Commercial multicolour fixed immunophenotyping panels can improve flow cytometric diagnostic immunophenotyping repertoire.
This study validated the commercially available, standardised Beckman Coulter lyophilised DURAClone RE panels to discriminate specific haematolymphoid subtypes.
We compared the diagnostic capability of the DURAClone acute leukaemia B (ALB), chronic leukaemia B (CLB), and plasma cells (PC) panels to the predicate second-line panels in Charlotte Maxeke Johannesburg Academic Hospital, Johannesburg, South Africa, from April to August 2020. Clinical diagnostic concordance between the in-house second-line immunophenotyping (the predicate method) and DURAClone was established. The ALB panels tested for precursor B-cell acute lymphoblastic leukaemia (
There was 100% clinical diagnostic concordance between the predicate and the test panels for second-line diagnostic investigation of B-ALL (with additional CD56), mature B-LPD (with additional discernment of CD81, ROR-1, CD79b and CD43) and PCD.
The DURAClone CLB exceeded the predicate second-line performance, offering extended second-line diagnostic discernment of mature B-LPD subtypes and discernment of CD5+ B-LPD from other non-CD5+ (or CD5–) B-LPD; likewise, the PC panels enabled discovery of PCD. While ALB testing offered no additional diagnostic advantage over existing predicate investigation, CD58 did offer additional information to discern haematogones from B-ALL.
Flow cytometry immunophenotypic analysis is a powerful diagnostic and research tool for investigating haemato-lymphoid malignancies, such as leukaemia, lymphoma, and plasma cell dyscrasias (PCD). Recently, multicolour flow cytometric immunophenotyping has become invaluable for improving detection and detailed immunophenotyping during diagnosis and detection of low frequencies of abnormal or aberrant cell populations during minimal residual disease assessment.
Commercially available, standardised, lyophilised, fixed and pre-titrated antibody panels,
Our laboratory previously established the positive impact on quality, laboratory workflow and diagnostic usefulness of the commercially available, fixed and standardised BC ClearLLab 10C
Existing predicate second-line investigation utilises locally established, in-house, non-standardised and manually assembled liquid fluorochrome-conjugated antibody panels. The primary aim of this study was to evaluate the fixed and standardised, pre-titrated commercially available DURAClone RE ALB, RE CLB and RE PC panels (Beckman Coulter, Mumbai, India) as a comprehensive and compact alternative to our existing second-line immunophenotypic investigation of haematological neoplasms. There were two parts to this study. First, we compared the DURAClone CLB and PC tubes to the laboratory’s in-house predicate 2–4-colour method for the following markers: combinations of CD10, CD23, FMC7 and CD22 to discern mature B-cell lymphoproliferative disorders, or CD19, CD38, CD56, CD200 and CD138 to characterise plasma cells. Secondly, we verified manufacturer-described expression for those additional markers that were not included in the older predicate 2–4-colour method, such as verifying the expression of CD43, CD79b, CD81 and ROR-1 in the CLB tube and CD38 and CD138 with CD27, CD81 and CD56 included in the DURAClone PC tube against expected expression in normal leucocyte population counterparts. We also validated the overall diagnostic outcome of the ALB tube; here, we asked whether the CD58 included in the ALB tubes offered any additional diagnostic advantage in identifying a B-ALL that had been identified in the first-line ClearLLab 10C investigation.
The University of the Witwatersrand Health Research Ethics Committee approved the study (Ethics Clearance number M1704129). The study’s objective was to validate the commercially available Beckman Coulter DURAClone PC, CLB and PC fixed-tube, pre-titrated panels as an alternative second-line diagnostic workup to our existing in-house second-line immunophenotypic investigation. The interpretation of flow cytometric data and consequent clinical diagnostic outcomes of the existing in-house, second-line, non-standardised panels were compared to the overall clinical diagnostic outcomes of the DURAClone analyses.
This study was a prospective observational cohort study conducted according to the Standards for reporting Diagnostics Accuracy.
Immunophenotypic testing of all samples at the Charlotte Maxeke Johannesburg Academic Hospital flow cytometry unit is two-tiered. Firstly, an initial first-line workup is performed with BC ClearLLab 10-C (Beckman Coulter, Miami, Florida, United States). The first workup identifies most acute myeloid and lymphoid leukaemia subtypes, distinguishes early and mature T-cell or B-cell lymphoproliferative disorders, and hints at the presence of a PCD (if a population of brightly expressing CD38 cells is noted). A limitation of this ClearLLab 10C system is that further immunophenotypic characterisation of mature B-cell lymphoproliferative disorders
After all routine testing with ClearLLab 10C and in-house second-line testing (
Possible reagent selection for method comparison according to suspected target population undertaken from April 2020 to August 2020 at an academic pathology service in Johannesburg, South Africa.
Possible target population | First-line analysis: ClearLLab 10C panels (BC, Peenya, India) | Second line analysis: in-house antibody panels | Second line analysis: DURAClone RE panel panels under evaluation (Beckman Coulter, Karnataka, India) |
---|---|---|---|
B-cell acute lymphoblastic leukaemia work-up or follow-up | ClearLLab 10C B, T, M1 and M2 panels | N/A | RE ALB |
Mature B-cell lymphoproliferative disorder work-up | ClearLLab 10C B, T, M1 and M2 panels | FMC7-FITC |
RE CLB |
Plasma cell dyscrasia work-upor follow-up | ClearLLab 10C B, T, M1 and M2 panels | CD19-FITC |
RE PC |
N/A, not applicable; FITC, fluorescein isothiocyanate; PE, phycoerythrin; ECD, phycoerythrin-Texas Red conjugate; APC, allophycocyanin; CD, cluster of differentiation.
Monoclonal Suppliers:
, Beckman Coulter, Marseille, France.
, Beckman Coulter Inc, Brea, California.
, Becton Dickinson, San Jose, California.
, Dako-Agilent, Santa Clara, California. & Invitrogen-Fisher Scientific Inc, Pittsburgh, Pennsylvania.
The NaviosTM (BC, Miami, Florida, United States) and the FACSCalibur flow cytometer (Becton Dickinson Biosciences, San Jose, California, United States) were used during this study. The NaviosTM flow cytometer was used to analyse the predicate ClearLLab 10C as well as the test DURAClone panels. Internal quality control on the NaviosTM included daily assessment of background contamination, cellular events carryover between tubes, and fluorospheres acquisition to verify the flow cytometer optical alignment and fluidics (Flow-Check Pro, BC, Lismeehan, Ireland). In addition, ClearLLab™ Normal and Abnormal process control cells (BC, Lismeehan, Ireland) were used to verify sample processing, acquisition, and analysis. ClearLLab 10C panel acquisition setup was achieved by applying target values set in the manufacturer manual and using BC FlowSet Pro beads to adjust the voltages that enabled optimal detection and separation of dim and bright antigens. After pilot testing (data not shown), the NaviosTM instrument ClearLLab 10C settings were deemed appropriate for the DURAClone multicolour data acquisition and analysis.
The FACSCalibur flow cytometer was used to acquire the predicate in-house second-line 2–4-colour fluorescence panels using CellQuest software (BD, San Jose, California, United States). Quality control performed on the BD FACSCalibur included daily assessment of background contamination, carryover, acquisition, and analysis of manufacturer-recommended 3-colour and APC Calibrite beads (Becton Dickinson Biosciences, San Jose, California, United States). Further acquisition and analysis of ImmunotrolTM process control (BC, Brea, California, United States) using four monoclonal antibodies, CD14 FITC, CD13 PE, CD45 PERCP and CD3 APC (all Becton Dickinson Biosciences, San Jose, California, United States), was done.
Sample preparation included lysing two or more 0.5 mL sample aliquots (dependent on the initial white cell count) with 14.5 mL isotonic ammonium chloride pH 7.1–7.4 (8.99% NH4Cl, 0.84% NaHCO3 and 0.0372% ethylenediaminetetraacetic acid; Merck, Darmstadt, Germany) in a 15 mL conical centrifuge tube for 15 min at room temperature. After incubation, samples were spun at 3000 g for 3 min, the supernatant was decanted, while the pellet was washed four times with 14 mL phosphate-buffered saline at pH 7.3 ± 2 (Oxoid Ltd, Basingstoke, United Kingdom) containing 0.09% sodium azide (NaN3) (Merck, Darmstadt, Germany) and 0.2% bovine serum albumin (Biowest, Nuaille, France). Following washing, samples were reconstituted to 0.5 mL with phosphate-buffered saline, and the white cell count was determined. Cells were then diluted (or concentrated) to achieve the recommended cellular concentration of ≤ 10 000 cells/µL for monoclonal antibody incubation. Subsequently, 100 µL of this cell concentrate, which contained approximately 106 cells, was added to the ClearLLab 10C and in-house panels (see
Possible reagent selection for method comparison according to suspected target population validated at the Charlotte Maxeke Johannesburg Academic Hospital flow cytometry laboratory in Johannesburg, South Africa, April 2020 to August 2020.
Fluorochrome DURAClone | FITC | PE | ECD | PC5.5 | PC7 | APC | APC-A700 | APC-A750 | PB | KO |
---|---|---|---|---|---|---|---|---|---|---|
RE ALB panel | CD58 | Blank | CD34 | CD10 | CD19 | Blank |
CD38 | CD20 | Blank | CD45 |
RE CLB panel | CD81 | ROR1 | Blank | CD79b | CD19 | CD5 | Blank | CD43 | CD20 | CD45 |
RE PC panel | CD81 | CD27 | Blank |
CD19 | CD200 | CD138 | Blank | CD56 | CD38 | CD45 |
FITC, fluorescein isothiocyanate; PE, phycoerythrin; ECD, phycoerythrin-Texas Red conjugate; PC5.5, phycoerythrin-cyanine5.5; PC7, phycoerythrin-cyanine7 conjugate; APC, allophycocyanin; APC-700, allophycocyanin-Alexa Fluor700; APC-750, allophycocyanin-Alexa Fluor750; PB, pacific blue; KO, krome orange.
, In two samples, liquid CD22 APC was added into the ‘blank’ channel;
, In one sample, liquid CD117 ECD was added.
All samples were acquired on a flow cytometer to acquire raw flow cytometric listmode data. For the predicate in-house panels acquired on the FACSCalibur, 5000 (.fcs data) events were collected. Paint-A-Gate software (Becton Dickinson Biosciences, San Jose, California, United States) was used for analysis of raw FACSCalibur flow cytometric data with a primary gating focus using CD45 and side scatter.
For both the ClearLLab 10C and DURAClone panels, at least 50 000 (listmode data) events were acquired on the Navios. Kaluza C™ version 1.1 (BC, Miami, Florida, United States) software was used to analyse all raw listmode data to facilitate clinical interpretation. Kaluza C™ ClearLLab B, T, M1 and M2, and DURAClone panel analysis protocols were developed according to manufacturer specifications.
Relative expression of CD81 relative to normal T-cells and granulocytes in the CLB tube validated at the Charlotte Maxeke Johannesburg Academic Hospital flow cytometry laboratory in Johannesburg, South Africa, April 2020 to August 2020. Each plot (a, b and c) shows CD81 expression (
Identification of plasma cells using the DURAClone PC tube validated at the Charlotte Maxeke Johannesburg Academic Hospital flow cytometry laboratory in Johannesburg, South Africa, April 2020 to August 2020. In histogram a, negative CD45 expression of plasma cells is shown (purple), with CD45 positive (background) expression noted amongst granulocytes (blue), monocytes (green) and mature lymphocytes (red). The target plasma cell population were primary-gated on bright CD138 and CD38 to reveal aberrant CD117 expression (b) and under-expression of CD81 (c) and CD27 (d).
A similar approach was used for each of the comparative second-line DURAClone RE ALB, RE CLB, and RE PC listmode data analysis but additionally incorporated manufacturer-recommended gating strategies for clinical interpretation of data.
Clinical diagnostic outcomes were collated into Microsoft Excel (Redmond, Washington, United States) spreadsheets. Marker expressions and specific clinical diagnoses noted for the DURAClone RE ALB, RE CLB, and RE PC panels were compared to the clinical diagnostic outcomes from the existing in-house 2–4-colour antibody panels described (
Twenty samples were tested with the DURAClone ALB panel. The DURAClone RE ALB evaluation on both routine diagnostic precursor B-ALL samples, and patient samples who were being followed up after therapy for precursor B-ALL, revealed 100% positive and negative agreement to ClearLLab 10C reported outcomes (
Comparison of predicate reagents and DURAClone RE ALB reagent during the assessment of normal and abnormal precursor B-cells undertaken at the Charlotte Maxeke Johannesburg Academic Hospital flow cytometry laboratory in Johannesburg, South Africa, April 2020 to August 2020.
DURAClone RE ALB | ClearLLab 10C and second-line in-house analysis |
Total ( |
|
---|---|---|---|
Disease detected – B-cell acute lymphoblastic leukaemia | No disease – normal haematogones | ||
Disease detected – B-cell acute lymphoblastic leukaemia | 11 (TP) | 0 (FP) | 11 |
No disease – normal haematogones | 0 (FN) | 9 (TN) | 9 |
Note: Percent agreement – Calculation details: Estimated diagnostic sensitivity true positive rate = 100% (TP/(TP+FN)); Estimated diagnostic specificity true negative rate = 100% (TN/(FP+TN)).
TP, true positive; TN, true negative; FP, false positive; FN, false positive.
The DURAClone RE CLB panel identified the B-cell target population using a combination of CD19 and CD5 or CD20, as well as surface CD81, ROR1, CD79b, and CD43 expression. Twenty samples were tested with the DURAClone CLB panel. The calculated sensitivity and specificity were 100%, with full diagnostic concordance noted across all cases evaluated (
Comparison of predicate reagents and DURAClone RE CLB reagent during the assessment of mature B-cell lymphoid proliferations undertaken at the Charlotte Maxeke Johannesburg Academic Hospital flow cytometry laboratory in Johannesburg, South Africa, April 2020 to August 2020.
Sample type | Number | Disease/haematologically malignant |
||
---|---|---|---|---|
ClearLLab 10C and second-line in-house diagnostics (diagnosis) | DURAClone RE CLB (diagnosis) | Agreement | ||
Peripheral blood | 7 | B-cell chronic lymphocytic leukaemia; CD23 positive | B-cell chronic lymphocytic leukaemia; ROR-1 positive | 100% diagnostic concordance |
Bone marrow | 9 | B-cell chronic lymphocytic leukaemia; CD23 positive | B-cell chronic lymphocytic leukaemia; ROR-1 positive | 100% diagnostic concordance |
1 | ~2% – 3% clonal B-cells; CD19 and CD20 positive with light chain restriction | CD19 and CD20 positive |
100% diagnostic concordance |
|
1 | Small cell, clonal mature B-cell lymphoproliferative disorder; CD5 negative | Small cell, mature B-cell lymphoproliferative disorder; CD5 negative, CD79b bright | 100% diagnostic concordance |
|
Pleural fluid | 1 | Follicular lymphoma; CD10 positive | Follicular lymphoma; CD10 and bright CD79b positive | 100% diagnostic concordance |
Note: Total –
, Disease was confirmed across both the predicate ClearLLab 10C with second-line in-house as well as with the index test, DURAClone RE CLB, in all instances.
Seventeen samples were tested with the DURAClone PC panel. A plasma cell population was identified by backbone markers CD138 and CD38 in the DURAClone RE PC panel, with subsequent determination of expression of CD19, CD27, CD45, CD56 and CD200. There was cross-panel marker equivalency to the existing in-house panels (
Comparison of predicate reagents and DURAClone RE PC reagent during assessment of normal and abnormal plasma cell populations undertaken at the Charlotte Maxeke Johannesburg Academic Hospital flow cytometry laboratory in Johannesburg, South Africa, April 2020 to August 2020.
Text method: DURAClone RE PC | Predicate method: ClearLLab 10C and second-line in-house analysis |
Total | |
---|---|---|---|
Disease detected – Plasma cell dyscrasia | No disease – normal or no plasma cells detected | ||
Disease detected |
15 (TP) | 0 (FP) | 15 |
No disease |
0 (FN) | 2 (TN) | 2 |
Note: Percent agreement – Calculation details: Estimated diagnostic sensitivity true positive rate = 100% (TP/(TP+FN)); Estimated diagnostic specificity true negative rate = 100% (TN/(FP+TN)); Diagnostic concordance = 100%.
TP, true positive; TN, true negative; FP, false positive; FN, false positive.
, Plasma cell dyscrasia – 14 bone marrow aspirate samples with 1% – 45% aberrant plasma cells) (1 peripheral blood sample diagnosed with plasma cell leukaemia, 45% aberrant plasma cells).
, 2 bone marrow aspirate samples with normal plasma cell immunophenotype, less than 1% – 4%.
The recently evaluated,
Complementary and supplementary multicolour panel options are published
The expression of CD10, CD19, CD34, CD38, CD20 and CD45 in the DURAClone RE ALB panel showed excellent cross-panel marker expression equivalency and concordant clinical diagnostic outcome when compared to ClearLLab 10C B-cell panel expression. CD58 together with CD38, CD10, CD19, CD34, CD20 and CD45 has potential for sensitive minimal residual disease assessment
The expression of CD5, CD19, CD20 and CD45 in the DURAClone RE CLB reagent showed excellent cross-panel marker expression, equivalent to ClearLLab 10C B panel expression, with 100% (
The CD38, CD138, CD45, CD19, CD56, and CD200 expressions in the DURAClone RE PC panel showed excellent cross-panel marker expression and equivalent diagnostic outcome and FACSCalibur/PAG analysis outcomes, with 100% (
Firstly, the sample size per panel evaluated is small, and further studies are needed to confirm the outcomes reported here. Secondly, the fluorochromes used in the panels are specifically designed for use on a BC Navios instrument. Therefore, the products could be used on alternative instruments only if the respective filter setups
This study confirms that the BC DURAClone ALB and CLB panels are suitable to provide additional second-line immunophenotypic workup of precursor B-ALL and mature B-cell lymphoproliferative disorders resepectively, at disease presentation and are suitable to supplement first-line ClearLLab 10C laboratory predicate method testing. The under-expression of CD81, CD79b, and positive CD43 and ROR1 expression, noted in the DURAClone CLB panel specifically assists in distinguishing B-cell CLL from mantle cell lymphoma and other CD5 negative B-cell LPDs.
The DURAClone RE PC panel is suitable for the second-line investigation of PCD. CD27, CD56, CD81 and CD200, together with CD19, CD38 and CD138 in a single analysis, were efficient in identifying aberrant plasma cells.
Although there was diagnostic and individual marker concordance, we did not find the ALB more useful for diagnosing B-ALL over our existing ClearLLab system (utilising the B-cell and M2 tubes). The inclusion of CD58
Lastly, the potential benefit of using commercialised multicolour lyophilised fixed panel preparations having stable, standardised antibody reagents reduces the risks of technical error, improves laboratory efficiency, and simplifies reagent inventory. In addition, the DURAClone ‘free fluorescence channels’ provide some additional flexibility for a laboratory to use their preferred markers to establish diseases (or not) of their choice.
D.K.G. and D.L. thank Mrs Merriam Machaba at Charlotte Maxeke Johannesburg Academic Hospital flow cytometry.
The authors declare that they have no financial or personal relationships that may have inappropriately influenced them in writing this article.
D.K.G. conceptualised, led and funded the study, provided project administration and project supervision as well as supervision of all laboratory testing and data analysis. D.L. and D.K.G. oversaw technical aspects and the setup of protocols for flow cytometric data acquisition and data analysis. L.S. and M.P. analysed flow cytometric data and recorded the outcomes in spreadsheets for analysis comparison. D.K.G. undertook final checking of data and compiled the final laboratory validation report. L.S. and D.K.G. translated the validation report into a first draft for publication format. All subsequent drafts were written by D.K.G.
Funding for this study was made possible with research incentive funds accrued to D.K.G. through the University of the Witwatersrand.
All outcomes are anonymised and published in the maunscript. Individal raw flow cytometric listmode data files are available on request from the corresponding author, D.K.G.
The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of any of the affiliated agencies of the authors.