Molecular monitoring of CML patients

EHA, Education Session, CML Stockholm, 14 June 2013 Molecular monitoring of CML patients Martin C. Müller Medical Faculty Mannheim Ruprecht-Karls-University Heidelberg Mannheim, Germany

Disclosures Research Support: Advisory Boards: Speaker s Bureau: Novartis, Bristol-Myers Squibb, ARIAD Novartis, Bristol-Myers Squibb, ARIAD Novartis, Bristol-Myers Squibb, ARIAD

Learning goals after this session you should be able to describe the benefits and limitations of molecular monitoring in CML patients interpret molecular results in terms of prognostic significance discuss the importance of harmonization of technologies according to the International Scale (IS) reasonably use derived techniques in case of resistance or unsatisfying response

Molecular Monitoring: What to measure, what to treat? BCR BCR-ABL BCR-ABL Philadelphia Chromosome ABL mrna Transcript Active Protein Quantitative Determination of BCR-ABL Transcripts using quantitative PCR (RQ-PCR) Inhibition of TK by TKI Adapted from Mensink et al. Br J Haematol 1998;102:768.

GCT9310 Monitoring Ph+ CML Disease Hematologic response (HR) Measure of blood counts and differentials Cytogenetic response (CyR) Chromosome banding analysis of marrow cell metaphases Molecular response Measurement of BCR-ABL transcript levels relative to a control gene Most sensitive measure of Ph+ CML disease burden 1 2 3 1. Maslak P. ASH Image Bank. Oct 2008. 8-00067. http://ashimagebank.hematologylibrary.org/. Copyright American Society of Hematology 2008. 2. Quintás-Cardama A, et al. Leukemia. 2007; 21(11): 2394-2396. Reprinted by permission from Macmillan Publishers Ltd: Leukemia. 2007;21: 2394-2396, copyright 2007. 3. Maslak P. ASH Image Bank. Dec 2001. 100202. http://ashimagebank.hematologylibrary.org/. Copyright American Society of Hematology 2001. 5

Monitoring of CML: European LeukemiaNet Recommendations* GCT9310 Hematologic response (HR) Cytogenetic response (CyR) Molecular response (MR) Every 15 days until CHR Month 3 and 6 and every 6 months until CCyR Every 3 months until MMR Every 3 months Every year Every 6 months CHR, complete hematologic response; CCyR, complete cytogenetic response; MMR, major molecular response. *Baccarani M, et al. J Clin Oncol. 2009;27:6041-6051. 6

Correlation Between Response and Disease Burden: Molecular Response Time Baccarani M, et al. J Clin Oncol. 2009;27(35):6041-6051. Radich JP. Blood. 2009;114:3376-3381.

Genetics of Chronic Myeloid Leukemia Chromosome 22 Chromosome 9 e1 1b m-bcr M-bcr e1 e2 b1 b5 5 3 BCR ABL 5 3 1a a2 a3 m -bcr e19 a11 e1a2 b2a2 b3a2 e19a2 p190 bcr-abl p210 bcr-abl p230 bcr-abl

Multiplex PCR Qualitative PCR before quantitative PCR! Please ensure that you know the transcript type of your patients! Only b3a2 and/or b2a2 transcripts can be compared according to the IS bp ladder b3a2 b2a2 b3a2, b2a2 e1a2 b2a3 bp ladder Cross et al., 1993

Molecular Analyses Quantitative Real-Time RT-PCR using LightCyclertechnology according to the International Scale (IS; Hughes et al., Blood 2006). Other housekeeping genes possible most appropriate: ABL, GUS, B2M, BCR Beillard et al., Leukemia 2003

Important for prognosis. - Depth of response - Time to achieve response

ELN recommendations 2009 on treatment and diagnostics of CML Optimal Response Suboptimal Response Failure Warning Baseline N/A N/A N/A N/A 3 Month CHR + mcyr (Ph+ < 65%) No CyR (Ph+ > 95%) No CHR High Risk CCA/Ph+ 6 Month PCyR (Ph+ < 35%) Less than PCyR (Ph+ > 35%) No CyR (Ph+ > 95%) N/A 12 Month CCyR (Ph+ = 0%) PCyR (Ph+ < 35%) Less than PCyR (Ph+ > 35%) N/A 18 Month MMR Less than MMR Less than CCyR (Ph+ > 0%) Less than MMR Anytime Stable/ Improving MMR Loss of MMR Mutations* Loss CHR Loss CCyR Mutations* CCA/Ph+ Molecular endpoints do not play a major role so far Baccarani M et al. JCO 27:6041 (2009).

Equivalence of cytogenetic with molecular endpoints? - CCyR 1% BCR-ABL IS - PCyR 10% BCR-ABL IS

New milestones at 6 months

Overall survival by CCyR at 6 months MD Anderson Cancer Center Jabbour E et al. Blood 2011;118:4541-4546 2011 by American Society of Hematology

Overall survival by CCyR at 6 months German CML-Study IV Hanfstein B, et al. Leukemia. 2012 Sep;26(9):2096-102.

Overall Survival (OS) BCR-ABL IS at 6 months 1% vs. 1-10% vs. >10% Hanfstein B, et al. Leukemia. 2012 Sep;26(9):2096-102.

New milestones at 3 months

Predictive value of BCR-ABL level at 3 months on Overall Survival: Cut off IS ~10% BCR-ABL/ABL<9.8% OS = 93.3% (n = 211) P <.001 BCR-ABL/ABL>9.8% OS = 56.9% (n = 68) Marin et al. J Clin Oncol. 2012 Jan 20;30(3):232-8 (adapted)

Overall Survival (OS) by BCR-ABL IS at 3 months 10% vs. >10% 10% >10% BCR-ABL IS n 5Y-OS 10% 501 95% >10% 191 87% p-value <0.001 Hanfstein B, et al. Leukemia. 2012 Sep;26(9):2096-102.

Overall Survival (OS) by Ph+ at 3 months 35% vs. >35% 35% >35% Ph+ n 5Y-OS 35% 336 95% >35% 124 87% p-value 0.036 Hanfstein B, et al. Leukemia. 2012 Sep;26(9):2096-102.

Patients alive (%) Overall Survival According to BCR-ABL Levels at 3 Months (DASISION Landmark Analysis) 100 dasatinib 100 mg QD 84% had 10% BCR-ABL 100 imatinib 400 mg QD 64% had 10% BCR-ABL 80 80 60 40 20 0 BCR-ABL level at 3 months 3-Year OS 1% 95.5% } P=.7342 >1-10% 96.5% } P=.0525 >10% 85.9% 10% vs >10% P=.0348 0 6 12 18 24 30 36 Months Subjects at risk 1% 112 112 110 109 106 104 85 29 >1-10% 86 85 84 83 83 79 66 25 >10% 37 37 35 34 33 27 22 9 42 60 40 20 0 BCR-ABL level at 3 months 3-Year OS 1% 100% } P=.3453 >1-10% 95.0% } P=.0110 >10% 87.8% 10% vs >10% P=.0036 0 6 12 18 24 30 36 42 Months Subjects at risk 1% 32 32 32 32 31 30 28 11 >1-10% 122 121 120 118 118 116 96 33 >10% 85 85 82 80 76 70 55 20 Saglio G, et al. Blood. 2012;120(21):[abstract 1675].

% Alive % Alive Overall Survival at 4 Years by BCR-ABL IS Levels at 3 Months (ENESTnd Landmark Analysis) nilotinib 300 mg BID 97.2% 96.5% 86.7% 100 90 80 70 OS rates at 4 years for 10% vs >10% 60 BCR-ABL at 3 months: 50 96.7% vs 86.7%; P =.0116 40 Pat Evt Cen 30 1% 145 5 140 20 >1% 10% 89 2 87 >10% 24 3 21 10 Censored observations 0 0 6 12 18 24 30 36 42 48 54 60 Time Since Randomization (Months) P =.6348 P =.0219 imatinib 400 mg QD 100 90 80 70 OS rates at 4 years for 10% vs >10% 60 BCR-ABL at 3 months: 50 98.9% vs 83.6%; P <.0001 40 Pat Evt Cen 30 1% 43 2 41 20 >1% 10% 133 1 132 >10% 88 14 74 10 Censored observations 0 0 6 12 18 24 30 36 42 48 54 60 Time Since Randomization (Months) 100% 95.3% 83.6% P =.0982 P <.0001 Evaluable patients Nilotinib 300 mg BID (N = 258) Imatinib 400 mg QD (N = 264) BCR-ABL at 3 months 1% N=145 >1 10% N=89 >10% N=24 1% N=43 >1 10% N=133 >10% N=88 % of patients 56% 34% 9% 16% 50% 33% Data cut-off: 27Jul2012. Hochhaus A, et al. Blood. 2012;120(21):[abstract 0167].

Evolving recommendations of the ELN Optimal Response Suboptimal Response Failure Warning Baseline N/A N/A N/A N/A 3 Month CHR + mcyr (Ph+ < 65%) No CyR (Ph+ > 95%) No CHR High Risk CCA/Ph+ 6 Month PCyR (Ph+ < 35%) Less than PCyR (Ph+ > 35%) No CyR (Ph+ > 95%) N/A 12 Month CCyR (Ph+ = 0%) PCyR (Ph+ < 35%) Less than PCyR (Ph+ > 35%) N/A 18 Month MMR Less than MMR Less than CCyR (Ph+ > 0%) Less than MMR Anytime Stable/ Improving MMR Loss of MMR Mutations* Loss CHR Loss CCyR Mutations* CCA/Ph+ Baccarani M et al. JCO 27:6041 (2009).

Evolving recommendations of the ELN Optimal Response 2009* Optimal Response 2013** Baseline 3 Month 6 Month 12 Month N/A CHR + mcyr (Ph+ < 65%) PCyR (Ph+ < 35%) CCyR (Ph+ = 0%) PCyR (Ph+ 35%) and/or 10% BCR-ABL IS CCyR (Ph+ 0%) and/or 1% BCR-ABL IS 0.1% BCR-ABL IS (MMR) 18 Month MMR Anytime Stable/ Improving MMR *Baccarani M et al. JCO 27:6041 (2009) * * Baccarani M et al. Blood 2013 (accepted for publication on 11 June 2013)

Standardization - why? Standardized BCR-ABL monitoring has shown to predict for response Hughes et al., NEJM 2003; Hughes et al., Blood 2010 Hanfstein et al., Leukemia 2012; Marin et al., JCO 2012 Molecular endpoints in recent clinical studies Saglio et al., NEJM 2010; Kantarjian et al., NEJM 2010 Huge variety of molecular methods leads to considerable variation of results Noncomparability of results Confusion of doctors and patients Suboptimal treatment decisions

The International Scale (IS) - Background Historical (IRIS trial; 2000): The mean BCR-ABL levels of 30 CML patients was defined as 100% in each of the three participating laboratories using BCR as a control. The value corresponding to MMR in each laboratory defined as 0.1% (reduction of 3 log from IRIS baseline, not from pre-treatment levels in the individual patient!)

Proposal of international CML experts for the expression of results in 2006 We propose that the value corresponding to a MMR in each laboratory be mathematically converted to: 0.10% Therefore the International Scale will be fixed to an absolute value as defined in the IRIS trial BCR-ABL local x CF local = BCR-ABL IS Hughes et al., Blood 2006

BCR-ABL BCR-ABL levels in 64 labs pre and post conversion CV 1.32 1.30 1.10 1.02 0.20 0.24 0.31 0.41 1000 100 10 1 0.1 MMR 0.01 0.001 0.0001 Level 1 Level 2 Level 3 Level 4 Level 1 Level 2 Level 3 Level 4 reference results Müller et al., Leukemia 2009 BCR-ABL L pre conversion BCR-ABL IS post conversion

Standardization of BCR-ABL quantification in Europe 2013 64 participating laboratories in 28 countries Iceland Sweden Finland Norway Estonia Portugal Ireland Spain Denmark Netherlands United Kingdom France Belgium Luxembourg Switzerland Germany Czech Rep Poland Slovakia Austria Hungary Slovenia Croatia Italy Latvia Lithuania Bosnia Serbia and Herzegovina. Albania Belarus Romania - Greece Bulgaria Ukraine Moldova Macedonia Montenegro Russia Turkey Countries with a ref lab with validated CF Countries using a ref lab in a different country Countries with a ref lab validation pending Israel

BCR-ABL How to standardize for deeper molecular responses? CV 1.32 1.30 1.10 1.02 0.20 0.24 0.31 0.41 1000 100 10 1 0.1 0.01 0.001 MMR MR 4 MR 5 0.0001 Level 1 Level 2 Level 3 Level 4 Level 1 Level 2 Level 3 Level 4 reference results Müller et al., Leukemia 2009 BCR-ABL L pre conversion BCR-ABL IS post conversion

Do we need to qualify for deep molecular responses? Isn t it enough to get a negative PCR result? BCR-ABL negative sample due to bad sample quality / low sensitivity?

Importance of reliable measurement of deep molecular responses Expected that more and more patients under first-line treatment with secondgeneration TKIs (Nilotinib, Dasatinib) will achieve deep molecular response Justification for treatment discontinuation

In order to safely stop TKI treatment PCR results need to be comparable (aligned to the International Scale, IS) very sensitive (low sensitivity leads to high number of false negatives)

Complete molecular response Definition for the ENEST1st trial Disease (detectable or undetectable) below a defined sensitivity level 4 log below standardized IRIS baseline

Definitions of Molecular Response Minimum sample quality required 100% [IRIS baseline] MR 4.0 ( 4 log reduction; 0.01% IS ) 10,000 ABL 10% copies 1% MR 4.5 ( 4.5 log reduction; 0.0032% IS ) 32,000 ABL 0.1% copies [IRIS MMR] 0.01% MR 5.0 ( 5 log reduction; 0.001% IS ) log reduction = reduction from IRIS baseline, not individual pretreatment levels 100,000 0.001% ABL copies BCR-ABL undetectable International Scale Cross et al. Leukemia. 2012

It s all about Sensitivity Improvement of sensitivity main factors: Number of leukocytes to use (10-20 Mio.) Please ensure that your lab provides you with RNA extraction method (Trizol vs kits) ABL values as a measure for sensitivity! cdna synthesis protocol (use 5µg instead of 1µg RNA) Optimize type and amount of RT enzyme No dilution of cdna

ELN Recommendations: Mutation Analysis in Case of Failure and Suboptimal Response Baccarani M, et al. J Clin Oncol 2009;27:6041 6051. Time Failure Suboptimal Response Warnings Diagnosis - -...and before changing to other BCR-ABL inhibitors or other therapy 3 mos No CHR no CyR High risk CCA in Ph+ cells 6 mos < CHR No CyR < PCyR 12 mos < PCyR < CCyR < MMR...and experiencing a 5 to 10-fold BCR-ABL increase while losing MMR 18 mos < CCyR < MMR Anytime CCA in Ph+ cells Loss of CHR Loss of CCyR Mutation (IMinsensitive) Loss of MMR Mutation (IMsensitive) Any transcript level CCA in Ph- cells

Most relevant BCR-ABL mutations Less sensitive to nilotinib Less sensitive to dasatinib Specific mutations Cortes J, et al. Leukemia 2008;22:2176 2183. le Coutre P, et al. Blood 2008;111:1834 1839. Hughes T, et al. J Clin Oncol 2009;27:4204 4210. Apperley JF, et al. J Clin Oncol 2009;27:3472 3479. Müller MC, et al. Blood 2009;114:4944 4953. Soverini S, et al. Blood 2011;118:1208-1215. Y253H E255K/V F359C/V T315I Q252H V299L T315A F317L Technical sensitivity of Sanger sequencing ~20%

Diagnostic requirements for management of TKI failure Check compliance, ask for side effects Blasts/precursors in PB/BM? AP/BC? BM aspiration, clonal cytogenetic evolution? PB/BM: BCR-ABL mutations? Relevant co-medication? (e.g. Medscape App, Interaction checker) Plasma level?

Conclusions (1/2) Ask for: transcript type (once per patient) ABL counts (sensitivity) lab is harmonized according to IS? Early deep responses predict for PFS and OS (any TKI), determined by Molecular Monitoring (and/or Cytogenetics) New ELN recommendations take into account molecular milestones for optimal response and failure, e.g. Optimal at 3 months: 10% BCR-ABL IS and/or PCyR (Ph+ 35%) Optimal at 6 months: 1% BCR-ABL IS and/or CCyR (Ph+ 0%) Molecular Monitoring is more sensitive and allows an early detection of potential relapse (or non-compliance) Can be performed using PB instead of BM

Conclusions (2/2) New Goal : Achievement of very deep molecular responses (MR 4, MR 5 ) allowing to consider treatment interruptions within clinical trials (STIM, EUROSKI, ENEST Path) Cure? Standardization of molecular methods is the prerequisite for comparability of results and represents a challenge for interested laboratories due to the permanent need to validate their conversion factor and improve sensitivity TKI failure: Disease phase? Blasts/precursors Cytogenetics: additional chromosomal abnormalities BCR-ABL mutation analysis Compliance? Side effects? Comedication/interactions?

Thank you! martin.mueller@umm.de