4-colour guide

Step 1 – Set up

Instrument optimisation

For accurate and reliable paroxysmal nocturnal haemoglobinuria (PNH) analysis it is recommended to perform the following:

1. Daily performance check

  • Use standard beads (Flow Check® beads, calibration beads, Cytometer Setup and Tracking [CS&T] beads).

2. Application setting

  • Photomultiplier tube (PMT) setting, time delay, colour compensation (needs to be adjusted each time the voltage is changed).

For PNH analysis we target specific cell populations (neutrophils, monocytes, red blood cells [RBCs]) and use optimally titrated monoclonal antibodies, which require specific application settings (PMT setting and colour compensation). Therefore, establish appropriate cytometer settings to see negative and positive cell populations ‘on scale’ and optimise for PNH analysis.

NOTE: Ideally, application setting should be done with labelled and non-labelled target cells (neutrophils, monocytes, RBCs) from a PNH patient; however, in the majority of cases application setting is performed using calibration beads or lymphocytes (eg CYTO-COMP™ cells).

Don’t forget to select the appropriate reagents for your instrument

Set-up and quality control for white blood cell (WBC [neutrophils]) assay

Prepare 3 tubes:

  1. 100 µl peripheral blood + CD15 + CD45
  2. 100 µl peripheral blood + CD15 + CD45 + CD24 + FLAER
  3. 100 µl PNH blood or mix tubes 1 and 2 (prepare this tube after acquisition of tubes 1–2 on flow cytometer)
FLAER, fluorescent aerolysin
  • Lyse RBC cells in stained samples using a commercial lysing agent, and wash with phosphate-buffered saline supplemented with 1% bovine serum albumin.

Step 1: Forward scatter (FS) and side scatter (SS) adjustment for WBCs

  • Run tube 1 and set light-scatter voltages in linear mode (lin:lin light scatter) to bring all WBCs on scale.
  • Ensure FS threshold is not excluding lymphocytes (lymphocytes can be used as internal controls).

 

FS, forward scatter; SS, side scatter

Step 2: Voltage adjustment of PMTs for WBCs

  • Follow vendor-specific protocol for setting up voltages and compensation settings to ensure that negative and positive populations are ‘on scale’ and properly compensated.
    • If FC500 or later in use, DO NOT use ‘negative offset’.
    • Irrespective of platform, ensure all compensation settings are set to zero before setting PMT voltage.
  • Run tube 1 to verify that the CD45+/CD15++ neutrophils (no FLAER or CD24 added to create ‘bona fide’ PNH cells) are clearly visible and on scale.
The purpose of setting up a tube with gating antibodies (CD15, CD45) but not GPI markers (FLAER, CD24) is to identify the location of the FLAER-negative/CD24-negative cells. This step can also be done with a PNH sample, but since PNH samples are rare, it can be done using a normal sample. If the negative population cannot be seen, it may be that the voltage is too low and/or that the compensation setting is suboptimal, ‘crushing’ the cells against the baseline. An alternative cell-based method of compensation using CD3 conjugates has been described and details can be found at http://www.cytometry.org/web/modules/module2.pdf.
FITC, fluorescein isothiocyanate; FLAER, fluorescent aerolysin; GPI, glycosylphosphatidylinositol; PE, phycoerythrin
  • Run tube 2 to verify the presence and location of normal CD45+/CD15++ neutrophils showing CD24 and FLAER expression.
In this tube, all markers are added, and the normal neutrophils should show bright expression for FLAER (~2 decades higher than negative population) and moderate expression for CD24 (~1.5 decades higher than negative population). CD24 should not be titrated too much and special attention should be paid to the compensation. Although the FLAER signal is detected in the same PMT as FITC, the compensation for FLAER is different (usually less) than that for FITC. This means that the compensation set-up process should include a FLAER-stained sample rather than an FITC-stained sample; otherwise there may be overcompensation.
FITC, fluorescein isothiocyanate; FLAER, fluorescent aerolysin; PE, phycoerythrin; PMT, photomultiplier tube

Step 3: Verify the location of PNH and normal populations

  • Run tube 3 to verify the position of normal FLAER+/CD24+ neutrophils, and FLAER- and CD24-deficient PNH neutrophils.
This test sample can either be a PNH-positive case (as in this example) or a mixture of tubes 1 and 2 (if no PNH-positive sample is available).
FLAER, fluorescent aerolysin; PE, phycoerythrin
Example of PMT and compensation settings for WBCs (BD FACSCanto)
BD, Becton Dickinson; PMT, photomultiplier tube; WBC, white blood cell

Set-up and quality control for RBC assay

Prepare 6 tubes:

1 Unstained RBCs Make 1:100 dilution of normal RBCs (eg add 10 µL peripheral blood to 990 µL PBS)
2 RBCs only Use 50 μL of 1:100 dilution of normal peripheral blood
3 RBCs + CD59-PE Use 50 μL of 1:100 dilution of normal peripheral blood and stain with appropriate amount of CD59-PE only
4 RBCs + CD235a-FITC Use 50 μL of 1:100 dilution of normal peripheral blood and stain with appropriate amount of CD235a-FITC only
5 RBCs + CD59-PE + CD235a-FITC Use 50 μL of 1:100 dilution of normal peripheral blood and stain with appropriate amount of CD59-PE and CD235a-FITC
6 PNH blood + CD59-PE + CD235a-FITC, or mix tubes 2 and 5 Prepare this tube after acquisition of tubes 1–5 on flow cytometer
FITC, fluorescein isothiocyanate; PBS, phosphate-buffered saline; PE, phycoerythrin; RBC, red blood cell

It is important to have an RBC-specific voltage set-up (FS log, SS log) and compensation to bring the CD59-negative RBCs on scale. The typical leukaemia and lymphoma WBC voltage/compensation set-up should not be used for the RBC assay. Voltages are typically higher for the PNH RBC assay.

Step 1: Light scatter and PMT setting for RBCs

  • Run tube 2 and check light scatter voltages in logarithmic scale (log:log light scatter) to bring RBCs on scale. Make sure the discriminator does not ‘cut off’ cells.
    • Adjust voltages so that the RBC population sits just beyond the centre of the plot and above any forward scatter threshold/discriminator.
  • Clear all compensation settings (as the RBC assay set-up involves only two colours, manual compensation is generally sufficient).
    • If FC500 or later in use, DO NOT use ‘negative offset’.
  • Adjust voltages of FL1 and FL2 to bring the population on scale.
The intersection point of the two red lines (on the right) provides a rough estimate for the x and y mean channels for the negative population. The red lines can also be used as a guide for the expected locations of the CD59+ only population and the CD235a+ only population.

 

Example of PMT setting for RBC analysis with the BD FACSCanto machine (compensation is set to 0); step 1.
BD, Becton Dickinson; FITC, fluorescein isothiocyanate; FS, forward scatter; PE, phycoerythrin; PMT, photomultiplier tube; RBC, red blood cell; SS, side scatter

Step 2: Manual compensation settings for RBCs (FL1 – FL2)

  • Run tube 3 and adjust compensation: FL1 minus FL2. Set-up and quality control for RBC assay
Typically the population shows some bleeding (spillover) into the FL1 channels. This needs to be compensated by ‘moving’ the population (see arrow) into the location where it shows CD59 staining only and is aligned using the y-mean channel of the negative unstained population from the first tube. The blue line (y-mean channel) provides a visual guide for the location of the CD59+ only population.
FITC, fluorescein isothiocyanate; PE, phycoerythrin

Step 3: Manual compensation settings for RBCs (FL2 – FL1)

  • Run tube 4 and adjust compensation: FL2 minus FL1.
Typically the population shows some bleeding (spillover) into the FL2 channels. This needs to be compensated by ‘moving’ the population (see arrow) into the location where it shows CD235a staining only and is aligned using the x-mean channel guide (blue line) of the negative unstained population from the first tube. Keep in mind that this CD235a brightly positive population may not be as tightly clustered as the negative population and may ‘fan out’ in the higher decades (‘trumpet effect’). Caution should be exercised to avoid overcompensation, where most of the cells are ‘crushed’ against the axis.
FITC, fluorescein isothiocyanate; PE, phycoerythrin

Step 4: Verify and fine-tune the RBC setting

Verification of adequate compensation is achieved using a normal sample stained with CD235a-FITC and CD59-PE and a PNH sample:

  • Run tube 5, which contains cells stained with both CD235a-FITC and CD59-PE, to verify the position of normal Type I RBCs.
    • The normal sample stained with CD235a-FITC and CD59-PE is also used to establish the Type II gate and Type III gate.
  • Run tube 6, which contains PNH cells stained with both CD235a-FITC and CD59-PE, to verify the position of PNH Type III, PNH Type II and normal Type I RBCs.
Running a PNH-positive sample should show the CD59-deficient Type III RBCs in the correct location (CD235a+ and CD59–). The dark blue line represents a rough guide for the mean channel of the CD235a+/CD59– Type III cells. Overcompensation will cause the Type III cells to be ‘crushed’ and not be visible, and undercompensation will cause this population to move into the Type II region and cause underestimation of the Type III cells.
Both the dual-parameter dot plot (CD235a/CD59) and the single-parameter (CD59) histogram may be used to determine the Type III and Type II clone sizes. The dot plots (left) typically show a better separation of the cell clusters and also identify staining and compensation artefacts. The single-parameter histograms may also be used to help with cursor settings, as with the light blue lines (usually with larger clone sizes).
FITC, fluorescein isothiocyanate; PE, phycoerythrin; RBC, red blood cell
These protocols were developed in close collaboration with Mrs Andrea Illingworth of Dahl-Chase Diagnostic Services in Bangor, ME, USA, Drs Thomas Matthes and Mathieu Hauwel of the Swiss Flow Cytometry School at the University Hospital of Geneva, Switzerland, and Dr Iuri Marinov of Hematology and Blood Transfusion in Prague, Czech Republic. Images were provided with permission from the netflow Steering Committee and Swiss Flow Cytometry School.

STEP 2

Step 2 – Stain and acquire

Principle

Cells are stained with lineage-specific antibodies to clearly identify populations of interest. Glycosylphosphatidylinositol (GPI)-anchored protein deficiency is tested using specific antibodies (eg CD59, CD24) and a GPI-binding reagent (fluorescent aerolysin [FLAER]). Negative controls are included in the following protocol, which is optional for routine day-to-day testing. Additionally, normal blood from healthy donors can be used as a positive control.

It is recommended for each antibody to be titrated to determine the optimal working concentration

Methods: white blood cells (WBCs)

To screen for the presence of paroxysmal nocturnal haemoglobinuria (PNH) clones, it is recommended to perform both the neutrophil and monocyte assay in order to get a more accurate reflection of the true PNH clone size, rather than assaying only one lineage.

  1. Invert the tube with the anti-coagulated whole blood several times to mix the sample thoroughly.
  2. Take four 100 µL aliquots of blood using reverse pipetting and transfer them to the bottom of four 5-mL fluorescence-activated cell sorting (FACS) tubes.
    • Be careful not to leave any blood on the edges of the tube.
  3. Add the antibody cocktail and mix by gently pipetting up-and-down followed by gentle vortexing.
    • Take care to ensure no aerosols are formed.
    Fluorochrome A488 PE PC5 (BC)
    APC (BD)
    PC7 (BC)
    PerCP (BD)
    Neutrophil cocktail Test FLAER CD24 CD15 CD45
    Controla x x CD15 CD45
    Monocyte cocktail Test FLAER CD14 CD64 CD45
    Controla x x CD64 CD45
    aThis control is not required for everyday testing
    BC, Beckman Coulter; BD, Becton Dickinson; FLAER, fluorescent aerolysin; PE, phycoerythrin
  4. Incubate at room temperature in the dark for 20–30 minutes.
    • Be careful not to expose the antibodies to light.
  5. Add appropriate amount of vendor-specific lysing solution (VersaLyse, FACSLyse, Immunoprep) and follow specific protocol.
  6. Add 2 mL phosphate-buffered saline (PBS) with 1% albumin (PBA) and spin for 3 minutes at 300 g.
  7. Discard supernatant.
  8. Re-suspend in 0.5–1 mL PBS.
  9. Vortex to dissociate aggregates.
  10. Acquire at least 50,000 CD15+ neutrophils and as many as possible CD64+ monocytes, using an appropriate acquisition template (see Step 3 – Analyse).
  11. If a PNH clone is identified in the neutrophils, set up two more tubes for the monocyte assay and follow steps 2–11.

Red blood cells (RBCs)

  1. Invert the tube with the anti-coagulated whole blood several times to mix the sample thoroughly.
  2. Add 1 mL PBS to a fresh, clean tube and add 10 µL peripheral blood (1:100 dilution).
  3. Using reverse pipetting, take two 100 µL aliquots of 1:100-diluted blood and transfer to the bottom of two 5 mL FACS tubes.
    • Be careful to avoid aerosols and not to leave any blood on the edges of the tube.

    Add the appropriately titrated antibody cocktail and mix by up and down pipetting followed by gentle vortexing.

    Fluorochrome FITC PE
    Erythrocyte cocktail Test CD235a CD59
    Controla CD235a x
    aThis control is not required for everyday testing
    FITC, fluorescein isothiocyanate; PE, phycoerythrin
  4. Incubate at room temperature in the dark for at least 20 minutes (up to 60 minutes).
    • Be careful not to expose the antibodies to light.
  5. Add 2 mL PBS and spin for 3 minutes at 300 g.
  6. Discard supernatant.
  7. Add 2 mL PBS and spin for 3 minutes at 300 g.
  8. Discard supernatant.
  9. Re-suspend the pellet in 1 mL PBS.
  10. ‘Rack’ the tube vigorously 3–4 times to dissociate aggregates (see Racking).
  11. Acquire at least 100,000 CD235a+ RBCs using an appropriate RBC acquisition template (see Step 3 – Analyse). If any PNH Type III events are detected, up to 1 million RBCs may need to be collected to increase confidence in the analysis.
    • Acquire within 15 minutes, as the staining intensity of CD235a fades rapidly.
These protocols were developed in close collaboration with Mrs Andrea Illingworth of Dahl-Chase Diagnostic Services in Bangor, ME, USA, Drs Thomas Matthes and Mathieu Hauwel of the Swiss Flow Cytometry School at the University Hospital of Geneva, Switzerland, and Dr Iuri Marinov of Hematology and Blood Transfusion in Prague, Czech Republic.

STEP 1STEP 3

Step 3 – Analyse

Dot plot layout for neutrophils

Create a template according to the following layout:

  1. Dot plot of TIME vs CD45 or SS on ungated events
  2. Density plot FS linear (lin)/SS lin for elimination of debris
  3. Dot plot CD45 Logicle (Log)/SS lin for leukocyte gating
  4. Dot plot CD15 log/SS lin for neutrophil gating
  5. Two-colour dot plot FLAER log/CD24 log
FLAER, fluorescent aerolysin; FS, forward scatter; SS, side scatter

Gating strategy for neutrophils

  • Perform analysis of the acquired sample following the described gating strategy:

Evaluation of PNH neutrophils by flow cytometry

Step 1: Gate WBCs on TIME vs SS lin to confirm absence of fluidic irregularities (not shown).
Step 2: Gate WBCs on FS lin vs SS lin.
Step 3: Gate WBCs on CD45 Log vs SS lin.

 

Step 4: Gate neutrophils on CD15 versus SS lin and backgate on CD45 Log vs SS lin and FS lin vs SS lin to discriminate immature neutrophils and eosinophils.

 

Step 5: Gate PNH neutrophils on FLAER versus CD24.
Step 6: Check compensation on FLAER versus CD24 gated on lymphocytes (Ly); (we should discriminate B, T, NK, Ba and PNH-Ly).
FLAER, fluorescent aerolysin; FS, forward scatter; PE, phycoerythrin; SS, side scatter; WBC, white blood cell

Dot plot layout for monocytes

Create a template according to the following layout:

  1. Dot plot of TIME vs CD45 or SS on ungated events
  2. Density plot FS lin/SS lin for eliminating debris
  3. Dot plot CD45 Log/SS lin for eliminating debris
  4. Dot plot CD64 Log/SS lin for gating monocytes
  5. Two-colour dot plot FLAER Log/CD14 Log
FLAER, fluorescent aerolysin; FS, forward scatter; SS, side scatter

Gating strategy for monocytes

  • Perform analysis of the acquired results according to the following gating strategy:

Evaluation of PNH monocytes by flow cytometry

Step 1: Gate WBCs on TIME vs SS lin to confirm absence of fluidic irregularities (not shown).
Step 2: Gate WBCs on FS lin vs SS lin.
Step 3: Gate WBCs on CD45 Log vs SS lin.

 

Step 4: Gate monocytes on CD64 Log vs SS lin and backgate on CD45 Log vs SS lin and FS lin vs SS lin to discriminate dendritic cells and apoptotic monocytes.

 

Step 5: Gate PNH monocytes on FLAER Log vs CD14 Log.
Step 6: Check compensation on FLAER vs CD14 gated on Ly (we should discriminate Ly, NK, Ba and PNH-Ly).
FLAER, fluorescent aerolysin; FS, forward scatter; PE, phycoerythrin; SS, side scatter; WBC, white blood cell

Dot plot layout for red blood cells (RBCs)

Create a template according to the following layout:

  1. Dot plot of TIME vs SS on ungated events
  2. Dot plot FS logarithmic (log)/SS log for RBC gating
  3. Dot plot CD235a log/SS log for RBC gating on CD235a
  4. Dot plot FS int log/CD235a log to visualise RBC aggregates/doublets
  5. Dot plot of CD235a/CD59 and CD59 single-parameter histogram to determine CD59 expression or CD59 absence on CD235a+ RBCs
  6. Dot plot CD59 log/CD235a log to visualise co-expression
FS, forward scatter; RBC, red blood cell; SS, side scatter

Dot plot layout for RBCs

This set-up of dot plots and gating strategy is based on the diagnostic dot plot CD235a/CD59, which allows for verification of whether the Type III gate is clean in case of a normal blood sample. There might be some events ‘straggling’ into the Type II RBC area, but this is most likely due to suboptimal staining of occasional normal RBCs. True Type II PNH RBCs have high CD235a staining (unlike in this case).
FITC, fluorescein isothiocyanate; FS, forward scatter; PE, phycoerythrin; RBC, red blood cell; SS, side scatter

Gating strategy for RBCs

  • Perform analysis of the acquired sample according the following gating strategy:

Evaluation of PNH RBCs by flow cytometry

Step 1: Gate RBCS on TIME vs SS log.
Step 2: Gate RBCs on FS log versus SS log.
Step 3: Gate RBCs on FS log versus CD235a log.

 

Step 4: Check for doublets. If >2%, rack rigorously again.

 

Step 5: Display RBCs on CD59 versus CD235a dot plot and determine Type III (CD59 negative) and Type II (CD59 partially negative) PNH RBCs.
FITC, fluorescein isothiocyanate; FS, forward scatter; PE, phycoerythrin; RBC, red blood cell; SS, side scatter
These protocols were developed in close collaboration with Mrs Andrea Illingworth of Dahl-Chase Diagnostic Services in Bangor, ME, USA, Drs Thomas Matthes and Mathieu Hauwel of the Swiss Flow Cytometry School at the University Hospital of Geneva, Switzerland, and Dr Iuri Marinov of Hematology and Blood Transfusion in Prague, Czech Republic. Images were provided with permission from the netflow Steering Committee and Swiss Flow Cytometry School.

STEP 2STEP 4

Step 4 – Report

Standardised terminology for PNH clones1,2

A glycosylphosphatidylinositol (GPI)-deficient population of cells is the terminology used to describe the absence of GPI-linked proteins on red blood cells (RBCs), neutrophils and monocytes. Paroxysmal nocturnal haemoglobinuria (PNH) clone size is determined by the size of the GPI-deficient population in the blood cell populations tested (ie neutrophils, monocytes and RBCs). PNH reporting should include both white blood cell (WBC) populations, since PNH monocyte clones are often larger than PNH neutrophil clones, as well as RBC clone size if possible, to inform a more accurate estimation of the true PNH clone size in WBCs. An important factor in determining clone size is the sensitivity of the assay, which should be determined by performing a ‘spiking’ assay, which describes the lower limit of detection.

GPI-deficient population Standardised terminology
>1% PNH clone
0.1–1% Minor population of PNH cells or minor PNH clone
<0.1% Rare cells with GPI deficiency or rare cells with PNH phenotype
GPI, glycosylphosphatidylinositol; PNH, paroxysmal nocturnal haemoglobinuria

Reporting PNH test results1,2

  • Report the PNH clone size (ie, proportion [%] of GPI-deficient cells) in each lineage tested:
    • For RBCs, report the total clone size as well as the proportions of Type II and Type III cells.
      • Type III RBCs (completely GPI-deficient; CD235a+, CD59-).
      • Type II RBCs (partially GPI-deficient; CD235a+, CD59-intermediate).
      • Total proportion of Type III plus Type II RBCs.
    • For WBCs, report the total clone size.
      • GPI-deficient neutrophils (CD15+, FLAER-, CD157-).
      • GPI-deficient monocytes (CD64+, FLAER-, CD157-).
  • To avoid confusion, do not report Type I RBCs with normal CD59 expression.
  • Avoid the use of ambiguous language, such as ‘positive’ or ‘negative’, when reporting test results (eg ‘the PNH test was negative’ may be interpreted as that the test was negative for a marker and, therefore, positive for PNH).
    • Clear language such as “deficiency of GPI-linked proteins” should be used.1 Laboratory results can then be combined with clinical findings to confirm or determine the clinical diagnosis.
  • Include details of which reagents were used, including all gating and diagnostic markers, to confirm that high-sensitivity flow cytometry was appropriately performed.
  • Report the lower limit of quantification (LLOQ) for both the neutrophil assay and the RBC assay. Also, report the recommended LLOQ of 0.05% or better for RBCs (100,000 gated cells) and 0.1% or better for neutrophils (50,000 gated cells).
  • Include results from current and previous assessments to allow clinicians to easily observe any trends in PNH clone size over time.
    • Including dot plots is optional, but may be useful to support the presence of a PNH clone and illustrate the assay quality.
  • Report any re-testing recommendations.

Interpretation of PNH test results

  • In the case of GPI-deficient populations/PNH clones, further testing relating to haemolysis may be indicated to determine the next steps in patient management.
  • If supporting clinical evidence is provided indicating aplastic anaemia or myelodysplastic syndrome, a recommendation for further testing should be included on the report.

For further information on reporting, see the following links from the Canadian PNH Network:

 

These protocols were developed in close collaboration with Mrs Andrea Illingworth of Dahl-Chase Diagnostic Services in Bangor, ME, USA, Drs Thomas Matthes and Mathieu Hauwel of the Swiss Flow Cytometry School at the University Hospital of Geneva, Switzerland, and Dr Iuri Marinov of Hematology and Blood Transfusion in Prague, Czech Republic.

STEP 3ASSAY VALIDATION

  • References
    1. Davis BH et al. CLSI H52-A2 Red Blood Cell Diagnostic Testing Using Flow Cytometry; Approved Guideline, 2nd ed. Wayne, PA: Clinical and Laboratory Standards Institute. 2014.
    2. Illingworth A et al. Cytometry B Clin Cytom 2018; 94: 49-66.

Assay validation

Improper validation of flow cytometry assays is one factor responsible for reported interlaboratory variability in paroxysmal nocturnal haemoglobinuria (PNH) testing. To ensure the high-quality detection of PNH clones, assay validation should be performed as part of routine and high-sensitivity PNH testing.

The PNH assay validation process should include the proper titration of antibodies, spiking experiments to determine assay sensitivity, and the correct use of controls for confirming antibody and instrument performance.

Antibody titration assay

To produce interpretable histograms and dot plots, the proper titration of individual antibodies is recommended before the preparation of antibody cocktails. Optimising antibody concentration reduces background noise due to non-specific binding and also significantly reduces costs. In principle, antibodies are diluted serially and used in a single-staining assay. Mean fluorescence intensity is plotted against antibody amount to draw the titration curve.

Protocol

This protocol is given for an antibody whose recommended working volume is 20 µL (for 100 µL of cells). This volume was selected as a starting point for subsequent dilutions. Adapt this protocol to any other antibody by adjusting the starting point to the manufacturer’s recommendation.

  1. Fill 10 flow cytometry measurement tubes with 100 µL cell suspension.
  2. Fill five microtubes with 10 µL phosphate-buffered saline (PBS).
  3. Dilute the commercial antibody serially in microtubes by carrying 10 µL each time to the next tube (mix well and change pipette tip). See table below.
  4. Add decreasing antibody volumes of commercial antibody stock (from 20 µL down to 6 µL) to the cell suspension.
  5. Add 8 µL diluted antibody to the cell suspension.
  6. Incubate, wash and acquire as usual.
Volumes, µL Decreasing volume Serial dilutions
Antibody 10 10a 10a 10a 10a
PBS 10 10 10 10 10
Stain added to 100 μL cell suspension 20 10 8 6 8 8 8 8 8
Final amount 20 10 8 6 4 2 1 0.5 0.25
aPipette from previous tube; dash indicates non-applicable
PBS, phosphate-buffered saline

 

PBS, phosphate-buffered saline

Determining quantity of antibody to use in assay

To determine optimal antibody quantities for flow cytometry assays, use the lowest concentration of antibody that results in at least 80% of the maximum signal intensity. By reducing the antibody concentration, background noise due to non-specific binding is reduced (as well as saving on cost).

  • CD235a (glycophorin A [GPA]): The plateau is not reached even when using 20 µL of antibody, which is the amount recommended by the manufacturer. Given the shape of the dilution curve, using less antibody is not advisable.
  • CD59: As low as 0.5 µL of antibody yields 85% of the maximum staining intensity. This would be perfectly acceptable. However, to avoid dilution effects for samples that are highly cellular, a safety margin must be applied. Between 1–2 µL of antibody should be used (eg 1.5 µL).

 


MFI, mean fluorescence intensity

RBC spiking assay

Every laboratory should perform a ‘spiking’ assay to assess the sensitivity of the PNH test

Sensitivity refers to the smallest number of cells that can specifically be detected with a particular assay. The typical acquisition number for neutrophils is 50,000 cells. The number of monocytes is typically lower and thus results in lower sensitivity than that achieved with neutrophils.

A ‘spiking’ assay, which is basically a serial dilution of PNH blood with normal blood, can verify the sensitivity of detecting PNH clones at lower dilutions.

In case PNH blood is not available, PNH clones can be mimicked by using only gating markers and not glycosylphosphatidylinositol (GPI)-linked markers in the staining solution. For example:

  • Red blood cells (RBCs), 2-colour panel: X/CD235a will show location of CD59-negative RBCs.
  • Neutrophils, 4-colour panel: X/X/CD15/CD45 will show the location for fluorescent aerolysin (FLAER)/CD24-negative neutrophils.
  • Monocytes, 4-colour panel: X/X/CD64/CD45 will show the location of FLAER/CD24-negative monocytes.

The detailed protocol below shows how to establish RBC assay sensitivity based on a normal blood sample. In this case, sensitivity is determined by diluting RBCs stained only for GPA (therefore negative for CD59) into RBCs stained for both GPA and CD59, thus mimicking PNH clones of increasing size.

Protocol

  1. Dilute 5 µL fresh blood in 495 µL PBS or any other flow cytometry measurement (FCM)-compatible buffer.
  2. Incubate 500 µL diluted blood with 100 µL anti-GPA (CD235a fluorescein isothiocyanate [FITC] clone KC16; use at 20 µL/100 µL cells) antibody and 10 µL anti-CD59 antibody (clone MEM43; use at 2 µL/100 µL cells) for 15 minutes in the dark.
  3. Incubate 110 µL diluted blood with 22 µL anti-GPA antibody for 15 minutes.
  4. Spin at 500 g for 3 minutes and carefully discard supernatant.
  5. Resuspend in 1 mL CellFixTM.
  6. Incubate at room temperature for 30 minutes in the dark.
  7. Add 2 mL PBS, then spin at 500 g for 3 minutes and carefully discard supernatant.
  8. Resuspend in 600 µL and 110 µL PBS, respectively.
  9. Rack vigorously (or pipette) to dissociate aggregates.
  10. Pipette 100 µL single-stained RBCs in an FCM tube and acquire 250,000 events (tube 1).
  11. Distribute 90 µL double-stained RBCs in tubes 2-6 and 100 µL in tube 7.
  12. Take 10 µL single-stained RBCs and dilute serially in tubes 2-6 by carrying 10 µL each time to the next tube.
  13. Acquire 1,000,000 events.
Tube CD235a CD59 CD235a PNH clone
1 0 100 100%
2 90 10 10%
3 90 10 1%
4 90 10 0.10%
5 90 10 0.01%
6 90 10 0%
7 100 0 0%

Data analysis

The tables below show the results of a spiking experiment in which a PNH blood sample was spiked with normal blood. The same experiment can be used to check the sensitivity of detecting PNH monocytes and neutrophils, although the number of monocytes and neutrophils is typically lower and thus results in lower sensitivity than that achieved with RBCs.

4-colour neutrophil assay sensitivity

Dilution PNH neutrophils Sensitivity, %
Neat 51,420 91.3
1:10 5799 9.4
1:100 573 0.94
1:1000 91 0.089
1:10,000 9 0.01

4-colour monocyte assay sensitivity

Dilution PNH monocytes Sensitivity, %
Neat 12,718 89.8
1:10 1227 4.1
1:100 112 0.4
1:1000 13 0.044
1:10,000 ND ND
ND, not detected

2-colour RBC assay sensitivity (based on 1,000,000 RBCs collected in data files)

Dilution Type III RBCs Sensitivity, %
Neat 348,626 34.86
1:10 17,665 1.77
1:100 1822 0.18
1:1000 203 0.02
1:10,000 18 0.0018
RBCs, red blood cells

Quality control on lymphocytes

In addition to optimising antibody concentrations and validating the sensitivity of the assay, quality controls should be put in place to confirm antibody and instrument performance.

Lymphocytes, although not recommended for the detection of PNH clones, can serve as excellent quality-control material.

Compensation check on lymphocytes

FLAER, fluorescent aerolysin; FS, forward scatter; PE, phycoerythrin; SS, side scatter

The top row shows gating dot plots for PNH testing, the middle row shows diagnostic dot plots for PNH testing, and the bottom row shows some additional ‘control’ dot plots using PNH lymphocytes (red box).

Starting from left to right, the first dot plot (FLAER/CD24) shows good signal-to-noise ratio for the FLAER-positive normal lymphocytes versus the FLAER-negative PNH lymphocytes; both populations are visibly ‘on scale’. The purpose of the PNH lymphocytes is not to assess the PNH clone (which is always much smaller than in the neutrophils), but rather to ensure that the negative cells are visible and not ‘crushed’. This dot plot also shows the FLAER+/CD24+ B-cells and includes confirmation that CD24 was added. This may be important for cases with no, or rare, PNH neutrophils, which could also be seen if the antibody were not added.

The second (FLAER/CD14-ECD) and third (FLAER/CD15-PC5) dot plots also serve as confirmation for antibody performance and instrument settings.

These protocols were developed in close collaboration with Mrs Andrea Illingworth of Dahl-Chase Diagnostic Services in Bangor, ME, USA, Drs Thomas Matthes and Mathieu Hauwel of the Swiss Flow Cytometry School at the University Hospital of Geneva, Switzerland, and Dr Iuri Marinov of Hematology and Blood Transfusion in Prague, Czech Republic. Images were provided with permission from the netflow Steering Committee and Swiss Flow Cytometry School.

STEP 4TESTING PITFALLS

Testing pitfalls

Sample processing pitfalls

Many of the issues encountered in paroxysmal nocturnal haemoglobinuria (PNH) testing can be remedied by careful sample preparation.

Washing

Washing is a prime example of how sample preparation can have a major impact on the results of PNH testing. Failing to wash red blood cell (RBC) samples following antibody incubation can lead to false-negative results.

Results obtained from unwashed RBCs versus those washed twice

PE, phycoerythrin; RBC, red blood cell

Washing the cells reduces the background noise and reveals a PNH population that would otherwise have gone undetected.

Andrea Illingworth explains the importance of washing RBCs (see figures above)

Racking

Incomplete re-suspension of sample cells can also lead to aggregates, which make results more difficult to interpret. Vigorous ‘racking’ (ie rubbing the tube along the top of a plastic tube rack) is required to reduce the percentage of aggregates in a sample.

Differences between samples that have been lightly vortexed versus vigorously racked

FITC, fluorescein isothiocyanate; GPA, glycophorin A; PE, phycoerythrin

Light vortexing leads to 29% aggregates, while vigorous racking reduces this to 0.5% aggregates.

Andrea Illingworth explains the importance of racking (see figure above)

Filtering

It is also very important to filter reagents to avoid artefacts during PNH testing.

Normal control sample gated on GPA positivity and stained for CD59

FITC, fluorescein isothiocyanate; GPA, glycophorin A; PE, phycoerythrin; RBC, red blood cell; SS, side scatter

The top panels show what appears to be a low (1.5%), if diffuse, population of CD59-deficient cells; however, when the gate is expanded to include intermediate glycophorin A (GPA)-stained cells (bottom panels), there appears to be a large population (14.7%) of CD59-deficient cells. These events do not represent CD59-deficient cells but are more likely the result of debris or unfiltered phosphate-buffered saline.

Other confounding factors

Andrea Illingworth discusses other confounding factors

Staining with CD59 and GPA

Importance of staining cells with CD59 and GPA

Staining samples for both CD59 and GPA allows the separation of true Type II cells from inadequately stained normal RBCs.

Analysis of a sample from a patient with a small Type III PNH clone (0.2%)

FITC, fluorescein isothiocyanate; GPA, glycophorin A; PE, phycoerythrin

The stream of CD59-positive, intermediate GPA-stained cells seen in the top panel may be misinterpreted as a 2.0% population of PNH Type II cells.

Andrea Illingworth on how CD235a (GPA) versus CD59 provides quality control

Separating Type II from Type III RBCs

CD59 is recommended over CD55 for staining RBCs.

CD55 and CD59 staining in RBCs

PE, phycoerythrin; RBC, red blood cell

The figure clearly demonstrates that CD59 (top panels) has a better signal-to-noise ratio than CD55, allowing separation of Type II from Type III RBCs. CD55 (bottom panels) is unable to separate Type II from Type III RBCs, so only one PNH population is evident.

Andrea Illingworth talks about CD55 and CD59 staining in RBCs (see figure above)

Compensation

Importance of optimising compensation

Optimising compensation is very important.

FITC, fluorescein isothiocyanate; PE, phycoerythrin

The top-left dot plot shows a clear Type III PNH population, which is also seen on the bottom-left single-parameter histogram. Comparing the top-right dot plot and the single-parameter histogram on the bottom right, the histogram appears to show a prominent Type II population, whereas the dot plot above clearly shows a compensation issue (‘stab-like’ appearance of population in the Type II gate).

Gating

Importance of correct gating for RBC testing

Example of incorrect gate setting

FITC, fluorescein isothiocyanate; PE, phycoerythrin

At first glance, the two top graphs seem to show similar results to the dot plot and histogram below. However, the percentage of Type III cells is very different. This is due to the Type III gate (top-left dot plot) not including all negative events (it is just one channel from the baseline that excludes a large portion of the negative events). This sample also had an overcompensation issue, which pushed the negative events further to the left.

False positives

Potential false positives when testing white blood cells

Similarities between the side scatter of true PNH populations versus those of normal cells

FLAER, fluorescent aerolysin; PE, phycoerythrin; SS, side scatter

The chart on the top left shows a true PNH clone (blue) and the normal population (pink); the chart on the bottom left highlights the overlap in the side scatter of these populations. On the right-hand side, we see an immature blast cell population, with the bottom-right panel demonstrating how these cells may be misinterpreted as a PNH clone.

Andrea Illingworth explains potential false positives in white blood cells (WBCs) [see figure above]

Potential false positives caused by inappropriate gating

Gating with CD15 can be useful for excluding debris that might otherwise be misinterpreted as a PNH population.

What can happen if only CD45 gating is applied

FLAER, fluorescent aerolysin; PE, phycoerythrin; SS, side scatter

The top row illustrates a gating strategy using CD15 to remove unwanted debris from the sample. In the bottom row, where only CD45 versus side scatter has been used for gating, it can be seen that some of the population falls into the area predefined as PNH population; in reality, this population is most likely platelets and other debris.

Andrea Illingworth talks about potential false positives caused by gating (see figure above)

Lineage-specific gating

Lineage-specific gating using CD33 or CD64 instead of CD45 achieves the purest population of monocytes, improving the accuracy of PNH testing.

Importance of lineage-specific gating of monocytes

FLAER, fluorescent aerolysin; SS, side scatter

Although dual-parameter analysis with CD14 and fluorescent aerolysin (FLAER) and gating strategies using CD33 or CD64 can improve the accuracy of PNH testing in monocytes, downregulation of CD14 as a result of monocyte differentiation into dendritic cells can lead to inaccurate analysis of monocytes, where a small number of cells may look glycosylphosphatidylinositol (GPI) negative. Therefore, it is recommended to test both monocytes and neutrophils for PNH.

Andrea Illingworth explains the importance of lineage-specific gating in monocytes (see figure above)

Images were provided with permission from Mrs Andrea Illingworth of Dahl-Chase Diagnostic Services in Bangor, ME, USA.

ASSAY VALIDATION


Netflow Portal


You are now leaving www.netflowconnect.com website

Alexion Pharmaceuticals, Inc. has no influence or control over these websites and shall not be held responsible in any way for any of the content held there.

Continue Cancel