Prognostic Markers can Guide Therapy? Alan K Burnett School of - - PowerPoint PPT Presentation

Prognostic Markers can Guide Therapy?

Alan K Burnett School of Medicine Cardiff University

NO(t yet)

Prognostic is Not Necessarily Predictive

Prognostic Factor Evolution

Multivariate analysis Validation on independent data Test intervention in randomised setting PREDICTIVE

Issues

• Is it independent and validated?
• Is the assay reliable?
• Is it treatment dependent?
• Is it practical e.g number of patients / cost
• Does it matter?
• Does is have consequential implications?
• Is the intervention validated?

Prognostic Factors

• Age
• Cytogenetics
• Mutation status
• Response to course 1
• Secondary disease
• WBC
• Male gender
• Performance score
• Co-morbidity
• LDH
• MRD

Impact of additional cytogenetic abnormalities: Survival of patients entered into AML 10 & 12 (n=3453)

100 50 75 25 1 2 3 4 5 Years from entry % still alive

Favourable only (n=478) Favourable + intermediate (n=331) Favourable + adverse (n=22) Intermediate only (n=2235) Adverse + intermediate (n=297) Adverse only (n=478)

71% 65% 59% 42% 17% 14%

Prognostic Factors Mutation status

• FLT3 ITD/TK
• NPM1
• cKIT
• WT1
• CEBPα (double mutant)
• EVI1
• MLL
• IDH1/IDH2
• Expression (BAALC/MN1)

Results of Cox regression

• The following variables are significant (in order of entry to the

model):

Variable Estimate χ2 p-value Cytogenetics 0.65082 102.7 <0.0001 Age 0.01325 29.16 <0.0001 Status post C1 0.19529 18.50 <0.0001 WBC 0.00169 11.92 0.0006 Male sex 0.16994 8.01 0.005 Secondary 0.22131 4.03 0.04

Risk Index

• The risk index for survival from CR is

therefore: 0.01325*age (in years) + 0.16994*sex (1=male, 0=female) + 0.22131*diagnosis (0=de novo, 1 secondary) + 0.65082*cytogenetics (1=favourable, 2=intermediate, 3 adverse) + 0.19529*status post C1 (1=CR, 2=PR, 3=NR) + 0.00169* WBC (x109/l)

Risk Index properties (adults)

Maximum Value 4.16 Minimum Value 1.27 Low index = low risk Appears to be evidence of two distinct populations Therefore cut the data at index=2 (about 15th percentile) Choose 75th centile for

• ther cut after looking at

splitting into four groups, so cut at 2.6667

1.50 2.25 3.00 3.75

icrnoapl

50 100 150

C

• u

n t

Outcome by index group

Numbers in Each Group

MRC Good MRC Standard MRC Poor Total New good 309 28 337 New standard 51 1289 42 1382 New poor 2 274 353 629 Total 362 1591 395 2347

Intermediate Risk: M-B Analysis

Poor Risk: M-B Analysis

What about Molecular Direction?

Prognostic Factors Mutation status

• FLT3 ITD/TK

30%

• NPM1`

25% (L)

• cKIT

25% (L)

• WT1

8-10%

• CEBPα (double mutant)

2%

• EVI1
• MLL
• IDH1/IDH2
• Expression (BAALC/MN1)

Survival by FLT3 Mutant Ratio

OS by ITD and NPM1

Intermediate Risk NPM1/FLT3 Genotype: Transplant

Transplant v not – OS (age <50)

Risk Group NPM1/FLT3 Genotype

Excluding the good risk genotype

Bad genotype – standard risk

Comparison of Bad Genotype who are Standard Risk.

Good/Standard Poor P-value Age (mean) 37 41 0.2 Sex (male %) 49% 79% .007 WBC (median) 24.1 88.5 .003 Secondary (%) 2% 17% .006 Adverse Cytos 0% 17% <.0001

Comparison of Good Genotype who are Poor Risk.

Good/Standard Poor P-value Age (mean) 41 52 <.0001 Sex (male %) 37% 70% .0002 WBC (median) 18.2 77.4 <.0001 Secondary (%) 2% 16% .0002 Adverse Cytos 0% 8% .0008

Regression Equation

Index= 0.707 * cytogenetics (1=interm./favourable, 2=adverse) + 0.275 * WBC group (1=<10.0, 2=10.0-99.9, 3=100+) + 0.191 * WHO PS + 0.0315 * age + 0.336 * AML type (1=de novo, 2=secondary)

AML14: Overall survival

What about MRD?

What about MRD?

• Specificity
• Reproducability
• Time point vs sequential
• Intervention implications
• Direct /indirect consequences
• How to validate the clinical value

Summary

• Not all prognostic factors are validated.
• A few change therapy e.g cytogenetics/

marrow response after induction/ FLT3 NPM1 mutations.

• Several are prognostic but only in a small

population

• While prognostic we need to do the

studies for predictive validation