The IMWG consensus response criteria for “fiow MRD-negative” requires a complete response and the absence of aberrant clonal plasma cells on BM aspirates using the EuroFlow-based NGF or an equivalent validated method with a minimum sensitivity of 1 in 10⁵ nucleated cells or higher [1]. This high-sensitivity NGF method includes an optimized two-tube 8-color antibody combination and adopts the bulk-lysis procedure as a reference for cell acquisition of ≥107. Moreover, it consists of six backbone fluorochrome-conjugated antibodies: CD38-FITC, CD56-PE, CD45-PerCPCy5.5, CD19-PECy7, CD138-BV421, and CD27-BV510. Each tube also contains additional fiuorochrome-conjugated antibodies: tube 1 includes CD117-APC and CD81-APC-C750, while tube 2 includes cytIgKappa-APC and cytIgLambda-APC-C750 [9]. This method incorporates automated gating and identification of plasma cells against a reference database of normal and patient BM. Since the IMWG response criteria require EuroFlow-based NGF or an equivalent validated method, several studies have reported alternative high-sensitivity fiow MRD methods [13-16]. For example, Roshal et al. performed a comparative study between a single 10-color tube and an 8-color 2-tube panel validated by the EuroFlow group [14]. A single-tube, 9-color panel assay was validated with CD200 added and surface staining only. Through an established validation process, this streamlined approach showed a detection level of 10^-5 and wide applicability [15]. Efforts to find a more efficient way for NGF MRD analysis include panel composition and the sample preparation method since it is one of the adjustable pre-analysis factors. Bayly et al. compared the post-lysis preparation technique with a modified pre-lysis technique from the EuroFlow method by reducing the number of washing steps [16]. A two-tube, 8-color panel was chosen with CD200 and CD3 added. The modified alternative approach showed equivalent performance, with no significant difference in the cellular composition between the samples. These validated approaches expand available options and contribute to the wider application of NGF MRD analysis.

Since fiow MRD consists of variations in panel selection, sample preparation, and data analysis, there is significant heterogeneity in the methodology, assessment, and reporting of the result. To adopt fiow MRD results as a surrogate endpoint across countries, laboratories, and various clinical trials, international harmonization efforts were proposed for data analysis and MM MRD using fiow cytometry [12, 17]. Soh et al. reported an international harmonized approach for a consensus algorithm by high-sensitivity fiow cytometry, including specific gating strategies, cell enumeration, and evaluation of sample adequacy [17]. A consensus gating protocol was designed, especially for the enumeration of “total analyzed cells,” which affects the calculation of the limit of detection (LOD) and lower limit of quantitation (LLOQ). In the harmonized gating approach, the first step was to exclude debris and doublets and identify singlets. The total number of analyzed cells included CD45⁺ leukocytes, CD45⁻ erythrocytes, and CD45^−/dim aberrant plasma cells. Next, total plasma cells expressing bright CD38⁺, CD138⁺, and CD45^+/− were identified using bivariate plots of CD38, CD138, and CD45. After which, considering the typical expression pattern of normal plasma cells, bright for CD38, positive for CD19, CD27, CD45, CD81, and negative for CD56, CD117, and polytypic cytoplasmic light chains, aberrant expression of abnormal plasma cells was considered. The presence of mast cells, erythroid precursors, myeloid precursors, or hematogones has been suggested as parameters of hemodilution in MRD samples [17]. When comparing the measurement of MM MRD between independent analysis using the in-house analysis method and the outcome after the adoption of the harmonized approach, the overall consistency between participants improved after adopting the consensus approach, especially in the enumeration of “total analyzed cells,” calculation of LOD and LLOQ, and in defining sample adequacy. Moreover, implementing a suitable harmonized approach diminishes interlaboratory variation, thereby enabling the application of fiow MRD assessment for the evaluation of MM disease status in a real-world setting.

As targeted immunotherapy using anti-CD38 monoclonal antibodies emerged, high-sensitivity fiow MRD approaches faced difficulties in plasma cell identification. While positive expression of CD38 and CD138 and decreased expression of CD45 are important in discriminating plasma cells from other lymphocytes, certain antibodies cannot be used as a gating marker occasionally. CD138 expression-downregulated subpopulations of abnormal plasma cells exist, especially in MM cells from the hypoxic bone marrow microenvironment [18]. Decreased CD38 positivity in the plasma cells was observed after immunotherapy and treatment with daratumumab [19]. Therefore, alternative plasma cell gating strategies have been proposed, including the CD38 multiepitope antibody incorporated in EuroFlow, the CD38 nanobody, the intracytoplasmic marker VS38c, CD319, and CD229 [19-22]. Moreover, VS38c recognizes a transmembrane protein on the rough endoplasmic reticulum and helps distinguish plasma cells from other hematopoietic cells, including polytypic B cells. Clonal plasma cells from patients with recent daratumumab exposure exhibit dim CD38 expression, while VS38c expression remains high. Antibody combinations, including alternative plasma cell markers CD319 and CD229, showed considerable resolution in discerning abnormal plasma cells, even in the plasma cells treated with an anti-CD319 monoclonal antibody. Although these alternative markers require each laboratory to redesign the current panel, clinical laboratories performing MM MRD analysis should consider the impact of these targeted immunotherapies when interpreting the results of treated specimens.

According to the IMWG response criteria, “sequencing MRD-negative” is defined as the absence of clonal plasma cells on BM aspirates using NGS, specifically via the LymphoSIGHT platform (currently ClonoSEQ, Adaptive Biotechnologies, WA) or an equivalent validated method with a minimum sensitivity of 1 in 10⁵ nucleated cells or higher, in addition to achieving a complete response [1, 23]. NGS-based MRD assessment focuses on the third complementarity-determining region (CDR3) at the junction of variable (V), diversity (D), and joining (J) immunoglobulin gene regions, particularly in the immunoglobulin heavy locus (IGH) gene [24]. To avoid the omission of clonotypes due to amplification bias and indistinct results against the background of polyclonal B cells, primer sets include the IGH frameworks 1, 2, and 3 (FR1, FR2, FR3), along with the upstream leader sequence region. Moreover, IGK (immunoglobulin kappa locus) and IGL (immunoglobulin lambda locus) can be used to track clonotypes. However, the two light chain genes are considered less reliable markers than IGH as light chain rearrangements involve only V and J regions and lack D segments [24].

Diverse genetic events result in patient-specific patterns through VDJ rearrangement, enabling the tracking of abnormal plasma cells. The sequence variations of clonal rearrangement provide genetic information for individual molecular profiling and are rarely affected by immunophenotypic changes due to immunotherapy [25-27]. Furthermore, analyzing serial samples allows for the investigation of clonal evolution. Moreover, NGS-based MRD can be applied to samples under various conditions, such as stored blood samples and formalin-fixed paraffin-embedded (FFPE) tissue. However, using good-quality FFPE samples with an appropriate fixation method, processing protocols, and a relatively short storage period is crucial to obtaining interpretable results for clinical diagnosis and monitoring [28]. In addition, NGS-based MRD assays can achieve a sensitivity of 10^-5 using a relatively low number of cell inputs. With the wide availability of NGS, many clinical laboratories are introducing NGS MRD testing. However, NGS-based MRD requires mandatory initial sample testing to identify patient-specific clones. A risk of trackable clone detection failure exists mainly due to sample quality, such as a low clonal plasma cell burden and hemodilution, and ineffective primer annealing problems caused by somatic hypermutation. These detection failure rates of trackable clones have been reported to be as high as 20%, especially when using the FR1 assay alone in patients with MM [28, 29]. However, the use of several methods can improve success rates. Firstly, incorporating more primer sets such as leader, FR2, and FR3 and combining them with the IGK assay enables clonal detection rates higher than 95% in all B-cell neoplasms, including MM. The addition of leader primers significantly increases the clonal detection rate, similar to the detection rate of assays using a combination of FR1, FR2, and FR3 [28]. Secondly, using an adequate-quality BM sample with a high tumor cell burden decreases the failure rate of identifying clonal sequences. Enrichment of CD138⁺ tumor cells is recommended.

ClonoSEQ has been approved by the United States Food and Drug Administration; however, it is not available in Korea as it requires samples to be sent to the central laboratory outside of Korea. On the contrary, the LymphoTrack assay (Invivoscribe, San Diego, CA) is a commercially available NGS-based clonality assay available in Korea for in-house sequencing and has been tested in MM [30]. The EuroClonality-NGS Consortium reported a standardized NGS panel assay to detect immunoglobulin and T-cell receptor gene rearrangements for clonality assessment [31, 32], and it is expected to be used for MRD analysis of various lymphoid neoplasms, including MM.

The prognostic values of MRD negativity using the ClonoSEQ assay have been proven in several clinical trials [33-35]. Patients who achieved CR and MRD negativity showed significantly better PFS than those who failed to reach CR or were MRD positive. Ching et al. demonstrated the ability of the clonoSEQ to detect MRD at a level of 10^-6 by serial dilution experiments, which showed a potential for a highly sensitive tool for the diagnosis and monitoring of lymphoid malignancies, including MM [23, 36]. Baseline identification of trackable clones and a validation study encompassing the establishment of LOD, linearity, precision, and reproducibility for the lymphoTRACK assay have been performed in previous studies [28, 37]. Medina et al. showed that MM plasma cells have different characteristics from other mature B-cell malignancies such as chronic lymphocytic leukemia and mantle cell lymphoma in the distribution of IGH gene repertoires [38]. In previous studies on Koreans with MM, the clonality detection rate was between 88–95% using the IGH FR1 assay and/or IGK assay and primer sets available in Korea [25, 30]. The profile of IGH and IGK gene usage in Koreans with MM was similar to those observed in previous studies conducted in the Western population. However, there were certain differences in the distribution of specific IGHV genes and the IGK rearrangement. These differences could be attributed to ethnicity-based variations, which need to be further investigated through future studies involving a larger number of Korean patients with MM [25].

As previously mentioned, the IMWG response criteria do not recommend a specific MRD method but emphasize the importance of using a validated assay with sufficiently high sensitivity [1]. Previous comparison studies between NGF and NGS-based MRD assays demonstrate a good correlation and a high concordance rate ranging from 80% to 93% [29, 39-42]. While there have been certain discordant cases, including MRD-positive cases detected only by NGF or NGS, most of these cases had MRD levels below 10^-5, suggesting a difference in sensitivity between the two methods or differences in the sampling procedure [28, 29, 39]. Sample quality may provide a few possible explanations, as it significantly impacts the results of both MRD methods, also related to the burden of plasma cells. Typically, the best-quality sample from the first pull of BM aspirate is used for cytomorphology, while subsequent samples with a risk of hemodilution are used for MRD analysis [29]. Differences in results may occur when comparing the highest-quality sample with other specimens that may have experienced hemodilution.

Moreover, the MRD status obtained from both methods is significantly associated with the PFS and OS of patients with MM [5, 39, 43]. In addition, Medina et al. evaluated the clinical significance of cases where both NGF and NGS assays showed negativity for MRD and discordant cases [39]. They observed no statistical differences in three-year PFS between the double-negative and discordant cases, although both subgroups had significantly better PFS rates compared to the double-negative cases. Ho et al. reported the risk of relapse in double-negative and discordant cases [29]. Approximately 20% of patients with double negativity using both NGF and NGS assays showed evidence of relapse within a median follow-up time of 36.3 months. Among the twelve discordant cases, three patients showed a relapse, and MRD was identified at a low level, near the limit of detection. When interpreting these low-level MRD results and evaluating their prognostic impact, sampling differences should be taken into consideration.

The characteristics of the NGF and NGS methods are compared in Table 1. NGF-based MRD analysis offers robust advantages, including increased sensitivity compared to conventional flow MRD, a fast turnaround time, does not require baseline sample testing, and can evaluate sample quality and the normal B-cell compartment by identifying hematopoietic cells other than plasma cells, such as mast cells, B-cell precursors, and erythroblasts. However, there are certain disadvantages. Additionally, it requires fresh samples, i.e., within 48 hours of obtaining the sample, to acquire viable cells, cannot be performed on FFPE tissue, and necessitates experienced personnel to carry out the procedure [9, 13, 44, 45]. On the contrary, NGS-based MRD assays can detect low levels of tumor cells and can be applied to various sample types other than blood. Moreover, it is minimally affected by immunotherapy, such as anti-CD38 monoclonal antibody treatment. Serial monitoring through NGS MRD enables the acquisition of valuable information regarding patient-specific clonotypes and clonal evolution. However, it requires baseline sample testing to identify patient-specific rearrangements, has a risk of detection failure for trackable clones, and needs a relatively longer turnaround time and is expensive [26, 39, 44]. However, the cost of MRD assessment using NGF or NGS assays varies between countries, hospitals, and laboratories and is greatly infiuenced by reimbursement policies [13, 26, 44]. Therefore, clinical laboratories should choose suitable MRD methods based on their availability and potential needs.

The field of MRD methods, procedures, assay sensitivity, evaluation, and reporting processes is highly heterogeneous. Standardization of the analytical aspects of NGF and NGS is necessary. However, achieving complete standardization can be challenging due to the need to unify methodologies. To promote international harmonization of MRD evaluation in clinical trials for MM, Costa et al. proposed several consensus statements [12]. The panelists provided recommendations on various topics, including analytical requirements, assay performance, BM sampling, and the timing of MRD testing in clinical trials. These consensus statements pertain to technical aspects and include the following criteria: the analytical performance of the assay must be validated with a defined limit of blank (LOB), LOQ, and LOD. MRD assays should achieve a LOD <10^-5 and should apply to over 90% of patients. If possible, sensitivity levels below 10^-6 should be reported. To ensure sufficient sensitivity, it is preferable to perform the MRD assay using the first pull of the BM aspirate.

The feasibility of MRD in peripheral blood (PB) was suggested [12]. MM is characterized by multiple lytic lesions and extramedullary involvement, suggesting the spread of abnormal plasma cells from the primary tumor to various BM niches through peripheral circulation [46]. High-sensitivity NGF can detect circulating tumor plasma cells (CTPCs) [47], allowing for monitoring of disease status using PB instead of BM specimens. This approach reduces the burden on patients, sparing them from painful procedures and facilitating periodic testing when needed. Furthermore, PB can be used in cases of poor BM aspirate quality due to patchy infiltration. Previous reports have highlighted the prognostic impact of CTPCs, although they have indicated that MRD detection in PB is less sensitive than in BM [48-50]. Nonetheless, MRD assessment with PB has the potential to complement BM-based MRD analysis. However, further investigation and cross-validation with BM-based MRD assays are required.