Measurement of BCR-ABL1 transcripts on the International Scale in the United States: current status and best practices

Abstract

Chronic myeloid leukemia (CML) results from the Philadelphia chromosome (Ph) translocation and expression of its fusion oncoprotein BCR-ABL1. BCR-ABL1 tyrosine kinase inhibitors (TKIs) are the standard therapy for Ph-positive CML. Achievement of deep molecular responses (typically defined as ≥4-log reduction in BCR-ABL1 RNA levels) is an emerging treatment goal becoming attainable for more patients due to the availability of second-generation TKIs. Deep molecular responses are associated with improved long-term outcomes and are required prior to attempting cessation of treatment in treatment-free remission clinical trials. The National Comprehensive Cancer Network and European LeukemiaNet recommend regular monitoring of BCR-ABL1 RNA levels using real-time quantitative polymerase chain reaction (RQ-PCR). However, BCR-ABL1 RQ-PCR is a complex laboratory-developed test; routine quantitative results from clinical diagnostic laboratories may differ from those used to establish the recommendations. Although an International Scale (IS) was developed for standardized reporting of BCR-ABL1 RNA levels, IS adoption has been slow in the United States, but is now used by the vast majority of laboratories. Here, we discuss the importance of molecular monitoring in CML, gaps between current and best molecular monitoring practices in the United States, and challenges and potential solutions for universal IS adoption in the United States.

Keywords:

• BCR-ABL1
• deep molecular response
• International Scale
• real-time quantitative polymerase chain reaction
• myeloid leukemias and dysplasias

Introduction

Chronic myeloid leukemia (CML) is a hematologic malignancy characterized by the presence of the pathognomonic Philadelphia chromosome (Ph), which results from the translocation of chromosomes 9 and 22.[1] Ph-positive (Ph+) cells express the oncoprotein BCR-ABL1, a constitutively active tyrosine kinase that drives disease pathogenesis.[1] Since receiving US Food and Drug Administration (FDA) approval in 2001,[2] the BCR-ABL1 tyrosine kinase inhibitor (TKI) imatinib has improved the prognosis for patients with Ph + CML compared with previously available therapeutic approaches.[3,4] Several more potent second-generation TKIs (i.e. nilotinib, dasatinib, bosutinib, and ponatinib) have since been developed and approved for the treatment of Ph + CML in chronic phase (CML-CP) as frontline and/or subsequent treatment options.[2] TKI therapy represents the standard frontline treatment for CML-CP per National Comprehensive Cancer Network (NCCN) and European LeukemiaNet (ELN) recommendations.[2,5]

Long-term follow-up data from the International Randomized Study of Interferon versus STI571 (IRIS) trial show that most patients with CML-CP treated with frontline imatinib achieved complete cytogenetic response (CCyR; defined as the absence of Ph + metaphases) and major molecular response (MMR; BCR-ABL1 ≤ 0.1% on the International Scale [BCR-ABL1^IS]).[6,7] However, in the Evaluating Nilotinib Efficacy and Safety in Clinical Trials–Newly Diagnosed Patients (ENESTnd; nilotinib 300 mg twice daily [BID] versus nilotinib 400 mg BID versus imatinib) and DASatinib versus Imatinib Study In treatment-Naive CML patients (DASISION; dasatinib 100 mg once daily versus imatinib) studies, fewer patients achieved MMR or deeper molecular responses of molecular response 4 (MR⁴; BCR-ABL1^IS ≤ 0.01%) or molecular response 4.5 (MR^4.5; BCR-ABL1^IS ≤ 0.0032%) with frontline imatinib than with the respective second-generation TKI.[8,9]

Achievement of MR^4.5 is associated with improved long-term outcomes compared with lesser levels of molecular response [10,11] and is an emerging treatment goal for patients with CML. In addition, sustained deep molecular response is an important eligibility criterion for attempting to stop TKI therapy in the context of treatment-free remission clinical trials;[12] both determination of eligibility and monitoring of patients while attempting treatment-free remission in these studies require sensitive, reproducible evaluations of BCR-ABL1 transcript levels. The increasing proportion of patients with CML who are able to achieve deep molecular response on second-generation TKIs and the potential for treatment-free remission highlight the need for standardization of laboratory practices for monitoring molecular responses (i.e. measurement of BCR-ABL1 transcripts by real-time quantitative polymerase chain reaction [RQ-PCR]).[13]

Nearly a decade ago, a National Institutes of Health (NIH) Consensus Group proposed the International Scale (IS) to standardize RQ-PCR monitoring of BCR-ABL1 transcript levels in patients with CML.[14] Subsequent international efforts resulted in the development and implementation of the IS;[15] however, adoption of the IS in clinical diagnostic laboratories in the United States has been slow. Although the vast majority of US laboratories now use IS-based monitoring, adoption remains incomplete. In this review, we give an overview of best practices for molecular monitoring of patients with CML, with an emphasis on the importance of BCR-ABL1 RQ-PCR testing standardized to the IS, and discuss challenges to universal IS adoption in the United States.

Rationale for development of the International Scale

Current NCCN and ELN recommendations call for regular molecular monitoring of patients with CML on TKI therapy using BCR-ABL1 RQ-PCR standardized to the IS.[2,5] Molecular response assessments are typically performed using peripheral blood or bone marrow samples. Using RQ-PCR, levels of BCR-ABL1 transcripts can be measured relative to those of a reference gene, such as ABL1 or glucuronidase beta (GUSB).[2,5,16] BCR-ABL1 RQ-PCR allows for a 2- to 3-log higher level of analytic sensitivity than that of cytogenetic and/or fluorescence in situ hybridization assessments, which, along with morphologic hematologic assessments, are also used to measure response to TKI treatment.[17] With optimized methodology, RQ-PCR can detect a single cell expressing BCR-ABL1 in a background of ≈100,000 normal cells.[18] However, BCR-ABL1 RQ-PCR is a nonautomated, multistep process [Figure 1] subject to many sources of variation, including the handling and processing of the blood samples, the RQ-PCR reaction and its components, and the normalization and reporting of results.[13]

Figure 1. Key steps in the BCR-ABL1 RQ-PCR assay.[44] cDNA, complementary DNA; RQ-PCR, real-time quantitative polymerase chain reaction. *Steps taking place outside the laboratory.

Sources of assay variation

Because RNA molecules are inherently prone to degradation, initial sample handling and transportation to the testing laboratory are key determinants of successful BCR-ABL1 RQ-PCR testing, but are difficult for the laboratory to control. Thus, any variation in sample processing, storage temperature, or travel duration can impact RNA stability and the integrity of BCR-ABL1 RQ-PCR results. There are numerous other opportunities for measurement imprecision and inaccuracy in the assay as well, including specimen collection (how and from what source [blood or bone marrow]); leukocyte and RNA preparation methods; complementary DNA (cDNA) synthesis heterogeneity; choice of enzymes, primers, and other RQ-PCR assay components; choice of calibrator materials; IS conversion methods; and data analysis and reporting methods. As levels of BCR-ABL1 may differ between peripheral blood and bone marrow at a given time, samples should be obtained from the same source in serial follow-up testing.[19]

The use of an RQ-PCR reference gene (and the reporting of BCR-ABL1 RNA levels as a ratio relative to reference gene RNA levels) can partially correct for differences in sample handling and preparation. However, the choice of reference gene may introduce additional sources of quantitative assay variability. The ideal reference gene for an RQ-PCR assay is one whose RNA is expressed at a level similar to that of BCR-ABL1, has degradation kinetics parallel to those of BCR-ABL1 RNA, does not vary between cell types or patients, and is stable across time and across disease and treatment status. One analysis of 14 candidate reference genes found the most stable to be ABL1, beta-2-microglubulin (B2M), and GUSB.[20] Subsequent work similarly identified GUSB as an optimal reference gene.[21,22] Although ABL1 is the reference gene used by the majority of US laboratories, the ratio of BCR-ABL1 to ABL1 is nonlinear when BCR-ABL1 transcript levels are high (BCR-ABL1^IS > 10%);[23] BCR-ABL1^IS values may thus be skewed at these high levels when ABL1 is used as the RQ-PCR reference gene.[13]

The IS is anchored by BCR-ABL1 to reference gene RNA ratios at two clinically significant threshold levels, which were defined using samples from patients in the IRIS trial [Figure 2].[14] The first anchor is the ‘baseline’ BCR-ABL1 RNA level (100% IS; against which all post-treatment samples can be compared); the quantitative value of the IS baseline was defined as the median pretreatment BCR-ABL1 to reference gene ratio among 30 newly diagnosed patients with CML-CP in the IRIS trial.[2,14,24] The second anchor is MMR, defined as a 1000-fold (3-log) posttreatment reduction in BCR-ABL1 RNA relative to the pretreatment median baseline. MMR was established in IRIS as a critical prognostic threshold (and treatment goal) at 12–18 months of imatinib therapy for predicting a prolonged subsequent remission.[7,25] BCR-ABL1 levels on the IS are reported as percentages; thus, BCR-ABL1^IS ≤ 0.1%, ≤ 0.01%, and ≤0.0032% are equivalent to MMR (also known MR³), MR⁴, and MR^4.5, respectively, with the superscript indicating the log reduction in BCR-ABL1 RNA from the IRIS baseline.[2,5]

Figure 2. Clinically relevant threshold levels of BCR-ABL1^IS.[16,17] BCR-ABL1^IS, BCR-ABL1 measured on the International Scale; IRIS: International Randomized Study of Interferon versus STI571 study; MR: molecular response; MMR: major molecular response.

Although the 30 samples from IRIS upon which the IS baseline level was defined have been exhausted, laboratories are able to calibrate their individual BCR-ABL1 RQ-PCR assays to the IS by either of two methods. The first method requires an exchange of a series of samples (20–30, spanning the entire analytic measurement range) with a reference laboratory that has already been calibrated to the IS and that maintains strict quality control standards for BCR-ABL1 RQ-PCR.[13,15,18] Following the sample exchange, a laboratory-specific conversion factor (CF) is derived using linear regression and bias analyses.[15] The CF is calculated as the antilog of the mean bias between the two laboratories.[15] After validation of the CF via a second sample exchange exercise, this laboratory-specific CF allows for conversion to the IS simply by multiplying the laboratory’s reported BCR-ABL1 to reference gene ratio by the CF.[15,23] In a study of 38 international laboratories, CFs calculated via sample exchange with the Adelaide laboratory (which served as the central reference laboratory for the IRIS trial) ranged from 0.18 to 13.5,[15] confirming significant interlaboratory variability of different testing methods. After these laboratories converted their assays to the IS using their laboratory-specific CF, approximately 90% of the laboratories had interlaboratory assay bias ≤1.2-fold.[15] The stability of CFs for a particular laboratory and the frequency with which CFs must be revalidated remain unknown.[13] Although the sample exchange method can effectively standardize individual laboratories to the IS, it is laborious, time-consuming, and logistically impractical for the routine clinical diagnostic laboratory.

The second method for standardization to the IS is to acquire a secondary BCR-ABL1 reference standard that has been accurately calibrated to the World Health Organization (WHO) primary reference standard for BCR-ABL1 RNA. This will likely be the predominant method by which laboratories calibrate and standardize IS-based assays in the future. This method of standardization has been made possible by the recent adoption and creation by the WHO of an International Genetic Reference Panel for BCR-ABL1 testing, consisting of four reference reagents (stably stored lyophilized cell line mixtures) assigned a fixed BCR-ABL1^IS value (≈10%, ≈1%, ≈0.1% [MMR], or ≈0.01%).[26] Because this WHO primary reference material is only available in limited quantities, practical routine laboratory quality control requires the subsequent creation of a secondary BCR-ABL1 reference standard that has been synthesized in large quantities, is stable after prolonged storage, and has been rigorously calibrated to the WHO primary reference standard. Several of these secondary BCR-ABL1 reference standards are now commercially available, separately and/or packaged into BCR-ABL1 testing kits, although they are not yet FDA approved.[27] The ideal secondary reference standard must be robustly calibrated (on the IS) to the WHO primary reference standard and must cover the entire broad range of expected BCR-ABL1 levels; it must also be composed of material that undergoes all steps of the RQ-PCR process (including RNA extraction) and be stable from time-dependent degradation, available in large quantities, and affordable. With broad availability of secondary reference standards, laboratories will be able to routinely monitor assay accuracy, precision, and time-dependent drift and, thus, may implement the same quality control procedures as they would for any other laboratory analyte in the clinical diagnostics arena.

Molecular monitoring in CML: importance and current recommendations

Regular molecular monitoring is crucial for assessing response to therapy as well as for the early identification of nonadherence, treatment resistance, or treatment failure in patients with CML. For example, a chart review analysis found that patients with frequent molecular monitoring per ELN guidelines during the first year of treatment were less likely to progress or die on study.[28] A separate chart review analysis found that patients who had 1–2 or 3–4 BCR-ABL1 RQ-PCR tests per year experienced a lower risk of progression than patients who had no BCR-ABL1 RQ-PCR tests.[29] Regular molecular monitoring in patients with CML is also associated with improved treatment compliance [30] and economic benefits.[31]

To assess leukemic burden, the NCCN recommends molecular monitoring via BCR-ABL1^IS RQ-PCR at diagnosis and then every 3 months after initiating treatment. Per the NCCN, upon achievement of CCyR (approximately equivalent to BCR-ABL1^IS 1%), BCR-ABL1^IS RQ-PCR should be performed every 3 months for 2 years, and every 3–6 months thereafter.[2] If a patient in MMR experiences a 1-log increase in BCR-ABL1^IS levels, the NCCN recommends BCR-ABL1^IS RQ-PCR be repeated in 1–3 months and the possible presence of kinase domain mutations be assessed.[2] For patients who have not achieved MMR with a 1-log increase in BCR-ABL1^IS transcripts, in addition to BCR-ABL1 kinase domain mutation analysis, the NCCN recommends bone marrow cytogenetics, which can detect additional chromosomal abnormalities in Ph + cells.[2]

By comparison, the ELN recommends that, once diagnosis is confirmed, patients receive BCR-ABL1^IS RQ-PCR testing every 3 months until achievement of MMR, and every 3–6 months thereafter. However, if BCR-ABL1^IS transcript levels increase >5-fold at a single time point and MMR is lost, molecular monitoring should occur more frequently and patients should be queried about compliance.[5] ELN recommends kinase mutation analysis for patients with confirmed loss of MMR on two consecutive tests, with one being BCR-ABL1^IS ≥ 1%.

The NCCN and ELN also provide guidance for interpretation of a patient’s molecular response to inform appropriate follow-up care. For example, the NCCN guidelines consider early molecular response (EMR; BCR-ABL1^IS ≤ 10%) at 3 months (and 6 months) a goal of treatment.[2] Failing to meet this threshold may be grounds for a switch in therapy, depending on the frontline TKI and other factors.[2] ELN recommendations similarly consider EMR at 3 months to be an optimal treatment response.[5] Molecular response milestones recommended by the NCCN and ELN for later time points are summarized in Table 1. Mutational testing is recommended by the NCCN for patients with inadequate initial response, loss of response, 1-log increase in BCR-ABL1^IS RNA with loss of MMR, or disease progression [2]; the ELN recommends mutational testing for patients meeting the criteria for treatment failure or warning, and in the case of disease progression.[5]

Table 1. Molecular response goals at specified times from treatment initiation, according to the NCCN and ELN.[2,5]

		Time from treatment initiation
	Molecular response	3 months	6 months	12 months
NCCN [2]	EMR	X^a	X^a
	BCR-ABL1^IS ≤ 1%  (or CCyR^b)			X^b
ELN [5]	EMR	X^c
	BCR-ABL1^IS < 1%		X^d
	MMR			X^e

BCR-ABL1^IS: BCR-ABL1 measured on the International Scale; CCyR: complete cytogenetic response; ELN: European LeukemiaNet; EMR: early molecular response (BCR-ABL1^IS ≤ 10%); MMR: major molecular response (BCR-ABL1^IS ≤ 0.1%); NCCN: National Comprehensive Cancer Network; Ph +; Philadelphia chromosome-positive; TKI: tyrosine kinase inhibitor.

^a For patients without EMR at 3 or 6 months, NCCN guidelines recommend evaluation of patient compliance, drug–drug interactions, and a mutational analysis. Follow-up care options include enrollment in a clinical trial, switch to an alternative TKI, dose escalation (at 3 months) if patient is on frontline imatinib (to a maximum of 800 mg daily), or continuation at the same dose of TKI or switch to an alternative TKI other than imatinib (at 3 months) if patient is on frontline nilotinib or dasatinib. For patients without EMR at 6 months, switch to an alternative TKI other than imatinib is recommended for all patients. Evaluation for hematopoietic stem cell transplantation depending on response to TKI therapy is recommended in all cases if the response milestones are not achieved.[2]

^b The NCCN guidelines consider CCyR (or BCR-ABL1^IS ≤ 1%) to be the optimal response milestone with ≥12 months of treatment. The preferred NCCN recommendation for patients with BCR-ABL1^IS ≤ 10% but >1% at 12 months is to switch to an alternative TKI other than imatinib, although continuation at the same dose of TKI and dose escalation of imatinib to a maximum of 800 mg daily are listed as alternative options. Bone marrow cytogenetics are recommended for patients without CCyR or MMR by 12 months.[2]

^c At 3 months, BCR-ABL1^IS > 10% is considered a warning response according to ELN recommendations; the ELN recommends that patients with warning responses receive more frequent monitoring to allow for timely changes in therapy, in case of treatment failure.[5] The ELN recommendations do not designate any molecular response level as indicative of treatment failure at 3 months (treatment failure at 3 months is defined as lack of complete hematologic response and/or >95% Ph + metaphases).[5]

^d At 6 months, BCR-ABL1^IS ≥ 1% to 10% is considered a warning response, and BCR-ABL1^IS > 10% is considered treatment failure according to ELN recommendations.[5] For patients with treatment failure, ELN recommendations state that patients should receive a different treatment to limit the risk of progression and death.[5]

^e At 12 months, BCR-ABL1^IS > 0.1% to 1% is considered a warning response, and BCR-ABL1^IS > 1% is considered treatment failure according to ELN recommendations.[5]

The NCCN and ELN recommendations for interpretation of early BCR-ABL1^IS transcript level reductions are based on clinical data showing that achievement of EMR predicted improved outcomes in patients with CML treated with imatinib [32–35] or with second-generation TKIs [8,9,36] in the frontline setting. In the German CML-IV study, BCR-ABL1^IS > 10% at 3 months (i.e. EMR failure) following initiation of imatinib therapy was associated with lower rates of 5-year progression-free survival (PFS) and overall survival (OS).[35] With 3 years of follow-up in the DASISION study of frontline dasatinib versus imatinib, achievement of EMR was associated with higher rates of PFS and OS with dasatinib or imatinib treatment; patients with EMR failure had a 6-fold higher risk of disease progression to accelerated phase/blast crisis.[9] Similarly, in the ENESTnd frontline nilotinib versus imatinib trial, EMR was predictive of higher 4-year rates of PFS in all treatment arms and OS in the nilotinib 300 mg BID and imatinib treatment arms.[8] In the Bosutinib Efficacy and Safety in Newly Diagnosed Chronic Myeloid Leukemia (BELA) study in patients with newly diagnosed CML, achievement of EMR was associated with higher rates of CCyR and MMR at 1 and 2 years from initiation of treatment with both bosutinib and imatinib.[36] Although achievement of EMR represents a good initial treatment response with both imatinib and second-generation TKIs, a higher proportion of patients achieved EMR with nilotinib, dasatinib, and bosutinib than with imatinib in these studies.[8,9,36]

Increasingly, evidence from clinical trials has also demonstrated the importance of a deep molecular response.[16,37] One study of 266 patients with CML-CP treated with frontline imatinib at the University Hospitals of Bordeaux and Lyon observed higher rates of event-free survival and failure-free survival in patients achieving MR^4.5 compared with patients achieving CCyR.[11] A separate study found a correlation between achievement of undetectable disease (assay sensitivity: 1 BCR-ABL1 transcript in 100,000 ABL1 copies; equivalent to MR⁵ on the IS) and transformation-free survival and OS (versus patients with BCR-ABL1^IS ≥ 0.1%) on imatinib, nilotinib, or dasatinib therapy.[38] In the German CML-IV study, there was a correlation between achievement of MR^4.5 and OS in patients treated with imatinib, and no patient who achieved MR^4.5 had progressed after a median follow-up of 3 years.[10] In addition, several treatment-free remission studies have been initiated in patients following treatment with imatinib and second-generation TKIs, with stringent molecular response criteria for eligibility. Extremely sensitive RQ-PCR assays are essential for these treatment-free remission trials, in which both cessation and reinitiation of therapy are solely dependent on documenting a very low level of BCR-ABL1 RNA (typically ≤ MR^4.5) at very frequent testing intervals. Thus, if the monitoring laboratory has an RQ-PCR assay whose limit of detection is suboptimal, a sample may be inappropriately reported as ‘undetectable BCR-ABL1’. For example, in a laboratory with a substandard limit of detection of MR^3.5 (BCR-ABL1^IS ≤ 0.032%), a sample with a true MR⁴ would be falsely called ‘undetectable’, and therapy could be inappropriately withdrawn under the false impression that the patient has a lower disease burden. Precision of the RQ-PCR assays is another crucial analytic parameter given that serial increases in already-low transcript levels are a criterion for therapy reinitiation in some treatment-free remission studies.

Molecular monitoring in the United States: current status

Despite the importance of regular molecular monitoring for patients with CML, very few community practitioners in the United States actually perform molecular monitoring per NCCN and ELN recommendations.[39] For example, a 2012 report revealed that only 16% of patients treated with frontline imatinib in the United States underwent molecular monitoring at 3 months from the start of therapy.[40] A retrospective study of 300 patients with CML-CP treated in community practices of the McKesson Specialty Health/US Oncology Network between July 2007 and September 2012 found that only half of patients had molecular monitoring at any time during the observation period, and only 23% had monitoring by an IS-standardized laboratory.[41] The adoption of IS-standardized reporting by US-based clinical diagnostic laboratories, although initially quite slow (with only 29% of laboratories self-reporting IS use in 2012), has been progressively increasing.[13] If variability in BCR-ABL1 RQ-PCR testing across laboratories is not standardized in the minority of US laboratories not using the IS, the results from these laboratories will not be translatable to consensus clinical actions. It is not possible to follow NCCN guidelines using results from non-IS laboratories because NCCN guidelines are based on IS-defined molecular response thresholds.

Without any FDA-approved reagents for BCR-ABL1 monitoring currently available, clinical diagnostic laboratories in the United States must develop, validate, and monitor the performance of the individual tests within their own laboratories using Clinical Laboratory Improvement Amendments criteria for analytic and clinical validity. As a result, these laboratory-developed tests use heterogeneous laboratory-specific reagents, methods, and equipment that will contribute to the interlaboratory variation in BCR-ABL1 testing.[23,42] One study of 38 North American laboratories published in 2007 revealed varied instrumentation platforms, reagents, and analysis methodology among participants.[43] Another study, compiled from unbiased College of American Pathologists (CAP) proficiency testing data, showed significant imprecision and inaccuracy across ≈ 100 participating laboratories in measuring blinded aliquots of a single 1-in-10,000 dilution of RNA from the CML cell line K562.[13] In that study, the average laboratory-reported BCR-ABL1 log reduction differed from the expected value (i.e. 4-log reduction) by 0.7 log, and results varied between laboratories by ≈ 1 log (i.e. interlaboratory standard deviation of 10-fold).[13] Without IS standardization, this high degree of heterogeneity of testing procedures may negatively impact patient care. A practical example illustrating this would be a patient with CML who changed monitoring laboratories during therapy, perhaps due to relocation or changes in employer insurance plans. Without standardization of BCR-ABL1 RQ-PCR testing methodology, the reported BCR-ABL1 transcript measurements at the second laboratory may not be comparable to the first, making appropriate clinical management more challenging, compromising accurate assessment of disease burden, and potentially resulting in misinformed patient management decisions.[44]

The slow uptake of standardization to the IS in the United States may be due to a variety of factors, such as an educational gap, an over-reliance on less sensitive non-PCR-based monitoring tools, cost considerations, or the clinician’s lack of access (or decision-making authority) to a trusted laboratory or pathologist partner. Variability in BCR-ABL1 measurement may also play a role. In comparison with the United States, molecular monitoring in Europe is conducted in large part by centralized academic laboratories that have full access to the patient’s clinical and laboratory data and have a standardized process for BCR-ABL1 testing and reporting. In contrast, based on survey results, half of the US practitioners reported sending their BCR-ABL1 testing samples to commercial, nonacademic laboratories [39] that must then interpret the stand-alone BCR-ABL1 data without the benefit of knowing the patient’s full clinical and laboratory record.

Challenges for International Scale adoption in the United States

Several challenges remain for universal adoption of the IS in the United States. For some laboratories, the current process for standardization to the IS may seem too logistically demanding. The process requires a significant fiscal and labor commitment, and may not be seen as the best investment of resources for laboratories that do not process a high volume of CML patient samples.[2] Additionally, laboratories may not be hearing strong objections from their physician clients as to the importance of IS standardization. This may be due to a low representation of patients with CML in the typical oncologist’s daily practice or knowledge gaps among both clinicians and pathologists. If demands for standardization were forthcoming, laboratories would likely be motivated from both a clinical care and economic perspective to invest in an IS recalibration project. However, the biggest obstacle to universal adoption of IS-based reporting is likely the current absence of FDA-approved reagent kits for BCR-ABL1 quantitation, which would be calibrated to the IS and clinically and analytically validated. Widespread commercial availability of such FDA-approved nationally standardized reagents with built-in IS assay calibrators would significantly reduce the initial financial cost and time commitment that may now prevent laboratories from investing in improvements to their assays. In contrast, future long-term operational costs for the laboratory will likely be significantly increased for these putative FDA-approved BCR-ABL1 testing kits compared with relatively inexpensive laboratory-developed reagents.

Currently, efforts are ongoing to address the remaining challenges for BCR-ABL1 standardization. As part of proposing the IS, the NIH Consensus Group put forth recommendations for BCR-ABL1 RQ-PCR methodology to reduce interlaboratory variability in BCR-ABL1 transcript level measurement.[14,24] In addition, several professional societies in the United States, including CAP, have developed guidelines for laboratory quantification of BCR-ABL1.[45] CAP also offers (and grades) semiannual blinded IS-based BCR-ABL1 proficiency testing challenges to participating laboratories. The repeated failure to accurately quantify these CAP proficiency testing samples will ultimately lead to inspection deficiencies and other punitive ramifications, including, if the failures are chronic, a loss of testing privileges. The European Treatment and Outcome Study (EUTOS) group is also evaluating the BCR-ABL1 test performance of several international laboratories in order to develop a set of recommendations on methodology, definitions, and other safeguards for quality assurance that will enhance interlaboratory data comparability.[16] As part of this effort, EUTOS has developed a network of national and regional reference laboratories standardized to the IS [46] and has published detailed technical recommendations for reproducible measurement of molecular responses, with particular emphasis on scoring deep molecular responses (including guidance on the use of sample replicates, the minimum number of reference gene RNA copies necessary to evaluate each level of response, and the impact of an assay’s limit of detection on the accuracy and precision of results).[37] There are also several diagnostic laboratory reagent manufacturers that are planning, operating, or analyzing data from clinical trials designed to assess standardized BCR-ABL1 test systems that, if successful, could result in the availability of standardized FDA-approved test methods. In the interim, several of these same reagent vendors can provide laboratories with IS-calibrated secondary reference standards that can be used to initiate or validate IS-based reporting of a laboratory’s own laboratory-developed BCR-ABL1 test.

Conclusion

Regular molecular monitoring is integral to the optimal management of patients with CML [2,5] and is associated with a number of clinical and other benefits.[28–31] The availability of multiple TKI treatment choices and the emergence of deep molecular responses as a treatment goal have increased the importance of accurate, precise, and sensitive measurement of BCR-ABL1. Although a number of challenges remain,[13] the widespread adoption of the IS in the United States is an important step toward global standardization of CML data reporting. International and US-focused efforts for standardization of BCR-ABL1 measurement are ongoing.

Potential conflict of interest

Disclosure forms provided by the authors are available with the full text of this article at http://dx.doi.org/10.1080/10428194.2016.1190974.