MRD evaluation of AML in clinical practice: are we there yet?

Abstract

MRD technologies increase our ability to measure response in acute myeloid leukemia (AML) beyond the limitations of morphology. When applied in clinical trials, molecular and immunophenotypic MRD assays have improved prognostic precision, providing a strong rationale for their use to guide treatment, as well as to measure its effectiveness. Initiatives such as those from the European Leukemia Network now provide a collaborative knowledge-based framework for selection and implementation of MRD assays most appropriate for defined genetic subgroups. For patients with mutated-NPM1 AML, quantitative polymerase chain reaction (qPCR) monitoring of mutated-NPM1 transcripts postinduction and sequentially after treatment has emerged as a highly sensitive and specific tool to predict relapse and potential benefit from allogeneic transplant. Flow cytometric MRD after induction is prognostic across genetic risk groups and can identify those patients in the wild-type NPM1 intermediate AML subgroup with a very high risk for relapse. In parallel with these data, advances in genetic profiling have extended understanding of the etiology and the complex dynamic clonal nature of AML, as well as created the opportunity for MRD monitoring using next-generation sequencing (NGS). NGS AML MRD detection can stratify outcomes and has potential utility in the peri-allogeneic transplant setting. However, there remain challenges inherent in the NGS approach of multiplex quantification of mutations to track AML MRD. Although further development of this methodology, together with orthogonal testing, will clarify its relevance for routine clinical use, particularly for patients lacking a qPCR genetic target, established validated MRD assays can already provide information to direct clinical practice.

Figure 1.

(A) Suggested management algorithm for patients with AML with a molecular MRD target. [1] ELN favorable risk patients with <4-log reduction in NPM1-mutant transcripts after first induction are shown to benefit from a CR1 allograft, and any positivity in the peripheral blood (PB) after second induction is associated with a very high risk for relapse. [2] MRD positivity > 200 copies per 10⁵ ABL (ie, molecular persistence) and serially increasing transcript levels after treatment (ie, molecular progression) reliably predict relapse. [3] At the end of treatment, patients with CBF AML with high or serially increasing transcript levels are destined to relapse (relevant thresholds are >500 copies [per 10⁵ ABL] of RUNX1/RUNX1T1 in the BM or >100 copies in the PB, and > 50 copies of CBFB/MYH11 in the bone marrow [BM] or >10 copies in the PB). Salvage according to (C) should be considered for these patients, although there is no evidence that this improves outcome. Conversely, patients with low copy numbers below these thresholds can be safely monitored according to (B). [4] Although CBF patients with an early unfavorable MRD response have a higher risk for relapse, there is insufficient evidence to warrant treatment change; however, this may prompt initiation of an early donor search. Salvage may be considered in cases with extremely poor early response or if there is an increase while on treatment (ie, molecular progression). [5] Although there is no evidence that standard-risk patients who remain MRD positive benefit from transplant, this is a reasonable approach and is adopted in the current NCRI AML19 and MyeChild01 protocols. (B) Suggested algorithm for sequential monitoring after treatment. Patients with conversion to MRD positivity, confirmed on a second sample with >1-log rise, should be diagnosed with molecular relapse and treated as shown. (C) Possible peri-transplant management strategy. [1] Patients with an NPM1 mutation without FLT3 ITD who have transcript levels below 1000 copies in the BM or 200 copies in the PB have a very good outcome after allograft, it is uncertain whether these patients benefit from salvage chemotherapy. [2] Patients with high levels of MRD after salvage, without an adequate response to donor lymphocyte infusion (DLI), as well as those to whom these standard therapies cannot be given should be considered for investigational approaches. Figure by Richard Dillon, NCRI Group.

Figure 2.

(A) qPCR is a common method for quantification of nucleic acid with real-time monitoring of the amplification of target of interest (eg, variant sequence shown with red X). Advantages include ubiquitous presence in most clinical laboratories, fast turnaround time, high sample throughput, and broad dynamic range. Disadvantages include limited number of suitable targets/assays available, relative lack of multiplexing ability, need to validate each target/assay individually, potential for false-negative results because of sample impurity, and limited ability to accurately discriminate between very low levels of target as seen in MRD. (B) Rather than performing the PCR reaction in “bulk,” digital PCR partitions the template of interest into individual compartments (top), improving the performance compared with qPCR because of the lower background error rate (lower right), elimination of template competition, and digital result output, allowing absolute quantification (lower left). Lack of deep multiplexing ability and the need to validate each target/assay individually remain limitations. (C) NGS has revolutionized the initial clinical diagnostic evaluation of AML by allowing for simultaneous evaluation of multiple target regions typically selected from those known to be often mutated in AML. NGS is useful for discovery of mutations present in the range from 5% to 100% of a sample (VAF). However, not all variants detected will be pathogenic somatic mutations, and care should be taken to consider the possibility of identification of homozygous or heterozygous germline variants, as well as loss-of-heterozygosity (LOH) events. Variant discovery below a VAF of 5% using panels designed for profiling variants at diagnosis is challenging because of the lack of sensitivity and high false-positive rates. Red asterisks represent low-level variant calls that should be regarded with particular caution as within the range of background error for conventional NGS. (D) NGS for AML MRD performed in recent high-quality research studies has typically included error correction (upper), by incorporation of UMIs, followed by consensus determination of true (red X) variants vs false positives introduced by the technique (green X) and/or bioinformatic approaches to model background error rates at each nucleotide position in those not having a variant and determine the probability that the observed variant is a true positive (red asterisk) (lower). Figure by Erina He, National Institutes of Health Medical Arts.

Figure 3.

(A) Strategy applied for flow cytometric AML MRD multicenter harmonization by the ALFA Intergroup. [1] Rationale of AML MRD flow panel design was based on simplicity, reproducibility, and cost. Tube 1 was a core combination for LAIP detected at diagnosis and/or by different-from-normal analysis, tube 2 was targeted to aberrancies of CD34⁺CD38⁻ cells (immunophenotypic LSCs), and tube 3 was an optional development tube for monocytic aberrancies. [2] Flow cytometer fluorescent settings were harmonized (“mirrored”) between Canto vs Navios cytometer platforms. Voltages were set to reach target mean fluorescence intensity (MFI) values by acquisition of rainbow calibration beads without compensation for fluorescent channels FL1 to FL8 on the Canto cytometers; these rainbow bead settings were transposed to Navios cytometers by applying MFI target = Canto target/256. Mirrored (superimposable) target peaks for both cytometers are shown for FL1 and FL8 fluorescent channels with an example of resulting comparable antibody profiles between cytometer platforms. (B) [3] Quality controls for reproducibility of staining profiles from harmonized cytometer settings/sample processing between cytometers/laboratories. Examples shown are for CD117 and CD38 expression intensity on CD34⁺ gated mononuclear cells of tube 1 from 10 shared BM samples stained and then acquired on Canto or Navios flow cytometers. Intensity profiles are similar between cytometer platforms for each sample. [4] External quality assessment for all harmonized steps from preanalytical to final gating analyses by distribution of a normal BM to 22 participating laboratories (cytometer platforms: 12 Cantos, 10 Navios). Example shows that strong reproducibility can be achieved in the detection of rare events (shown for CD34⁺CD38⁻) among 22 participating laboratories. Figure by Christophe Roumier and Adriana Plesa, ALFA.

Figure 4.

CD34/CD38 expression pattern of blasts from 5 diagnostic AML samples showing the variability in the frequency of the most immature leukemia cells (CD34⁺CD38⁻; orange) compared with normal BM. CN, normal cytogenetics. Figure by Adriana Plesa and Christophe Roumier, ALFA.