New Paradigms in Personalized Medicine and Drug Discovery for Cancer - - PowerPoint PPT Presentation

▶
new paradigms in personalized medicine and drug discovery
SMART_READER_LITE
LIVE PREVIEW

New Paradigms in Personalized Medicine and Drug Discovery for Cancer - - PowerPoint PPT Presentation

Feb 24, 2023 250 likes •740 views

New Paradigms in Personalized Medicine and Drug Discovery for Cancer d D Di f C Dan Theodorescu MD PhD University of Virginia Charlottesville, Virginia, USA Cancer of the Bladder Cancer of the Bladder The Clinical Challenge of Metastatic

slide-1
SLIDE 1

New Paradigms in Personalized Medicine d D Di f C and Drug Discovery for Cancer

Dan Theodorescu MD PhD

University of Virginia Charlottesville, Virginia, USA

slide-2
SLIDE 2

Cancer of the Bladder Cancer of the Bladder

The Clinical Challenge of Metastatic Disease The Clinical Challenge of Metastatic Disease

Overall Survival of 405 patients with T4b or N2-3 or M1 urothelial carcinoma randomized to Gem-Cis or MVAC

von der Maase H et al J Clin Oncol; 23:4602 4608 2005

www mritutor com

von der Maase, H. et al. J Clin Oncol; 23:4602-4608 2005

Metastatic disease = Death

www.mritutor.com

Visceral Metastasis curable only in ~5-7%

Needs

Individualize “personalize” therapy

www.tju.com

Individualize “personalize” therapy Better drugs / drug combinations

slide-3
SLIDE 3

Challenges of Individualized cancer therapy Challenges of Individualized cancer therapy

We have prognostic markers of outcome We don’t have predictive biomarkers of We don t have predictive biomarkers of treatment response in majority of tumors

Could our single drugs or drug combinations th t 5 10% f ti t if li d t that cure 5-10% of patients, if applied to specific patient subsets this result in improved cure rates ?

Treatment Treatment

Chemotherapy Regimens Treatment Response Biomarkers Tumor Sample MVAC GC Regimens Optimized Regimen Selection In aggregate

Cancer Cancer

Biomarkers GT In aggregate Cures Likely >10%

slide-4
SLIDE 4

Challenges of Drug Discovery Challenges of Drug Discovery

Discovery (2-10 Years)

5,000 - 10,000

Compound Success Rate in Drug Discovery ( ) Preclinical Testing Laboratory and animal testing Phase I Determine safety and dosage Phase II

Screened 250 Enter preclinical testing

Phase II Efficacy and side effects Phase III Adverse reactions to long-term use FDA R i /A l

Enter preclinical testing 5 Enter clinical testing

2 4 6 8 10 12 14 16

FDA Review/Approval Additional Post-market Testing Years

1

2 4 6 8 10 12 14 16

~$880 million / successful drug ~$880 million / successful drug

1 Approved by FDA

Time Time—12 years! 12 years!

slide-5
SLIDE 5

Drug Discovery and Clinical Medicine Drug Discovery and Clinical Medicine Common Problem Common Problem—The The “Tumor “Tumor-

  • Drug Disconnect”

Drug Disconnect”

NCI-60 Cell Panel

Human Cell Lines L k i (6) M l (7) B t (8)

Drug Discovery and Development

Leukemia (6) Melanoma (7) Breast (8) Ovarian (6) CNS (6) Lung (9) Prostate (2) Colon (7) Kidney (8). HTS Drugs Screening (eg. NCI) >100,000 chemical compounds lt f 45K il bl bli ll

Clinical

575A

Sensitive

1 1 1 1 1 2 2 2 3 3 3

results of 45K are available publically

Clinical Practice

1 1 2 2 3 3

Poor Predictability of Drug Action in Patients

slide-6
SLIDE 6

Classic Solution to the “Tumor Classic Solution to the “Tumor-

  • Drug Disconnect”

Drug Disconnect”…..does …..does not not help efficiency of drug discovery! help efficiency of drug discovery!

Discovery (2-10 Years) Preclinical Testing Laboratory and animal testing Phase I

Tumor Sample Taken 3 Drug Chemotherapy Regimen “ABC”

Tumor Profiling

ase Determine safety and dosage Phase II Efficacy and side effects Phase III

Regimen ABC Responders Non Responders Biomarker Development

Adverse reactions to long-term use FDA Review/Approval Additional Post-market Testing

Treatment Response Biomarkers Development

Clinical Selection of patients

Post-market Testing Years Biomarker Development Clinical use with Response Biomaker

Clinical Use of ABC Selection of patients that respond to ABC

$$$$ and 1 $$$$ and 1-2 yrs! 2 yrs!

2 4 6 8 10 12 14 16

$$$$ $$$$ y Limited utility Limited utility

slide-7
SLIDE 7

Comprehensive solution to the “Tumor Comprehensive solution to the “Tumor-

  • Drug Disconnect”

Drug Disconnect” Addressing Addressing both both Drug Discovery and Individualized Therapy Drug Discovery and Individualized Therapy

Inspiration….The Rosetta Stone

Hieroglyphic: script for important / religious documents Demotic Egyptian: common script of Egypt Greek: language of the rulers of Egypt NCI-60 Panel Cell Lines

575A

Sensitive

1 1 1 2 2 2 3 3 3

t Control

Idea….

NCI 60 Panel Cell Lines Gene expression profile

1 1 2 2 2 3 3 3

Dose Percent

Patient Bladder Tumor

slide-8
SLIDE 8

The Idea: COXEN “ The Idea: COXEN “CO CO-e eX Xpression pression E Extrapolatio xtrapolatioN N”

Uses in Drug Discovery and Individualized Therapy Uses in Drug Discovery and Individualized Therapy

NCI-60 Cell Line Panel

575A

Sensitive

1 1 1 1 1 2 2 2 2 3 3 3 3

Expression Profiling IC50 for 45,345 Compounds

2 2 3 3

Human Bladder Cancer Cell Lines Bladder Cancer patient samples

COXEN COXEN

Gene Expression Model (GEM) for each Test Compound p ( ) p GEM Score Evaluation on Bladder Cancer patient tissues or cells

Hi h S

575A

Sensitive

1 1 1 2 2 3 3

wth

COXEN Score for each drug across all patients

High Score Low Score

0.0 0.2 0.4 0.6 0.8 1.0

Predicted responders Predicted non-responders (P-value = 0.021)

Fraction Disease-Free (p-value = 0.021)

COXEN Score for each patient for specific drug

High Score Low Score

1 1 1 2 2 2 3 3 3 575A

Sensitive

1 1 1 1 1 2 2 2 2 2 3 3 3 3 3

Growth Grow

Drug Discovery

Low Score

50 100 150 Survival time

Time (months)

Individualized Therapy

2 3

[Drug]

slide-9
SLIDE 9

The Idea: COXEN “ The Idea: COXEN “CO CO-e eX Xpression pression E Extrapolatio xtrapolatioN N”

Uses in Drug Discovery and Individualized Therapy Uses in Drug Discovery and Individualized Therapy

Discovery (2-10 Years)

COXEN COXEN

Preclinical Testing Laboratory and animal testing Phase I Determine safety and dosage Phase II Efficacy and side effects Phase III Adverse reactions to long-term use FDA Review/Approval FDA Review/Approval Additional Post-market Testing Biomarker Development

2 4 6 8 10 12 14 16

Years Clinical use with Response Biomaker

slide-10
SLIDE 10

COXEN Applied to COXEN Applied to Individualized Therapy Individualized Therapy

COXEN COXEN

COXEN provides treatment response biomarkers without the need for tissue from patients treated with h th i ! Discovery (2-10 Years) Preclinical Testing

COXEN COXEN

chemotherapy regimens!

  • Uses in vitro data
  • Can develop biomarkers

f d g Laboratory and animal testing Phase I Determine safety and dosage Phase II Efficacy and side effects for any drug combinations within days with minimal effort ! y Phase III Adverse reactions to long-term use FDA Review/Approval Additional Post-market Testing Biomarker Development

2 4 6 8 10 12 14 16

Years Clinical use with Response Biomaker

slide-11
SLIDE 11

Can COXEN predict effectiveness of cisplatin Can COXEN predict effectiveness of cisplatin and paclitaxel in and paclitaxel in Bladder cancer cells Bladder cancer cells ?

Cisplatin (GI50) Cisplatin (GI50) Paclitaxel (GI50) Paclitaxel (GI50)

umuc9 X253jp slt4p3 X253jbv umuc14 Cisplatin normalized log(GI50) & MiPP prediction scores

sensitive log(GI50) (p-value = 0.016)

Sensitive: Actual GI50

BLA-40 Cell line

sensitive log(GI50) iti di t d

htb9 crl2742 umuc2 ku7 X253jbv BLA-40 Cell line

(p-value = 0.006)

Sensitive: Actual GI50

umuc14 crl7833 fl3p10 ku7 umuc3 umuc3e rt4 crl7193

sensitive log(GI50) sensitive predicted resistent log(GI50) resistent predicted resistant resistant

Sensitive: Actual GI50 Sensitive: COXEN Prediction Resistant: Actual GI50 Resistant: COXEN Prediction

sensitive predicted resistent log(GI50) resistent predicted

53jb scaber umuc6 umuc1 X253jp vmcub1 jon cubIII

Sensitive: COXEN Prediction Resistant: Actual GI50 Resistant: COXEN Prediction

crl2169 ht1197 X575a cubIII mghu3 jon kk47 crl7193 j82 psi bc16.1 ht1197 rt4 kk47 1 bc16.1

  • 1.5
  • 1.0
  • 0.5

0.0 0.5 1.0 1.5 2.0

Standardized log(GI50) Standardized COXEN Score

  • 1.5
  • 1.0
  • 0.5

0.0 0.5 1.0 1.5

umuc1

Standardized log(GI50) Standardized COXEN Score

…..No bladder cell lines were on NCI60 panel

slide-12
SLIDE 12

Conclusion: Conclusion: COXEN COXEN Predicts In Vitro Predicts In Vitro Chemotherapy Responses Chemotherapy Responses

But WAIT!........most human cancers treated with combination chemotherapy

Can COXEN predict effectiveness of known chemotherapeutic drug combinations in bladder cancer cell lines? cell lines? Can COXEN predict treatment responses of known drug combinations in bladder cancer patients?

slide-13
SLIDE 13

Can COXEN Predict Combination Chemotherapy Can COXEN Predict Combination Chemotherapy Responses ? Responses ?

Approach

Use 40 bladder cancer cell lines (BLA-40) Evaluate common “doublet” drug combinations used in patients

NCI-60 Panel Cisplatin Gemcitabine Taxol

COXEN COXEN

575A

Sensitive

1 1 1 1 1 2 2 2 2 2 3 3 3 3 3

C

BLA-40 Taxol

2 3

Validation of COXEN predictions on BLA-40

In vitro evaluation of combinations in BLA-40 cells COXEN Scores of cisplatin + gemcitabine cisplatin + paclitaxel gemcitabine + paclitaxel

Validation of COXEN predictions on BLA-40

compare

in BLA 40 cells g p activity in the BLA-40

IC50

slide-14
SLIDE 14

Results: Results: COXEN predicts effectiveness of COXEN predicts effectiveness of combination combination therapy in therapy in bladder cancer cells bladder cancer cells ?

CIS + PAC

Resistant

Resistant 100

12 3

gate

CIS + GEM PAC + GEM 80 Correct Class

12 3 11 4

50 Surrog

60 40 Correct Class Misclassified

IC5

20

COXEN Score

Sensitive

0.0 0.2 0.4 0.6 0.8 1.0 Sensitive

Non-responder Responder

COXEN Score

slide-15
SLIDE 15

Can Can COXEN COXEN Predict Combination Chemotherapy Predict Combination Chemotherapy Response in Response in bladder cancer patients bladder cancer patients?

Reference Set (Used for Model Development)

MSKCC (N=105) and UVA (N=58) MSKCC (N 105) and UVA (N 58) Tissues profiled prior to undergoing TURBT or cystectomy No follow up information used Pathological Information: Pathological Information:

Stage N(%) UVA (N=58) MSKCC (N=105) T0 5 (8) 3 (3) ( ) ( ) Tis, G3 5 (8) Ta, G1 1 (2) Ta, G2 10 (17) 2 (2) Ta, G3 19 (33) T1, G2 3 (5) 13 (12) , ( ) ( ) T1, G3 3 (5) 12 (11) T2, G2 1 (2) 1 (1) T2, G3 3 (5) 10 (10) T3, G2 4 (4) T3, G3 3 (5) 48 (46) T3, G3 3 (5) 48 (46) T4, G2 1 (2) T4, G3 4 (7) 11 (10)

slide-16
SLIDE 16

Can Can COXEN COXEN Predict Combination Chemotherapy Predict Combination Chemotherapy Response in Response in bladder cancer patients bladder cancer patients?

Validation Sets

St di ith li i l t th d fili i f ti Studies with clinical response to therapy and gene profiling information NCI60-Drug sensitivity information on panel available Completely independent from Training/COXEN model derivation Als (Denmark)(Clin Cancer Res 2007;4407 13(15):4407) Treatment MVAC (N=16) or GC (N=14) M0 M ti t th th M0 or M+ patients, no other therapy Outcome: Overall survival T k t (J ) (Cli C R 2005 11(7) 2625 ) Takata (Japan) (Clin Cancer Res 2005;11(7): 2625 ) Neoadjuvant MVAC (N=45) followed by surgery or XRT Outcome: Tumor size reduction/Downstaging, Overall survival

slide-17
SLIDE 17

COXEN prediction of treatment outcome in COXEN prediction of treatment outcome in patients treated with MVAC or GC patients treated with MVAC or GC

Als (Denmark)(Clin Cancer Res 2007;4407 13(15):4407)

Treatment MVAC (N=16) or GC (N=14)

PARAMETER N(%)

Follow-up for patients at risk (mo) Median (range) 81.8 (56.7-98.0)

Overall Survival

Age (y) Median (range) 61.5 (49-74) Sex Male 24 (80) Female 6 (20) PS (ECOG)

urviving

0.6 0.8 1.0

M0 M1 P=0.208

( ) 0-1 27 (90) >2 3 (10) Hemoglobin Normal 15 (50) Low 15 (50) P-alkaline phosphatase

Proportion Su

0.2 0.4

P-alkaline phosphatase Normal 22 (74) Elevated{dagger} 8 (26) Stage M0 T4b, N2-3 15 (50) M1

0.0 Survival Time (Months) 12 24 36 48 60 72 84 96

Time (months)

M1 Extra pelvine lymph node 6 (20) Visceral organs 9 (30)

slide-18
SLIDE 18

COXEN prediction of treatment outcome in patients COXEN prediction of treatment outcome in patients using using combination combination drug GEM for MVAC or GC drug GEM for MVAC or GC

8 1.0

p y

P = 0.0469

p = 0.039

MVAC (N=16)

MSKCC & UVA

2 0.4 0.6 0.8 || | 0.0 0.2

Survival Time (Months)

12 24 36 48 60 72 84 96

Predicted Responders (5) Predicted Nonresponders (9)

COXEN COXEN

Gene Expression Model Evaluation of Model on

0.8 1.0

p y

|| |

GC (N=14)

Als (Denmark)

Cells or Tumors (PCR)

575A

Sensitive

1 1 1 1 1 2 2 2 2 2 3 3 3 3 3

0.2 0.4 0.6

Predicted Responders (4) Predicted Nonresponders (10)

P = 0 0303

p = 0.030

NCI-60 Panel Cisplatin Gemcitabine Methotrexate

0.0

Survival Time (Months)

12 24 36 48 60 72 84 96 P = 0.0303

p

Doxorubicin Vinblastine

slide-19
SLIDE 19

COXEN prediction of treatment outcome in COXEN prediction of treatment outcome in patients treated with neoadjuvant MVAC patients treated with neoadjuvant MVAC

Takata (Japan) (Clin Cancer Res 2005;11(7): 2625)

Neodjuvant MVAC (N=45)

Overall Survival

PARAMETER N(%) Surviving (%)

Follow-up for patients (mo) Median (range) 27 (2-56) Age (y) Median (range) 67 (53-77) Sex

0.6 0.8 1.

Moreno Orntoft Takada Als

Proportion S

Sex Male 33 (73) Female 12 (27) Stage M0 T2a, N0 1 (2)

0.2 0.4

T2b, N0 8 (18) T2b, N2 1 (2) T3a, N0 5 (11) T3b, N0 30 (67)

0. Survival Time (Months) 12 24 36 48 60 72 84 96

Time (months)

slide-20
SLIDE 20

COXEN prediction of treatment outcome in COXEN prediction of treatment outcome in patients treated with neoadjuvant MVAC patients treated with neoadjuvant MVAC

MSKCC & UVA

ction

Tumor Size Reduction vs. COXEN Score NCI-60 Panel

575A 1 1 2 2 3 3

COXEN Score nt Tumor Reduc

Methotrexate Vinblastine

575A

Sensitive

1 1 1 1 2 2 2 2 3 3 3 3

COXEN COXEN

C Percen

Patient Number

Doxorubicin Cisplatin

Gene Expression Model Evaluation of Model on Cells or Tumors (PCR)

Downstaging vs. COXEN Score

  • re

Takata (Japan)

Cells or Tumors (PCR)

COXEN Sco

Downstaging defined as ≤pT1 or ≤T1 after two courses of MVAC Downstaged NO Downstage

slide-21
SLIDE 21

COXEN prediction of treatment with neoadjuvant MVAC COXEN prediction of treatment with neoadjuvant MVAC Impact of the COXEN GEM score cutoff Impact of the COXEN GEM score cutoff

6 0.8 1.0

A

0.0 0.2 0.4 0.6

Predicted Responders (3) Predicted Non-Responders (42) P-Value = 0.228

10 20 30 40 50 Survival time

Lower COXEN Score

B

cutoff

B

Conclusion:

COXEN Score thresholding can provide patient cohorts more (A) or less (B) likely to respond to therapy depending on clinical requirements

slide-22
SLIDE 22

COXEN prediction of treatment with neoadjuvant MVAC COXEN prediction of treatment with neoadjuvant MVAC

COXEN GEM vs. Conventional GEM prediction COXEN GEM vs. Conventional GEM prediction

Tumor Sample Taken MVAC Chemotherapy treatment

Tumor Profiling

MSKCC & UVA NCI-60 Panel

575A

Sensitive

1 1 1 1 1 2 2 2 2 2 3 3 3 3 3

Takata (Japan)

MVAC Chemotherapy treatment Responders Non Responders Biomarker Development

COXEN

MVAC Drugs

N=87 N=18

COXEN GEM |

1.0

Predicted Responders (32) Predicted Nonresponders (13)

Takata (Japan) N=45 | |

1.0

Predicted Responders (27) Predicted Nonresponders (18)

Takata (Japan) N=45 Takata GEM A C

se Free se Free

Evaluation of Model on Test Sets | |

12 24 36 48 60 72

| | | ||||| | |||

0.0 0.5 P = 0.00474

| |

12 24 36 48 60 72

| | || ||||| | |||

0.0 0.5 P = 0.00117 Months Months Proportion Diseas Proportion Disea 5 1.0

Predicted Responders (7) Predicted Nonresponders (7)

Als (Denmark) N=14 | | | || | |||

5 1.0

Predicted Responders (18) Predicted Nonresponders (9)

Takata (Japan) N=27 D F | | |

1.0

Predicted Responders (20) Predicted Nonresponders (7)

E Takata (Japan) N=27 || |

1.0

Predicted Responders (4)

Als (Denmark) N=14 B ||

12 24 36 48 60 72 84 96

|

0.0 0.5 P = 0.73

||

12 24 36 48 60 72

| || | |||

0.0 0.5 P = 0.0777

|

12 24 36 48 60 72

| | | |||

0.0 0.5 P = 0.0527 12 24 36 48 60 72 84 96 0.0 0.5

Predicted Responders (4) Predicted Nonresponders (10)

P = 0.0198

slide-23
SLIDE 23

COXEN Applied to Other Cancers COXEN Applied to Other Cancers

Question…… Question……

Can COXEN predict clinical outcome in other p cancer types beyond bladder cancer?

slide-24
SLIDE 24

Can COXEN Algorithm Predict Combination Can COXEN Algorithm Predict Combination Chemotherapy Responses in Chemotherapy Responses in patients patients?

Studies Data Search

Collect studies with clinical response to therapy and gene profiling Collect studies with clinical response to therapy and gene profiling information (same criteria as single drug breast trials) Drug sensitivity information on NCI60 panel

Breast cancer: 5 studies Breast cancer: 5 studies

patients with stage I-III breast cancer Adj Tam or Neoadjuvant Docetaxel single agent Neoadjuvant paclitaxel and fluorouracil-doxorubicin- Neoadjuvant paclitaxel and fluorouracil-doxorubicin- cyclophosphamide (T/FAC), Overall survival Outcome: Pathological response

Ovarian: 2 studies Ovarian: 2 studies

Carbo-Paclitaxel or Cisplatin Chemotherapy Outcome: Overall survival

Analysis: Similar to that shown for BLA 40 combination Analysis: Similar to that shown for BLA-40 combination chemotherapy but instead of in vitro validation, we would validate

  • ur predictions by the clinical outcome
slide-25
SLIDE 25

Validation: Validation: Can Can COXEN COXEN predict predict patient treatment patient treatment

  • utcome
  • utcome in breast cancer clinical trials ?

in breast cancer clinical trials ?

Primary tumor response to neoadjuvant docetaxel (DOC 24) Survival following adjuvant tamoxifen (TAM-60)

responder res. tumor size responder predicted score

Responder: Tumor size Sensitive: COXEN Prediction

1.0

Predicted responders

neoadjuvant docetaxel (DOC-24) adjuvant tamoxifen (TAM-60)

1 2

non-responder res. tumor size non-responder predicted residual size XEN Score score

(p-value = 0.033)

Non-responder: Tumor size Non-responder: COXEN Prediction

0.6 0.8

ease-Free

dardized tumor tandardized COX

.2 0.4

Predicted non-responders

(P-value = 0 021)

raction Dise

(p value = 0 021)

  • 1

Stan St

50 100 150 0.0

(P-value = 0.021)

Fr

(p-value = 0.021)

Survival time

Time (months)

slide-26
SLIDE 26

Validation: Validation: Can Can COXEN Algorithm Predict COXEN Algorithm Predict …..T/ …..T/FAC FAC…. Responses in …. Responses in breast cancer patients breast cancer patients?

14 16 18 Non Responder Responder

ents

4 6 8 10 12

Number of Patie

2

N COXEN Score

70% 80% 90% 100% Non Responder Responder

f being a n-Responder

10% 20% 30% 40% 50% 60%

Likely hood of sponder or Non

0% 0.05 0.10 0.12 0.15 0.20 0.20 0.25 0.27 0.30 0.34 0.35 0.40 0.41 0.45 0.49 0.50 0.55 0.58

COXEN Score Res

slide-27
SLIDE 27

Validation: Validation: Can Can COXEN COXEN Algorithm Predict Algorithm Predict Clinical Responses in Clinical Responses in Ovarian cancer patients Ovarian cancer patients?

P-VALUE = 0.002 P-VALUE < 0.001

  • N=119 advanced-stage serous ovarian cancers
  • Treated with platinum-based chemotherapy
  • Reference: J Clin Oncol. 2007;25(5):517-25
  • N=85 advanced-stage serous ovarian cancers
  • Treated with neoadjuvant platinum-based

chemotherapy

  • Reference: PLoS ONE. 2007 May 16;2(5):e441
slide-28
SLIDE 28

Conclusion Conclusion

COXEN offers predictive ability for:

Known single and combination chemotherapeutic drugs in bladder cancer cell lines Treatment responses of known single and combination drugs in bladder cancer patients. Clinical outcomes for several major cancer types (breast, bladder and ovarian) Cli i l t di ti i il (b tt ?) th Clinical outcomes prediction similar (better?) than conventional (using patients) GEM derivation

slide-29
SLIDE 29

COXEN Applied to COXEN Applied to Drug Discovery Drug Discovery

COXEN can discover new drugs for bladder cancer that have a high

Discovery (2-10 Years) Preclinical Testing

COXEN COXEN

probability of working in patients.

  • Uses in vitro data

g Laboratory and animal testing Phase I Determine safety and dosage Phase II Efficacy and side effects

Uses in vitro data

  • Can identify drugs with

high likelihood of success in patients weeks

y Phase III Adverse reactions to long-term use FDA Review/Approval

patients weeks after initial synthesis with minimal cost!

Additional Post-market Testing Biomarker Development

2 4 6 8 10 12 14 16

Years Clinical use with Response Biomaker

slide-30
SLIDE 30

COXEN in drug discovery COXEN in drug discovery

Computational screening of 45,000 compounds Computational screening of 45,000 compounds

S

slide-31
SLIDE 31

COXEN in drug discovery COXEN in drug discovery

Screening results Screening results

Compound Library (N=45,678) COXEN Screening Compounds Effective in Bladder Cancer N=858 COXEN Score

1 858

Cisplatinum Carboplatin Adriamycin 5FU/ Pemetrexed

234 456

Rank: Drugs Currently Used in Bladder Cancer Carboplatin Gemcitabine Paclitaxel Methotrexate Vinblastine Drugs Currently Used in Bladder Cancer 233 Compounds better than Cisplatin in Bladder Cancer

slide-32
SLIDE 32

COXEN in drug discovery COXEN in drug discovery

Validation of screening results Validation of screening results

Identification of 115 novel putative anticancer compounds for human bladder cancer with COXEN SCORES > 90 NCI Repository Validation of NSC 637993 NCI Repository availability

  • f top 8 candidates

Top candidate: NSC 637993

COXEN Scores of NSC 637993 activity in the BLA-40

potency on BLA-40

In vitro evaluation of NSC 637993 activity in BLA-40 cells

compare

slide-33
SLIDE 33

Validation of new drug effectiveness in Validation of new drug effectiveness in human bladder cancer cells human bladder cancer cells

FL3

Sensitive

1 1 1 1 1 2 2 2 2 2 3 3 3 3 3 VMCUB2

Sensitive

1 1 1 1 1 2 2 2 2 2 3 3 3 3 3 HT1197

Sensitive

1 1 1 1 1 2 2 2 2 2 UMUC6

Sensitive

1 1 1 1 1 2 2 2 2 2 UMUC2

Resistant

1 1 1 1 1 2 2 2 2 2 UMUC9

Resistant

1 1 1 1 1 2 2 2 2 2

nts

VMCUB3

Sensitive

1 1 1 1 1 2 2 2 2 2 3 3 3 3 3 T24T 1 2 2 3 JON

Sensitive

1 1 1 1 1 2 2 2 2 2 3 3 3 3 3 BC16 1 1 1 2 2 UMUC14

Sensitive

1 1 1 1 1 2 2 2 2 2 MGH-U3 1 1 2 2 TCCSUP

Sensitive

1 1 1 1 1 2 2 2 2 2 CUBIII 1 1 2 2 253JLava

Resistant

1 1 1 1 1 2 2 2 2 2 UMUC13D 1 2 T24

Resistant

1 1 1 1 1 2 2 2 2 2 HT1376 1 1 2 2

NSC 637993

  • f cell coun

T24T

Sensitive

1 1 1 1 2 2 2 2 3 3 3 3 SLT4

Sensitive

1 1 2 2 3 3 BC16.1

Sensitive

1 1 1 2 2 2 HU456

Sensitive

1 1 1 2 2 MGH U3

Sensitive

1 1 1 2 2 2 2 MGH-U4

Sensitive

1 1 2 2 CUBIII

Sensitive

1 1 1 1 2 2 2 KU7

Resistant

1 1 2 2 UMUC13D

Resistant

1 1 1 1 2 2 2 2 253J-BV

Resistant

1 1 2 2 HT1376

Resistant

1 1 1 2 2 2 2 253J-P

Resistant

1 1 2 2

NSC 637993

Percent o

1 1 1 2 2 2 3 3 3 RT4

Sensitive

1 1 1 1 1 2 2 2 2 2 3 3 3 3 3 1 1 2 2 2 J82

Sensitive

1 1 1 1 2 2 2 1 1 1 2 2 2 UMUC3

Sensitive

1 1 1 2 2 2 1 1 1 2 2 2 UMUC1

Resistant

1 1 1 2 2 2 1 1 1 2 2 2 KK47

Resistant

1 1 1 2 2 2 1 1 1 2 2 2 CRL7833

Resistant

1 1 2 2 2

40 human bladder cancer cell lines (BLA-40)

20 60 100

575A

Sensitive

1 1 1 1 1 2 2 2 2 2 3 3 3 3 3 1 1 2 2 HTB9

Sensitive

1 1 1 1 1 2 2 2 2 2 1 1 1 2 2 2 VMCUB1

Sensitive

1 1 1 1 1 2 2 2 2 2 1 1 2 2 CRL2742

Resistant

1 1 1 1 1 2 2 2 2 2 1 1 2 2 2 SW1710

Resistant

1 1 1 1 1 2 2 2 2 2 1 1 1 2 2 2 CRL7193

Resistant

1 1 1 1 1 2 2 2 2 2

Log base 10 of molar concentration of NSC 598354

  • 8 -7 -6 -5 -4

637993

slide-34
SLIDE 34

Specificity (for bladder cancer) of 637993 Specificity (for bladder cancer) of 637993

Breast Leukemia Ovarian

NCI-60

! " " #

100

BLA-40

ent of control

! ! # # # $ $ $ $

50

CNS Melanoma Prostate

Perc

! ! " " " # # # $ $

log base 10 of molar concentration

  • 100
  • 50
  • 8
  • 7
  • 6
  • 5
  • 4

Renal NSLC Colon

Sensitive cell lines at dose concentration 10-6

slide-35
SLIDE 35

NSC NSC 637993 and C1311 637993 and C1311

NSC 637993 (CID 367849) No data in vivo or patients Dead end? Discussions with DTP staff led to chemists in Poland who described an entire family of compounds… C1311 (CID 132127) Analog of NSC 637993 Analog of NSC 637993 Top hit in COXEN screen member of imidazoacridinone anticancer drug family St t i l l l t d t it t d l t Structure is closely related to mitoxantrone and losoxantrone Orally bioavailable Effective in xenograft models of breast and colon cancer In Phase 2 trials in breast and colon cancer and IBD, MS

slide-36
SLIDE 36

NSC NSC 637993 and C1311 637993 and C1311

In vitro effect of C1311 on human bladder cancer cells human bladder cancer cells

80 100 HT1197

No drug

40 60 HT1197 253J‐BV KU7 UMUC6

ber at 48 hrs vs. N

20 0.001 0.01 0.1 1 10 100 0.001 0.01 0.1 1 10 100 253J‐P UMUC3 T24T

Percent Cell Num

In vivo attainable

C1311 NSC637993

Drug Concentration (μM)

concentration

slide-37
SLIDE 37

COXEN in drug discovery COXEN in drug discovery

Conclusion: COXEN can discover new drugs for g bladder cancer B t WAIT! t t h t t d But WAIT!........are not most human cancers treated with combination chemotherapy? Question: What do we need to figure out how to use new drugs in rational combinations? Answer: We need to know their mechanism of action and molecular target! action and molecular target!

slide-38
SLIDE 38

Mechanism of action of new drug Mechanism of action of new drug

Concept of Synthetic Lethality Concept of Synthetic Lethality

slide-39
SLIDE 39

Exploring the Mode Exploring the Mode-

  • of
  • f-
  • Action of bioactive compounds by

Action of bioactive compounds by Chemical Chemical-

  • Genetic Profiling and SGA in Yeast

Genetic Profiling and SGA in Yeast

Nature Methods, 2006, 3, 601-605

Yeast is in low abundance Yeast is in normal abundance

slide-40
SLIDE 40

Validating the Yeast Mode Validating the Yeast Mode-

  • of
  • f-
  • Action of in human bladder

Action of in human bladder cancer xenografts cancer xenografts

800 1000 600 800 No Drugs C1311[C]

ume (mm3)

200 400 C1311[C] Taxol[T] C+T

Tumor Volu

** * ** *

1 2 3 4 5 6 7 1 2 3 4 5 6 7 T24T UMUC3

Weeks from Subcutaneous Inoculation (μM)

slide-41
SLIDE 41

Plans for the 115 COXEN hits in bladder Plans for the 115 COXEN hits in bladder

115 Drug Leads

IC50 for Lead Human Bladder Cancer

575A

Sensitive

1 1 1 1 1 2 2 2 2 2 3 3 3 3 3

A

IC50 for Lead Compounds Human Bladder Cancer Cell Panel (BLA‐40)

B

Chemical Genomic Yeast Mutant panel

Clinical

Rational combinations

C

Chemical Genomic Profiling and SGA Yeast Mutant panel

Clinical Trials

Best Novel Agents

Xenograft response for Lead Compounds ADME / Tox

slide-42
SLIDE 42

Conclusion Conclusion

COXEN Biomarkers

Predicted clinical outcomes in patients in 4 cancer types Discovered new drugs in bladder cancer Predictions are equivalent to patient developed biomarker panels p Prediction for targeted agents are superior to target analysis readouts

and can do all this from ONLY IN VITRO DATA ....and can do all this from ONLY IN VITRO DATA

NCI-60 Panel Cell Lines

575A

Sensitive

1 1 1 2 2 2 3 3 3

t Control

NCI 60 Panel Cell Lines Gene expression profile

1 1 2 2 2 3 3 3

Dose Percen

Patient Tumor

slide-43
SLIDE 43

Clinical Applications Clinical Applications

“Personalized Therapy”: Match patient’s tumor with drug treatment

COXEN li d t ti t ti d t COXEN applied to patient tissue removed at surgery COXEN provides recommendations for:

Best FDA approved chemotherapy drugs and targeted agents Best FDA approved chemotherapy drugs and targeted agents Best drug combination regimen:

Established combinations: GC, MVAC etc… Novel combinations with FDA approved agents

Discovery of new compounds (and Drug Rescue) for most cancer types

By virtue of the algorithm design:

Discovered drugs should be effective in patients Improve compound attrition rate in clinical trials

Significantly reduced discovery timelines

slide-44
SLIDE 44

Clinical Applications Clinical Applications

Personalized Therapy Personalized Therapy

Pre Tx Tumor

FFPE T

Applications

Neoadjuvant d Harvest

Tumor

RNA Extraction Sample Profiling Adjuvant Metastatic Gene Expression Model (GEM) Score Calculation Sample Profiling MVAC Score GC Score Non Responder Responder Non Responder Responder

MVAC Therapy Option GC Therapy Option Is patient a GC Therapy Responder? Is patient a MVAC Therapy Responder?

Yes No Yes No

Empiric MVAC or GC Selection or Clinical Investigations (i.e. COXEN “Miniscreen”)

slide-45
SLIDE 45

Novel Trial Designs: 2 Birds 1 Stone Novel Trial Designs: 2 Birds 1 Stone

Personalized Therapy + New Drugs Evaluation Personalized Therapy + New Drugs Evaluation

R

A+B Standard R i

Outcome

Tumor

A N D O

Regimen C O M P COXEN GEM Sample A+B X+Y

O M I Z

O t

P A R E GEM Assigned Regimen New 1+2 New 5+6

E

Outcome

Stratification Factors

KPS: good (> 70) v poor (70) TNM staging: M0 v M1

  • Alk. phos. group: normal v high

Disease: measurable v nonmeasurable

Effective New Agents

Disease: measurable v nonmeasurable Number of sites: ≤3 v > 3 Visceral metastasis: no v yes

von der Maase, H. et al. J Clin Oncol; 23:4602, 2005

COXEN Drug Discovery Phase I/II Single Agent Trials

slide-46
SLIDE 46

Acknowledgments Acknowledgments

Theodorescu Lab

Paul Williams PhD Dima Havaleshko MD

Funding

NCI Sooyoung Cheon PhD Michael Harding PhD Yimin Wu PhD

Pathology

Henry Frierson Chris Moskaluk

Yeast Biology

D B k Chris Moskaluk

Bioinformatics and Statistics

Jae Lee Mark Conaway Dan Burke Stefan Bekiranov

Small Animal Imaging

Mark Conaway Stuart Berr

Supercomputer Center

Andrew Grimshaw John Karpovich

slide-47
SLIDE 47

Thank you to my PMH/ Thank you to my PMH/OCI OCI teachers and mentors teachers and mentors

slide-48
SLIDE 48

Liberty is to the collective body what health is to Liberty is to the collective body, what health is to every individual body…….Without health no pleasure can be tasted by man…….without liberty, no happiness can be enjoyed by society.

Thomas Jefferson

Founder University of Virginia