International Standardisation for BCR-ABL RQ-PCR in CML

How close are we to achieving the ultimate goal of global standardization
for BCR-ABL measurement by RQ-PCR in CML?

Over recent years various studies have demonstrated that serial analysis of BCR-ABL mRNA levels by real-time quantitative PCR (RQ-PCR) accurately reflects the level of leukemic inhibition induced by therapy and provides an effective monitoring strategy for patients with CML.

Since the introduction of BCR-ABL kinase inhibitor therapy, molecular monitoring has become particularly relevant since most de novo treated patients achieve a complete cytogenetic response (CCR) soon after initiating therapy. Once CCR is achieved, residual disease can be tracked at the molecular level.

The IRIS trial was the first formal demonstration of the dramatic superiority of imatinib over interferon based regimens. RQ-PCR analysis for the trial was centralised in three centres – Adelaide, London and Seattle – that used different laboratory procedures. It was observed that reproducible differences in median BCR-ABL values were being measured between the three centres at specific timepoints which prompted the need for an urgent alignment of their respective results. In the absence of any independent reference materials, the decision was made that each centre would measure the level of disease in a common set of 30 pretreatment samples, and that patient results would be normalised to this standardised baseline. Reanalysis of the data showed improved comparability of results between the three laboratories and the standardised baseline was used to normalise all IRIS RQ-PCR results. Thus, major molecular response (MMR) is defined as a three log reduction from the IRIS standardised baseline and not a three log reduction from pretreatment material for each case. MMR corresponds to a level of disease approximately 1 log below that at which CCR is achieved and is considered as an important therapeutic milestone.

There are a range of RQ-PCR techniques in use throughout the world and therefore variation in BCR-ABL levels reported by different laboratories is inevitable. Even laboratories using the same technique may report different values. Whilst this variation is not necessarily a major drawback when tracking an individual patient’s response in a single centre, it does limit the accuracy with which MMR can be gauged and makes the comparison of BCR-ABL values between laboratories difficult. One problem for laboratories trying to establish the level representing MMR has been in the determination of an appropriate BCR-ABL value that represents the baseline. This is not a simple task and imprecision can occur depending on the RQ-PCR technique in use, the patients selected for analysis and the control gene that is employed to normalise results of the assay.

To overcome these difficulties an international reporting scale (IS) has been established which abolishes the requirement to determine a baseline value. The IS expresses detectable disease as a percentage and, critically, this scale is essentially identical to that used in the IRIS trial, with 100% IS defined as the standardised baseline and 0.1% IS corresponding to MMR. Over the past few years pilot studies initiated by the Adelaide laboratory have established that individual testing labs can align their reporting scale to the IS by exchange of samples and derivation of a laboratory specific conversion factor (CF). Whilst this process works well, it is not possible for a single reference laboratory to standardise all other testing laboratories in the world and the concept of regional or national reference laboratories has been developed, particularly in Europe through the European Treatment and Outcome Study (EUTOS) within the European LeukemiaNet. The aim is that selected laboratories who have established a validated CF with Adelaide can then exchange samples with other laboratories and thereby propagate traceable CFs to local centres.

Whilst the development of CFs is a major step forward, it is obvious that this approach represents a pragmatic compromise between what is desirable and what is currently achievable. Ideally, any testing laboratory should be able to readily access reference standards that enable them to convert patient results directly to the IS, provided that their assay conforms to accepted quality assurance criteria. A program to develop such reagents is  well underway.

Although standardisation of BCR-ABL RQ-PCR testing is a complex and expensive process, we believe that it is both desirable and achievable. Our goal is to engage as many testing laboratories and clinical centres as possible in this process. The International BCR-ABL standardization group is open to all who are interested and meets regularly at the annual American Society of Hematology and European Hematology Association meetings. The group aims to improve the quality and comparability of BCR-ABL RQ-PCR testing through promulgation of the IS as well as providing a forum for presentation and discussion of other BCR-ABL related laboratory issues. Please contact us if you would like to be added to the mailing list.

Dr Susan Branford
Adelaide University (Australia)

Scientific advisor of the iCLMf

Dr Nick Cross
University of Southhampton (UK) 

Scientific advisor of the iCMLf