Development and Clinical Validation of Large Fusion Panel for Pediatric and Adult Cancers

Fengqi Chang, Ph.D.

Children’s Hospital of Philadelphia

Perelman School of Medicine at the University of Pennsylvania

08-08-2016

Rationales

• Gene fusion: one of the most common types of driver mutation in cancer

• Identification of oncogenic gene fusions: important for disease diagnosis, risk stratification and therapeutic decision making.

• FISH and RT-PCR: labor intensive, time consuming, and low throughput.

• Whole transcriptome sequencing (RNA-Seq): cost, sensitivity, efficiency, TAT, starting material; significant computational/bioinformatics resources.

Targeted, semi-quantitative RNA-sequencing methods emerge as an better solution to comprehensively sample known or novel fusions with high accuracy and sensitivity in clinical setting.

• The Fusion Panel includes 106 major fusion partner genes and can

detect 586 different known fusions and many more novel fusions in

both solid and liquid tumors in a single reaction.

• The panel uses Archer Anchored Multiplex PCR (AMP™) technology.

• The panel works on blood, BM, fresh/frozen tissues, and FFPE

tissues.

• CHOP FusionPanel also provides a reference of gene expression.

The CHOP FusionPanel

Nucleotide extractionRNA extraction:

RiboPureRneasy MiniAgencourt FormaPure

Library constructionArcherDX Universal RNA Kit

SequencingMiSeq/NextSeq

2X150bp

Analysis PipelineArcher Analysis Software

Fusion annotation & interpretationJbrowse: visualizationInternal database: common read- throughs

and false positivesDatabases:

Atlas of Genetics and Cytogenetics inOncology and Haematology

Mitelman database of Chromosome Aberrations and Gene Fusions in Cancer

TCGA fusion Gene DatabaseFusionCancerCOSMIC FusionsPubMedDr. Google

Fusion confirmationSanger sequencingReal-time quantitative PCRNested PCRFISH

Clinical report

QC check

QC check

QC checkQC check

Workflow and Analysis Pipeline

Sensitivity of CHOP FusionPanel

100% sensitivity:

Identified at least one causal fusion in all 30 positive cases

Detected NO fusion in 2 normal controls

• 32 validation samples from COG were tested blindly

Validation of Fusion Panel

Detection limit of CHOP FusionPanel

Strong evidence fusions:* >5 unique fusion supporting reads** >3 unique start sites

Validation of Fusion Panel - Cont’

• All 586 gene specific primers work properly in the multiplex PCR assay (reads ranging

from 9 to 16,000)

Validation of Fusion Panel - Cont’

Gene Expression Evaluation

* RNA Unique Molecular Bins: the RNA reads with unique molecular barcodes and random start site after deduplication.

der(Y)

XX

der(14)

Clinical Application

Positive rate: 34.2% (39/114)

EWSR1 exon7 CREB1 exon6

EWSR1 exon11 HLF exon4

ETV6 exon5 RUNX1 exon4

NUP214 exon32 ABL1 exon3

ETV6 exon5 RUNX1 exon3

PAX7 exon7 FOXO1 exon2

EWSR1 exon7 FLI1 exon8

MTAP exon7 BRAF exon9

JAZF1 exon4 TAX1BP1 exon16

PAX3 exon6 FOXO1 exon2

ASPSCR1 exon7 TFE3 exon6

PAX5 exon8 SOX5 exon4

TCF3 exon16 PBX1 exon3

EWSR1 exon7 FLI1 exon6

BRAF exon9

BRAF exon9

NOTCH1 exon30 ROS1 exon34

KMT2A exon10 ARHGEF12 exon12

CIC exon20 DUX4 exon1

WFS1 exon1 PLAG1 exon3

IGH CRLF2

KIAA1549 exon16 BRAF exon9

C11orf95 exon3 RELA exon2

CRTC1 exon1 MAML2 exon2

COL1A1 exon25 PDGFB exon2

RBPMS exon5 NTRK3 exon14

KIAA1549 exon13

KIAA1549 exon15

KIAA1549 exon10 BRAF exon11

A case of B-ALL: treated as standard risk

NUP214-ABL1 fusion identified on day 7 - Ph-like ALL

Tyrosine kinase inhibitor added to the treatment

Protein domain architectures

KIAA1549 exon16 BRAF exon9

As of 8/5/2016, 114 FusionPanels tested

• A 12 year old boy with high risk Pre B-cell acute lymphoblastic leukemia (ALL)

• AALL1131/VHR standard arm A Phase III Randomized Trial for Newly Diagnosed High Risk B-Lymphoblastic Leukemia (B-ALL)

• Day 8 MRD: 43.3%• Day 29 MRD: 40%

Identification of Novel Fusions

46,XY,der(6)t(6;14)(p23;q32)del(6)(q15q23),der(9)inv(9)(p?q?34)del(9)(p21p13),del(9)(q22q31),der(14)t(6;14)(p23;q32)[14].ish der(6)(IGH+),der(14)(IGH)/46,XY[4]

1

0.61

0.02 0.011 0.00056610.00384160.00001740

0.2

0.4

0.6

0.8

1

1.2

M1 M2 M3 M4 M5 M6 M7

Real Time qPCR showed nearly two log

reduction after two months of ruxolitinib

treatment

Gel picture of RT-PCR3 forward primers which locatedat exon 8 (F3), exon 9 (F2) andexon 10 (F1) of GENE1 and 3reverse primers which located atexon 19 (R1), exon 20 (R2) andexon 22 (R3) of JAK2 weredesigned.

Identification of Novel Fusions

A personalized qPCR test

Conclusions

• We have developed a customized Fusion panel for pediatric and adult cancers

• Clinical validation showed 100% sensitivity and 1% detection limit

• CHOP FusionPanel is compatible with different sample types (FFPE samples)

• High detection yield (34.2%) in different types of cancers

• CHOP FusionPanel allows for the detection of actionable fusions that provide

accurate diagnosis, prognosis and new treatment options for cancer patients

• Enable the development of personalized test to monitor treatment response and

minimal residual diseases

DGD Cancer Genomic Diagnostics

Marilyn M. LiMinjie LuoFumin LinLuanne WainwrightDonna WilmothAdam GleasonTammy GrouGozde AkgumusDaniel J GalloMichele ThiessXiaonan ZhaoJunxia TangJorune BalciunieneDerek A Anderson

DGD Bioinformatics

Mahdi SarmadyKajia CaoChao Wu

CHOP Cancer CenterPathology Dept.

Steve HungerJohn MarisBruce PawelMariarita Santi-ViciniYael P. MosseAngela J. WaandersSarah TasianAnd others

Other DGD members