Clinically, the evidence of the malignant clone was and continues to be diagnosed by the presence of M protein, in addition to fluorescent in situ hybridization (FISH) studies to detect the potential underlying gene abnormalities (29, 30). Nonetheless, the requirement for tools to monitor the clonal mutation(s) and disease progression at the cellular level revealed the necessity of immunophenotyping techniques. For instance, from a clinical decision-making point of view, IPT has enabled cellular monitoring to detect signs of disease aggravation such as the absence of CD45 expression in MM (28), the overexpression of FOXP3 and CTLA4 in the BM (31), and the overexpression of CD229 on myeloma clonogenic precursors (32). By revealing the patterns of alterations at the cellular level, IPT can provide useful information that could aid in better clinical decision-making.

Additionally, myeloma cells from different patients are heterogeneous, creating unique genotypic patterns (33). The characteristics of each genetic variant affect how the aberrant markers are expressed and consequently determine the phenotype. In clinical hematology, the immunophenotypic and genotypic characterization of peripheral blood and BM-derived myeloma cells is frequently used for initial diagnosis, determination of aberrant expression, and tracking the efficacy of therapeutic regimens (34). Myeloma cells can be recognized using common aberrant markers included in the World Health Organization (WHO) classification of hematological malignancies. However, this list is not all-inclusive. For example, it has been shown that both healthy PCs and malignant myeloma cells exclusively express the plasma cell marker Syndecan-1 (CD138) (35). Recently, it was discovered that Syndecan-1 influences CD4+ T-cell modulation (36). In lymphoid and myeloid lineages, CD38, a transmembrane glycoprotein with ectoenzymatic activity, is typically expressed during different cell maturation and activation processes (37, 38). Additionally, it is expressed on both healthy and cancerous plasma cells (39). PCs are the only population that expresses BCMA (B-cell maturation antigen) (40). The basic phenotypic profile of myeloma cells is distinguished from normal PCs by a CD38-dim CD138-bright expression pattern, together with cytoplasmic light chain kappa or lambda restriction (41). Negative or low expression of CD45, negative CD19 expression in 95% of cases, and negative or low expression of CD27 and CD81 are examples of aberrant expression patterns of MM cells, compared to the phenotype of normal plasma cells (42). Multivariate research has not been able to determine these indicators’ independent prognostic relevance, despite the possibility that some expression patterns are linked to more serious conditions.

Immunophenotyping plays a crucial role in MM, as it provides valuable information about the characteristics of myeloma cells (43). By examining the expression patterns of specific markers, such as CD138, CD38, and BCMA, immunophenotyping helps to accurately identify and differentiate myeloma cells from normal cells. Given the heterogeneity of myeloma cells, it is particularly important to address the need to differentiate them from other plasma cell disorders. Moreover, immunophenotyping enables the monitoring of disease progression and response to therapy (42, 44, 45). Changes in the expression of certain markers can indicate treatment efficacy or disease relapse. It allows clinicians to assess minimal residual disease (MRD), which refers to the presence of small number of residual malignant cells that may not be detectable by conventional diagnostic methods. MRD assessment through immunophenotyping provides valuable insights into treatment outcomes, helps guide therapeutic decisions, and potentially improves patient prognosis. In addition, immunophenotyping contributes to the development of personalized treatment approaches in MM (46). By identifying specific marker expression profiles, clinicians can tailor treatment strategies to target the unique characteristics of each patient’s disease, potentially leading to improved treatment responses and long-term outcomes (47). However, research is still ongoing to establish the prognostic significance of various markers and refine the classification of myeloma cells.

The expression of one marker, CD24, on the surface of clonal PCs is directly correlated with overall survival (48). The expression of another marker, CD45, and its role in disease etiology has also been studied (43). In untreated or relapsed MM or SMM, CD45 expression has been reported to be inversely correlated with disease progression and negatively correlated with high-grade angiogenesis (49). Furthermore, the expression of CD45 on PCs is negatively correlated with their expression of CD138, CD56, and CD54 (49). In terms of marker expression, the expression of two additional markers, MPC-1 and CD49e, was restricted to the mature myeloma cells. Researchers identified immature, intermediate, and mature cells as MPC-1−CD45−/+ CD49e−, MPC-1⁺ CD45^−/+ CD49e⁻, and MPC-1⁺ CD45⁺ CD49e⁺, respectively (50). Importantly, these morphological and phenotypic classifications do not have established prognostic significance. Nonetheless, understanding the immunophenotypic and genotypic characteristics of myeloma cells is crucial for diagnosing and monitoring the disease.

CD138 (syndecan-1) expression is significantly elevated on the surface of myeloma cells. Elevated CD138+ expression on myeloma cells is associated with enhanced proliferation, prolonged survival, and suppressed apoptosis. This positive correlation is driven by amplified IL-6R signaling, highlighting the role of CD138 in myeloma cell dynamics (51). Nevertheless, low expression of CD138 is associated with tumor fibrosis in bone marrow and is correlated with heparin-binding growth factors that contribute to the pathogenesis of myeloma (52). Blocking CD138 in a myeloma mouse model renders myeloma more vulnerable to bortezomib chemotherapy, resulting in a dramatic decrease in tumor size (51). Combining anti-CD138 antibodies with proteosome inhibitors (bortezomib) or immune modulatory drugs (IMiDs) such as lenalidomide, creates a potentially powerful new strategy to fight myeloma, boosting the immune system’s ability to engulf and destroy both sensitive and resistant cancer cells (53).

An aberrant paraprotein called monoclonal protein is typically produced in patients with MM (54). Both the presence of neoplastic PCs in the BM and the detection of monoclonal protein in the blood or urine are used to establish a diagnosis of MM (54). Although neoplastic myeloma cells constitute the majority of the PCs in patients with MM, some of these patients also have normal or reactive plasma cells. A study by Rawstron et al. revealed that peripheral blood myeloma PCs had much lower levels of CD56 and Syndecan-1 expression compared to levels in bone marrow samples from the same patients (55). These cells were found in 75% (38/51) of patients at presentation, 92% (11/12) of relapsing patients, and 40% (4/10) of stem cell harvests (55). Jeong et al. discovered that a simplified immunophenotyping panel consisting of CD56/CD19/CD138/CD38/CD45 is beneficial in clinical practice for discriminating neoplastic myeloma cells from reactive PCs in a study of 70 individuals with MM (62 newly diagnosed vs. 8 treated) (56). Notably, CD19 expression in neoplastic PCs was negative in both untreated and treated patients, independent of therapy (56). In this regard, it is possible to infer that CD19 is the most helpful marker for distinguishing neoplastic PCs from reactive plasma cells. However, because the immunophenotype of NK cells is frequently identified as CD38+/CD56+/CD19- by flow cytometry, re-gating of cells with dim or negative CD45 expression is required to exclude NK cells. In a study of 132 patients with MM, Robillard et al. discovered that malignant PCs accounted for a median of 97% of total CD38+/CD138+ plasma cells (range: 76.5% to 100%) (1). Their research also found that PCs missing CD19 and expressing either CD56 or CD28 were the most common immunophenotypic population in MM samples (1). Furthermore, Klimienė et al. discovered that a decrease in adhesion molecule expression in MM patients’ BM may contribute to the abundance of circulating PCs in the peripheral blood of patients with relapsed/refractory disease, resulting in the absence of CD49d, CD49e, CD56, CD138, and CD58 markers in circulating peripheral blood plasma samples (57).

Several studies correlated the aberrant expression of markers to clinical outcomes: CD117, HLA-DR, and CD33 were found to be independent prognostic factors for decreased progression-free survival (58–61), and the expression of CD200 and CD307 are related to decreased overall survival and poor prognosis, correspondingly (62–65). According to Sanoja-Flores et al., circulating malignant PCs in patients with MM display decreased levels of activation/differentiation-associated antigens such as CD27, Vs38c, and Ki67 (a proliferation marker) (66). Other maturation-associated antigens previously reported to be aberrantly expressed in MM, such as CD19, CD20, and CD45, were found to have no significant differences in the phenotypic profile of peripheral blood versus BM neoplastic plasma cells (66). These findings suggest that antigen expression levels in circulating and BM PCs differ in MM patients, and more research is needed to fully characterize these differences ( Figure 2 ).

To clarify immune cell subset accumulation and its influence on immune system homeostasis, researchers have employed a range of immunological models, encompassing diverse methodologies and approaches beyond immune cell markers. These models aid in unraveling the complex dynamics and effects of specific immune cell subsets on overall immune system balance. For example, the accumulation of terminally differentiated effector memory T-cells (TEMRA), which are identified as CD3⁺, CD4⁺/CD8⁺, CD127^+/dim, CD27^-, CD45RA⁺, and CCR7^-, may indicate the exhaustion or malfunction of some immune subsets such as NK cells. Another example is related to the difference in the pattern expression of mature NK cells and early NK cells, which are CD56⁺/CD16⁺ and CD56⁺⁺/CD16^-, respectively. Such findings are important in understanding how immune cells can contribute to tumor progression and for developing new strategies for treating cancer ( Figures 3 , 4 ).

Recent research has shown that studying peripheral blood mononuclear cells (PBMCs) in patients with cancer is important for understanding how immune exhaustion mechanisms contribute to tumor progression. Gating PBMCs from patients with MM is similar to the normal gating strategy for healthy donors; however, the detection of circulatory myeloma cells is the main goal in the study of MM patient PBMCs as opposed to regular PBMCs. Herein, we review the marker expression of PBMCs in MM patients. Please refer to Table 1 for a detailed breakdown of the 43 markers in the CyTOF panel.

In recent years, different sophisticated single-cell analysis methods that can be used for IPT have evolved. Moldavan is widely regarded as having made the first attempt to count or quantify cells in suspension (1934) (82). Tsutomu Nomizu et al. later found in 1994 that it was feasible to nebulize, dry, and ignite single cells in a hot plasma to form ion clouds detectable by emission spectrometry; this experiment was the first true mass experiment of single cells (83). In 2007, Scott D. Tanner, inspired by flow-cytometry innovations, invented mass cytometry, also known as Cytometry by Time-of-flight (CyTOF), which is the most promising technology for high-dimensional and high-throughput single-cell analysis at the protein level (84). Later, in 2008, Tanner and colleagues depicted that the tandem attachment of a flow cytometer to an inductively coupled plasma mass spectrometry (ICP-MS) machine together with tagging antibodies to metal tags would allow higher multiplexity (85).

CyTOF may be used to investigate cell phenotypes and function, signaling networks, apoptosis, cell cycle, and a variety of other complicated biological processes. Both CyTOF and flow cytometry (FCM) are founded on the notion of multiplexed single-cell analysis utilizing labeled antibodies. However, there are several key distinctions between the two technologies ( Table 1 ). In place of traditional fluorescent labels, CyTOF employs non-biologically accessible metal isotopes with succinct mass spectrometry parameters. Flow cytometry, on the other hand, detects cellular targets using fluorophore-labeled antibodies (86). This change of labeling method allows mass cytometry to examine up to 50 variables simultaneously, avoiding the difficulties associated with overlapping emission spectra, which are common in FCM research (87).

A typical FCM test detects 8–10 distinct markers, while bigger panels are becoming feasible for researchers to develop (88). With the use of next generation flow cytometry (NGF), researchers can now study up to 12 markers in a single staining sample to detect minimal residual disease in MM (70). However, this capacity is less than that of CyTOF, which can identify at least 43 markers in a single sample. In CyTOF staining, the cells are stained by incubating them with a cocktail of probes or antibodies labeled with non-radioactive and non-biological heavy metal isotopes (89). The single-cell suspension is next passed through an argon (Ar) plasma, which atomizes and ionizes the sample, transforming each cell into a cloud containing ions of the elements present within or on the cell surface (90). A high-pass optic extrudes low-mass ions obtained from each cloud, resulting in a cloud containing sole ions linked with the isotope-conjugated probes. The ions are then dispersed by m/z in the time of flight (TOF) chamber before being strengthened and converted into electrical signals (91). Flow cytometry acquisition, by contrast, entails passing intact fluorescently labeled cells through a series of lasers and collecting the emitted light, using specific detectors (92). CyTOF’s exceptional qualities have positioned it as a powerful tool for analyzing considerably diverse clinical samples, while also playing a crucial role in the diagnosis and research of malignancies.

CyTOF detects more markers per sample compared to FCM, whereas FCM has faster acquisition times and preserves cell integrity, allowing them to be recovered for future use (93). Because a CyTOF instrument uses just one detector, no extra tuning or calibration is required for each experiment (94). Moreover, concurrent analysis may be performed from a single tube without the need for single-stained or autofluorescence controls, which is a significant advantage in cases when in vivo or clinical research materials are scarce. For example, researchers may profile 30 immune markers from just 300 µL of blood using the Standard BioTools Maxpar^® Direct™ Immune Profiling Assay™, a validated, dry-format antibody panel for use on the CyTOF XT™ and Helios™ mass cytometry systems (95). Additionally, mass cytometry data processing does not necessitate post-acquisition data normalization, and although correction of spectrum overlap is essential in flow cytometry, CyTOF requires less compensation due to the use of heavy metal tags rather than fluorochromes. Another advantage of CyTOF is that the metal tags are exceedingly durable and can endure fixation, permeabilization, and freezing without affecting the signal. This feature enables the simultaneous staining of cell surface and intracellular markers, as well as the storage and shipment of labeled samples for multi-site investigations. The use of fluorophores, alternatively, may alter cellular and proteomic changes in the cells, which may influence the immunophenotypic signature as well as the intracellular and functional targets.

On the other hand, one present disadvantage of CyTOF is that it is a relatively new technique when compared to FCM. Heavy metal-conjugated antibodies are less frequently commercially accessible than fluorophore-conjugated antibodies (96). As a result, if a large mass cytometry panel is sought, a major amount of a researcher’s effort may be allocated to self-reagent production, such as in-house antibody conjugation. Furthermore, the need to sort the subpopulations discovered by high-dimensional phenotyping for further downstream research, such as single-cell RNAseq or functional studies, is widespread and addressable with full-spectrum FCM rather than CyTOF (97). Flow rates in CyTOF (500 events/s) are substantially slower than in fluidic or acoustic flow cytometers (3,000 and 35,000, respectively) (98). Also, the Helios™ CyTOF cytometers are fairly expensive (98). When this cost is combined with the additional expense of conjugating antibodies to heavy metal tags, many researchers cannot accommodate the expenditures.

However, CyTOF is beneficial in clinical scenarios, where a large number of variables on a small sample size may be examined (99). Indeed, CyTOF analysis has been used in several clinical research studies throughout the world to examine various areas of human illness to better understand and enhance the effectiveness of clinical treatments. The most significant barrier that remains is inherent in the complexity of the data processing, which necessitates extensive biostatistical and bioinformatic abilities, which frequently complicates its implementation in a clinical situation. Table 2 summarizes the advantages and disadvantages of FCM and CyTOF (99).

Based on our experience with and the guidelines for use of flow cytometry (126) for cleaning up the non-specific signals generated from unstable signals, beads, cell doublets, non-viable cells, and other unknown noises, we adopted these methods for cleaning up the fcs data files of CyTOF samples: After data normalization, the next clean-up step is to gate on the stable signal events ( Figure 5A ), followed by a selection of positive intercalator signal (Ir191/Rh103), while excluding the negative beads population (Ce140) ( Figure 5B ). The outcome events of the previously selected gates may be further refined by a selection of singlets population ( Figures 5C, D ). In exceptional cases when using frozen samples, monocytes are shown to be aggregated and appear in the doublet’s region; therefore, it is better to include these doublets and subsequently perform cleaning at each step of sequential gating. The last step is the exclusion of non-viable cells ( Figure 5E ) before starting the gating of backbone markers (usually CD45) for analysis ( Figure 5F ).

The need for the use of next-generation flow cytometry and CyTOF has increased recently in immunophenotyping research, notably in the analysis of larger panels of markers (130). The use of larger panels enables a more thorough characterization of immune subsets and offers a deep comprehension of the intricate interactions between various cell types under both healthy and diseased circumstances (122).

CD3, a crucial T cell marker, is found to be expressed abnormally on neoplastic plasma cells (131). This overlap between different markers, either normally or abnormally expressed, makes comprehensive immunophenotypic profiling and functional analysis very challenging and requires that all plasma cell subpopulations be considered. In addition to these considerations, alternate markers or gating techniques can also be investigated. Unusual and less frequent cases, such as MM that is CD138-negative, might render typical flow cytometry procedures difficult (132). Additionally, the use of certain monoclonal antibodies, such as daratumumab (a CD38-targeted monoclonal antibody), during therapy may affect plasma cells’ CD38 positivity (133). For a precise diagnosis and evaluation of the effectiveness of the treatment, such variations require careful observation and reporting. To minimize diagnostic errors, pathologists should also be attentive to aberrant expression patterns and ensure they address any discrepancies or imbalances in their interpretations. Aberrant expression patterns can be recognized and addressed with quality control procedures and staying current with new research discoveries.

Heterogeneous marker distribution in myeloma can also result in missed diagnoses, particularly when inadequate sample size or sampling errors are involved. To reduce the likelihood of misdiagnoses, it is crucial to consider the total morphological picture and combine other diagnostic methods, including imaging or genetic analysis. Due to the intricate modifications needed for color correction in flow cytometry, simultaneous multi-color immunophenotyping might be difficult. Data quality can be affected, and variability can be introduced by the operator’s subjective manipulation of instrument controls. In clinical flow cytometry laboratories, using automated technology for color adjustment during data gathering can lower operator-dependent variability and raise uniformity. Making the right choices for color compensation can be aided by using software such as Bagwell and Watson’s automated technique.

To ensure the quality and reliability of the data acquired, researchers are faced with new hurdles when handling larger antibody panels in CyTOF. The lack of analytical tools that can successfully accommodate these larger panels is one of the main issues (143). The number of parameters that a conventional flow cytometry analysis software can manage is usually limited, and the computing power required to process the data is another limiting factor (144). It is crucial to have new processes to contend with the challenges of larger panels. To address the increased complexity of the data produced by more extensive antibody panels, the field requires the development of advanced algorithms and data processing methods (145). Improvements to the physical infrastructure, such as increasing processing and storage capacity, are among the other requirements for an efficient analysis of the data associated with large antibody panels (146). Furthermore, it is essential to recruit reliable validation strategies in addition to new analytical instruments. This approach entails ensuring that the antibody staining is accurate and reliable, optimizing the panel design to reduce spectral overlap and maximizing the resolution, and setting the right quality control methods in place along the experimental workflow.

Spectral unmixing flow cytometry offers advantages in studying the immune map of multiple myeloma to characterize the immunohematology complexities. Using 40 markers will help to provide a deep comprehensive view of the hematopoietic and immune cell landscapes at the single-cell level, such as plasma cell populations, changes in immune cell profiling, and their microenvironment. This multiplexing in addition to different unbiased analysis algorithms may quantify rare cell types, discern differences in cell abundance and phenotype across patients, and find any subtle changes in the immune landscape that lead to either disease progression or treatment response (147). With over 40 markers, diagnostic panels are critical to MM management. These panels help distinguish MM from other plasma cell disorders, identify high-risk MGUS and smoldering MM, and are essential for post-treatment MRD assessment (42). Both spectral unmixing flow cytometry or CyTOF will help to capture detailed immunophenotypic profiles that can also contribute to the definition of antigenic profiles that impact prognosis, as well as the identification of new therapeutic targets, potentially leading to more personalized treatment approaches (42).

The precise quantitative and qualitative data required to analyze marker expression kinetics, population similarity, and sample comparisons at the minor population level can only be obtained through the implementation of several crucial algorithms. The cells can be visualized in a 3D projection alongside the overall stained markers and then represented in a 2D figure in an unbiased manner (147). Using machine learning algorithms and this tSNE programming helps to improve the tSNE plotting. As a result of several developments, optimized tSNE (Opt-NSE) emerges as an automated solution for analysis to improve the visualization of large sets of data (148, 149). Based on tSNE, vast comparative algorithms are created with regard to events clustering (such as FlowSOM, SPADE) and statistical characterization (such as CITRUS) (150–152).