Literature Review Paper

450 Abstracts

tumor cell search in PBL and/or CSF using mutation specific

primers was negative for all seven patients.

Discussion: If the region on chromosome 22 is especially liable

to double stranded DNA breaks, compiling breakpoints in a single

map could reveal that in the future. The negative results for re-

sidual tumor cell search in PBL and/or CSF of the patient may

have two possible reasons: either the detection limit of the method

was not sufficiently low to detect existing residual tumor cells or

the patient’s PBL or CSF did not contain any tumor cells

respectively. Future research objectives can further optimize the

method, enhance the sensitivity and perhaps establish stan-

dardized methods on the ground of these findings and introduce

them to clinical pathology. Supported by the “F€ordergemeinschaft

Kinderkrebszentrum Hamburg e.V.”

Detection of an ATRT Brain Tumor Gene Deletion Diagnostic in Circulating Blood using Next Gen Sequencing

Madhavi Chakravadhanula a,b, Waibhav Tembe b,

Christophe Legendre b , David Carpentieri

a , Winnie S. Liang

b ,

Kimberly J. Bussey b , John Carpten

b , Michael E. Berens

b ,

Ratan D. Bhardwaj a,b

a Barrow Neurologic Institute at Phoenix Children’s Hospital, Phoenix,

Arizona; b Translational Genomics Institute, Phoenix, Arizona

Background: Circulating biomarkers are emerging as novel

diagnostic tools to detect cancer non-invasively. In the cancer

genome, mutations in somatic chromosomes serve as specific

biomarkers that may be detected in the circulation. In this study,

we describe focal deletions found in a patient with Atypical

Teratoid Rhabdoid Tumor (ATRT), a highly aggressive early

childhood pediatric tumor.

Methods: First, we used Magnetic Resonance Imaging (MRI)

and histopathology to study the tumor anatomy; next we used

whole genome sequencing (Next Gen Sequencing) to discover

the presence of mutations in the tumor tissue as well as in the

blood of the patient.

Results and Discussion: Three focal deletions were detected in

the tumor tissue of the patient. Bioinformatics interrogation of

sequencing data indicated that blood DNA showed two of the three

deletions seen in the tumor. Also about 20% of the sequencing

reads from circulating blood DNA matched the reads from tumor

DNA at the chromosome 22 locus containing the SMARCB1 gene.

Circulating, tumor-specific DNA aberrations can be a promising

biomarker (diagnostic and prognostic) for ATRT patients. The

unusually high percentage of tumor DNA detected in the blood

indicates that eithercirculating brain tumor cells lyse in the blood or

that contents of ruptured brain tumor cells traverse what may be a

compromised blood-brain barrier in this patient.

Involvement of SMARCB1 in Inherited Predisposition to Schwannoma, Meningioma and Rhabdoid Tumours

Miriam J. Smith, Kristen D. Hadfield, William G. Newman,

D. Gareth Evans

Genetic Medicine, Manchester Academic Health Sciences Centre

(MAHSC), St Mary’s Hospital, University of Manchester,

Manchester, UK

Meningiomas are benign tumors of the meninges and make

up one third of all primary CNS tumors in adults. Schwannomas

are benign nerve sheath tumors that can occur on nearly all

nerves but are most predominant on the vestibular nerve.

Around 1 in 300 people develop a symptomatic meningioma in

their lifetime and 1 in 500 develop a schwannoma. Rarely both

meningiomas and schwannomas occur at multiple sites, usually

occurring in patients with neurofibromatosis 2 (NF2). NF2 ac-

counts for around 3-5% of all cases of schwannoma and 2-3% of

people with meningiomas. However, not all multiple site cases

are accounted for by NF2 germline mutations. The SWI/SNF

chromatin remodeling complex gene, SMARCB1, is a cause of

schwannomatosis disease. SMARCB1 mutations account for

around w50% of familial cases of schwannomatosis and w10% of sporadic cases. We have identified SMARCB1 germline mu-

tations in 14/31 (45%) families with multiple affected individuals

and 15/123 (12%) sporadic cases. The presumed pathogenic 3’

untranslated region mutation c.*82C>T was found in 6 families.

Many schwannomatosis related SMARCB1 mutations occurred

in extreme ends of the gene and were presumed hypomorphic.

Meningiomas occur rarely in schwannomatosis disease (only in

2/55 individuals with SMARCB1 mutations). Sequencing of

SMARCB1 in 37 cases of multiple cranial meningiomas and 14

cases of spinal meningioma (5 familial) failed to identify any

germline mutations. We will also present data on germline

testing of around 100 cases of malignant rhabdoid tumour.

Expression of Mutant SMARCB1 Proteins in Schwannomas of Schwannomatosis Patients

Theo Hulsebos a , Susan Kenter

a , Volkmar Hans

b ,

Wim Verhagen c, Frank Baas a, Uta Flucke d, Pieter Wesseling d,e

a Department of Genome Analysis, Academic Medical Center,

Amsterdam; b Department of Neuropathology, Evangelisches

Krankenhaus, Bielefeld (Germany); c Department of Neurology and

Clinical Neurophysiology, Canisius Wilhelmina Hospital, Nijmegen; d Department of Pathology, Radboud University Medical Centre,

Nijmegen; e Department of Pathology, Free University Medical

Centre, Amsterdam, Netherlands

Background: Most, but not all, constitutional mutations in the

SMARCB1 gene of schwannomatosis patients are missense

mutations, in-frame deletions or insertions, or splice-site muta-

tions. These are supposed to result in the synthesis of an aberrant

SMARCB1 protein with altered activity. This is in marked contrast

to rhabdoid tumors, in which the two copies of the SMARCB1 gene

are found to be inactivated by mutation and deletion, resulting in

absence of SMARCB1 protein expression in the tumor.

Results and Discussion: Here, we studied the consequences of

typical germline SMARCB1 mutations at the RNA and protein

level in the schwannomas of schwannomatosis patients. We

specifically studied the effects of c.30delC, c.34C>T, and

c.46A>T, which are nonsense codon-generating mutations in

exon 1 of SMARCB1 and predicted to result in the absence of

SMARCB1 protein expression because of nonsense-mediated

mRNA decay. Interestingly, SMARCB1 nonsense mutations have

not been reported to occur in exon 1 of rhabdoid tumors. We could

demonstrate, for c.30delC and c.34C>T, that the respective

mRNAs were present and, therefore, available for translation in

the schwannomas of the patients. We hypothesized that these

mRNAs were prevented from degradation by translation re-initi-

ation at the AUG codon at positions 79-81 (amino acid residue 27).

To test this hypothesis, we introduced by site-directed mutagen-

esis each mutation in a SMARCB1 expression vector and