Immunophenotypic Profile of Multiple Myeloma: A Tertiary Care Centre Experience

Introduction

Plasma cell dyscrasias (PCDs) results from clonal expansion of PCs that secrete single homogenous monoclonal immunoglobulin called an M protein.^[1] The diagnostic criteria included the presence of > 10% clonal bone marrow PCs on morphology or an extramedullary plasmacytoma along with one or more myeloma defining events defined by the International Myeloma Working Group (IMWG).^[2] Multiparametric flow cytometry (FCM) has been the forerunner in diagnosing and monitoring in many hematological neoplasms mostly because of the high sensitivity, specificity along with the ability to provide results within a few hours.^[3] The importance of FCM in diagnosing, monitoring, and prognosticating PCDs, especially MM, have progressively increased in recent years. FCM is useful in differentiating PCDs from reactive plasmacytosis or other B cell neoplasms with extensive plasmacytic differentiation.^[3] FCM has been useful in differentiating APCs from NPCs based on the different surface antigen expressions as well as the clonality analysis.^[4] APCs characteristically show under-expression of CD19, CD45, CD27, CD81, CD38, and overexpression of CD56, CD117, CD28, CD33, and CD200.^[3,4] NPCs predominantly express CD19, CD45, CD27, and CD81 and are negative for CD56, but a varying minor population amongst NPCs can have altered expression of surface antigens.^[3,4] The enumeration of APCs and NPCs is also useful in differentiating MM from monoclonal gammopathy of unknown significance (MGUS) and smoldering MM (SMM) based on the percentage of NPCs in a total PC compartment in the bone marrow.^[5] The most useful application of enumerating NPCs is in SMM, where FCM can differentiate between an MGUS-like profile from an MM-like phenotypic profile.^[6] Newly diagnosed MM cases (NDMM) who have > 5% NPCs from all bone marrow PCs have a lower bone marrow PC burden, higher hemoglobin level, longer progression-free survival (PFS), and longer overall survival (OS) compared with patients who had less than 5% NPCs.^[7] However, different FCM panels incorporating different antigens for plasma cells have been used.^[8] Scarce studies have been published evaluating the sensitivities and specificities of individual and combination of antigen-differentiating APCs from NPCs.^[9] Literature on objective evaluation of CD38 and CD138 in differentiating APCs and NPCs is also rare.^[10] Enumeration of PCs by FCM has been particularly difficult owing to discrepancy with bone marrow (BM) morphology, hemodilution, and cell loss secondary to sample processing.^[9,11–13] FCM was found to have greater sensitivity and limit of detection compared with morphology supplemented with IHC despite being limited to more hemodilute second aspirate.^[13] Despite discrepancies in PCs enumeration, there has been significant correlation between FCM and BM PCs numbers.^[14–16] Complete absence of CD19 and/or CD45 along with dim expression of CD38 and aberrant expression of CD56 has been used to identify APCs in most MM cases; however, precise enumeration of APCs has been challenging, especially in the evaluation of minimal residual disease (MRD).^[9] Therefore, a combination of highly sensitive and specific markers differentiating APCs and NPCs is necessary.

We evaluated sensitivities and specificities of CD19, CD45, CD81, CD27, CD56, and CD117 to differentiate APCs from NPCs. Objective evaluation of CD38 and CD138 in terms of median fluorescence intensity (MFI) was done in APCs and NPCs. PCs enumeration by FCM was correlated with BM morphology. Correlation of total PC percentage and NPCs percentage with baseline characteristics of newly diagnosed MM cases was also evaluated.

Patients and Methods

BM aspirate samples of 48 newly diagnosed MM cases were evaluated for diagnostic FCM. The diagnosis of MM was done in these cases according to IMWG criteria.^[2] Ten nonmyeloma cases with reactive bone marrows (solid tumor/Hodgkin's lymphoma staging marrow/immune thrombocytopenic purpura [ITP]/chronic kidney disease/ATTR amyloidosis) were recruited for controls.

A 3-tube, 6-color multiparameter FCM panel was used for plasma cell immunophenotyping (►Table 1).

BM aspirate samples were received in EDTA anticoagulant and processed within 12 hours. An ammonium chloride-based bulk lysis/pre-lysis protocol (RBC lysis buffer, BioLegend, San Diego, CA) was used for all samples. Antibodies against surface antigens and intracytoplasmic light chain antigens were stained according to previously described protocols.^[17] For cytoplasmic light chain staining permeabilizing solution (BD Perm/wash, BD biosciences, San Jose, CA) was used after surface staining. In all washing steps aspiration of supernatant was done in the place of simple decantation of tube to minimize cell loss. A minimum of 0.5 million events were acquired on BD FACSCanto II 3-laser flow cytometer (BD Biosciences, San Jose, CA) immediately after sample processing in all cases.

A sequential gating strategy was used for immunophenotyping of PCs. BD FACSDiva v6.0 software (BD biosciences, San Jose, CA) was used in all the cases for analysis.

Results

We studied 48 MM cases with a male predominance (males = 33, females = 15). The mean age in our cohort was 58.4 years (range, 40–79 years). Staging of the disease in MM patients was done according to the published guidelines as the international staging system (ISS) (n = 48)/revised international staging system (RISS) (n = 41).^[1] Most patients in our cohort were in ISS-3 (31 cases, 65%) or RISS-2 (26 cases, 63%). The demographics, baseline characteristics including staging are shown in ►Table 2.

Discussion

FCM PC enumeration can accurately discriminate between MGUS and MM.^[6,18] The number of residual polyclonal PCs (NPCs) is a useful discriminating marker between MGUS and MM at diagnosis.^[6,19] MGUS usually has more than 5% NPCs within total bone marrow PCs.^[18] The proportion of NPCs in patients with PCDs is a powerful prognostic factor in all stages of the disease as well as during follow-up post-therapy.^[6,19] MGUS and SMM with abnormal to total PC ratio more than 95% have a higher risk of progression to symptomatic MM with time to progression at 5 years of 25% versus 5% and 64% versus 8%, respectively.^[6] MM patients with more than 5% NPCs in BM have a lower frequency of immuneparesis (42% vs. 83%, p = 0.003) and a greater response to autologous stem cell transplantation (ASCT) (complete remission [CR] rate after ASCT of 64% vs. 33%, p ≤ 0.001).^[20] These findings suggest that FCM is very useful in defining subgroups of patients with better prognoses, irrespective of achievement of CR or not. Because the exact enumeration of APCs and NPCs have prognostic significance, instituting a multiparametric FCM panel is important to differentiate APCs from NPCs. A classical immunophenotype of APCs is described as CD138 positive, dim CD38 (compared with NPCs), CD19 negative, CD45 dim to negative and aberrant CD56/CD117/CD20.^[8] CD200 and CD33 are expressed on APCs in a smaller subset of myeloma cases.^[8,9] CD28 is usually negative in NPCs but is overexpressed in APCs.^[8,9] The antigen expression in plasma cell leukemias (PCLs) vary from only MM cases. PCLs are more likely to express CD20 and tend to be negative for CD56 and CD117.^[8] Subpopulations of NPCs are known to show aberrancies in CD19, CD45, and CD56 expressions.^[9,19] Tembhare et al, in their patient cohort, found small sub-populations of NPCs showing abnormalities in CD19, CD45, or CD56 expression in 68% of cases.^[9] These normal variations in NPCs may pose significant difficulty in APC enumeration and NPCs versus APCs differentiation.^[21] Classically, CD38 and CD138 along with scatter properties and CD45 are being used for PC identification by FCM. However, with increasing use of anti-CD38 targeted therapies, such as daratumumab, additional PC gating marker should be used for monitoring of MM cases post therapy. Signaling lymphocyte activation molecules (SLAM) family proteins such as CD229 and CD319 have been increasingly used in combination with CD138 for PC identification post anti-CD38 therapy minimal residual disease analysis.^[22]

In lines with previous studies, our study confirmed that there is a significant PCs loss when enumerated by FCM but a significant correlation exists between enumerations by the two methodologies. The significant loss can be attributed to either sample dilution, patchy nature of the disease or cell loss in processing. We found CD19 to be 100% sensitive and CD81 to be 100% specific to identify APCs. Combination of CD19/CD45/CD56 though gives a 100% specificity, it is less sensitive in identifying APCs owing to presence of similar aberrancies in subset of NPCs. Next, 80% of control cases evaluated for immunophenotype of NPCs in our cohort showed abnormal antigenic expression. None of the control cases showed abnormality in CD81 in NPCs. Incorporating CD81 into the panel increased the sensitivity to 93.62% for accurate discriminations of APCs from NPCs. Individually, specificities of CD19 and CD56 (60% and 40%) were much lower than that of CD81 (100%). CD27 showed a decent sensitivity of 89.36%, however, a low specificity (70%). CD38 MFI was significantly lower in APCs in the majority of cases in comparison to NPCs. CD38 MFI can be used a useful discriminator between APCs versus NPCs wherever possible. The majority of our cases had < 5% NPCs by FCM, similar to previous published literature.

Conclusion

FCM at diagnosis is highly sensitive for differentiating APCs from NPCs. Along with CD19, CD45, CD56, CD117, and CD27, CD81 increases the specificity for APC and NPC identification and enumeration. A combination of multiple markers increases the accuracy of APC identification. CD38 MFI calculation in APC and NPC subset could help as an adjunct to different surface markers in APC enumeration. Though there is discrepancy in morphology and FCM PC enumeration, FCM should be routinely incorporated in MM cases at diagnosis owing to its prognostic significance at diagnosis. However, due to limitations in study sample, larger studies are required to confirm the findings.

Close

Fig. 1

MM immunophenotyping gating strategy. (A)Time vs. CD38 gate. (B) Singlet gate (FSC-A vs. FSC-H). (C) Non-debris cells (NDCE) gate (FSC-A vs. SSC-A). (D) CD45 vs. SSC plot. (E) PC gate (CD38 vs. CS138) (F). Refined PC gate on NDCE (CD38 vs. CD45). (G) CD19 vs. CD45 plot on refined PC. (H) CD19 + CD56-(purple) NPCs and CD19-CD56+ (pink) APCs (I) CD19 + CD81+ NPCs and CD19-CD81-APCs. (J) CD27 + CD117-NPCs and CD27-CD117 + APCs (K) Kappa-restricted APCs. (l) Polyclonal NPCs.

Export to PPT