European LeukemiaNet laboratory recommendations for the diagnosis and management of chronic myeloid leukemia

Abstract

From the laboratory perspective, effective management of patients with chronic myeloid leukemia (CML) requires accurate diagnosis, assessment of prognostic markers, sequential assessment of levels of residual disease and investigation of possible reasons for resistance, relapse or progression. Our scientific and clinical knowledge underpinning these requirements continues to evolve, as do laboratory methods and technologies. The European LeukemiaNet convened an expert panel to critically consider the current status of genetic laboratory approaches to help diagnose and manage CML patients. Our recommendations focus on current best practice and highlight the strengths and pitfalls of commonly used laboratory tests.

Fig. 1. BCR::ABL1 fusions.

The genomic configurations of the ABL1 (11 exons) and BCR genes (23 exons) are shown at the top. In the great majority of CML patients the genomic breakpoints fall in the regions indicated by the two horizontal double blue arrows, i.e. 5’ of ABL1 exon 2 and between exons 13 and 15 of BCR, giving rise to e13a2 and/or e14a2 BCR::ABL1 mRNA fusions. The approximate positions of recurrent variant breakpoints are indicated by the red vertical arrows giving rise to mRNA fusions that involve different BCR exons and/or ABL1 exon 3. Some of these atypical variants are illustrated (e1a2, e6a2, e8a2, e19a2, e13a3, e14a3) but other variants are seen in occasional cases. BCR exon 8 is out of frame with ABL1 exon 2; e8a2 fusions typically break within BCR exon 8 or additional intron-derived sequences are retained to maintain the reading frame. Diagrams are illustrative and not to scale.

Fig. 2. Low level expression of e1a2 BCR::ABL1 in e13a2/e14a2 cases.

Top panel: illustration of low level e1a2 mRNA transcripts generated by alternative splicing in a patient expressing high level e14a2 BCR::ABL1. Bottom panel: example showing how an apparent e1a2 signal may be generated from intact e14a2 mRNA (depending on the positions of primers, probe, cDNA quality and amplification kinetics). Misinterpretation of low-level BCR::ABL1 products may lead to incorrect assignment of transcript BCR::ABL1 type, and incorrect molecular follow up on treatment.

Fig. 3. Model of two pathways to CML.

In most cases, a normal hematopoietic stem cell acquires the BCR::ABL1 fusion leading to the development of CP CML (top pathway). Acquisition of additional mutations and epigenetic changes eventually precipitate a block in differentiation and transition to BP. Some of the more commonly mutated genes at BP are indicated: TP53, RUNX1, ASXL1 and MECOM are associated with myeloid BP; IKZF1 and CDKN2A/B are associated with lymphoid BP. In some cases, BCR::ABL1 is acquired on a background of CH (either as a CH subclone or independently of the CH clone), for example CH driven by mutations in DNMT3A, TET2, ASXL1 or JAK2 (bottom pathway). These pre-existing mutations remain detectable during remission on TKI therapy. In contrast, for the top pathway any additional mutations found pre-treatment become undetectable at remission. The dotted line indicates a potential route to transformation from the CH clone (which may also develop ACAs) to a BCR::ABL1-negative myeloid neoplasm such as MPN or MDS. Figure created using BioRender.

Fig. 4. The International Scale for BCR::ABL1 mRNA measurement.

IS values are expressed as percentages and/or molecular response (MR) levels relative to the IRIS standardized baseline. Testing laboratories use either (i) RT-qPCR or RT-dPCR to measure the ratio of BCR::ABL1 mRNA to that of a reference gene (ABL1, GUSB, BCR) and convert to the IS by multiplying the raw result by a laboratory-specific conversion factors (CF), or (ii) a validated kit/system that directly outputs results on the IS.

Fig. 5. BCR::ABL1 TKD mutations.

Map of ABL1 indicating the mutations reported in the literature to be associated with resistance to ATP-competitive TKIs (imatinib, dasatinib, nilotinib, bosutinib and ponatinib).

Fig. 6. Practical algorithm indicating when BCR::ABL1 TKD mutation testing should be performed based on BCR::ABL1 transcript levels at each timepoint during therapy.

For example, at 6 months on treatment a patient who has not achieved (N) ≤ 1% BCR::ABL1^IS should be considered for BCR::ABL1 TKD testing. If ≤ 1% BCR::ABL1^IS has been achieved (Y) then TKD testing is not required and the patient should be reassessed at 12 months. TKD mutation testing is not generally recommended for any patient with ≤ 0.1% BCR::ABL1^IS, in part because the sensitivity of detection is low and amplification of the large fragment required for BCR::ABL1-specific mutation detection may be challenging.

Fig. 7. Schematic representation of the strategies commonly employed to amplify and sequence the TKD, either by Sanger sequencing or by NGS.

After reverse transcription of RNA to cDNA, a first step of amplification is performed using a forward primer mapping to BCR sequences close to the breakpoint (usually in exon 12 or 13, to amplify both e13- and e14- transcript variants) and a reverse primer mapping to ABL1 sequences immediately 3’ prime of the sequencing encoding the TKD (usually exon 10). The first amplicon may then be fragmented, indexed and sequenced by NGS or subjected to a second step of amplification using a series of internal primer pairs generating amplicons of suitable length for either Sanger sequencing or NGS.