Overview
- Multiparameter flow cytometry (MFC) is a cornerstone technology in diagnostic and therapeutic haematology.
- Modern instruments simultaneously measure forward scatter (FSC), side scatter (SSC), and fluorescent emissions across 8-12 or more parameters per cell, enabling rapid, objective immunophenotyping of thousands of cells in minutes.
- MFC serves four major functions: lineage assignment and subclassification of acute leukaemias; characterisation of chronic lymphoproliferative and myeloproliferative neoplasms; detection and quantification of measurable residual disease (MRD); and monitoring of specific clinical states such as paroxysmal nocturnal haemoglobinuria (PNH).
- Understanding technical principles, gating hierarchies, and disease-specific immunophenotypic signatures is essential for RCPA Fellowship-level competency.
Technical Principles
Instrument Parameters and Panel Design
| Parameter | Biological correlate | Practical use |
|---|---|---|
| FSC (forward scatter) | Cell volume / size | Distinguishes populations by size |
| SSC (side scatter) | Internal granularity / complexity | Separates granulocytes from lymphocytes |
| Fluorochrome conjugates | Surface, cytoplasmic, nuclear antigens | Lineage and maturation marker identification |
- Modern MFC panels exploit spectral separation of fluorochromes (e.g. FITC, PE, PerCP, APC, PE-Cy7, BV421) across multiple laser lines (405 nm, 488 nm, 638 nm).
- Panel design requires careful consideration of antigen density and fluorochrome brightness: high-density antigens (e.g. CD45) can be paired with dimmer fluorochromes, whereas low-density or intracellular antigens require brighter fluorochromes.
- Compensation matrices must be accurately constructed to correct for spectral overlap before acquisition.
For intracellular markers, including terminal deoxynucleotidyl transferase (TdT), cytoplasmic CD3 (cyt.CD3), cytoplasmic CD79a (cyt.CD79a), myeloperoxidase (MPO), and cytoplasmic immunoglobulin (cyt.Ig), cells must undergo membrane permeabilisation prior to staining.
Pre-analytical Considerations
- Specimen: peripheral blood, bone marrow aspirate, or tissue cell suspension
- Process within 24 hours of collection (preferably <6 hours for optimal viability); fresh cells required, stored or fixed specimens are not suitable for MRD
- Red cell lysis (ammonium chloride-based) or density gradient separation
- Instrument flush prior to acquisition to eliminate carry-over from previous samples
- Record FSC-height, FSC-area, and time-of-aspiration parameters to exclude doublets and air artefacts
Gating Strategy
A hierarchical, sequential gating strategy is the foundation of reproducible MFC analysis.
Sequential Gating Hierarchy
- Debris exclusion gate: FSC-A vs SSC-A, exclude debris, dead cells, and aggregates below the main cell cloud
- Doublet discrimination: FSC-H vs FSC-A (or SSC-H vs SSC-A), exclude doublets that cause false parameter readings
- CD45 vs SSC gate (leucocyte differential gate): The single most informative gating step for bone marrow and blood. CD45 expression versus SSC separates distinct populations:
| Population | CD45 expression | SSC |
|---|---|---|
| Lymphocytes | Bright | Low |
| Monocytes | Moderate-bright | Moderate |
| Granulocytes | Moderate | High |
| Myeloid/lymphoid blasts | Dim or negative | Low |
| Erythroblasts | Negative | Low |
- The CD45/SSC gate is superior to FSC/SSC alone because it permits isolation of blast populations that would otherwise be obscured by erythroblasts.
- This is well established as the preferred approach for bone marrow blast gating in acute leukaemia.
- Lineage-specific sub-gates: Once blasts or populations of interest are isolated, further gating uses lineage markers (e.g. CD19 for B cells, CD3 for T cells, CD13/CD33 for myeloid cells)
- Back-gating verification: Putative MRD or rare populations must be back-gated onto the FSC/SSC plot to confirm biological plausibility and exclude artefacts
PNH Gating (Example of Sequential Gating in Practice)
- For PNH neutrophil clone detection, a three-step sequential gate is applied: (i) CD45-positive leucocytes → (ii) SSC vs CD15 to isolate neutrophils → (iii) FSC vs SSC confirmation → final plot of CD24 vs FLAER to identify GPI-deficient neutrophil clones.
- Both percentage and absolute values should be reported.
- Pitfalls include lipidaemia causing poor population separation, resolved by removing plasma and replacing with PBS prior to re-testing.
B-Cell Acute Lymphoblastic Leukaemia (B-ALL)
Immunophenotypic Classification
- B-ALL blasts universally express CD19, HLA-DR, and TdT.
- The most specific early B-lineage cytoplasmic markers are cyt.CD79a and cyt.CD22.
- CD45 is frequently negative or dim, facilitating isolation on the CD45/SSC gate.
Five immunological subtypes correspond to sequential stages of B-cell ontogeny (note: CD10-negative normal early B-cell progenitors are controversial):
| Subtype | CD10 | CD20 | cyt.IgM | sIg |
|---|---|---|---|---|
| B-I / Pro-B / Early B | Negative | Negative | Negative | Negative |
| B-II / Common | Positive | Positive/− | Negative | Negative |
| B-III / Pre-B | Positive | Positive/− | Positive | Negative |
| B-IV / Mature B | Positive/− | Positive | Negative | Positive (κ or λ) |
Note: Mature B-ALL (B-IV) is TdT positive or negative.
Genotype-Immunophenotype Correlations in B-ALL
| Molecular/cytogenetic lesion | Immunophenotypic signature |
|---|---|
| t(9;22) / BCR::ABL1 | Common B-ALL (B-II); CD25 often positive; CD66c co-expression |
| KMT2A rearrangements | Pro-B / early B phenotype; CD15 and NG2 often positive |
| ETV6::RUNX1 | Common B-ALL; CD27 co-expression |
| Hyperdiploid (>50 chromosomes) | Common B-ALL; CD21 expression |
| TCF3::PBX1 | Pre-B phenotype (B-III) |
| BCR::ABL1-like (Ph-like) | Variable; CD25+, CRLF2 overexpression assessable by MFC |
| iAMP21 | Common B-ALL phenotype |
- Aberrant myeloid antigen co-expression (CD13, CD33, CD117) occurs in approximately 20-30% of B-ALL cases.
- Per WHO 2022 criteria, this does not alter lineage assignment but should be documented as it may influence MRD tracking.
T-Cell Acute Lymphoblastic Leukaemia (T-ALL)
Immunophenotypic Classification
- T lineage is established by cyt.CD3 (most specific marker), TdT, and CD7, which are present in most cases.
- Surface CD3 confirms mature T-cell differentiation.
- Additional markers, CD2, CD5, CD1a, CD4, CD8, CD34, are expressed in a pattern reflecting thymic development.
- Weak cyt.CD79a expression can occur in some T-ALL cases (particularly ETP-ALL) and does not indicate B lineage.
| Subtype | CD34 | CD1a | sCD3 | CD4/CD8 | CD10 |
|---|---|---|---|---|---|
| ETP-ALL | Often + | Negative | Negative | CD4−/CD8− | Variable |
| Pro-T | Positive | Negative | Negative | CD4−/CD8− | Negative |
| Pre-T | Negative | Negative | Negative | CD4−/CD8− | Negative |
| Cortical T | Negative | Positive | Negative | CD4+/CD8+ (double positive) | Positive |
| Medullary T | Negative | Negative | Positive | CD4+ or CD8+ (single positive) | Negative |
- Early T-cell Precursor ALL (ETP-ALL) is a clinically distinct high-risk subtype defined by: CD1a negative, CD8 negative, CD5 weak or negative, with co-expression of ≥1 myeloid or stem cell marker (CD117, CD34, HLA-DR, CD13, CD33, CD11b, or CD65).
- ETP-ALL must be flagged given its aggressive biology, inferior outcome with standard therapy, and distinct management implications.
- MPO negativity must be confirmed to exclude AML or MPAL.
Acute Myeloid Leukaemia (AML)
Immunophenotypic Approach
- AML blasts occupy the dim CD45 / low SSC gate.
- Core myeloid antigens are CD13, CD33, CD117, and MPO.
- CD34 marks immature progenitors but is absent in monocytic and more mature subtypes.
- Aberrant phenotypes (leukaemia-associated immunophenotypes, LAIP) are detectable in >90% of AML cases at diagnosis and are the basis of MFC-MRD monitoring.
| AML subtype | Key immunophenotypic features |
|---|---|
| AML with minimal differentiation (M0) | CD13+, CD33+, CD117+, MPO+ (flow/EM); CD34+; no monocytic markers |
| AML without maturation | CD13+, CD33+, MPO+, CD34 variable |
| AML with maturation | CD13+, CD33+, MPO+, CD15 and CD11b emerge |
| Acute myelomonocytic (M4) | CD13+, CD33+, MPO+, CD14+, CD64++, CD11b+ |
| Acute monocytic (M5) | CD14+, CD64++, CD11c+, CD36+; MPO often weak/negative |
| Acute erythroid | CD71+, CD235a (glycophorin A)+, CD117+; CD34 variable |
| Acute megakaryoblastic (M7) | CD41+, CD61+, CD36+; CD34 variable |
| APL (PML::RARA) | CD34−, HLA-DR−, CD13+, CD33++, CD64 weak, CD11b−; characteristic high SSC |
| Acute basophilic leukaemia | CD13+, CD33+, CD9+, CD11b+, CD22+, CD123+ |
Platelet/RBC fragment adhesion to blasts can cause non-specific CD41/CD61 positivity; correlation with morphology and immunohistochemistry is mandatory.
Mixed Phenotype Acute Leukaemia (MPAL)
Per WHO 2022, lineage assignment requires:
- Myeloid: MPO positivity OR monocytic differentiation (CD64++, CD11c+, NSE+, CD14+)
- T lineage: cyt.CD3
- B lineage: CD19 strong, OR CD19 weak with ≥2 of cyt.CD79a / cyt.CD22 / CD10
MPAL is defined when the same blast population expresses markers meeting criteria for two or more lineages simultaneously.
MRD Detection by Flow Cytometry
Principles and Strategies
MFC-MRD exploits two complementary approaches:
- LAIP (leukaemia-associated immunophenotype): Aberrant antigen combinations identified at diagnosis are tracked at follow-up. Requires a diagnostic baseline sample.
- DfN (different from normal): At follow-up, any population not conforming to normal regenerating haematopoietic progenitors is flagged. This approach accommodates phenotypic shift, the well-recognised phenomenon whereby leukaemic cells at relapse may alter antigen expression relative to diagnosis. Thorough knowledge of normal and regenerating bone marrow immunophenotypes is essential for DfN interpretation.
- Modern 8- to 10-colour panels applied at both diagnosis and follow-up allow detection of aberrant cells using sequential gating in most patients.
- Cell viability information and assessment of normal haemopoiesis are additional advantages of MFC over molecular MRD methods.
Comparative Sensitivity of MRD Methods
| Method | Applicability | Sensitivity | Specimen |
|---|---|---|---|
| Multiparameter flow cytometry | ~95% of ALL; >90% of AML (LAIP detectable) | $1 \times 10^{-4}$ | Fresh cells |
| RQ-PCR for Ig/TCR rearrangements | ~90% of ALL | $1 \times 10^{-5}$ | DNA |
| RQ-PCR for fusion gene transcripts (e.g. BCR::ABL1) | Depends on frequency; BCR::ABL1 ~20-25% of adult ALL | $1 \times 10^{-5}$ | RNA |
| High-throughput sequencing (Ig/TCR) | ~90% | $10^{-5}$ to $10^{-6}$ | DNA |
| Digital droplet PCR (ddPCR) | Mutation/fusion gene targets | $10^{-5}$ to $10^{-6}$ | DNA/RNA |
MFC is not the most sensitive platform, but its near-universal applicability (~95% of ALL cases have an informative immunophenotype), ability to provide cell viability data, and assessment of normal haemopoietic reconstitution make it indispensable.
Technical Requirements for MRD
| Parameter | Minimum standard | Optimal standard |
|---|---|---|
| Sensitivity | $1 \times 10^{-4}$ | $1 \times 10^{-5}$ to $10^{-6}$ |
| Events acquired | ≥500,000 leucocytes | ≥1,000,000 leucocytes |
| Events in MRD cluster | ≥30-50 | ≥50-100 |
| Specimen | Fresh bone marrow aspirate (preferred) | , |
| Time to processing | <24 hours | <6 hours |
Always perform: instrument flush before acquisition; back-gating of candidate MRD events; record FSC-height, FSC-area, and time parameters to exclude artefacts.
MRD in AML
- Flow cytometric MRD in AML uses LAIP or DfN strategies on bone marrow aspirate.
- LAIP is detectable by MFC in 86-89% of adult AML patients at diagnosis.
- A clinically validated MRD cut-off of 0.035% residual leukaemic cells (after induction or consolidation) has been used in EORTC/GIMEMA studies to discriminate MRD-negative from MRD-positive cases.
- A threshold of ≥0.1% residual leukaemic cells is also widely applied.
- MRD negativity after induction and after consolidation is independently prognostic for overall survival in patients with favourable or intermediate cytogenetic risk.
- Sequential MRD monitoring requires sensitivity of at least $1 \times 10^{-3}$ (ELN 2022 minimum) and is applicable to the vast majority of AML patients.
- MRD status is now recognised as an independent predictor of relapse risk superior to cooperating mutations at early post-remission time points.
Key clinical applications:
- Risk stratification to guide alloHSCT decisions
- Pre-transplant MRD positivity identifies patients benefiting from myeloablative conditioning (MAC) over reduced-intensity conditioning (RIC), demonstrated in the BMT CTN 0901 trial using ultra-deep DNA-based NGS MRD
- Post-transplant monitoring for early relapse detection
- Molecular MRD (NPM1 RT-qPCR, RUNX1::RUNX1T1, CBFB::MYH11, PML::RARA) complements MFC in genetically defined subgroups; NPM1 RT-qPCR is preferred in NPM1-mutated AML; serial PML::RARA PCR is the standard in APL
MRD in ALL
- MRD is the single most powerful independent prognostic factor in ALL (childhood and adult).
- MRD-guided therapy is standard of care.
| MRD level (end of induction) | Clinical significance |
|---|---|
| $<10^{-4}$ (MRD negative) | Favourable; de-escalation may be considered in paediatric ALL |
| $10^{-4}$ to $10^{-3}$ | Intermediate; close monitoring warranted |
| $>10^{-3}$ | High relapse risk; consider intensification or alloHSCT |
- In childhood AML, MRC studies showed 3-year RFS of 64% for MRD <0.1% vs 14% for MRD >0.5% at end of induction.
- COG AAML0531 data (4-colour MFC) showed DFS at 3 years of 34% (MRD >0.1%) vs 60% (MRD undetectable) at end of induction.
- In Ph+ ALL, BCR::ABL1 qPCR is the preferred molecular MRD modality.
- MRD in ALL is present in up to 30-50% of patients in traditional morphological CR.
- Pre-HCT MRD status is a critical determinant of post-transplant outcome.
MRD in Multiple Myeloma
- Gating strategy for plasma cells uses a combination of CD138, CD38, CD45, and light scatter characteristics, the recommended four-colour minimum.
- CD38/CD45 alone (older method) risks excluding CD45+ plasma cells that may represent the majority of neoplastic cells.
| Marker | Normal plasma cells | Neoplastic plasma cells |
|---|---|---|
| CD19 | Positive | Negative |
| CD56 | Negative | Positive |
| CD27 | Strong | Weak or negative |
| CD81 | Strong | Weak or negative |
| CD28 | Negative/weak | Strongly positive |
| CD200 | Negative/weak | Strongly positive |
| CD117 | Negative | Aberrantly positive (subset) |
| CD20 | Negative | Aberrantly positive (subset) |
- No single marker reliably distinguishes all neoplastic from normal plasma cells; a panel approach is mandatory.
- At least 100 neoplastic plasma cell events should be acquired for MRD.
- The IMWG response criteria incorporate MRD negativity at $10^{-5}$ sensitivity as a response milestone.
- Current evidence supports $10^{-6}$ as the future standard for bone marrow MRD negativity (requires ≥3 million cells analysed; $10^{-7}$ would require 30 million cells, practically limiting).
- MRD testing at sensitivity $\leq 10^{-4}$ is not considered informative for myeloma.
Emerging Technologies
Next-Generation Flow (NGF)
- EuroFlow consortium has standardised 8-colour panels enabling $10^{-5}$ to $10^{-6}$ sensitivity for myeloma MRD and has defined reference ranges for normal bone marrow populations.
- NGF protocols provide reproducible results across centres using standardised acquisition and analysis software (e.g. Infinicyt).
Spectral Flow Cytometry
- Captures full spectral emission rather than using bandpass filters, enabling 25-40 simultaneous parameters with improved resolution of overlapping fluorochromes.
- Unmixing algorithms replace conventional compensation matrices.
- Entering clinical laboratory practice with substantially expanded panel design capability.
Mass Cytometry (CyTOF)
- Uses heavy metal isotope-conjugated antibodies detected by time-of-flight mass spectrometry, enabling simultaneous measurement of 40-50 parameters per cell without spectral overlap.
- Limitations: lower throughput, cells cannot be recovered post-acquisition, specialised infrastructure required.
- Primarily a research tool.
Digital PCR and NGS-MRD
ddPCR offers improved sensitivity and quantitative accuracy over conventional qPCR for molecular MRD targets (e.g. NPM1, fusion gene transcripts). Error-corrected deep NGS enables mutation burden tracking at $10^{-5}$ to $10^{-6}$ sensitivity across 13+ commonly mutated AML genes. These molecular platforms complement MFC rather than replacing it, since MFC uniquely provides cell viability data and assessment of normal haemopoietic reconstitution.
Single-Cell and Multi-Omic Integration
- Platforms combining flow cytometric sorting with single-cell RNA sequencing (scRNA-seq) or CITE-seq (cellular indexing of transcriptomes and epitopes by sequencing) allow simultaneous protein and transcriptomic profiling.
- These technologies are defining novel leukaemic stem cell populations and new LAIP candidates, and are expected to enter routine clinical practice within the next 5-10 years.
Quality Assurance and Pitfalls
| Pitfall | Consequence | Mitigation |
|---|---|---|
| Inaccurate pipetting | Error in cell concentration and antigen quantification | Calibrated pipettes; internal bead controls |
| Incorrect gating | False-positive or false-negative MRD calls | Back-gating verification; independent review |
| Phenotypic shift | Loss of LAIP at relapse; missed MRD | Use DfN strategy at follow-up in parallel with LAIP |
| Haemodilution of BM aspirate | Underestimation of disease burden | Assess spicule adequacy; report cellularity |
| Platelet/RBC fragment adhesion to blasts | Non-specific CD41/CD61 positivity on myeloblasts | Correlation with morphology and immunohistochemistry |
| Lipidaemia | Poor population separation | Remove plasma, replace with PBS; repeat |
| Sample age | Antigen degradation, increased cell death | Process within 24 hours (MRD: fresh cells only) |
| Carry-over from previous sample | False-positive events | Instrument flush before each acquisition |
Clinical Integration and Regulatory Considerations
- MFC-MRD is incorporated into clinical trial endpoints and routine practice guidelines.
- ELN 2022 recommendations for AML include MRD assessment as part of response criteria, with validated MRD negativity defined as absence of leukaemic cells at a sensitivity of $\geq 10^{-3}$ by validated flow or molecular techniques; MRD negativity is emerging as a surrogate therapeutic endpoint supplementing and potentially replacing morphological CR criteria.
- MRD is routinely performed in reference laboratories using standardised protocols (EuroFlow or equivalent).
- Results must report MRD as a percentage of total leucocytes or total nucleated cells, with the analytical sensitivity explicitly stated.
- Laboratories should operate under accredited quality management systems (NATA accreditation in Australia; RCPA Quality Assurance Programs participation) and participate in external quality assurance schemes.