OUP user menu

Use of a FLAER-Based WBC Assay in the Primary Screening of PNH Clones

D. Robert Sutherland MSc, Nancy Kuek, Juan Azcona-Olivera PhD, Tanya Anderson, Erica Acton, David Barth MD, Michael Keeney
DOI: http://dx.doi.org/10.1309/AJCPMRDZZFQM7YJ4 564-572 First published online: 1 October 2009


Diagnosis of paroxysmal nocturnal hemoglobinuria (PNH) with flow cytometry traditionally involves the analysis of CD55 and CD59 on RBCs and neutrophils. However, the ability to accurately detect PNH RBCs is compromised by prior hemolysis and/or transfused RBCs. Patients with aplastic anemia (AA) and myelodysplastic syndrome (MDS) can also produce PNH clones. We recently described a multiparameter fluorescent aerolysin (FLAER)-based flow assay using CD45, CD33, and CD14 that accurately identified PNH monocyte and neutrophil clones in PNH, AA, and MDS. Here, we compared the efficiency of this WBC assay with a CD59-based assay on RBCs during a 3-year period. PNH clones were detected with the FLAER assay in 63 (11.8%) of 536 samples tested, whereas PNH RBCs were detected in only 33 (6.2%), and always with a smaller clone size. The FLAER assay on WBCs is a more sensitive and robust primary screening assay for detecting PNH clones in clinical samples.

Key Words:
  • Fluorescent aerolysin assay
  • FLAER assay
  • Flow cytometry
  • Paroxysmal nocturnal hemoglobinuria clones

Paroxysmal nocturnal hemoglobinuria (PNH) is a rare acquired hematopoietic stem cell disorder caused by a somatic mutation of the X-linked phosphatidylinositol glycan complementation class A (PIG-A) gene.14 Depending on whether this mutation is a missense or nonsense variant (or both), there is a partial or absolute defect in the biosynthesis and expression of glycophosphatidylinositol (GPI)-anchored cell-surface structures.5,6 PNH clones are also frequently found in aplastic anemia (AA)/hypoplastic anemia and myelodysplastic syndrome (MDS), in which there is an immune-mediated failure of normal hematopoiesis.7 Clinical features of PNH include intravascular hemolysis that leads to hemoglobinuria, bone marrow failure, and thrombosis, with the latter being a major cause of morbidity and mortality.8 Two GPI-linked molecules, CD55 and CD59, have an important role in the control of complement-mediated hemolysis of RBCs, and, although their complete or partial absence is specific for PNH,9,10 CD59 deficiency alone seems to be largely responsible for hemolysis and other clinical symptoms of this disease.11

Diagnosis and monitoring of PNH samples currently rely on the analysis of CD55 and CD59 expression on RBC and granulocyte lineages by flow cytometry.1214 Flow cytometric analysis of the RBC lineage in a patient who has not received a transfusion can additionally be used to enumerate the type III (complete deficiency) and type II (partial deficiency) clones and type I (normal expression) cells.14 However, hemolysis of type III RBCs and/or the recent transfusion of normal RBCs can significantly reduce the ability to detect PNH clones in the RBC lineage.13 In addition, significant technical issues remain associated with reliably delineating bona fide, rare PNH RBCs among “unstained” events in a stained sample. With respect to granulocytes, severe neutropenia in some patients makes it difficult to identify the target population.

An alternative approach has more recently been developed based on the bacterial toxin, aerolysin. Aerolysin binds to the GPI moiety of GPI-linked structures on normal cells, causing their lysis15,16; PNH clones, lacking GPI-linked proteins, are not lysed.17 A fluorochrome-conjugated derivative, FLAER, has subsequently been developed that binds to but does not lyse normal cells,18 and a number of studies have confirmed that even simple FLAER-based flow cytometric assays are more sensitive and accurate than assays based on individual antibodies to GPI-linked structures.1820

To improve our ability to also detect rare PNH clones in closely related diseases such as AA and MDS, we recently developed a more sophisticated multiparameter, single-tube FLAER-based assay.21 In addition to improving the sensitivity and accuracy of PNH clone detection in monocyte and granulocyte lineages, this assay also has usefulness in detecting other hematologic abnormalities. In this study, we compared the ability to detect PNH clones in the granulocyte and monocyte lineages using our FLAER assay with the more traditional CD59-based assay on RBCs on more than 500 samples submitted for PNH testing from November 2005 until December 2008. We also generated stabilized blood from a patient with PNH that remained stable for up to 12 weeks and that was suitable for quality control for the FLAER assay and the RBC assay.

Materials and Methods

FLAER Assay for WBCs

Samples were prepared and stained essentially as described previously.21 Briefly, RBCs from 100 μL of anticoagulated (EDTA) peripheral blood were lysed with Optilyse C (Beckman Coulter, Miami, FL), washed, and resuspended in 100 μL of phosphate-buffered saline (PBS)/3% albumin. The sample was stained with 10 μL of CD33-PE (phycoerythrin; clone D3HL60.251), 5 μL of CD45-ECD (Texas red; clone J33), 3 μL of CD14-PC5 (PE-cyanin 5; clone RMO52) (Beckman Coulter), and 10 μL of FLAER “working” solution21 (Protox Biotech, Victoria, Canada) for 15 minutes at room temperature. After washing, the sample was analyzed on an FC500 cytometer (Beckman Coulter).

RBC Assays

Single Color

For this assay, 10 μL of whole blood was diluted with 5 mL of PBS and 50 μL added to each of 2 tubes using the reverse pipetting technique. Next, 10 μL of CD59-FITC (fluorescein isothiocyanate; clone P282E, Beckman Coulter) was added to the second tube, the first serving as an unstained control sample. After 15 minutes at room temperature, the cells were washed, resuspended in 1 mL of PBS and 40,000 RBCs, identified by light scatter (log or linear forward vs log side scatter) acquired in listmode. All samples reported in this study were tested with this assay.

Dual Color (CD59-FITC:CD235a-PE)

A dual-color assay was developed based on adding PE-conjugated antiglycophorin A (CD235a; clone 11E4B-7-6, Beckman Coulter) to the single-color assay described above. The stock reagent was diluted 1:50 and 5 μL added to the CD59-FITC, as above.

Dual Color (CD59-PE:CD235a-FITC)

To obviate the aggregation of RBCs that occurred with CD235a-PE (see “Sensitivity of the RBC Assays” in the “Results” section), an alternative cocktail was developed based on 2 μL of neat CD235a-FITC (clone 11E4B-7-6) and 10 μL of CD59-PE (clone p282 [H19], BD Biosciences, San Jose, CA).

Quality Control of the FLAER Assay

A PNH patient sample was stabilized within 24 hours of sample draw using a previously described process.22 The staining and flow cytometric analysis of the fresh and stabilized samples from the same patient was performed with the FLAER assay for WBCs and the various RBC assays. Repeated testing was performed at selected intervals up to 12 weeks after stabilization to verify the stability of this material.


We have previously shown that the use of FLAER in combination with CD33, CD45, and CD14 allows the simultaneous detection of PNH clones in monocyte and neutrophil lineages.21 To assess the performance of this assay in a clinical laboratory setting, we used this assay to screen 536 samples submitted to our laboratory for PNH testing from November 2005 to December 2008.

In the study, 63 samples (11.8%) from 33 individual patients were shown to contain PNH clones in the monocyte and granulocyte lineages using the FLAER assay. In almost all samples tested, the monocyte and granulocyte lineages showed similar proportions of PNH clones vs normal cells. An example of the FLAER assay performed on a fresh PNH patient sample is shown in Image 1A containing 81% PNH granulocytes and 73% PNH monocytes.

In contrast, the single-parameter CD59 assay detected PNH RBC clones in only 33 of the 63 samples identified with the FLAER assay. An example of the single-color CD59 assay on the same fresh PNH patient sample, shown in Image 2A , contained approximately 17% type III RBCs, 7% type II RBCs, and 75% normal/type I RBCs by CD59-FITC and CD59-PE assays.

Image 1

Analysis of WBC paroxysmal nocturnal hemoglobinuria (PNH) clones with fluorescent aerolysin (FLAER) and multiparameter flow cytometry. A, Fresh PNH sample stained with FLAER, CD33, CD45, and CD14. Of monocytes from R2, 72.7% exhibit a CD14–/FLAER– PNH phenotype (lower left); 81% of neutrophils from R3 exhibit a CD14–/FLAER– PNH phenotype (lower middle). About 14% of the lymphocytes in this sample were also unstained by FLAER (lower right). B, Poststabilized sample stained as in A with FLAER, CD33, CD45, and CD14. Of monocytes from R2, 72.6% exhibit a CD14–/FLAER– PNH phenotype (lower left); 80.4% of neutrophils from R3 exhibit a CD14–/FLAER– PNH phenotype (lower middle). About 15.5% of the lymphocytes in this sample were also unstained by FLAER (lower right). C, Fresh sample from an aplastic anemia sample stained as in A containing small PNH WBC clones without detectable PNH RBC clones. Of monocytes from R2, 3.1% exhibit a CD14–/FLAER– PNH phenotype (lower left); 2.1% of neutrophils from R3 exhibit a CD14–/FLAER– PNH phenotype (lower middle). When the PNH neutrophils from region R5 are back-scattered to the CD45 vs side-scatter (SS) plot (top left), they exhibit slightly higher CD45 staining (black dots) than the normal neutrophils. ECD, Texas red; Lin, linear scale; PC5, PE-cyanin 5; PE, phycoerythrin.

For all patients in whom PNH clones were detectable by both assays, the proportions of PNH clones vs normal cells in the RBC lineage was always lower than that detected in monocyte and/or granulocyte lineages Table 1 . This relationship held even for patients who had not received transfusions, such as the example shown in Image 2A. For samples with a clinical diagnosis of PNH (Table 1), this discrepancy was marked (CD59 type II or type III cells, mean, 29%; range, 1%–96%; FLAER positive, mean, 85%; range, 10%–99%).

As shown in Table 2 , 26 patient samples contained PNH clones by the FLAER assay but failed to show detectable numbers of PNH RBC clones above our threshold of 3% negative cells (type II plus type III) with the single-parameter CD59-FITC assay. The clinical diagnoses of this set of patients were far more heterogeneous and in the AA and MDS cases, almost all patients had received prior RBC transfusions.

Stabilized PNH Sample

To generate suitable material for quality control and proficiency testing of the FLAER assay, we stabilized blood from the same PNH patient whose sample freshly stained with FLAER, CD14, CD33, and CD45 is shown in Image 1A. As shown in Image 1B , staining of the stabilized blood sample clearly demonstrates the qualitative and quantitative aspects of the FLAER assay to be unaffected by the stabilization process. Similarly, the numbers of type II and type III RBC clones were unaffected by the stabilization process (data not shown). Significant differences were not detected up to 12 weeks after stabilization, at which point our pilot batch of stabilized material was exhausted.

Sensitivity of the FLAER Assay

Previous single-parameter FLAER-based assays have been claimed to be capable of detecting PNH WBC clones in the 0.5% to 1% range. With multiparameter FLAER-based assays such as described herein, the sensitivity should be further improved. To confirm that small numbers of FLAER-negative neutrophils indeed represent PNH clones, we were able to take advantage of the fact that PNH granulocytes express slightly higher levels of CD45 than their normal counterparts in the same stained sample Image 1C . With this extra Boolean gate, we were able to confirm that the small numbers of FLAER-negative neutrophils detected in some samples (Table 2) were PNH clones. Although this gating step is of most use in samples containing 5% or fewer PNH neutrophils, the phenomenon is consistently noted regardless of the size of the PNH neutrophil clone (data not shown).

Image 2

Analysis of RBC clones by single- and dual-color flow cytometry. A, Fresh paroxysmal nocturnal hemoglobinuria (PNH) sample diluted 1:500 and stained with CD59-FITC (top) or CD59-PE (bottom). Type III, type II, and type I (normal) RBCs are gated in regions R2, R3, and R4, respectively (for CD59-FITC), and R5, R6, and R7, respectively, for CD59-PE. B, Fresh PNH sample as in Image 1A stained with CD235-PE and CD59-FITC. Note CD235PE-induced RBC aggregation and poor resolution between type I (normal) RBCs (R5) and PNH clones (R4). Aggregates can be detected in plot 1 as a diagonal smear in region R1, also as a second peak of brightly stained CD235+ events in plot 3, and as a diagonal smear in region R5 of plot 4.

C, Fresh PNH sample as in Image 1A stained with CD235-FITC and CD59-PE. Note the lack of RBC aggregation in regions R1, R3, and R5 and better signal-to-noise of the FITC-conjugated CD235. However, note the poor resolution between type I (normal) RBCs (R5) and type II and type III clones (R4). FITC, fluorescein isothiocyanate; FS, forward scatter; FSC-H, forward scatter height, log scale; SSC-H, side scatter height, log scale.

Sensitivity of the RBC Assays

It is problematic to demonstrate that the presence of small numbers of “negative” events in a single antibody conjugate–stained sample represent bona fide PNH clones and are not due to aerosols, other technical errors, or the presence of noncellular events and debris. During the course of this study, a number of samples were analyzed that contained more than our threshold 3% CD59– RBCs in the absence of any detectable PNH clones with the FLAER assay. On repeating both assays, the FLAER assay always generated the same results, but the RBC assay was shown to be in error, with none of these samples showing more than 3% CD59– cells on repeated testing (data not shown). To develop a more robust assay and exclude non-RBC events from analysis, we initially added an antiglycophorin A conjugate (CD235a-PE) to the basic CD59-FITC assay. We found that we needed to titrate the CD235a extensively to reduce the aggregation of RBCs. Even after diluting the commercial reagent 1:50 and using only 5 μL of it, we could not entirely prevent all aggregation while maintaining a reliable level of specific staining. This compromised our ability to accurately delineate type I RBCs from PNH clones. An example of the CD235a-PE/CD59-FITC assay performed on the same fresh PNH patient is shown in Image 2B .

We then tested the same antiglycophorin A antibody in FITC-conjugated form alongside a PE conjugate of a different CD59 clone from another vendor. An example of this assay on the same fresh patient sample is shown in Image 2C . At 2 μL of neat CD235a-FITC, RBC aggregation was essentially undetectable, while still giving a good staining signal. In the presence of CD235-FITC, however, the CD59-PE conjugate was not very effective in delineating type II from type III RBCs (Image 2C). Regardless, the ability of this 2-color assay to distinguish normal (type I) RBCs from PNH clones was maintained, with approximately 75% of the CD235+ events representing type I cells, in agreement with single-color analysis (Image 2A).


We recently described a single-tube, multiparameter, flow cytometric assay based on FLAER, CD14, CD33, and CD45 that allows the simultaneous detection of PNH clones in monocyte and neutrophil lineages.21 The assay could detect small PNH clones (0.5%–1%) in samples up to 48 hours after sample draw and required only 40 minutes from sample receipt to result availability. The assay could be performed on a wide range of 4-color instruments, and specialized software was not required for acquisition or analysis. It is interesting that because of the antibody combination used, the FLAER assay proved useful in detecting other previously undiagnosed hematologic abnormalities in samples submitted for PNH screening.21

In this study, we assessed more than 530 fresh patient samples submitted to our clinical flow facility for PNH testing. We compared the ability of the FLAER assay to detect PNH clones in the neutrophil and monocyte lineages with a simple CD59-based assay for PNH RBCs. Of 63 samples demonstrated to contain PNH clones with the FLAER assay, just more than half of them contained detectable numbers of CD59– RBCs. No cases of congenital CD59 deficiency were detected during the course of our study. Among samples containing both PNH WBC and RBC clones, the size of the PNH clone was always larger with the FLAER assay than with the RBC assay. The explanation for this difference in bona fide PNH cases is that the FLAER assay is unaffected by hemolysis and/or prior transfusion of RBCs. In some of our cases, this relationship held even when the patient had not undergone any previous RBC transfusions. Thus, the FLAER assay on WBCs more reliably quantifies the size of the PNH clone than does the CD59 RBC assay, a result in keeping with other FLAER-based1820 and non–FLAER-based studies comparing granulocyte and RBC clone sizes.1214

Our data also suggest that the identification of PNH WBCs with the FLAER assay represents a far more efficient and cost-effective primary screening test for most clinical flow laboratories that perform PNH testing than RBC-based assays. This is most significant for laboratories that perform large numbers of tests for this rare disease. Although testing for the presence of PNH RBC clones may provide additional diagnostic information, studies have shown the probability of a thromboembolic event is directly related to the size of the PNH WBC (granulocyte) clone.23,24 In rare cases of PNH in which only neutrophils can be detected or when the sample comes from a neutropenic person, it is advantageous that the assay is capable of detecting PNH clones in the monocyte or neutrophil lineage.

View this table:
Table 1

With respect to testing for PNH RBCs, it is important to note that good assay design is of great importance to the generation of reliable data. As indicated, our initial assessment of PNH RBC clones relied on a simple screen for CD59 staining, but this assay identified several samples as containing PNH clones in the absence of similar indications from the FLAER assay. Repeating both assays on these samples invariably confirmed the CD59 assay to be in error. The major cause of these discrepancies seems to be technical. The generation of aerosols during pipetting and/or inadequate mixing of diluted blood samples with CD59 are the most likely explanations. The increase in the blood dilution factor from 1:150 to 1:500 and the adoption of gentle reverse-pipetting techniques using 50-μL rather than 20-μL volumes greatly reduced the frequency of disparate results between the assays but did not eliminate it totally.

To improve the fidelity of the RBC assay and eliminate the remaining false-negatives, we added PE-conjugated antiglycophorin A (CD235a) to the CD59-FITC assay. Unfortunately, CD235a-PE caused massive agglutination of the RBCs, except at very dilute concentrations, at which point, the level of specific staining vs unstained cells was suboptimal. Because glycophorin A is a highly negatively charged, mucin-like molecule, we reasoned that the use of the same monoclonal antibody labeled with a negatively charged fluorochrome such as FITC might promote less agglutination than the PE-conjugated version. This turned out to be the case, and we were able to determine a concentration of the FITC conjugate that did not promote agglutination while still exhibiting a good signal/background staining ratio. By changing the cocktail to CD235a-FITC and CD59-PE (BD Biosciences, San Jose, CA), we were able to detect and distinguish normal type I cells from type II and type III PNH clones within the CD235a+ fraction in PNH samples. Furthermore, during the study, we did not detect any samples that contained PNH RBCs with the CD235-FITC/CD59-PE combination when the FLAER assay failed to detect PNH WBCs. However, we subsequently analyzed a case that failed to show PNH clones by the FLAER assay (even after repeated testing) but showed a large type II RBC population with the single- and dual-color RBC assays (data not shown). This turned out not to be a case of congenital CD59 deficiency but one of cold agglutinins.

View this table:
Table 2

In the development of any clinical assay, it is of value to have available suitable material for quality control and proficiency testing purposes. This is especially important when assessing an assay for a disease as rare as PNH. In this study, we assessed the suitability of a stabilized PNH sample for this assay and found that the data generated on this material was the same as that generated on the fresh sample with our FLAER assay. Similarly, the fresh and stabilized material yielded almost identical data with the RBC assays. Subsequent analyses undertaken during a 12-week period demonstrated this material to be stable with similar data being generated with the FLAER assay and the various RBC assays (not shown).

We compared our multiparameter FLAER-based assay for PNH clone detection in monocyte and neutrophil lineages with more traditional methods used to detect PNH clones in the RBC lineage in more than 530 patient samples. Our data demonstrate the clear superiority of the FLAER assay over the RBC assays in the primary screening of PNH, AA, and MDS samples. Therefore, to optimize laboratory efficiency, we recommend that the RBC assay should be performed only in the context of finding PNH WBCs with the initial FLAER test.


Upon completion of this activity you will be able to:

  • discuss the role of the fluorescent aerolysin (FLAER)-based assay in the detection of paroxysmal nocturnal hemoglobinuria (PNH) clones.

  • compare the sensitivity of the FLAER assay in WBCs vs the CD59 assay on RBCs in screening for PNH.

  • list the technical issues involved in the screening of RBCs for PNH clones.

The ASCP is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The ASCP designates this educational activity for a maximum of 1 AMA PRA Category 1 Credit™ per article. This activity qualifies as an American Board of Pathology Maintenance of Certification Part II Self-Assessment Module.

The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose.

Questions appear on p 641. Exam is located at www.ascp.org/ajcpcme.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
View Abstract