retinoblastoma with no
evidence of RB1 mutations and with afunctional pRb, that likely initiated by amplification of the
MYCN oncogene.
Our international study of 1,068 unilateral retinoblastomas revealed a distinct RB1+/+MYCNA
subtype comprising 1·4% of unilateral, non-familial retinoblastomas. At least two previously
described tumours may also be RB1+/+MYCNA retinoblastomas,25 although RB1 genetic status was
not defined. Despite the low incidence of RB1+/+MYCNA retinoblastoma, RB1+/+MYCNA tumours
were independently discovered and characterized in the Toronto and Amsterdam labs, with different
patient cohorts and technologies.
We also identified 14 RB1+/+ retinoblastomas without MYCN amplification. aCGH showed
genomic gains and losses distinct from either RB1-/- or RB1+/+MYCNA retinoblastomas, but we
lacked RB1+/+ tumours for gene expression or protein studies. In particular, these RB1+/+
retinoblastomas showed increased frequent gain of 19p and q, 17p and q, 2p, and the telomeric end
of 9q. The RB1+/+ group appears heterogeneous and merits further study.
Our paper highlights how molecular diagnostics can identify novel malignancies that elude
histopathological recognition. Although RB1+/+MYCNA retinoblastomas were not previously
recognized as distinct, they resemble large nucleolar neuroblastomas with MYCN-amplification and
poor outcome.21 Like MYCN-amplified neuroblastomas,14 RB1+/+MYCNA tumours showed less
complex genomic copy-number alterations than tumours without MYCN amplification, suggesting
that MYCNA may be the critical driver of malignancy. However, the early age of diagnosis of
RB1+/+MYCNA tumours may allow less time to accumulate genomic alterations. Although whole
15
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
genome sequencing recently suggested that point mutations (other than in the RB1 gene) are few in
RB1-/- retinoblastoma,26 loss of RB1 induces mitotic changes and lagging chromosomes,27 leading to
genomic instability. RB1 loss in retinoblastoma and pre-malignant retinoma8, 16 is also associated
with specific common DNA copy-number changes, that are less frequent in RB1+/+MYCNA
retinoblastomas.
Are RB1+/+MYCNA tumours truly retinoblastoma? RB1+/+MYCNA retinoblastomas arise in
developing retina and express markers of embryonic retina, meeting the definition of retinoblastoma
as ‘a blast cell tumour arising from the retina’.19, 20, 28 We demonstrate intact RB1 genes in primary
RB1+/+ MYCNA retinoblastomas, with and pRB with strong nuclear N- and C-terminal pRb antibody
staining, in primary RB1+/+MYCNA retinoblastomas. The Three cell lines derived from
RB1+/+MYCNA tumours expressed full length RB1 mRNA, and inactive hyperphosphorylated and
active un-hypophosphorylated pRb that binds to E2F1, indicating normal functional pRb18. The
possibility of normal cell contamination masking an RB1 mutations in the RB1+/+ MYCNA tumours is
ruled out by the cell lines that maintain the RB1+/+ genotype of the primary tumour, the very high
level MYCN amplification, histology with the only normal cells being vascular, and strong
histological and biochemical evidence that pRb is functional.
The very early presentation of RB1+/+MYCNA retinoblastomas, lack of RB1 mutations, functional
pRb and high MYCN-amplification in a relatively copy-number stable genome, suggests that these
retinoblastomas arise by somatic MYCN oncogene amplification in a retinal progenitor cell. This is
supported by one RB1+/+MYCNA tumour in which MYCN amplification was the only identified
genomic copy-number change. The RB1+/+MYCNA retinoblastomas are already large in very young
children, so they likely initiate earlier in retinal development than RB1-/- retinoblastomas. In
children with an RB1 mutation of equivalent age, the tumours are usually much smaller (figure 4F).
How MYCN-amplification is initiated, and whether MYCN-amplification alone suffices to initiate
these retinoblastomas remains to be formally demonstrated.
16
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
In retinal development the N-myc protein can support cell division in the absence of activating
E2fs; presumably unregulated MYCN expression associated with high- level gene amplification in
RB1+/+MYCNA tumours promotes cell division indirectly through inactivation of pRb.29 The relative
genomic stability of RB1+/+MYCNA retinoblastomas suggests that anti-N-Myc therapeutic agents30
may avoid emergence of drug resistance acquired through progressive genomic rearrangements in
RB1-/- retinoblastoma. Our preliminary MYCN knock down experiments resulted in rapid death of
RB1+/+MYCNA cell lines. We predict that children less than one year of age with extra-ocular
retinoblastoma around the world23, may have RB1+/+MYCNA retinoblastomas that might benefit from
anti-N-Myc therapy. This idea is consistent with our (BG) anecdotal observation during review of
retinoblastoma pathology slides in Kenya, of RB1+/+MYCNA histology in an orbital recurrence,
confirmed to have 40 copies of MYCN.
Young age at diagnosis of unilateral retinoblastoma is frequently interpreted as an indication of
heritable retinoblastoma, and is currently cited as a reason to try to cure the cancer without
enucleation. However, attempts to salvage an eye with a large RB1+/+MYCNA retinoblastoma could
be dangerous; in our study, prompt removal of the unilateral affected eyes was curative with no
adverse outcomes. The young patients with RB1+/+MYCNA tumours in our study had large,
aggressive tumours, including invasion into the optic nerve. At similar young ages, the usual
hereditary RB1-/- tumours are very much smaller, detected only by active surveillance (figure 4F).
While our data predict that 18·.4% of children diagnosed with non-familial, unilateral
retinoblastoma at age six months or younger will have RB1+/+MYCNA retinoblastoma (table S4B), if
size of tumour size is also considered, the risk will be higher, and RB1+/+MYCNA tumours may be
clinically predictable, facilitating prompt removal of these eyes with good outcomes for the
children.
Standard care for unilateral non-familial retinoblastoma is identification of the RB1 mutant
alleles in tumour, and examination of blood to determine whether either mutant allele is germline.
17
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Our study suggests that when no RB1 mutation is detected in a retinoblastoma tumour, determining
MYCN copy-number may assist on-going care. The diagnosis of RB1+/+MYCNA retinoblastoma
strongly suggests non-hereditary disease with normal population risks for retinoblastomas in the
other eye and other cancers later inthroughout life.
Our findings challenge the long-standing dogma that all retinoblastomas are initiated by RB1
gene mutations (figure 4G). RB1+/+MYCNA retinoblastoma provides an intriguing contrast to
classical retinoblastoma, of immediate importance to patients.
Panel: Research in Context
Systematic Review
We systematically searched the published, peer-reviewed literature on PubMed
(http://www.ncbi.nlm.nih.gov/pubmed/) using the search terms “retinoblastoma”, “initiation” and
“genetics”; “retinoblastoma tumour genetics”; “retinoblastoma development”; and “retinoblastoma
initiation”. We reviewed publications with a main focus on genetic initiation and development of
human retinoblastoma. We found no data that challenged Knudson’s 1971 conclusion that two rate-
limiting events1, later shown to be loss of both RB1 gene alleles,2-5, 7 are essential but not necessarily
sufficient for development of retinoblastoma. We found no data to suggest another form of
retinoblastoma.
Interpretation
Our collaborative studies identify a previously unrecognized disease: retinoblastoma with normal
RB1 genes, apparently driven by MYCN oncogene amplification. This newly recognised form of
retinoblastoma has immediate clinical implications for patients. RB1+/+MYCNA retinoblastomas are
diagnosed by molecular study of the tumour after removal of the eye of very young children with
unilateral non-familial disease. The children with RB1+/+MYCNA retinoblastoma and their relatives
are predicted to be at normal population risk for other cancers. Attempts to salvage the eye on the
18
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
assumption of heritable disease in young children may incur high treatment morbidity and failure to
cure these aggressive oncogene-driven retinoblastomas.
19
1
2
3
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Tables
Test Site Total RB1+/+ Proportion RB1+/+Number
RB1+/+MYCNA
Proportion
RB1+/+MYCNA
Canada 441 7 1·6% 5 1·1%
Germany 400 12 3·0% 4 1·0%
France 150 5 3·3% 2 1·3%
New Zealand 30 2 6·7% 1 3·3%
The Netherlands 47 3 6·4% 3 6·4%
Total 1068 29 2·7% 15# 1·4%
*Fisher's exact tests indicates that percentage of unilateral tumours with RB1+/+ is not related to site (p = 0·08).
Table 1. Frequency of RB1+/+ unilateral retinoblastoma at five international diagnostic labs
20
1
2
3
4
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Figures
Figure 1: Copy numbers of genes commonly altered in RB1-/-
retinoblastoma
(A) Box-plot of genomic copy-numbers determined by QM-PCR for indicated
genes in unilateral retinoblastoma, grouped by RB1 mutation status. T33 is an
outlier of the RB1+/- group, showing an RB1+/+ MYCNA -like profile. Gray line, 2-
copies. (B) Heat-map for copy number for the profile genes (red, increased, and
blue, decreased copy-number; gray, 2 copies; white, not tested; n, number in
each group). Recently discovered RB1+/+ MYCNA tumours not in (A) are indicated
by black triangles.
Figure 1: The 5-gene copy number signature of retinoblastoma
(A) Box-plot of genomic copy-numbers determined by QM-PCR for indicated genes in unilateral
retinoblastoma, categorized by RB1 mutation status. On each boxplot, vertical line marks the
maximum and minimum copy-numbers observed while the box bounds second and third quartiles,
and horizontal line within the box represents the median (11 RB1+/+MYCNA tumours, median 54
MYCN copies; 14 RB1+/+tumours, median 3·1 MYCN copies). T33 is an outlier of the RB1+/-group,
showing an RB1+/+MYCNA-like profile; gray line, 2-copies. (B) Heat-map for copy-number by QM-
PCR for the profile genes (red, increased, and blue, decreased copy-number; gray, 2 copies; white,
not tested; n, number in each group) including more recently discovered RB1+/+MYCNA tumours
(black triangles) not included in (A). (C) MYCN copy-number assessment in 30 RB1-/- and 3 RB1+/+
tumours studied by MLPA, showing high MYCN copy-number in the RB1+/+ tumours.
21
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
22
1
2
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Figure 2: Genomic copy number alterations
(A) aCGH on 12 RB1+/+ MYCNA (including T33) (red), 12 RB1+/+ (blue), 13 RB1+/-
(brown), and 11 RB1-/- (green) tumours; gains, right, and losses, left of
chromosome; minimal commonly gained/lost regions in RB1-/- tumours boxed;
*normally occurring copy number variations. Tumour T33 shows loss of most of
13q. (B) The minimal amplicon of 513 kbp is defined by two MYCNA tumours
(pink band); MYCN copy number by QM-PCR, red italics; aCGH individual
probes, green bars.
Figure 2: Fewer genomic copy-number alterations in RB1+/+MYCNA than
RB1-/-tumours
(A) aCGH on 12 RB1+/+MYCNA (including T33), 12 RB1+/+, 13 RB1+/-, and 11 RB1-/-tumours; gains,
right; losses, left; minimal commonly gained/lost regions in RB1-/- tumours boxed; *normally
occurring copy-number variations. The RB1+/- MYCNA tumour T33 shows loss of most of 13q; this
may not be an initiating event. (B) The minimal amplicon of 513 kbp is defined by two MYCNA
tumours (pink band); MYCNcopy-number by QM-PCR, red italics; aCGH individual probes, green
bars. (C) Boxplot of bp altered shows fewer changes in RB1+/+MYCNA than RB1-/-tumours (p =
0·033; t-test with Welch’s adjustment); vertical line marks the maximum and minimum copy-
numbers observed, box bounds first and third quartiles, and horizontal line within the box represents
the median. (D) Fewer aCGH clones are altered in RB1+/+ and RB1+/+MYCNA, than RB1-/-tumours;
each class has more unique clones altered than in common.
23
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
24
1
2
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Figure 3: Expression of Rb1 and MYCN
(A) Staining of adjacent retina and RB1+/+ MYCNA or RB1 mutant retinoblastoma
for N-Myc protein and pRB (C-terminus antibody); T, tumour; INL, inner nuclear
layer retina. (B) Western blots with pRb antibody that recognizes both hypo-
and hyperphosphorylated pRb, phospho-Rb (Ser795) antibody, and E2F1
antibody. (C) Cell lysates were immunoprecipitated with antibodies to mouse
IgG (negative control), pRb or E2F1, and western blots performed with
antibodies to pRb and E2F. (D) Real-time RT-PCR for RB1, MYCN and KIF14 in
human fetal (FR) and adult (HR) retina, RB1+/+ MYCNA , and RB1-/- primary
tumours and cell lines; triplicate measurements normalized against GAPDH,
relative to FR; MYCN DNA copy-numbers in italics; #, KIF14 not done.
Figure 3: RB1+/+MYCNA tumours express pRb and MYCN
25
1
2
3
4
5
6
7
8
9
10
11
12
13
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
(A) h
26
1
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
27
12
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Figure 4: Clinical features of children with RB1+/+MYCNA tumours in very
young children are clinically distinct
(A) Children with RB1+/+MYCNARB1+/+ MYCNA retinoblastoma are
diagnosed significantly younger than children with RB1-/-RB1-/- tumours
(p<0·0001, Wilcoxon rank sum test). (B) The Knudson plot of
proportion not yet diagnosed vs age at diagnosis, using birth as a
surrogate for initiation, fits showing a two-hit curve (blue) but notand a
one-hit curve (red) for children with unilateral non-familial RB1-/-RB1-/-
diseasetumours, as expected; or for RB1+/+MYCNA RB1+/+ MYCNA
retinoblastomatumours, the data points for the 12 children younger than 10
months most closely approximate the one-hit curve (red), but those diagnosed
at older ages deviate toward the two-hit curve; scatterplot does not
distinguish identically aged children. (C) Fundus image of an a large
RB1+/+ MYCNA RB1+/+MYCNA unilateral tumour, extending from optic
nerve (white arrow) to anterior border of retina (double arrows) in a 4
month-old child with characteristic calcification on ultrasound, and
round nuclei with prominent large multiple nucleoli on pathology, in
comparison to (D) RB1-/-RB1-/- tumour showing classic Flexner-
Wintersteiner rosettes and nuclear molding; hematoxylin-eosin
staining. (E) RB1+/+ MYCNA RB1+/+MYCNA retinoblastoma in an 11 month-
old child (A2) with extra-ocular extension into the optic nerve (arrows)
(2·5x, hematoxylin-eosin staining). (F) In comparison, in 3·5 month-old
child with heritable RB1-/-RB1-/- retinoblastoma, a small tumour a small
tumour is present in the inner nuclear layer of the retinashown to be in the
inner nuclear layer of the retina on optical coherent tomography (OCT).
28
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
(G) Schema of data establishing RB1+/+MYCNA RB1+/+ MYCNA
retinoblastoma as a novel disease; months (m), data figures (f) and
tables (t) indicated in grey on left.
29
1
2
3
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
30
1
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
Contributors
Diane E Rushlow recognized the initial connection between MYCNA amplification and RB1
mutation status, performed literature search and QM-PCR analysis and supervised RB1 mutation
analysis, coordinated collaborations with the other sites and was the major contributor to manuscript
preparation. Berber M Mol recognized the RB1 and MYCN mutation status of the Amsterdam
samples by MLPA and SNP array analysis, experiments and performed SNP array data
analysis,performed immunohistochemistry, imaging and Western blots and co-
immunoprecipitations. Jennifer Y Kennett performed aCGH and analysed aCGH data. Berber M
Mol determined MYCN status by MLPA experiments and performed SNP array data analysis,
immunohistochemistry imaging and Western blots. Stephanie Yee performed analysis of aCGH
data, the MYCNA alignment, immunohistochemistry and reverse transcriptase PCR. Sanja Pajovic
performed literature search, reverse transcriptase PCR and immunohistochemistry. Brigitte L
Thériault performed literature search and RNA expression studies. Nadia L Prigoda-Lee performed
literature search, statistical analysis and contributed to figure and manuscript preparation. Clarellen
Spencer performed immunohistochemistry. Helen Dimaras and Timothy W Corson performed
literature searches, assisted in data analysis and conceptualization of discussion, and contributed to
figure and manuscript preparation. Renee Pang performed statistical and bioinformatic analyses on
the aCGH data. Christine Massey performed statistical analyseis on age of diagnosis data. Roseline
Godbout discovered and characterised the first RB1+/+MYCNA tumortumour with MYCN
amplification and normal pRb (RB522) (now RB1+/+ MYCNA ) long before anyone believed her; she
provided the cell line and data and the Western blot showing functional proteinpRb. Zhe Jiang and
Eldad Zacksenhaus performed Western blots to demonstrate functional phosphorylated pRb.
Katherine Paton and Annette C Moll provided clinical images and material, and conceptual
discussion. Claude Houdayer and Anthony Raizis provided RB1 mutation analysis, and clinical
31
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
features. William Halliday recognized and characterized the unique histological features of the
RB1+/+MYCNA retinoblastomas and prepared digital images for publication. Wan L Lam supervised
Jennifer Kennet and the aCGH experiments. Paul C Boutros performed detailed and novel analysis
of the aCGH data, and statistical analyses throughout the project, and supervised Renee Pang.
Dietmar Lohmann performed literature search, provided RB1 mutation analysis, and contributed to
figure construction and development of concepts. Josephine C Dorsman coordinated the Amsterdam
study, and, with Berber Mol who she supervised, recognized the RB1 and MYCN mutation status of
the Amsterdam samples, and supervised Berber Mol. Brenda L Gallie supervised overall, performed
literature search, provided critical guidance on all components of the project, and contributed
extensively to figure and manuscript preparation. All authors contributed to manuscript preparation.
Conflicts of interest
BLG is was part-owner of Solutions by Sequence. All other authors declare that they have no
conflicts of interest.
Acknowledgments
This study was conducted with the support of the Ontario Institute for Cancer Research to PCB
through funding provided by the Government of Ontario. SY was funded by the Vision Science
Research Program of the University Health Network and the University of Toronto. RP was funded
in part by a Great West Life Studentship from Queen’s University School of Medicine. BMM was
funded by a grant from CCA/V-ICI/ Avanti-STR (to JCD, J. Cloos and ACM), the Dutch research
was also funded in part by KIKA (JCD, H. te Riele, J. Cloos, ACM). We thank Leslie MacKeen for
the montage of RetCam images in figure 3B. We thank Dr. Valerie White of U. British Columbia
for providing clinical and pathological details and images. We thank Cynthia Vandenhoven for the
clinical images in figure 4F. We thank members of the VU University Medical Center/The
Netherlands Cancer Institute, Institut Curie, Toronto retinoblastoma teams and other, wise
32
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
colleagues for useful discussions. We thank the children and families who donated tissues for these
studies for the benefit of future families.
33
1
23
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
REFERENCES
1. Knudson AG. Mutation and cancer: statistical study of retinoblastoma. Proceedings of the National Academy of Science, USA 1971; 68(4): 820-3.
2. Friend SH, Bernards R, Rogelj S, Weinberg RA, Rapaport JM, Albert DM, et al. A human DNA segment with properties of the gene that predisposes to retinoblastoma and osteosarcoma. Nature 1986; 323(6089): 643-6.
3. Cavenee WK, Hansen MF, Nordenskjold M, Kock E, Maumenee I, Squire JA, et al. Genetic origin of mutations predisposing to retinoblastoma. Science 1985; 228(4698): 501-3.
4. Lohmann DR, Gallie BL. Retinoblastoma: Revisiting the model prototype of inherited cancer. Am J Med Genet 2004; 129C(1): 23-8.
5. Rushlow D, Piovesan B, Zhang K, Prigoda-Lee NL, Marchong MN, Clark RD, et al. Detection of mosaic RB1 mutations in families with retinoblastoma. Hum Mutat 2009; 30(5): 842-51.
6. Lohmann D, Gallie B, Dommering C, Gauthier-Villars M. Clinical utility gene card for: Retinoblastoma. Eur J Hum Genet 2011; 19(3).
7. Houdayer C, Gauthier-Villars M, Lauge A, Pages-Berhouet S, Dehainault C, Caux-Moncoutier V, et al. Comprehensive screening for constitutional RB1 mutations by DHPLC and QMPSF. Hum Mutat 2004; 23(2): 193-202.
8. Dimaras H, Khetan V, Halliday W, Orlic M, Prigoda NL, Piovesan B, et al. Loss of RB1 induces non-proliferative retinoma: increasing genomic instability correlates with progression to retinoblastoma. Human molecular genetics 2008; 17(10): 1363-72.
9. Corson TW, Gallie BL. One hit, two hits, three hits, more? Genomic changes in the development of retinoblastoma. Genes Chromosomes Cancer 2007; 46(7): 617-34.
10. Ishkanian AS, Malloff CA, Watson SK, DeLeeuw RJ, Chi B, Coe BP, et al. A tiling resolution DNA microarray with complete coverage of the human genome. Nat Genet 2004; 36(3): 299-303.
11. Watson SK, deLeeuw RJ, Horsman DE, Squire JA, Lam WL. Cytogenetically balanced translocations are associated with focal copy number alterations. Hum Genet 2007; 120(6): 795-805.
12. Myllykangas S, Bohling T, Knuutila S. Specificity, selection and significance of gene amplifications in cancer. Semin Cancer Biol 2007; 17(1): 42-55.
13. O'Neill S, Ekstrom L, Lastowska M, Roberts P, Brodeur GM, Kees UR, et al. MYCN amplification and 17q in neuroblastoma: evidence for structural association. Genes Chromosomes Cancer 2001; 30(1): 87-90.
14. Mosse YP, Diskin SJ, Wasserman N, Rinaldi K, Attiyeh EF, Cole K, et al. Neuroblastomas have distinct genomic DNA profiles that predict clinical phenotype and regional gene expression. Genes Chromosomes Cancer 2007; 46(10): 936-49.
15. Chen D, Gallie BL, Squire JA. Minimal regions of chromosomal imbalance in retinoblastoma detected by comparative genomic hybridization. Cancer Genet Cytogenet 2001; 129(1): 57-63.
16. Sampieri K, Amenduni M, Papa FT, Katzaki E, Mencarelli MA, Marozza A, et al. Array comparative genomic hybridization in retinoma and retinoblastoma tissues. Cancer Sci 2009; 100(3): 465-71.
17. Spencer C, Pajovic S, Devlin H, Dinh QD, Corson TW, Gallie BL. Distinct patterns of expression of the RB gene family in mouse and human retina. Gene Expr Patterns 2005; 5(5): 687-94.
18. Templeton DJ, Park SH, Lanier L, Weinberg RA. Nonfunctional mutants of the retinoblastoma protein are characterized by defects in phosphorylation, viral oncoprotein association, and
34
1
23456789
10111213141516171819202122232425262728293031323334353637383940414243444546
MYCN amplified retinoblastoma with normal pRb RETINOBLASTOMA
nuclear tethering. Proceedings of the National Academy of Sciences of the United States of America 1991; 88(8): 3033-7.
19. Murakami A, Yajima T, Sakuma H, McLaren MJ, Inana G. X-arrestin: a new retinal arrestin mapping to the X chromosome. FEBS Lett 1993; 334(2): 203-9.
20. Terry J, Calicchio ML, Rodriguez-Galindo C, Perez-Atayde AR. Immunohistochemical Expression of CRX in Extracranial Malignant Small Round Cell TumorTumours. The American journal of surgical pathology 2012; 36(8): 1165-9.
21. Tornoczky T, Semjen D, Shimada H, Ambros IM. Pathology of peripheral neuroblastic tumortumours: significance of prominent nucleoli in undifferentiated/poorly differentiated neuroblastoma. Pathol Oncol Res 2007; 13(4): 269-75.
22. Flexner S. A peculiar glioma (neuroepithelioma?) of the retina. Johns Hopkins Hosp Bull 1891; 2: 115.
23. Dimaras H, Kimani K, Dimba EA, Gronsdahl P, White A, Chan HS, et al. Retinoblastoma. Lancet 2012; 379(9824): 1436-46.
24. Chantada GL, Casco F, Fandino AC, Galli S, Manzitti J, Scopinaro M, et al. Outcome of patients with retinoblastoma and postlaminar optic nerve invasion. Ophthalmology 2007; 114(11): 2083-9.
25. Lillington DM, Goff LK, Kingston JE, Onadim Z, Price E, Domizio P, et al. High level amplification of N-MYC is not associated with adverse histology or outcome in primary retinoblastoma tumours. Br J Cancer 2002; 87(7): 779-82.
26. Zhang J, Benavente CA, McEvoy J, Flores-Otero J, Ding L, Chen X, et al. A novel retinoblastoma therapy from genomic and epigenetic analyses. Nature 2012; 481(7381): 329-34.
27. Manning AL, Longworth MS, Dyson NJ. Loss of pRB causes centromere dysfunction and chromosomal instability. Genes Dev 2010; 24(13): 1364-76.
28. Kobayashi M, Takezawa S, Hara K, Yu RT, Umesono Y, Agata K, et al. Identification of a photoreceptor cell-specific nuclear receptor. Proceedings of the National Academy of Sciences of the United States of America 1999; 96(9): 4814-9.
29. Chen D, Pacal M, Wenzel P, Knoepfler PS, Leone G, Bremner R. Division and apoptosis of E2f-deficient retinal progenitors. Nature 2009; 462(7275): 925-9.
30. Hook KE, Garza SJ, Lira ME, Ching KA, Lee NV, Cao J, et al. An integrated genomic approach to identify predictive biomarkers of response to the aurora kinase inhibitor PF-03814735. Molecular cancer therapeutics 2012; 11(3): 710-9.
35
123456789
101112131415161718192021222324252627282930313233
34
35