The Journal of Molecular Diagnostics, Vol. 15, No. 5, September 2013

jmd.amjpathol.org

REVIEWBCR-ABL1 RT-qPCR for Monitoring the Molecular Responseto Tyrosine Kinase Inhibitors in Chronic Myeloid LeukemiaRichard D. Press,* Suzanne Kamel-Reid,y and Daphne Ang*

From the Department of Pathology and Knight Cancer Institute,* Oregon Health & Science University, Portland, Oregon; and the Department of Pathology,y

The University Health Network, Toronto, Ontario, Canada

Accepted for publication

C

a

P

h

April 8, 2013.

Address correspondence toRichard D. Press, M.D., Ph.D.,Department of Pathology,L113, Oregon Health & ScienceUniversity, 3181 SW SamJackson Park Rd, Portland,OR 97201. E-mail: pressr@ohsu.edu.

opyright ª 2013 American Society for Inve

nd the Association for Molecular Pathology.

ublished by Elsevier Inc. All rights reserved

ttp://dx.doi.org/10.1016/j.jmoldx.2013.04.007

The pathognomonic genetic alteration in chronic myeloid leukemia is the formation of the BCR-ABL1 fusion gene, which produces a constitutively active tyrosine kinase that drives leukemictransformation. Targeted tyrosine kinase inhibitor treatment with imatinib, nilotinib, dasatinib,bosutinib, and ponatinib is the cornerstone of modern therapy for this hematologic malignancy.Real-time quantitative RT-PCR (RT-qPCR, also RQ-PCR) of BCR-ABL1 RNA is a necessary laboratorytechnique for monitoring the efficacy of tyrosine kinase inhibitor therapy and quantitativelyassessing minimal residual disease. The molecular response measured by BCR-ABL1 RT-qPCR assistsin identifying suboptimal responses and can help inform the decision to switch to alternativetherapies that may be more efficacious (or to pursue more stringent monitoring). Furthermore, thetyrosine kinase inhibitoremediated molecular response provides valuable risk stratification andprognostic information on long-term outcomes. Despite these attributes, informed, universal,practical utilization of this well-established monitoring test will require heightened efforts by themolecular diagnostics laboratory community to adopt the standardized reporting units of theInternational Scale. Without widespread adoption of the International Scale, the consensus majormolecular response and early molecular response treatment thresholds will not be definable, andoptimal clinical outcomes for patients with chronic myeloid leukemia may not be achieved.(J Mol Diagn 2013, 15: 565e576; http://dx.doi.org/10.1016/j.jmoldx.2013.04.007)

Supported by Novartis Pharmaceuticals Corporation for medical editorialassistance.

Disclosures: R.D.P. has received honoraria from and served on advisoryboards for Novartis and Ariad, has received grant support from Cepheid,and has received honoraria and grant support from and serves on anadvisory board for Asuragen. S.K.-R. has received honoraria and grantsupport from Novartis and Bristol-Myers Squibb.

Virtually all patients with chronic myeloid leukemia (CML),a hematologic cancer characterized by the overproductionof immature and mature myeloid cells in the peripheralblood, spleen, and bone marrow, carry the Philadelphiachromosome (Ph), a reciprocal translocation between theAbelson gene (ABL1) on chromosome 9 and the breakpointcluster region gene (BCR) on chromosome 22.1,2 Theresulting fusion gene, BCR-ABL1, produces a constitutivelyactive chimeric tyrosine kinase that is required to initiate,propagate, and maintain the leukemic phenotype in patientswith CML. Tyrosine kinase inhibitor (TKI) therapy targetsthe BCR-ABL1 kinase and over the past decade has becomethe recommended first-line treatment approach for patientsnewly diagnosed in the initial chronic phase (CP) of CML.The success of TKI therapy in CML has furthermorebecome the scientific paradigm for molecularly targetedtherapy of other cancers in the 21st century.3

stigative Pathology

.

Since the advent of BCR-ABL1 TKIs, patients with CMLhave had significantly improved prognosis, as reflected bylonger median overall survival (OS) and lower rates ofdisease progression, compared with previous therapeuticregimens.4e6 Imatinib was the first TKI approved by the USFood and Drug Administration for CML, based on the phase3 International Randomized Study of Interferon and STI571(IRIS) clinical trial. Long-term follow-up of CML-CPpatients treated with first-line imatinib showed a progres-sion-free survival (PFS) rate of 93% at 5 years7 and an

Press et al

estimated OS rate of 85% at 8 years (93% when only CML-related deaths were considered).8

Although most patients treated with imatinib havedurable responses, resistance to imatinib does occur in somepatients, most commonly through development of mutationsin the BCR-ABL1 kinase domain that abrogate imatinib-induced inhibition.9e12 Dasatinib and nilotinib, morepotent TKIs than imatinib, were initially approved by theFood and Drug Administration for the treatment of CMLafter imatinib failure or intolerance. More recently, bothdasatinib and nilotinib have demonstrated, in separate phase3 trials, superior response rates and higher rates of PFS asfirst-line therapy, compared with imatinib.13e17 In addition,rates of major molecular response (MMR) were significantlyhigher with dasatinib or nilotinib versus imatinib over eachof 3 years of follow-up.13,14,16,17 Two other TKIs, bosutiniband ponatinib, were recently approved for second- and third-line treatment of CML.18,19

Targeted therapy with TKIs dramatically reduces theleukemic cell burden in CML.20e25 The low levels ofminimal residual disease (MRD) typically achieved duringTKI therapy demand a sensitive monitoring assay for reliabledetection and quantitation.26 Real-time quantitative RT-PCR(RT-qPCR, also RQ-PCR) has proven to be an effective andclinically validated laboratory method to quantify BCR-ABL1transcript levels and assess molecular responses. Further-more, given the predictive value of molecular response forlong-term clinical outcomes and as a direct biomarker fordrug efficacy and response, RT-qPCR has become an integralcomponent of CML disease management.22e25

As BCR-ABL1 RT-qPCRebased molecular monitoringevolves, a number of factors have been identified aspotential sources of confusion, thus limiting this assay’spractical clinical utility. In the quest to improve the labo-ratory utility of CML molecular monitoring, the followingquestions need to be addressed:

1. What are the clinically relevant BCR-ABL1 thresholdlevels for predicting TKI treatment responses and long-term PFS?

2. What are the clinically relevant BCR-ABL1 threshold levelsfor predicting suboptimal response, loss of response, andevolving drug resistance?

3. What laboratory-specific factors limit the precision,accuracy, and sensitivity of the BCR-ABL1 RT-qPCRassay?

4. How can the BCR-ABL1 RT-qPCR assay be standardizedto the International Scale (IS) of measurement by whichclinically relevant threshold levels have already beendefined?

This review discusses the current use of molecularmonitoring in the management of patients with CMLreceiving TKIs, its practical utility for measuring treatmentresponse (and nonresponse), and the remaining challengesfor improving molecular monitoring. The lessons learnedfrom the extensive research on molecular monitoring of

566

BCR-ABL1 will benefit, not only patients with CML, butalso patients with other cancers, for which a variety ofmolecularly targeted therapies and companion diagnosticsare now becoming available.

What Are the Clinically Relevant BCR-ABL1Threshold Levels for Predicting TKI TreatmentResponses and Long-Term PFS?

Monitoring Response to Therapy

The clinical utility of laboratory methods for monitoring TKItherapeutic efficacy, including hematologic, cytogenetic, andPCR-based techniques, depends largely on their limits ofdetection. Complete hematologic response is based onnormalization of peripheral blood counts [NationalComprehensive Cancer Network (NCCN) Clinical PracticeGuidelines in Oncology (NCCN Guidelines) ChronicMyelogenous Leukemia Version 3.2013. http://www.nccn.org/professionals/physician_gls/pdf/cml.pdf, last accessedFebruary 24, 2013; user registration required].27 Cytogeneticresponse is based on the percentage of Ph-positive (Phþ)metaphase cells observed in a bone marrow sample, typicallyby Giemsa staining of metaphase chromosome spreads.Complete cytogenetic response (CCyR), a clinically impor-tant prognostic response threshold in patients with CMLreceiving TKIs, is defined as the absence of detectable Phþ

chromosomes (NCCN Guidelines).28,29 Because a minimumof 20 metaphase cells is recommended for cytogenetic eval-uation (NCCN Guidelines), the limit of detection of thistechnique is relatively low (1:20, or 5%). Another limitationof cytogenetic testing is the need for an invasive bone marrowbiopsy to obtain culturable metaphase cells. Cytogenetictesting does have, however, the capability to detect otherchromosomal abnormalities besides the Ph chromosome, andthis may have prognostic relevance. Ancillary interphase cellfluorescence in situ hybridization (FISH), with DNA probesfor BCR and ABL1, is also commonly used to monitor CMLresponse to treatment. Because 100 to 500 interphase cells aretypically examinedwith FISH, this approach ismore sensitive(1:500 to 1:100, or 0.2 to 1%) than metaphase cytogenetics.Depending on the probes used, however, it can have a highfalse-positive rate. FISH lacks practical utility for MRDmonitoring in the majority of TKI-treated CML patients, andtherefore is not recommended by the NCCN for routinemonitoring of TKI treatment response (NCCN Guidelines).Instead, guidelines from the NCCN, the European Leu-

kemiaNet (ELN),27 and the National Institutes of Health21

recommend serial BCR-ABL1 RT-qPCR assays at regular3- to 6-month intervals for routine MRD monitoring ofCML patients receiving TKI therapy. Molecular monitoringinvolves extraction of RNA from a bone marrow orperipheral blood specimen and subsequent RT-qPCR tomeasure transcript levels of BCR-ABL1 relative to those ofa reference gene. Because the analyte of RT-qPCR, RNA, islabile and degradation prone, many pre-analytical variables

jmd.amjpathol.org - The Journal of Molecular Diagnostics

Monitoring Molecular Response in CML

can affect sample quality and quantitative results, includingsample source (peripheral blood versus bonemarrow), samplestorage (temperature, type of tube), sample transport (durationof travel, temperature), and sample stabilization (lysis buffer,storage buffer). The reference gene in RT-qPCR serves asa control for overall RNA quality (with respect to degrada-tion) and, assuming equivalent reference gene expression inall hematopoietic cells, for the number of input cells per PCRreaction. For example, in our laboratory, when the level ofreference gene RNA falls two SD below the mean, RT-qPCRfor the BCR-ABL1 and reference gene are repeated, and if thisrepeat analysis again shows low reference gene RNA levels,the sample is reported as inadequate (new sample requested).Compared with cytogenetic testing, PCR-based molecularmonitoring offers exquisite analytical sensitivity, 100 to 1000times greater than FISH or bonemarrow cytogenetic analysis.It is applicable to bone marrow and peripheral blood samples,andwith a short turnaround time, provides quantitative resultsthat are associated with validated clinical response thresh-olds.30 Disadvantages of RT-qPCR include its lack of meth-odological and reporting standardization, its need forspecialized laboratories and equipment, and the variability ofanalytical and reporting systems.31 Table 1 summarizescurrent NCCN testing recommendations for monitoring TKItreatment response in patients with CML.

Key Levels of Molecular Response

Response to TKI therapy is assessed by the change in BCR-ABL1 transcript levels from a standardized pretreatmentbaseline level. The standardized baseline, defined as 100% on

Table 1 Recommended Testing Parameters for Patients on TKI Therap

Test Recommended test specifications

Bone marrowcytogenetics

Analyze �20 metaphase cells

FISH Use peripheral bloodUse dual probes for BCR and ABL1 genes

RT-qPCR Use an RT-qPCR (IS) assay with sensitivityof �4.5 log below the standardized baselin

Use peripheral blood or bone marrow

BCR-ABL1 KDmutational analysis

None provided

*Level of response that warrants further patient evaluation and considerationCCyR, complete cytogenetic response; CML, chronic myeloid leukemia; FISH, flu

MMR, major molecular response.

The Journal of Molecular Diagnostics - jmd.amjpathol.org

the IS, was established in the IRIS study of imatinib treatmentin which the three participating laboratories independentlydetermined the median BCR-ABL1 RT-qPCR value of thesame set of 30 CML-CP samples collected before imatinibtreatment initiation.32 Thus, 100% BCR-ABL1 (IS) becamethe standardized baseline against which future BCR-ABL1RNA levels were compared, and any treatment-associatedreduction in BCR-ABL1 would represent an absolute reduc-tion that is independent of individual patient-specificBCR-ABL1 baseline values. MMR, a �3-log reduction inBCR-ABL1 transcript levels from this standardized baseline,is defined as �0.1% on the IS (NCCN Guidelines), andrepresents a clinically relevant level of molecular response,based on the observation that patients who achieved MMRhad a minimal risk of disease progression and improved long-term PFS compared with patients who did not.32

Achievement of MMR has been incorporated into ELNpractice guidelines to define an optimal level of response at12 to 18 months (Table 2),27,33 and failure to achieve thismolecular response milestone is considered a suboptimalresponse that calls for careful reevaluation of the currenttherapeutic regimen.27,33 Analysis of IRIS data at 7 yearsdemonstrated that patients who had achieved MMR by 12 or18 months had significantly superior event-free survival andOS, and more durable CCyR, compared with those who hadnot.24 Another study showed that patients on imatinib with�2-log reduction in BCR-ABL1 levels at the time of CCyRhad longer PFS than patients not achieving that level ofmolecular response.25 In the German CML-Study IV ofpatients on imatinib, MMR at 12 months predicted signifi-cantly better PFS and OS at 3 years.23 Further analysis of this

y per NCCN Guidelines

Recommended test frequency

At diagnosisAt 3 months, if RT-qPCR (IS) is not availableAt 12 months, if neither CCyR nor MMR is achievedAt 18 months, if no MMR and no CCyR at 12 months�1-log increase in BCR-ABL1 level without MMR

At diagnosis, if collection of bone marrow is not feasibleNot recommended for monitoring response to treatment

eAt diagnosis

Every 3 months for responding patientsAfter CCyR is achieved, every 3 months for 3 years, thenevery 3 to 6 months thereafter

If �1-log increase in BCR-ABL1 level with MMR, repeat in1 to 3 months

If inadequate* response to first-line therapy or any signof loss of response, defined as hematologic orcytogenetic relapse, or �1-log increase in BCR-ABL1 leveland loss of MMR

If disease progression to accelerated or blast phase CML

of a change in therapy.orescence in situ hybridization; IS, international scale; KD, kinase domain;

567

Table 2 Expected Levels of Response to First-line TKI Therapy (per NCCN and ELN27,33 Guidelines)

Time ontherapy

NCCN Guidelines (v3.2013) ELN Guidelines (2009, updated 2013)*

Adequate response Inadequate response Optimal responseSuboptimal response(“warning”)

Failure (therapychange indicted)

3 Months �10% BCR-ABL1(IS) or �PCyR

>10% BCR-ABL1 (IS)or <PCyR

CHR and �PCyR; �10%BCR-ABL1 (IS)

Ph+ 36e95%; >10%BCR-ABL1 (IS)

<CHR; No CyR

6 Months NS NS CCyR; <1% BCR-ABL1 (IS) Phþ 1e35%; 1e10%BCReABL1 (IS)

<PCyR; >10%BCR-ABL1 (IS)

12 Months CCyR �PCyR or cytogeneticrelapse

MMR 0.1e1% BCR-ABL1 (IS) <CCyR; >1%BCR-ABL1 (IS)

18 Months CCyR �PCyR or cytogeneticrelapse

Anytime NS NS Stable or improving MMR Loss of CHR, MMR, orCCyR; KD mutations;clonal progression

*After this article was accepted, an updated version of the consensus European LeukemiaNet (ELN) recommendations for management of chronic myeloidleukemia was published.33 This table reflects those changes. Also see Note Added in Proof.

CCyR, complete cytogenetic response (0% Phþ); ELN, European LeukemiaNet; IS, international scale; mCyR, minor cytogenetic response; MMR, majormolecular response; No CyR, no cytogenetic response; NS, not specified; PCyR, partial cytogenetic response; Ph, Philadelphia chromosome.

Press et al

study also showed better outcomes for patients with lowerBCR-ABL1 transcript levels at 3 months after TKI initiation;the 5-year OS and PFS rates were significantly longer forpatients with �10% BCR-ABL1 (IS) at 3 months than forpatients with >10% BCR-ABL1 (IS).34 Attainment of a �1-log BCR-ABL1 transcript reduction, or �10% BCR-ABL1(IS), at 3 months after TKI initiation is now defined as anearly molecular response. The practical prognostic value ofearly molecular response has been independently confirmedin a British CML cohort, where patients who failed to attainthis level of response had significantly shorter OS and PFS at8 years than patients who did attain this early molecularresponse landmark.35 On the basis of these data, the mostrecent NCCN CML guidelines specifically recommenda change in treatment for patients who fail to reach 10% BCR-ABL1 (IS) after 3 months of TKI therapy (Table 2).

Molecular response to nilotinib or dasatinib has also beenshown to predict long-term outcomes. For CML-CP patientsreceiving first-line dasatinib, an early molecular response waspredictive of improved long-term outcomes at 2 years.35

Similarly, CML-CP patients receiving first-line nilotinibwho achieved early molecular response had higher rates ofmolecular response and improved PFS andOS at 2 and 3 yearsthan patients who did not achieve early molecular response.36

In the second-line setting, patients receiving nilotinib afterimatinib failure who achieved �10% BCR-ABL1 (IS) at 3months had higher 24-month event-free survival rates thanpatients with >10% BCR-ABL1 (IS) at 3 months.37

Complete molecular response (CMR) has been defined asa level of BCR-ABL1 RNA that is undetectable by RT-qPCR.Because the absence of detectable BCR-ABL1 may be due topre-analytical factors, such as sample degradation, or to tech-nical factors, such as the variability in the limits of detection ofheterogeneous laboratory-developed assays, the definition ofCMR is very much sample- and laboratory-dependent. To

568

address these issues, it has been proposed that a finding ofundetectable BCR-ABL1 be qualified by the log limit ofdetection of the assay as defined by the transcript level of thereference gene [eg, CMR4 Z undetectable BCR-ABL1 ina sample in which BCR-ABL1 RNA would have been detect-able, had it been present at a level above 0.01% IS (4 log belowbaseline)].38 In the Oregon Health & Science University(OHSU) laboratory, the cutoff sample quality threshold fordefining CMR is a 4.7-log reduction from baseline [0.002%BCR-ABL1 (IS)], as defined by the reference gene level.Approximately 10% of samples submitted to the OHSUlaboratory have reference gene transcript levels below this 4.7-log threshold such that even though BCR-ABL1 may beundetectable in these samples, they are not reported as havingachieved CMR. By comparison, NCCN Guidelines defineCMR as undetectable BCR-ABL1 RNA with a minimalanalytical assay sensitivity of 4.5 log below the standardized ISbaseline.Given that the achievement of �10% BCR-ABL1 (IS) at 3

months and �0.1% BCR-ABL1 (IS) (ie, MMR) at 18months predict good prognosis, would achieving undetect-able BCR-ABL1 predict even better clinical outcomes? Insupport of this hypothesis, achieving CMR has been shownto be associated with more durable remissions and signifi-cantly longer relapse-free survival than achieving MMRwithout CMR.39 Other studies have shown that thefrequency of CMR increases with longer treatment times,and that the achievement of MMR at 12 months predictssubsequent CMR.40 In the Stop Imatinib (STIM) study,a prospective study evaluating the feasibility of stoppingimatinib treatment in patients with durable CMR, 12 monthsafter discontinuation of imatinib, 41% of patients with�12 months of follow-up maintained CMR.41 Thus, 59% ofpatients regained PCR-detectable BCR-ABL1 RNA after ima-tinib discontinuation. The long-term prognostic significance of

jmd.amjpathol.org - The Journal of Molecular Diagnostics

Monitoring Molecular Response in CML

undetectable BCR-ABL1 is actively being researched to betterunderstand its implications.

What Are the Clinically Relevant BCR-ABL1Threshold Levels for Predicting SuboptimalResponse, Loss of Response, and Evolving DrugResistance?

If achieving MMR signals a good long-term outcome, doesloss of MMR suggest impending resistance or relapse?42,43

In general, rising BCR-ABL1 levels signal the need for morefrequent patient follow-up, to either detect early resistanceor to exclude poor adherence to TKI therapy.44 Availabledata suggest that a rising level of BCR-ABL1 RNA duringTKI therapy can indeed predict increased risk of subsequentdisease progression, even if MMR has been achieved.39

Although most patients with a TKI-induced CCyR anda subsequent rise in BCR-ABL1 RNA levels remain inCCyR, there is a persistent increased risk of both relapse anda newly developed BCR-ABL1 kinase domain (KD) muta-tion,39,43,45 which, by itself, is a proven risk factor fordisease progression.4 ELN guidelines consider a rise inBCR-ABL1 levels at any time during treatment to constitutea warning of a possible suboptimal response (Table 2).27,33

At the very least, rising BCR-ABL1 levels should warranta repeat RT-qPCR within the next 1 to 3 months (NCCNGuidelines), sooner than the recommended routine testinginterval, the goal being early detection of true diseaseprogression, and possible intervention before full-fledgedcytogenetic or hematologic relapse.

How high must BCR-ABL1 RNA levels rise to be ofclinical concern? Cutoff values to define a clinicallysignificant rise in BCR-ABL1 RNA levels vary by laboratoryand therefore have been difficult to standardize. Suggestedtranscript rise thresholds that should prompt clinical actionhave included a 0.5-log increase,39 a 1-log increase (NCCNGuidelines), and a twofold (0.3-log) increase.47 A subse-quent study using a receiver-operating characteristic curveanalysis found that a 2.6-fold (0.41-log) increase in BCR-ABL1 levels was the optimal cutoff for predicting thepresence of a concomitant KD mutation.43 The negativepredictive value of this 2.6-fold cutoff was 97%, suggestingthat only 3% of patients with transcript rises <2.6-fold willhave a mutation that is missed. This study also showed thathad the NCCN Guidelineerecommended 10-fold (1-log)transcript rise cutoff been used as a trigger for KD mutationanalysis, the diagnostic sensitivity would have been poor(26%), with many confirmed mutations missed. Recom-mendations from the ELN suggest direct mutation testingwhen primary treatment with imatinib fails, when theincrease in BCR-ABL1 transcripts leads to loss of MMR, orwhenever there is a suboptimal response (eg, lack of MMRafter 18 months of imatinib).12 Updated NCCN Guidelinesrecommend mutation testing when the initial responseis inadequate [no partial cytogenetic response or >10%

The Journal of Molecular Diagnostics - jmd.amjpathol.org

BCR-ABL1 (IS) at 3 months, no CCyR at 12 or 18 months],when there is any sign of loss of response (hematologic orcytogenetic relapse, or 1-log increase in BCR-ABL1 and lossof MMR), or when CML progresses to advanced stages ofdisease.

BCR-ABL1 KD Mutations Directly InformTherapeutic Choices

The emergence of BCR-ABL1 KD mutations has beenshown to be associated with an increased likelihood ofsubsequent disease progression in patients with TKI-treatedCML.43,46 Although the switch to an alternative TKI, trig-gered by the finding of a KD mutation, has never beenshown to directly improve long-term outcomes ina prospective trial, ELN and NCCN guidelines neverthelessspecifically recommend a switch to certain TKI agents whenparticular mutations are detected.12 Due to the diversity ofmutations, full BCR-ABL1 KD mutation screening is donein most laboratories by a direct Sanger DNA sequencingtechnique that has a detection limit of approximately 20%mutant allele.11,48 A significant majority of BCR-ABL1 KDmutations cluster to one of four hot spots: the ATP-bindingP-loop (amino acids 248 to 256); the imatinib-bindingregion (amino acids 315 to 317); the catalytic domain(amino acids 350 to 363); and the activation (A)-loop(amino acids 381 to 402).11 Differential sensitivity to ima-tinib, dasatinib, nilotinib, bosutinib, and ponatinib has beendemonstrated by these diverse mutant BCR-ABL1 kinasesin in vitro studies.49e52 Because there is often, but notalways, a good correlation between mutation-specific in vi-tro resistance and in vivo clinical responses for some, butnot all, KD mutations and TKIs, the identification of thespecific mutation can help to inform the optimal manage-ment strategy.11,12,49 In particular, the presence of thecommon T315I mutation suggests that ponatinib, and noother TKI, may be effective.19 In addition, NCCN and ELNguidelines suggest a switch to nilotinib (not dasatinib) forpatients with the V299L, T315A, or F317L/V/I/C mutations;and a switch to dasatinib (not nilotinib) for patients with theY253H, E255K/V, or F359V/C/I mutations.

Aside from point mutations, the BCR-ABL1 KD alsocommonly develops insertion/deletion mutations, includinga 35-bp intronic insertion at the exon 8 to 9 junction, anL248V mutation with deletion of 81 bp of exon 4, an exon 7deletion, and several others.53e55 Although the clinical anddrug resistance significance of most of these insertion-deletion mutations is still unclear, the very common 35-bpintronic insertion after exon 8 does not appear to mediateTKI resistance, in vitro or in vivo.56 The BCR-ABL1 KD alsocarries some common single nucleotide polymorphisms thatappear to be wholly benign, including three nonsynonymous(K247R, F311V, Y320C), and three synonymous (T240T,T315T, E499E) variants, each of which has no known effecton TKI binding or drug resistance.57

569

Table 3 Sources of Heterogeneity among US Laboratories inBCR-ABL1 RT-qPCR Assessment

Source of heterogeneity in BCR-ABL1 RT-qPCR measurement

Equipment and reagentsNo predominant real-time PCR instrument (brand and model)used in most laboratories

Multiple sources of PCR primers and probes: commerciallydesigned versus laboratory designed

Different primers for reverse transcription: random hexamersversus gene-specific primers

Multiple manufacturers/distributors of reverse transcriptase andTaq polymerases

MethodologyNo predominant RNA extraction method used across alllaboratories

Variability in RT-qPCR protocols: commercially available kitversus laboratory-developed assayNo predominant laboratory-developed assay used across alllaboratories

No predominant reference gene used across all laboratories:ABL1, ACTB, B2M, BCR, G6PD, GAPDH, GUSB, TBP

Multiple types of material used to generate standard curve (orno standard curve generated): dilutions derived from plasmidDNA, RNA, cDNA, cell lines

Multiple sources of patient material used in RT-qPCR replicates:RNA, cDNA, cells, cell extracts/lysates

Reported resultsNo predominant method for reporting RT-qPCR results usedacross all laboratories; BCR-ABL1 level reported as:Copies per unit of RNARatio, relative to control gene, laboratory median or mean ofdiagnostic samples, patient’s previous result, diagnosticpatient sample, diluted or undiluted K562 cells, or other

Log reduction from pre-treatment baselinePercentage on the IS

Table is compiled from Zhang et al58 and the CAP MRD Survey.

BCR-ABL Log-Reduction from Baseline

Nu

mb

er o

f C

AP

L

ab

orato

ries

45

40

35

30

25

20

15

10

5

0

-2 -1 0 1 2 3 4 5 6 7

N: 300 labs

Mean: 3.3

SD: 0.99

Median: 3.4

25-75%: 2.7 – 3.8

ABL (n = 187)

Non-ABL (n = 111)

Reference Gene

p < 0.0001

BCR-ABL Log-Reduction from Baseline

-2 -1 0 1 2 3 4 5 6 7

Figure 1 Distribution of BCR-ABL1 RT-qPCR log-reduction values inlaboratories participating in the CAP MRD proficiency testing survey. Once ineach of the years 2009, 2010, and 2011, approximately 100 laboratoriesreceived two blinded samples for BCR-ABL1 RT-qPCR analysis: a baselinesample containing undiluted K562 cell-line RNA and a post-treatmentsample containing K562 RNA (dilution 1:10,000 in non-CML cell RNA).Each laboratory measured the BCR-ABL1 RNA ratio by its own validatedmethod, and reported back to CAP the relative log reduction of the post-treatment sample compared with the baseline sample. A: Histogram of thecombined log-reduction data from the cumulative 3-year testing period (nZ300). B: Box-whisker plot of the log-reduction values split by the referencegenes used, either ABL1 (nZ 187) or any other non-ABL1 reference gene (nZ 111). Information regarding reference gene use was not available for twolaboratories. The arrow on the x axis indicates the expected target log-reduction value (4.0). The vertical line in the middle of each box is themedian value of the distribution. The notch around the median representsthe 95% CI for the median. The vertical lines at the end of each box are the25% and 75% values of the distribution. The vertical bars outside of eachbox are the 10% and 90% values of the distribution.

Press et al

What Laboratory-Specific Factors Limit thePrecision, Accuracy, and Sensitivity of the BCR-ABL1 RT-qPCR Assay?

The dearth of high-quality, affordable, standardized reagentsremains a substantial barrier to wider adoption of IS-standardized BCR-ABL1 RT-qPCR assays. Assay-relatedheterogeneity affects all aspects of the multistep RT-qPCRprocedure (Table 3), which contributes to the considerablevariability in BCR-ABL1 quantitative values reported by USclinical diagnostic laboratories [College of AmericanPathologists (CAP) Surveys 2012 and Anatomic PathologyEducation Programs. MRD participant summary report,hereinafter referred to as CAP Survey].58 Several studieshave indicated a need for improvement in BCR-ABL1 RT-qPCR assay reproducibility (precision) between laborato-ries and even within laboratories.58e62 For example, ina study of 38 laboratories that used blinded shared samples,BCR-ABL1 RT-qPCR results varied by 1.6 to 3 log for the

570

same sample.58 A more comprehensive and less biasedsource of real-world data on BCR-ABL1 RT-qPCR analyt-ical assay performance is the CAP MRD proficiency surveywhereby blinded CML RNA samples are sent biannually toparticipating clinical molecular diagnostic laboratories aspart of a proficiency testing program required to maintainlaboratory accreditation (CAP Survey). Over each of thepast 3 years (2009, 2010, 2011), one CAP MRD survey peryear has included the same 1 in 10,000 (4-log) dilution ofCML cell-line K562 RNA, along with a comparative samplecontaining undiluted K562 RNA. In these surveys, CAP hasrequired that laboratories report a relative log reduction ofthe BCR-ABL1 RNA ratio in the diluted (post-treatment)sample divided by the undiluted (baseline) sample. Insteadof the expected 4-log difference between samples, and

jmd.amjpathol.org - The Journal of Molecular Diagnostics

0.5

N = 29 shared samples

Before Conversion:

13/29 within 2-fold18/29 within 3-fold26/29 within 5-foldBias = 0.35 logs (P < 0.001)

Conversion Factor = 2.22 (anti-log bias)

IS units (OHSU) = (BCR-ABL ratio)* 2.22

MMR level at OHSU (0.1% IS) = 0.045%

Bias = 0.35

X = Y

0.0

-0.5

Ad

elaid

e IS

(lo

g)

-1.0

-1.5

-2.0

-2.5 -2.0 -1.5 -1.0

BCR-ABL Ratio (log)

-0.5 0.0 0.5

-2.5

0.5

N = 29 shared samples

After Conversion:

15/29 within 2-fold27/29 within 3-fold29/29 within 5-foldBias = 0.025 logs (1.06-fold)(P = 0.69)No significant bias

Validated CF = 2.22

X = Y

0.0

-0.5

Ad

elaid

e IS

(lo

g)

-1.0

-1.5

-2.0

-2.5 -2.0 -1.5 -1.0

BCR-ABL IS (log)

-0.5 0.0 0.5

-2.5

Figure 2 IS CF determination at the Oregon Health & Science University (OHSU; Portland, Oregon). Twenty-nine samples were shared between OHSU andthe IS-validated Hughes & Branford laboratory (Adelaide, Australia). Each laboratory independently measured BCR-ABL1 RNA in these samples, and a log-loglinear regression analysis (bold line) was used to determine the OHSU laboratory’s CF to the IS (anti-log of bias). A: Before conversion to the IS. B: Afterconversion to the IS, using the laboratory-specific CF of 2.22. The X Z Y line (thin and gray) indicates a perfect correlation between the methods.

Pe

rc

en

ta

ge

o

f C

AP

L

ab

s

30

25

20

15

10

5

0

2009AYear:

CAP labs, N:

2009B 2010A 2010B 2011A 2011B 2012A

100 108 122 129 136 143 156

2012B

154

Figure 3 Percentage of CAP laboratories using BCR-ABL1 RT-qPCR ISreporting. Surveys are repeated early (A) and late (B) each calendar year.Data are self-reported by the laboratories subscribing to the CAP MRDsurvey.

Monitoring Molecular Response in CML

despite efforts to improve assay standardization and preci-sion by eliminating method-dependent variability from thelog-reduction result calculation, the distribution of reportedlog-reduction values (n Z 300) showed considerable assayimprecision and inaccuracy (Figure 1A). The overall meanlog reduction of 300 reported results from three combinedCAP surveys (2009, 2010, 2011) was 3.3 log (median, 3.4log), 0.7 log removed from the expected 4-log value. Therewas substantial interlaboratory variability (�0.99 log SD) inthis cumulative 3-year distribution that was not considerablydifferent from that observed in the individual surveys [�0.8log SD in 2009 (n Z 87), �1.0 log in 2010 (n Z 108), and�0.9 log in 2011 (n Z 105)]. This high degree of inter-laboratory imprecisiondeven without accounting for vari-ability stemming from RNA extraction and other methodbiasesdclearly precludes the practical ability to meaning-fully interpret serial BCR-ABL1 RNA levels when thosemeasurements are not all performed in the same laboratory,and reaffirms the critical need for improved assay stan-dardization. Furthermore, these data suggest that manyclinical laboratories are inaccurate in their quantification ofBCR-ABL1 RNA, with 32% of laboratories reporting log-reduction values >1 log (10-fold) removed from the ex-pected 4-log value.

A significant and much-debated source of variability inthe BCR-ABL1 RT-qPCR assay is the choice of referencegene used to calculate the relative transcript ratio.58,63 Ina direct exploration of this issue, Wang et al63 examinednine reference genes and deemed GUSB as the most suitablefor BCR-ABL1 RT-qPCR. In another study, Zhang et al58

found BCR to be a more suitable reference gene thanGUSB. Nonetheless, of the seven reference genes used byapproximately 100 CAP-surveyed laboratories, the twomost commonly used are ABL1 and G6PD, togetheraccounting for over 80% of laboratories (CAP Survey).

The Journal of Molecular Diagnostics - jmd.amjpathol.org

Although ABL1 is by far the most popular reference gene, itis actually a rather poor choice given that the PCR primerstypically used for quantitating ABL1 also cross-react andamplify BCR-ABL1.58,63,64 Thus, the measured level ofABL1 actually reflects levels of BCR-ABL1 plus ABL1, andfluctuates in samples with varying leukemia burdens. Thesesample-specific fluctuations in reference gene RNA levelsare completely contradictory to the primary defining char-acteristic of an ideal RT-qPCR reference gene. Thus, theresulting BCR-ABL1/ABL1 RNA ratio is nonlinear andfalsely low, particularly when the leukemia burden is high.63

When the ABL1 reference gene is used for monitoringduring TKI treatment, the log-reduction from baselinecalculation will be skewed artificially low, and the resultingIS value therefore will be falsely high. Consistent with thisprediction, the blinded CAP survey data showed that for the187 laboratories using ABL1 as the reference gene, the log-reduction values (mean, 2.9 log � 0.07 SE) were signifi-cantly lower (by 0.9 log; P < 0.0001) than for the 111

571

Press et al

laboratories using any other non-ABL1 reference gene(mean, 3.8 log � 0.08 SE) (Figure 1B). In other words, theaverage laboratory using the ABL1 reference gene falselyoverquantitated BCR-ABL1 RNA by >1 log (>10-fold)relative to the expected 4-log result. Although this ABL1reference geneedependent assay bias is perhaps exacer-bated in this CAP survey that utilizes a cell line (K562) witha known BCR-ABL1 gene amplification, an analogous (butperhaps lower magnitude) overquantitation phenomenonwould also be expected to occur with any other samplesource.

How Can the BCR-ABL1 RT-qPCR Assay BeStandardized to the IS of Measurement, byWhich Clinically Relevant Threshold LevelsHave Already Been Defined?

Inadequate clinical laboratory adoption of a standardizedreporting scale for BCR-ABL1 RT-qPCR remains a signifi-cant barrier to the real-world implementation of theconsensus molecular monitoring clinical practice guidelinesrecently put forth in both the US and Europe (NCCNGuidelines).27 International efforts to better standardizeBCR-ABL1 RT-qPCR reporting have been ongoing for manyyears, and have recently resulted in at least two notableachievements: the creation of an IS of quantitativemeasurement for BCR-ABL1 RT-qPCR21 and the availabilityof standardized reference material calibrated to the IS andvalidated by the World Health Organization (WHO).65

The original samples used in the IRIS study to establish theanchor values (ie, the pretreatment median standardizedbaseline and MMR) for the IS are no longer available, butquantitative traceability to those original values has beenprovided by the Hughes & Branford laboratory in Adelaide,Australia, which participated in the IRIS study and hasmaintained strict assay quality control standards for the pastdecade.26,31 Because the BCR-ABL1 IS anchor valuesrepresent absolute, not relative, values and are not dependenton the pretreatment baseline value of any particular patient,

Figure 4 The pathway for use of certified reference standards for BCR-ABL1certified by the WHO. Based on White et al64. The actual assigned IS values for the0.1%, and 0.01% target values, and depend on the reference gene that was used

572

laboratories around the world, using heterogeneous methodsfor RT-qPCR (including different reference genes), can adoptthe IS without fundamental technical changes to local stan-dard operating procedures.31 Two specific methods areavailable for harmonizing a laboratory’s local BCR-ABL1RT-qPCR assay to the IS: sample sharing with a laboratoryalready harmonized to the IS to establish a mathematicalconversion factor (CF),31 and local assay calibration withreference material that is calibrated to the IS.65

The sample-exchange process for conversion to the ISinvolves an exchange of samples with a validated referencelaboratory whose BCR-ABL1 RT-qPCR assay is alreadyaligned to the IS. The reference BCR-ABL1 RT-qPCR datafrom this reference laboratory are compared with thosegenerated in a field laboratory, and a CF is mathematicallyderived from simple linear regression and bias analyses.31

This CF is a constant fixed number by which the RT-qPCR ratio generated in the field laboratory is multipliedto convert to the IS. In a study seeking to align the BCR-ABL1 RT-qPCR values of 38 field laboratories from aroundthe world to that of the IS-calibrated Adelaide laboratory,calculated CFs ranged from 0.18 to 13.5 in an initial roundof sample exchanges. In our laboratory, the CF to the IS was2.22 (Figure 2). With a second round of sample exchanges,the majority of field laboratories were able to validate theirCFs, with performance characteristics such that <10% ofpatients who achieved MMR would be misclassified.31

Although this sample exchange approach successfullyattained its goal of minimizing interlaboratory BCR-ABL1reporting heterogeneity, several issues remain to beaddressed: i) For laboratories that exchange RNA (and notraw cell lysates), how should different RNA preparationmethods be controlled for? ii) How stable are the CFs? iii)How often should the sample exchange procedure berepeated? and iv) What degree of change to the technicalprotocol (eg, different PCR reagent manufacturing lots)should trigger a recalculation of the CF? Furthermore, thesample exchange process is challenging, time consuming,and expensive. The practical difficulty of this process isreflected in serial data from the CAP MRD BCR-ABL1

molecular monitoring, illustrating traceability back to primary standardsWHO BCR-ABL1 reference material are slightly different from the 10%, 1%,.

jmd.amjpathol.org - The Journal of Molecular Diagnostics

Monitoring Molecular Response in CML

proficiency testing survey (CAP Survey), which shows that,despite the proven analytical and clinical advantages of theIS, only a minority of clinical laboratories have successfullyswitched to this reporting method (Figure 3). Althoughadoption of the IS among clinical laboratories has increasedover the past 3 years, only 45 of 154 (29%) laboratoriessurveyed in late 2012 used IS-based reporting (up from 4%in 2009). This trend of increased adoption of IS-basedreporting, likely due to the recent commercial availabilityof BCR-ABL1 RT-qPCR assay reagents that are directlycalibrated to the IS, will likely continue. In the interim,however, for CML patients being serially monitored in those71% of laboratories not using the IS, it is unclear how theefficacy of TKI therapy is being clinically assessed, giventhe consensus IS-defined BCR-ABL1 RNA level thresholdsfor early response and MMR.

The availability of internationally validated, traceable,widely available universal reference materials would greatlyfacilitate the conversion of BCR-ABL1 testing laboratoriesto the IS. Toward this goal, a WHO International GeneticReference Panel was recently established and validatedfor quantitation of BCR-ABL1 RNA with any of threecommonly used control genes (ABL1, BCR, and GUSB).65

These extensively tested reference materials consist ofa lyophilized admixture of K562 and HL60 cell lines at fourdifferent dilution levels, with BCR-ABL1 RNA levels thatare close to clinical thresholds (eg, MMR) and directly tiedto the IS.65 Unfortunately, only a limited supply of thisprimary reference material was created, and it is envisionedthat reference laboratories, commercial reagent vendors, andother organizations will use this primary material to createvalidated large-scale secondary reference materials to beused as routine calibrators in individual testing laboratories(Figure 4).65 Several commercial reagent vendors nowmarket such IS-standardized reagents or calibrators. Inaddition, two of these vendors have each recently begunclinical trials with the aim of providing Food and DrugAdministrationeapproved assay kits that will automaticallyincorporate IS-based reporting units for BCR-ABL1.

Conclusions

It is now widely accepted that molecular monitoring by serialBCR-ABL1 RT-qPCR provides essential information on TKItreatment response in patients with CML. In particular, theachievement of MMR at 18 months of TKI therapy is recog-nized as a consensus indicator of successful treatment responseand subsequent good prognosis. More recently, the NCCNguidelines have included the achievement of early molecularresponse, defined as�10% BCR-ABL1 (IS) at 3 months post-TKI initiation, as an important responsemilestone that directlyinforms diseasemanagement decisions at this early time point.Additional data suggest that even lower levels of MRD,namely CMR (undetectable BCR-ABL1 using an RT-qPCRassay with sensitivity of �4, �4.5, or �5 log), may afford

The Journal of Molecular Diagnostics - jmd.amjpathol.org

even better prognosis. The failure to achieve optimal thera-peutic landmarks, as defined by BCR-ABL1 (IS) values,conveys an increased risk of poor outcome, and may indicatethe need to switch the TKI agent. Furthermore, increasinglevels of BCR-ABL1 RNA indicate a rising leukemia burdenand confer an increased risk of disease progression and theappearance of drug-resistantmutations inBCR-ABL1. Regularmonitoring (every 3 to 6 months) with accurate, reproducible,and standardizedBCR-ABL1RT-qPCR assays is thus requiredthroughout the course of TKI therapy. Imprecision and inac-curacy in measuring BCR-ABL1 transcripts can have seriousadverse clinical consequences, and treatable problems of earlyrelapse or resistance may go unnoticed.

Efforts to facilitate accurate and precise measurements ofBCR-ABL1 RNA were enhanced recently by the validationof WHO reference standards. It is hoped that the recentavailability of validated secondary reference materials basedon the primary WHO reference standards will encouragemore laboratories to adopt the standardized IS for reportingof BCR-ABL1 assay results. This should facilitate theaccurate interlaboratory comparison of data, and theconsistent identification of patients who have achievedclinically relevant therapeutic milestones such as earlymolecular response or MMR (or even CMR). Improvedreliability of molecular monitoring will enhance the confi-dence in these measurements to the benefit of clinicians,laboratories, pathologists, patients, regulators, and payers.

Note Added in Proof

While this manuscript was in preparation for print, theEuropean LeukemiaNet (ELN) published updated recom-mendations for management of chronic myeloid leukemia.33

These new ELN recommendations vary from the 2009 ELNrecommendations primarily by the addition of specific time-dependent molecular monitoring thresholds for defining“Optimal,” “Warning” (previously called “Suboptimal”),and “Failure” categories of TKI responses. In particular, the2013 ELN recommendations update the information inTable 2 of this manuscript to include:

� An “Optimal” response is defined by a 3-month post-TKIBCR-ABL qPCR value below 10% IS, a 6-month post-TKI BCR-ABL RT-qPCR value below 1% IS, and a 12-month post-TKI BCR-ABL RT-qPCR value below 0.1%IS (ie, MMR).

� A “Failed” response (meaning a change of therapy isrecommended) is defined by a 6-month post-TKI BCR-ABL qPCR value above 10% IS and a 12 month post-TKIBCR-ABL qPCR value above 1% IS.

Acknowledgments

We thank Drs. Anna Lau and Patricia Segarini (PercolationCommunications LLC) for their medical editorial assistance,

573

Press et al

Dr. Lawrence Jennings (Lurie Children’s Hospital, Chicago,IL) for his insightful preview of the manuscript, andMariana Ovnic for the illustrations in Figure 4.

References

1. Nowell PC, Hungerford DA: Chromosome studies on normal andleukemic human leukocytes. J Natl Cancer Inst 1960, 25:85e109

2. Rowley JD: Letter: a new consistent chromosomal abnormality inchronic myelogenous leukaemia identified by quinacrine fluorescenceand Giemsa staining. Nature 1973, 243:290e293

3. Horne SD, Stevens JB, Abdallah BY, Liu G, Bremer SW, Ye CJ,Heng HH: Why imatinib remains an exception of cancer research.J Cell Physiol 2012, 228:665e670

4. Bjorkholm M, Ohm L, Eloranta S, Derolf A, Hultcrantz M, Sjoberg J,Andersson T, Hoglund M, Richter J, Landgren O, Kristinsson SY,Dickman PW: Success story of targeted therapy in chronic myeloidleukemia: a population-based study of patients diagnosed in Swedenfrom 1973 to 2008. J Clin Oncol 2011, 29:2514e2520

5. Hehlmann R: How I treat CML blast crisis. Blood 2012, 120:737e7476. Kantarjian H, O’Brien S, Jabbour E, Garcia-Manero G, Quintas-

Cardama A, Shan J, Rios MB, Ravandi F, Faderl S, Kadia T,Borthakur G, Huang X, Champlin R, Talpaz M, Cortes J: Improvedsurvival in chronic myeloid leukemia since the introduction of ima-tinib therapy: a single-institution historical experience. Blood 2012,119:1981e1987

7. Druker BJ, Guilhot F, O’Brien SG, Gathmann I, Kantarjian H,Gattermann N, Deininger MW, Silver RT, Goldman JM, Stone RM,Cervantes F, Hochhaus A, Powell BL, Gabrilove JL, Rousselot P,Reiffers J, Cornelissen JJ, Hughes T, Agis H, Fischer T, Verhoef G,Shepherd J, Saglio G, Gratwohl A, Nielsen JL, Radich JP,Simonsson B, Taylor K, Baccarani M, So C, Letvak L, Larson RA:Five-year follow-up of patients receiving imatinib for chronicmyeloid leukemia. N Engl J Med 2006, 355:2408e2417

8. Deininger M, O’Brien SG, Guilhot F, Goldman JM, Hochhaus A,Hughes TP, Radich JP, Hatfield AK, Mone M, Filian J, Reynolds J,Gathmann I, Larson RA, Druker BJ: International randomized studyof interferon versus STI571 (IRIS) 8-year follow up: sustainedsurvival and low risk for progression or events in patients with newlydiagnosed chronic myeloid leukemia in chronic phase (CML-CP)treated with imatinib (abstract). Blood (ASH Annual MeetingAbstracts) 2009, 114. Abstract 1126

9. Agrawal M, Garg RJ, Cortes J, Quintas-Cardama A: Tyrosine kinaseinhibitors: the first decade. Curr Hematol Malig Rep 2010, 5:70e80

10. Diamond JM, Melo JV: Mechanisms of resistance to BCR-ABLkinase inhibitors. Leuk Lymphoma 2011, 52(Suppl 1):12e22

11. Jones D, Kamel-Reid S, Bahler D, Dong H, Elenitoba-Johnson K,Press R, Quigley N, Rothberg P, Sabath D, Viswanatha D, Weck K,Zehnder J: Laboratory practice guidelines for detecting and reportingBCR-ABL drug resistance mutations in chronic myelogenousleukemia and acute lymphoblastic leukemia: a report of the Associ-ation for Molecular Pathology. J Mol Diagn 2009, 11:4e11

12. Soverini S, Hochhaus A, Nicolini FE, Gruber F, Lange T, Saglio G,Pane F, Muller MC, Ernst T, Rosti G, Porkka K, Baccarani M,Cross NC, Martinelli G: BCR-ABL kinase domain mutation analysisin chronic myeloid leukemia patients treated with tyrosine kinaseinhibitors: recommendations from an expert panel on behalf ofEuropean LeukemiaNet. Blood 2011, 118:1208e1215

13. Saglio G, Kim DW, Issaragrisil S, le Coutre P, Etienne G, Lobo C,Pasquini R, Clark RE, Hochhaus A, Hughes TP, Gallagher N,Hoenekopp A, Dong M, Haque A, Larson RA, Kantarjian HM:Nilotinib versus imatinib for newly diagnosed chronic myeloidleukemia. N Engl J Med 2010, 362:2251e2259

14. Kantarjian H, Shah NP, Hochhaus A, Cortes J, Shah S, Ayala M,Moiraghi B, Shen Z, Mayer J, Pasquini R, Nakamae H, Huguet F,

574

Boque C, Chuah C, Bleickardt E, Bradley-Garelik MB, Zhu C,Szatrowski T, Shapiro D, Baccarani M: Dasatinib versus imatinib innewly diagnosed chronic-phase chronic myeloid leukemia. N Engl JMed 2010, 362:2260e2270

15. Kantarjian HM, Hochhaus A, Saglio G, De Souza Santos FP,Flinn IW, Stenke L, Goh Y-T, Rosti G, Nakamae H, Gallagher NJ,Hoenekopp A, Blakesley R, Larson RA, Hughes TP: Nilotinib versusimatinib for the treatment of patients with newly diagnosed chronicphase. Philadelphia chromosome-positive, chronic myeloid leukaemia:24-month minimum follow-up of the phase 3 randomised ENESTndtrial. Lancet Oncol 2011, 12:841e851

16. Hochhaus A, Shah NP, Cortes JE, Baccarani M, Bradley-Garelik MB,Dejardin D, Kantarjian H: Dasatinib versus imatinib (IM) in newlydiagnosed chronic myeloid leukemia in chronic phase (CML-CP):DASISION 3-year follow-up. J Clin Oncol 2012, 30:6504

17. Larson RA, Hochhaus A, Hughes TP, Clark RE, Etienne G, Kim DW,Flinn IW, Kurokawa M, Moiraghi B, Yu R, Blakesley RE,Gallagher NJ, Saglio G, Kantarjian HM: Nilotinib vs imatinib inpatients with newly diagnosed Philadelphia chromosome-positivechronic myeloid leukemia in chronic phase: ENESTnd 3-yearfollow-up. Leukemia 2012, 26:2197e2203

18. Cortes JE, Kantarjian HM, Brümmendorf TH, Kim DW, Turkina AG,Shen ZX, Pasquini R, Khoury HJ, Arkin S, Volkert A, Besson N,Abbas R, Wang J, Leip E, Gambacorti-Passerini C: Safety and effi-cacy of bosutinib (SKI-606) in chronic phase Philadelphia chromo-some-positive chronic myeloid leukemia patients with resistance orintolerance to imatinib. Blood 2011, 118:4567e4576

19. Cortes JE, Kantarjian H, Shah NP, Bixby D, Mauro MJ, Flinn I,O’Hare T, Hu S, Narasimhan NI, Rivera VM, Clackson T,Turner CD, Haluska FG, Druker BJ, Deininger MW, Talpaz M:Ponatinib in refractory Philadelphia chromosome-positive leukemias.N Engl J Med 2012, 367:2075e2088

20. Baccarani M, Saglio G, Goldman J, Hochhaus A, Simonsson B,Appelbaum F, Apperley J, Cervantes F, Cortes J, Deininger M,Gratwohl A, Guilhot F, Horowitz M, Hughes T, Kantarjian H,Larson R, Niederwieser D, Silver R, Hehlmann R: Evolving conceptsin the management of chronic myeloid leukemia: recommendationsfrom an expert panel on behalf of the European LeukemiaNet. Blood2006, 108:1809e1820

21. Hughes T, Deininger M, Hochhaus A, Branford S, Radich J, Kaeda J,Baccarani M, Cortes J, Cross NC, Druker BJ, Gabert J, Grimwade D,Hehlmann R, Kamel-Reid S, Lipton JH, Longtine J, Martinelli G,Saglio G, Soverini S, Stock W, Goldman JM: Monitoring CMLpatients responding to treatment with tyrosine kinase inhibitors:review and recommendations for harmonizing current methodologyfor detecting BCR-ABL transcripts and kinase domain mutations andfor expressing results. Blood 2006, 108:28e37

22. Branford S, Hughes TP, Rudzki Z: Monitoring chronic myeloidleukaemia therapy by real-time quantitative PCR in blood is a reliablealternative to bone marrow cytogenetics. Br J Haematol 1999, 107:587e599

23. Hehlmann R, Lauseker M, Jung-Munkwitz S, Leitner A, Muller MC,Pletsch N, Proetel U, Haferlach C, Schlegelberger B, Balleisen L,Hanel M, Pfirrmann M, Krause SW, Nerl C, Pralle H, Gratwohl A,Hossfeld DK, Hasford J, Hochhaus A, Saussele S: Tolerability-adapted imatinib 800 mg/d versus 400 mg/d versus 400 mg/d plusinterferon-a in newly diagnosed chronic myeloid leukemia. J ClinOncol 2011, 29:1634e1642

24. Hughes TP, Hochhaus A, Branford S, Muller MC, Kaeda JS,Foroni L, Druker BJ, Guilhot F, Larson RA, O’Brien SG,Rudoltz MS, Mone M, Wehrle E, Modur V, Goldman JM, Radich JP:Long-term prognostic significance of early molecular response toimatinib in newly diagnosed chronic myeloid leukemia: an analysisfrom the international randomized study of interferon versus STI571(IRIS). Blood 2010, 116:3758e3765

25. PressRD,LoveZ,TronnesAA,YangR,TranT,Mongoue-Tchokote S,Mori M, Mauro MJ, Deininger MW, Druker BJ: BCR-ABL mRNA

jmd.amjpathol.org - The Journal of Molecular Diagnostics

Monitoring Molecular Response in CML

levels at and after the time of a complete cytogenetic response (CCR)predict the duration of CCR in imatinib mesylate-treated patients withCML. Blood 2006, 107:4250e4256

26. Cross NC: Standardisation of molecular monitoring for chronicmyeloid leukaemia. Best Pract Res Clin Haematol 2009, 22:355e365

27. Baccarani M, Cortes J, Pane F, Niederwieser D, Saglio G, Apperley J,Cervantes F, Deininger M, Gratwohl A, Guilhot F, Hochhaus A,Horowitz M, Hughes T, Kantarjian H, Larson R, Radich J,Simonsson B, Silver RT, Goldman J, Hehlmann R: Chronic myeloidleukemia: an update of concepts and management recommendationsof European LeukemiaNet. J Clin Oncol 2009, 27:6041e6051

28. Faderl S, Talpaz M, Estrov Z, Kantarjian HM: Chronic myelogenousleukemia: biology and therapy. Ann Intern Med 1999, 131:207e219

29. Kantarjian H, Sawyers C, Hochhaus A, Guilhot F, Schiffer C,Gambacorti-Passerini C, Niederwieser D, Resta D, Capdeville R,Zoellner U, Talpaz M, Druker B, Goldman J, O’Brien SG, Russell N,Fischer T, Ottmann O, Cony-Makhoul P, Facon T, Stone R, Miller C,Tallman M, Brown R, Schuster M, Loughran T, Gratwohl A,Mandelli F, Saglio G, Lazzarino M, Russo D, Baccarani M, Morra E:Hematologic and cytogenetic responses to imatinib mesylate inchronic myelogenous leukemia. N Engl J Med 2002, 346:645e652

30. Akard LP, Wang YL: Translating trial-based molecular monitoringinto clinical practice: importance of international standards andpractical considerations for community practitioners. Clin LymphomaMyeloma Leuk 2011, 11:385e395

31. Branford S, Fletcher L, Cross NC, Muller MC, Hochhaus A,Kim DW, Radich JP, Saglio G, Pane F, Kamel-Reid S, Wang YL,Press RD, Lynch K, Rudzki Z, Goldman JM, Hughes T: Desirableperformance characteristics for BCR-ABL measurement on aninternational reporting scale to allow consistent interpretation ofindividual patient response and comparison of response rates betweenclinical trials. Blood 2008, 112:3330e3338

32. Hughes TP, Kaeda J, Branford S, Rudzki Z, Hochhaus A,Hensley ML, Gathmann I, Bolton AE, van Hoomissen IC,Goldman JM, Radich JP: Frequency of major molecular responses toimatinib or interferon alfa plus cytarabine in newly diagnosed chronicmyeloid leukemia. N Engl J Med 2003, 349:1423e1432

33. Baccarani M, Deininger M, Rosti G, Hochhaus A, Soverini S,Apperley JF, Cervantes F, Clark RE, Cortes JE, Guilhot F, Hjorth-Hansen H, Hughes T, Kantarjian HM, Kim DW, Larson RA, LiptonJH, Mahon FX, Martinelli G, Mayer J, Muller MC, Niederwieser D,Pane F, Radich JP, Rousselot P, Saglio G, Saubele S, Schiffer C,Silver R, Simonsson B, Steegmann J-L, Golman JM, Hehlmann R:European LeukemiaNet recommendations for the management ofchronic myeloid leukemia. Blood http://dx.doi.org/10.1182/blood-2013-05-501569

34. Hanfstein B, Muller MC, Hehlmann R, Erben P, Lauseker M,Fabarius A, Schnittger S, Haferlach C, Gohring G, Proetel U,Kolb HJ, Krause SW, Hofmann WK, Schubert J, Einsele H,Dengler J, Hanel M, Falge C, Kanz L, Neubauer A, Kneba M,Stegelmann F, Pfreundschuh M, Waller CF, Branford S, Hughes TP,Spiekermann K, Baerlocher GM, Pfirrmann M, Hasford J, Saussele S,Hochhaus A: Early molecular and cytogenetic response is predictivefor long-term progression-free and overall survival in chronicmyeloid leukemia (CML). Leukemia 2012, 26:2096e2102

35. Marin D, Ibrahim AR, Lucas C, Gerrard G, Wang L, Szydlo RM,Clark RE, Apperley JF, Milojkovic D, Bua M, Pavlu J,Paliompeis C, Reid A, Rezvani K, Goldman JM, Foroni L:Assessment of BCR-ABL1 transcript levels at 3 months is the onlyrequirement for predicting outcome for patients with chronic myeloidleukemia treated with tyrosine kinase inhibitors. J Clin Oncol 2012,30:232e238

36. Hochhaus A, Guilhot F, Kathrin Haifa AA, Rosti G, Nakaseko C, DeSouza CA, Larson R, Kantry HM, Hoenekopp A, Blakesley R, Yu R,Saglio G, Hughes T. Early BCR-ABL transcript levels predict futuremolecular response and long-term outcomes in newly diagnosedpatients with chronic myeloid leukemia in chronic phase: analysis

The Journal of Molecular Diagnostics - jmd.amjpathol.org

of ENESTnd 3-year data. 17th Congress of the European Hema-tology Association, 2012 June 15-17, Amsterdam, The Netherlands,Abstract 0584

37. Branford S, Kim D-W, Soverini S, Gottardi E, Beppu L, Muller MC,Haque A, Shou Y, Woodman RC, Kantarjian HM, Martinelli G,Radich JP, Saglio G, Hughes TP, Hochhaus A: Molecular response at3 months on nilotinib therapy predicts response and long-termoutcomes in patients with imatinib-resistant or -intolerant chronicmyeloid leukemia in chronic phase (CML-CP) (abstract). Blood(ASH Annual Meeting Abstracts) 2009, 114:3292

38. Cross NC, White HE, Muller MC, Saglio G, Hochhaus A: Stan-dardized definitions of molecular response in chronic myeloidleukemia. Leukemia 2012, 26:2172e2175

39. Press RD, Galderisi C, Yang R, Rempfer C, Willis SG, Mauro MJ,Druker BJ, Deininger MW: A half-log increase in BCR-ABL RNApredicts a higher risk of relapse in patients with chronic myeloidleukemia with an imatinib-induced complete cytogenetic response.Clin Cancer Res 2007, 13:6136e6143

40. Branford S, Seymour JF, Grigg A, Arthur C, Rudzki Z, Lynch K,Hughes T: BCR-ABL messenger RNA levels continue to decline inpatients with chronic phase chronic myeloid leukemia treated withimatinib for more than 5 years and approximately half of all first-linetreated patients have stable undetectable BCR-ABL using strictsensitivity criteria. Clin Cancer Res 2007, 13:7080e7085

41. Mahon FX, Rea D, Guilhot J, Guilhot F, Huguet F, Nicolini F,Legros L, Charbonnier A, Guerci A, Varet B, Etienne G, Reiffers J,Rousselot P: Discontinuation of imatinib in patients with chronicmyeloid leukaemia who have maintained complete molecular remis-sion for at least 2 years: the prospective, multicentre Stop Imatinib(STIM) trial. Lancet Oncol 2010, 11:1029e1035

42. Breccia M, Alimena G: The significance of early, major and stablemolecular responses in chronic myeloid leukemia in the imatinib era.Crit Rev Oncol Hematol 2011, 79:135e143

43. Press RD, Willis SG, Laudadio J, Mauro MJ, Deininger MW:Determining the rise in BCR-ABL RNA that optimally predictsa kinase domain mutation in patients with chronic myeloid leukemiaon imatinib. Blood 2009, 114:2598e2605

44. Marin D, Bazeos A, Mahon FX, Eliasson L, Milojkovic D, Bua M,Apperley JF, Szydlo R, Desai R, Kozlowski K, Paliompeis C,Latham V, Foroni L, Molimard M, Reid A, Rezvani K, deLavallade H, Guallar C, Goldman J, Khorashad JS: Adherence is thecritical factor for achieving molecular responses in patients withchronic myeloid leukemia who achieve complete cytogeneticresponses on imatinib. J Clin Oncol 2010, 28:2381e2388

45. Kantarjian HM, Shan J, Jones D, O’Brien S, Rios MB, Jabbour E,Cortes J: Significance of increasing levels of minimal residual diseasein patients with Philadelphia chromosome-positive chronic myelog-enous leukemia in complete cytogenetic response. J Clin Oncol 2009,27:3659e3663

46. Khorashad JS, deLavallade H, Apperley JF, Milojkovic D, Reid AG,Bua M, Szydlo R, Olavarria E, Kaeda J, Goldman JM, Marin D:Finding of kinase domain mutations in patients with chronic phasechronic myeloid leukemia responding to imatinib may identify thoseat high risk of disease progression. J Clin Oncol 2008, 26:4806e4813

47. Branford S, Rudzki Z, Parkinson I, Grigg A, Taylor K, Seymour JF,Durrant S, Browett P, Schwarer AP, Arthur C, Catalano J, Leahy MF,Filshie R, Bradstock K, Herrmann R, Joske D, Lynch K, Hughes T:Real-time quantitative PCR analysis can be used as a primary screento identify patients with CML treated with imatinib who have BCR-ABL kinase domain mutations. Blood 2004, 104:2926e2932

48. Terasawa T, Dahabreh I, Trikalinos TA: BCR-ABL mutation testingto predict response to tyrosine kinase inhibitors in patients withchronic myeloid leukemia. PLoS Curr 2010, 2:RRN1204

49. Branford S, Melo JV, Hughes TP: Selecting optimal second-linetyrosine kinase inhibitor therapy for chronic myeloid leukemiapatients after imatinib failure: does the BCR-ABL mutation statusreally matter? Blood 2009, 114:5426e5435

575

Press et al

50. O’Hare T, Eide CA, Deininger MW: Bcr-Abl kinase domain muta-tions, drug resistance, and the road to a cure for chronic myeloidleukemia. Blood 2007, 110:2242e2249

51. O’Hare T, Shakespeare WC, Zhu X, Eide CA, Rivera VM, Wang F,Adrian LT, Zhou T, Huang WS, Xu Q, Metcalf CA 3rd, Tyner JW,Loriaux MM, Corbin AS, Wardwell S, Ning Y, Keats JA, Wang Y,Sundaramoorthi R, Thomas M, Zhou D, Snodgrass J, Commodore L,Sawyer TK, Dalgarno DC, Deininger MW, Druker BJ, Clackson T:AP24534, a pan-BCR-ABL inhibitor for chronic myeloid leukemia,potently inhibits the T315I mutant and overcomes mutation-basedresistance. Cancer Cell 2009, 16:401e412

52. Redaelli S, Piazza R, Rostagno R, Magistroni V, Perini P, Marega M,Gambacorti-Passerini C, Boschelli F: Activity of bosutinib, dasatinib,and nilotinib against 18 imatinib-resistant BCR/ABL mutants. J ClinOncol 2009, 27:469e471

53. Sherbenou DW, Hantschel O, Turaga L, Kaupe I, Willis S, Bumm T,Press RD, Superti-Furga G, Druker BJ, Deininger MW: Character-ization of BCR-ABL deletion mutants from patients with chronicmyeloid leukemia. Leukemia 2008, 22:1184e1190

54. Gruber FX, Hjorth-Hansen H, Mikkola I, Stenke L, Johansen T: Anovel Bcr-Abl splice isoform is associated with the L248V mutationin CML patients with acquired resistance to imatinib. Leukemia 2006,20:2057e2060

55. Laudadio J, Deininger MW, Mauro MJ, Druker BJ, Press RD: Anintron-derived insertion/truncation mutation in the BCR-ABL kinasedomain in chronic myeloid leukemia patients undergoing kinaseinhibitor therapy. J Mol Diagn 2008, 10:177e180

56. O’Hare T, Zabriskie MS, Eide CA, Agarwal A, Adrian LT, You H,Corbin AS, Yang F, Press RD, Rivera VM, Toplin J, Wong S,Deininger MW, Druker BJ: The BCR-ABL35INS insertion/truncation mutant is kinase-inactive and does not contribute to tyro-sine kinase inhibitor resistance in chronic myeloid leukemia. Blood2011, 118:5250e5254

57. Hughes TP, Branford S: Monitoring disease response to tyrosinekinase inhibitor therapy in CML. Hematology Am Soc Hematol EducProgram 2009:477e487

58. Zhang T, Grenier S, Nwachukwu B, Wei C, Lipton JH, Kamel-Reid S: Inter-laboratory comparison of chronic myeloid leukemiaminimal residual disease monitoring: summary and recommenda-tions. J Mol Diagn 2007, 9:421e430

576

59. Beillard E, Pallisgaard N, van der Velden VH, Bi W, Dee R, van derSchoot E, Delabesse E, Macintyre E, Gottardi E, Saglio G,Watzinger F, Lion T, van Dongen JJ, Hokland P, Gabert J: Evaluationof candidate control genes for diagnosis and residual disease detectionin leukemic patients using ‘real-time’ quantitative reverse-transcriptase polymerase chain reaction (RQ-PCR)da EuropeAgainst Cancer program. Leukemia 2003, 17:2474e2486

60. Muller MC, Erben P, Saglio G, Gottardi E, Nyvold CG, Schenk T,Ernst T, Lauber S, Kruth J, Hehlmann R, Hochhaus A: Harmoniza-tion of BCR-ABL mRNA quantification using a uniform multifunc-tional control plasmid in 37 international laboratories. Leukemia2008, 22:96e102

61. Muller MC, Saglio G, Lin F, Pfeifer H, Press RD, Tubbs RR,Paschka P, Gottardi E, O’Brien SG, Ottmann OG, Stockinger H,Wieczorek L, Merx K, Konig H, Schwindel U, Hehlmann R,Hochhaus A: An international study to standardize the detection andquantitation of BCR-ABL transcripts from stabilized peripheral bloodpreparations by quantitative RT-PCR. Haematologica 2007, 92:970e973

62. Yamada MF, Fujiwara T, Ishikawa I, Kohata K, Katoh C,Miyamura K, Harigae H: Interlaboratory comparison of quantitativeRT-PCR based detection for minimal residual disease in leukemias:a standardization approach in Japan. Tohoku J Exp Med 2008, 214:97e104

63. Wang YL, Lee JW, Cesarman E, Jin DK, Csernus B: Molecularmonitoring of chronic myelogenous leukemia: identification of themost suitable internal control gene for real-time quantification ofBCR-ABL transcripts. J Mol Diagn 2006, 8:231e239

64. Lee JW, Chen Q, Knowles DM, Cesarman E, Wang YL: beta-Glucuronidase is an optimal normalization control gene for molecularmonitoring of chronic myelogenous leukemia. J Mol Diagn 2006, 8:385e389

65. White HE, Matejtschuk P, Rigsby P, Gabert J, Lin F, Lynn Wang Y,Branford S, Muller MC, Beaufils N, Beillard E, Colomer D,Dvorakova D, Ehrencrona H, Goh HG, El Housni H, Jones D,Kairisto V, Kamel-Reid S, Kim DW, Langabeer S, Ma ES, Press RD,Romeo G, Wang L, Zoi K, Hughes T, Saglio G, Hochhaus A,Goldman JM, Metcalfe P, Cross NC: Establishment of the first WorldHealth Organization International Genetic Reference Panel forquantitation of BCR-ABL mRNA. Blood 2010, 116:e111ee117

jmd.amjpathol.org - The Journal of Molecular Diagnostics