Leveraging the transcriptome for the early detection of

Leveraging the transcriptome for the early detection of

Leveraging the transcriptome for the early detection of prostate cancer Arul M Chinnaiyan, M.D., Ph.D. Scott A. Tomlins, M.D., Ph.D. EDRN BDL MiPS= Mi Prostate Score (PCA3+ TMPRSS2-ERG + urinary PSA+ serum PSA) Test will be employed in our Team Projects (MRI-Biomarker and Upgrading Study) Project Overview Identify and validate novel prostate cancer and aggressive prostate cancer specific transcriptomic biomarkers Determine the utility of prognostic transcriptomic tissue based biomarkers

Determine the utility of diagnostic transcriptomic urinary based biomarkers Project Overview Identify and validate novel prostate cancer and aggressive prostate cancer specific transcriptomic biomarkers Determine the utility of prognostic transcriptomic tissue based biomarkers Determine the utility of diagnostic transcriptomic urinary based biomarkers We have PCa early detection tests None specifically use markers of aggressive prostate cancer We have PCa early detection tests

None specifically use markers of aggressive prostate cancer Multiplex RNAseq Isolate RNA (and DNA) Ampliseq RNA (RT, multiplex PCR, Ligate adaptors/barcode) NGS

Potential advantages <1-10ng FFPE or urine RNA vs. >>10ng qRT-PCR for many targets Multiplexing is easier by sequencing than most qRT-PCR Digital data (sequence information) Combinatorial priming for gene fusions/splice variants Correlation vs. RNAseq 21 bladder cancer cell lines Bladder cancer targeted RNAseq panel Standard RNAseq (UMHSCC sequencing core) Hovelson et al. European Urology 2018 306 transcript targeted RNAseq assay Diagnostic genes Subtyping genes (all ETS fusions) AR splice variants NePC genes

lncRNAs Expressed mutations Decipher Oncotype DX Prolaris (CCP) Salami et al., Manuscript submitted. Polaris Decipher Oncotype Dx 306 transcript targeted RNAseq assay (Tissue) 84 transcript targeted RNAseq assay for urine (MiPS-NGS)

TMPRSS2:ETS (fusions, many isoforms) PCA3, SChLAP1, other lncRNAs KLK3, other PCa genes, AR splice isoforms Expressed variants/mutations (HOXB13, SPOP) MiPS-NGS development ~90% informative rate with optimized extraction Same post-DRE whole urine sample as PCA3/MiPS Highly reproducible r of raw reads = 0.96 (range 0.84-1.00) Confirmed expression of key targets via qRT-PCR

Identified additional prostate specific urine housekeeping genes beyond PSA (KLK3) Can MiPS-NGS replace MiPS? Can NGS for TMPRSS2:ERG, PCA3 and KLK3 replace transcription mediated amplification assessment of the same markers? MiPS-NGS on banked urine samples from men who underwent clinical MiPS (n=20) Replaced TMA with NGS derived values for above biomarkers in the clinical models MiPS-NGS risk score MiPS vs. MiPS-NGS risk scores r = 0.74 100%

80% 60% 40% 20% 0% 0% 20% 40% MiPS PCa risk score

60% Benign. 30 negative biopsy cores. Fusion-negative 80% 100% Can MiPS-NGS replace and supplement MiPS? Archived urine samples from 66 men with no cancer/Gleason score 6 (GG1) on bx/RRP 43 men with Gleason score 4+3=7 (GG3) on bx/RRP Diagnostic performance of MiPS (but using NGS) vs. serum PSA alone Trained and tested new MiPS based logistic regression models Trained on 2/3 cohort (n=73), tested on remaining 1/3 (n=36)

Random forest variable selection model used to build 29 transcript regularized logistic regression model Trained on 2/3 cohort (n=73), tested on remaining 1/3 (n=36) Random forest selected 29 targets T2:ERG TDRD1 ERG T2:ERG T2:ERG T2:ERG SCHLAP1

T2:ERG T2:ERG T2:ERG SCHLAP1 ARLNC1 GG>=3 B9/GG1 PRCAT104 T2:ERG

T2:ERG T2:ERG T2:ERG T2:ERG Performance of 29 transcript model Training (n=73) Serum PSA MiPS MiPS-NGS 29 trans model Testing (n=36)

AUC 0.670 0.714 0.831 0.899 Serum PSA MiPS MiPS-NGS 29 trans model AUC 0.625 0.757 0.714 0.812 Summary

Developed a post-DRE, whole urine-based targeted RNAseq assay for PCa early detection Highly correlated to current clinically validated assay assessing T2:ERG and PCA3 (MiPS) Improves upon the performance of serum PSA Likely additional value in multigene model NGS platform allows for evolution of content Research Highlights MiPS is offered as a CLIA/CAP test. PCA3 is FDA approved. Established MiPS-NGS assay (as the nextgen test) Discovered/characterized novel lncRNA biomarkers (PCAT14 (Neoplasia 2016), ARLNC1 (Nature Genetics 2018), and THOR (Cell 2017)) Identified CDK12 a biomarker for sensitivity to immune

checkpoint inhibitors (Cell 2018) Wu, Cieslik, Robinson et al, Cell 2018 Challenges Issues with Commercial Partner (Hologic/Gen-Probe) for urine PCA3 and T2-ERG assays Requirement of post-DRE not ideal (need more sensitive assays) Advent of prostate MRI While early diagnosis is possibly with a urine test, developing a test that picks up aggressive prostate cancer is the key Outlook/Future Plans

Establish MiPS NGS as the new urine MiPS test (will be used as part of EDRN CORE project) Refine content of MiPS NGS Carry out MiPS-NGS validation studies as part of the EDRN. Identify a new commercial partner to develop and market MiPS and MiPS-NGS Acknowledgements MCTP Andi Cani Kevin Hu Sumin Han Daniel Hovelson Albert Liu Komal Kunder

UMHS John Wei Javed Siddiqui Simpa Salami Ganesh Palapattu Todd Morgan Collaborators Pete Nelson (Univ. Washington) Nathalie Rioux-Leclercq (Rennes Hospital) Shahrokh Shariat (Med. Univ. of Vienna Mark Rubin (WCMC)Pete Nelson (FHCRC) Karen Knudsen (TJU ) Seth Sadis (ThermoFisher) Ryan Dittamore (Epic Biosciences) Gert Attard (Royal Marsden) ThermoFisher Ampliseq R&D Team

Challenges Co-PI, Scott Tomlins recruited to industry (however we have maintained 5% effort at U-M) Issues with Commercial Partner (Hologic/Gen-Probe) for urine PCA3 and T2-ERG assays Requirement of post-DRE not ideal (need more sensitive assays) Advent of prostate MRI While early diagnosis is possibly with a urine test, developing a test that picks up aggressive prostate cancer is the key Patents, Licenses, and International Collaborations

Secured PCA3 license from Hologic; T2-ERG IP returned to U-M (considering new commercial partners) MiPS in the U-M CLIA Pathology labs as offered as urine test U-M has secured a supply contract from Hologic for T2-ERG urine assay reagents International collaborations with institutions in China are under discussion Outlook/Future Plans Establish MiPS NGS as the new urine MiPS test (will be used as part of EDRN CORE project) Refine content of MiPS NGS Carry out MiPS-NGS validation studies as part of the EDRN.

Identify a new commercial partner to develop and market MiPS and MiPS-NGS Publications (acknowledging EDRN) 1. Merdan S, Tomlins SA, Barnett CL, Morgan TM, Montie JE, Wei JT, Denton BT. Assessment of long-term outcomes associated with urinary prostate cancer antigen 3 and TMPRSS2:ERG gene fusion at repeat biopsy. Cancer. 2015 November 15;121(22):4071-9. 2. Udager AM, DeMarzo AM, Shi Y, Hicks JL, Cao X, Siddiqui J, Jiang H, Chinnaiyan AM, Mehra R. Concurrent nuclear ERG and MYC protein overexpression defines a subset of locally advanced prostate cancer: Potential opportunities for synergistic targeted therapeutics. Prostate. 2016 Jun;76(9):845-53. 3. Shukla S, Zhang X, ., Bui HH, Siddiqui J, Jing X, Cao X, Dhanasekaran SM, Feng FY, Chinnaiyan AM, Malik R. Identification and Validation of PCAT14 as Prognostic Biomarker in Prostate Cancer. Neoplasia. 2016 Aug;18(8):489-99. 4. Mani RS, Amin MA, Li X, , Chinnaiyan AM. Inflammation-Induced Oxidative Stress Mediates Gene Fusion Formation in Prostate Cancer. Cell Reports 2016 Dec 6;17(10):2620-2631. 5. Wang X, Qiao Y, Chinnaiyan AM. Development of Peptidomimetic Inhibitors of the ERG Gene Fusion Product in Prostate Cancer. Cancer Cell. 2017 Apr 10;31(4):532-548.e7.

6. Blattner M, Liu D, Robinson BD, Huang D, Poliakov A, Gao D, Nataraj S, Deonarine LD, Augello MA, Sailer V, Ponnala L, Ittmann M, Chinnaiyan AM, Sboner A, Chen Y, Rubin MA, Barbieri CE. SPOP Mutation Drives Prostate Tumorigenesis In Vivo through Coordinate Regulation of PI3K/mTOR and AR Signaling. Cancer Cell. 2017 Mar 13;31(3):436-451. 7. Robinson DR, Wu YM, , Chinnaiyan AM. Integrative clinical genomics of metastatic cancer. Nature. 2017 August 17;548(7667):297-303. 8. Hosono Y, , Chinnaiyan AM. Oncogenic Role of THOR, a Conserved Cancer/Testis Long Noncoding RNA. Cell. 2017 December 14;171(7):1559-1572.e20. 9. Zhang Y, Pitchiaya S, , Chinnaiyan AM. Analysis of the androgen receptor-regulated lncRNA landscape identifies a role for ARLNC1 in prostate cancer progression. Nature Genetics. 2018 June;50(6):814-824. 10. Quigley DA, , Chinnaiyan AM, Maher CA, Small EJ, Feng FY. Genomic Hallmarks and Structural Variation in Metastatic Prostate Cancer. Cell. 2018 Jul 26;174(3):758-769.e9. 11. Wu YM, PCF/SU2C International Prostate Cancer Dream Team, Robinson DR, Chinnaiyan AM. Inactivation of CDK12 Delineates a Distinct Immunogenic Class of Advanced Prostate Cancer. Cell. 2018 Jun 14;173(7):17701782.e14. Publications (acknowledging EDRN) 1.

Merdan S, Tomlins SA, Barnett CL, Morgan TM, Montie JE, Wei JT, Denton BT. Assessment of long-term outcomes associated with urinary prostate cancer antigen 3 and TMPRSS2:ERG gene fusion at repeat biopsy. Cancer. 2015 November 15;121(22):4071-9. 2. Udager AM, DeMarzo AM, Shi Y, Hicks JL, Cao X, Siddiqui J, Jiang H, Chinnaiyan AM, Mehra R. Concurrent nuclear ERG and MYC protein overexpression defines a subset of locally advanced prostate cancer: Potential opportunities for synergistic targeted therapeutics. Prostate. 2016 Jun;76(9):845-53. 3. Shukla S, Zhang X, ., Bui HH, Siddiqui J, Jing X, Cao X, Dhanasekaran SM, Feng FY, Chinnaiyan AM, Malik R. Identification and Validation of PCAT14 as Prognostic Biomarker in Prostate Cancer. Neoplasia. 2016 Aug;18(8):489-99. 4. Mani RS, Amin MA, Li X, , Chinnaiyan AM. Inflammation-Induced Oxidative Stress Mediates Gene Fusion Formation in Prostate Cancer. Cell Reports 2016 Dec 6;17(10):2620-2631. 5. Wang X, Qiao Y, Chinnaiyan AM. Development of Peptidomimetic Inhibitors of the ERG Gene Fusion Product in Prostate Cancer. Cancer Cell. 2017 Apr 10;31(4):532-548.e7. 6. Blattner M, Liu D, Robinson BD, Huang D, Poliakov A, Gao D, Nataraj S, Deonarine LD, Augello MA, Sailer V, Ponnala L, Ittmann M, Chinnaiyan AM, Sboner A, Chen Y, Rubin MA, Barbieri CE. SPOP Mutation Drives Prostate Tumorigenesis In Vivo through Coordinate Regulation of PI3K/mTOR and AR Signaling. Cancer Cell. 2017 Mar 13;31(3):436-451. 7. Robinson DR, Wu YM, , Chinnaiyan AM. Integrative clinical genomics of metastatic cancer. Nature. 2017 August

17;548(7667):297-303. 8. Hosono Y, , Chinnaiyan AM. Oncogenic Role of THOR, a Conserved Cancer/Testis Long Noncoding RNA. Cell. 2017 December 14;171(7):1559-1572.e20. 9. Zhang Y, Pitchiaya S, , Chinnaiyan AM. Analysis of the androgen receptor-regulated lncRNA landscape identifies a role for ARLNC1 in prostate cancer progression. Nature Genetics. 2018 June;50(6):814-824. 10. Quigley DA, , Chinnaiyan AM, Maher CA, Small EJ, Feng FY. Genomic Hallmarks and Structural Variation in Metastatic Prostate Cancer. Cell. 2018 Jul 26;174(3):758-769.e9. 11. Wu YM, PCF/SU2C International Prostate Cancer Dream Team, Robinson DR, Chinnaiyan AM. Inactivation of CDK12 Delineates a Distinct Immunogenic Class of Advanced Prostate Cancer. Cell. 2018 Jun 14;173(7):17701782.e14. PCa urine RNA NGS assay uMiPS_v1 High concordance with MiPS (r=0.83) T2:ERG T1E4 not always most common isoform Paired samples

cluster together, median r=0.99 (range 0.77-0.99) Assay technical parameters and reproducibility Mean Mapped Reads Mean On Target Reads 2,588,665 99.9% DNase Untreated DNA contamination Urine #173 RNA sequencing 10,000 100

r=0.95 1 100 10,000 No DNase Mean Read Length 122 bp DNAse - treated 1,000,000 1

Mean end-to-end reads 83.0% 1,000,000 uMiPS_v2 High quality reproducible sequencing DNase treatment does not affect read count in sample with low DNA contamination Pearson mean r = 0.96 (range 0.841.00) Current informative sample rate >90% uMiPS_v2 Concordance with MiPS PCA3 Score

T2:ERG_T1E4 Score 10 Sequencing 8 8 6 6 4 4 2 2 r = 0.85 0

3 log2 transformed 4 5 MIPS 6 7 0 r = 0.96

0 1 2 3 4 5 6 7 MIPS

MiPS and Sequencing assays are highly concordant 8 9 MiPS-NGS hg risk score MiPS hg vs. MiPS-NGS hg risk scores r = 0.81 80% 60% 40% 20%

Benign. 30 negative biopsy cores. Fusion-negative 0% 0% 20% 40% MiPS hg Risk Score 60% 80% RT-qPCR assay Extreme Design Cohort: Benign/GG1 vs. GG 3,

4, 5 Equally Transcripts PCAT4 Score KLK2Expressed Score 250 ns ns 100 (x 1000) 200

150 50 100 Benign/GG 1 GG 3,4,5 Benign/GG 1 GG 3,4,5 Expression of housekeeping genes not associated with tumor grade Extreme Design Cohort: TMPRSS2:ERG T2:ERG_T1E4 Score

Isoform Source Grade Group p=0.07 MIPS T2-ERG_T1E4 400 300 T2:ERG_All Score UR175 PBX 4

4000 p=0.07 347 TMPRSS2-ERG_T1E4 T2ERG_All 448 3,924 TMPRSS2-ERG_T3E4 TMPRSS2-ERG_T2E4 TMPRSS2-ERG_T4E4 TMPRSS2-ERG_T4E5 2,473

681 0 0 3000 2000 200 Isoform Source Grade Group MIPS T2-ERG_T1E4 100 0

Benign/GG 1 GG 3,4,5 UR202 RRP 3 N/A TMPRSS2-ERG_T1E4 T2ERG_All 34 1,230 TMPRSS2-ERG_T3E4 TMPRSS2-ERG_T2E4 TMPRSS2-ERG_T4E4

TMPRSS2-ERG_T4E5 628 567 0 0 1000 0 Benign/GG 1 GG 3,4,5 Additional T2:ERG isoforms may be clinically useful Performance of MiPS/MiPS-NGS hg

Ex design cohort (n=109) Serum PSA MiPS MIPShg AUC 0.660 0.729 0.710 using clinical MiPS models but uMIPS NGS quantified KLK3, TMPRSS2:ERG, and PCA3 values

Recently Viewed Presentations

  • Why Is Asset Allocation Important?

    Why Is Asset Allocation Important?

    No accelerated vesting at Duke Energy. Eight proposals go to vote. No accelerated vesting at DirecTV, 31%. No accelerated vesting at Windstream, 22%. Equity retention at Comcast, 11%. Independent Chair at Abbott Laboratories, 19%. Cumulative Voting at Autonation, 15%. Equity...
  • What turning points determine our individual pathways to ...

    What turning points determine our individual pathways to ...

    The Kite Runner. The House on Mango . Street. A Separate Peace (In what ways is it? Is it not?) ... Common imagery and symbols -fruit (Especially apples! Ignorance is bliss— knowledge ends ignorance, knowledge ends bliss.) -gardens/forest -falling -flowers
  • Inventory Management - University of Arizona

    Inventory Management - University of Arizona

    Dependent demand items. Items that are subassemblies or component parts to be used in the production of finished goods. Independent demand items. Items that are ready to be sold or used. Inventories are a vital part of business: necessary for...
  • Module 4. Introduction to Statistical Process Control

    Module 4. Introduction to Statistical Process Control

    Introduction to Statistical Process Control Module 4 History of Statistical Process Control Quality Control in Industry Shewhart and Bell Telephones Deming & Japan after WWII Use in Health Care & Public Health The Run Chart The Count Sunday 12 Monday...
  • Stille Nacht - Frau Holmquist

    Stille Nacht - Frau Holmquist

    When Gruber sat down, his fingers began playing the simple melody he had written for Mohr's Christmas poem. Deeply impressed, Mauracher took the music and words of "Stille Nacht" back to his own Alpine village, Kapfing. There, two well-known families...
  • Data Modeling using XML Schemas

    Data Modeling using XML Schemas

    Revision Instructor: Mohamed Eltabakh [email protected] * Midterm Regulations Midterm is Feb. 7th (11:00am - 12:30pm) Covers until before the Views You answer in the same exam paper Closed-Book, One sheet is allowed for your notes * Topics Covered in Midterm...
  • Regional Haze Rule

    Regional Haze Rule

    SAGO. JOSH. PORE. SAGA. DOME. KAIS. LAVO. ... Arial Times New Roman Symbol Default Design Microsoft Excel Chart Microsoft Office Excel Chart Microsoft Photo Editor 3.0 Photo REGIONAL HAZE Progress Report CALIFORNIA CLASS 1 AREAS OVERVIEW BASIC REQUIREMENTS MONITOR DATA...
  • Getting Started with the Raspberry Pi - Seattle Robotics

    Getting Started with the Raspberry Pi - Seattle Robotics

    GPIO also awkward to get at from inside Linux. No motor drivers, or high current outputs. Really no standard I/O of any kind other than USB and Ethernet! Doesn't really have enough processing power to do vision applications. ... Getting...