OUP user menu

Distinct Expression Patterns of CD123 and CD34 on Normal Bone Marrow B-Cell Precursors (“Hematogones”) and B Lymphoblastic Leukemia Blasts

Nagwa M. Hassanein MD, Felisa Alcancia MT, Kathryn R. Perkinson MT, Patrick J. Buckley MD, PhD, Anand S. Lagoo MD, PhD
DOI: http://dx.doi.org/10.1309/AJCPO4DS0GTLSOEI 573-580 First published online: 1 October 2009


We compared the expression of CD123, the α chain of the interleukin-3 receptor, on normal B-cell precursors in bone marrow (“hematogones”) from 75 specimens and on leukemic blasts in 45 newly diagnosed B-acute lymphoblastic leukemias (B-ALL) cases. We found that the less mature hematogones (dim CD45+) that express CD34 lack CD123 expression, whereas the more mature hematogones (moderate CD45+) lack CD34 but always express CD123. In contrast with this discordant pattern of CD34 and CD123 expression in hematogones, blasts in 41 (91%) of 45 cases of B-ALL showed concordant expression of the 2 antigens: 80% (36 of 45) cases expressed both antigens, whereas 11% (5 of 45) expressed neither. We found that these distinct patterns of CD34/CD123 expression on hematogones (discordant) and B-ALL blasts (concordant) remain stable after chemotherapy and are useful in differentiating small populations of residual blasts from hematogones that may be simultaneously present.

Key Words:
  • Hematogones
  • CD123
  • Flow cytometry
  • B-ALL
  • Lymphoblasts
  • Minimal residual disease
  • IL-3 receptor
  • B lymphoblastic leukemia

Hematogones are the normal bone marrow precursors of mature B-lymphocytes with morphologic and immunophenotypic properties that overlap those of lymphoblasts. They were originally identified by their characteristic morphologic features in bone marrow smears.1 Later, their relatively distinct flow cytometric immunophenotype was identified.2 They are found in low numbers in most normal adult marrow specimens analyzed by flow cytometry but occur in larger numbers in most healthy infants and young children and in a variety of diseases in children and adults.3 Hematogones may be particularly prominent in the regeneration phase following chemotherapy or bone marrow transplantation and in patients with autoimmune and congenital cytopenias, neoplasms, and AIDS.4

B-acute lymphoblastic leukemia (B-ALL) is the most common acute leukemia in children and also can occur in adults.5 It is an aggressive but potentially curable disease in which monitoring the immediate and early response to therapy is of critical importance for optimal management. Current management protocols require assessment of residual leukemic cells at defined intervals after initiation of chemotherapy. Increased numbers of hematogones are often present in these specimens and may cause problems in interpretation because they share many morphologic and immunophenotypic features with neoplastic lymphoblasts of B-ALL.6 A flow cytometric test that can reliably separate hematogones from B-ALL would be helpful in cases of suspected recurrent B-ALL.

CD123 is the α chain of the human interleukin (IL)-3 receptor and is essential for the formation of the high-affinity heterodimeric IL-3 receptor.7 IL-3 is known to be involved in proliferation and maturation of human B cells.8,9 In this retrospective study, we examined the expression of CD123 on bone marrow hematogones, mature B lymphocytes, and B lymphoblasts from acute lymphoblastic leukemia. We report the novel finding of distinct and opposing patterns of CD123 expression on hematogones when compared with B-ALL blasts. We show that this expression pattern is stable for individual cases of B-ALL and is useful to distinguish residual leukemic blasts from hematogones.

Materials and Methods

Patients and Samples

After obtaining approval from the institutional review board the pathology Cerner database (Cerner, Kansas City, MO) was searched to identify cases of B-ALL. From the records in the Clinical Flow Cytometry Laboratory, cases in which CD123 was analyzed were identified. This antibody was introduced in the laboratory for routine testing in August 2006, and since then, it has been a component of the antibody panel used on all specimens in which acute leukemia is suspected. Only in cases in which the specimen was insufficient for complete analysis (such as after treatment or in cerebrospinal fluid samples), the antibody was not tested. By cross-referencing these 2 lists, 45 cases of B-ALL in which flow cytometry was performed on bone marrow samples at initial diagnosis, including analysis of CD123, were selected for further analysis. Through a review of laboratory records, 75 flow cytometric analyses that included CD123 and in which the final diagnosis was not B-ALL were also identified. These 75 bone marrow specimens were obtained from patients with the following conditions: postchemotherapy acute myeloid leukemia (9 cases), post–stem cell transplantation for acute myeloid leukemia (AML; 8 cases), evaluation of peripheral blood cytopenia or other blood abnormality (45 cases), myelodysplastic syndrome (3 cases), and staging for lymphoma (10 cases). The listmode flow cytometry data from all identified cases were reanalyzed for hematogones.

Finally, mature B cells from 100 additional bone marrow samples from patients without a diagnosis of B-ALL were analyzed for expression of CD123. Out of these, 50 cases showed identifiable hematogones and the remaining 50 did not. The diagnoses in these cases were as follows: cases containing mature B lymphocytes with hematogones present: 12 cases of AML, 23 postchemotherapy bone marrow specimens, 1 case of chronic myelogenous leukemia, 8 cases of myelodysplastic syndrome, and 6 cases of cytopenia; cases showing mature B lymphocytes without detectable hematogones: 5 cases of myelodysplastic syndrome, 25 cases of AML, 9 cases of cytopenia, and 11 postchemotherapy marrow specimens.


Fluorochrome-conjugated antibodies to the following antigens (antibody clone shown in parentheses) were used to profile hematogones and B-ALL cells at indicated quantities per test: fluorescein isothiocyanate (FITC)-conjugated isotype control antibodies for IgG1/IgG2a (x40/x39), CD34 (8G12), CD10 (W8E7), CD71 (L01.1), and terminal deoxynucleotidyl transferase (HTdT-134) were used at 10 μL; myeloperoxidase (5B8) was used at 20 μL. Phycoerythrin-conjugated isotype control antibodies for IgG2a/IgG1 (x40/x39), CD117 (95C3), CD7 (M-T701), CD20 (L27), and CD33 (P67.6) were used at 10 μL per test, and CD123 (9F5) was used at 5 μL. CD45 (2D1) was labeled with peridinin chlorophyll protein–cyanine 5.5 and used at 10 μL. Allophycocyanin-conjugated antibodies were HLA-DR (L243), CD19 (SJ25C1), CD22 (SHCL1), and CD56 (NCAM 16.2), used at 2.5 μL, while IgG1 (not applicable) and CD38 (HB7) were used at 1.25 and 5 μL, respectively. A combined reagent containing FITC-labeled κ (TB282) and phycoerythrin-labeled λ (11552) was used at 10 μL. All monoclonal antibodies were purchased from Becton Dickinson, San Jose, CA, except CD7, which was from Invitrogen, Carlsbad, CA.

These antibodies were used in combination (cocktails) to form 4-color panels. The following combinations were used for hematogone analysis, with the antibodies in each combination being conjugated with FITC, phycoerythrin, peridinin chlorophyll protein–cyanine 5.5, and allophycocyanin, respectively: (1) isotype control IgG1/IgG2a, isotype control IgG2a/IgG1, CD45, and isotype control IgG1; (2) CD10, CD7, CD45, and CD19; (3) CD34, CD117, CD45, and HLA-DR; and (4) blank, CD123, CD45, and CD56. Additional combinations used for analysis of B-ALL cases were as follows: (1) CD71, CD33, CD45, and CD38; (2) κ, λ, CD45, and CD19; (3) cytoIgG1, cytoIgG2a, CD45, and CD22; (4) TdT, CD20, CD45, and CD22; and (5) myeloperoxidase, blank, CD45, and CD19.

Staining and Acquisition

The number of cells in the bone marrow specimens was counted, and the cell concentration was adjusted to 1 × 106/mL in each tube. Staining was performed on 5 × 105 cells per tube using the 4 color combinations of the conjugated antibodies listed in the preceding section for 15 minutes in the dark at room temperature. RBCs were lysed using FACSLyse solution (Becton Dickinson) for 10 minutes and centrifuged at 1,200 rpm for 5 minutes. The supernatant was aspirated, and the pellet was resuspended and washed with 2.0 mL of phosphate-buffered saline twice before being resuspended in 0.5% formaldehyde and examined. An isotype-matched negative control sample (BD Biosciences, San Jose, CA) was used in all cases to assess background fluorescence intensity.

Stained cells were acquired on a FACSCalibur flow cytometer (BD Biosciences) that was set up daily using validated quality assurance procedures. At least 10,000 events were acquired that crossed the preset threshold for forward scatter.

Data Analysis

Stored listmode data were analyzed using the CellQuest software program (BD Biosciences). An inclusion gate was first set on viable cells based on forward light scatter and side light scatter. An isotype control was used for quadrant adjustment to subtract autofluorescence and nonspecific binding. The expected positive and negative staining populations in each specimen were used to confirm appropriate staining as validated for the bone marrow specimens in the laboratory.

The following gating strategies were used for analysis of hematogones: First, the population was identified as hematogones by gating on the CD19+/CD10+ events and then back-gating them on the side scatter-vs-CD45 histogram. The location of the hematogones on side scatter vs CD45 was used to anchor this population in subsequent analysis. Next, the hematogones identified from this anchor were subdivided using CD34 expression into CD34+ and CD34– subpopulations. These subpopulations were “back-gated” to the CD45 and side scatter histogram to “anchor” them for further analysis. Finally, the CD123 expression on the subpopulations of hematogones (CD34+ vs CD34−) was determined. In cases of B-ALL, the dominant population of leukemic blasts was identified in the CD45-vs-side scatter histogram, and the expression of CD34 and CD123 on this population was then assessed. For analysis of mature B cells, the CD19+ B cells were divided into CD10+ (hematogone) and CD10− (mature B-cell) subsets. These 2 gated populations were anchored by back-gating to the CD45-vs-side scatter histograms. Further analysis for CD123 expression was carried out on these anchored populations.


Expression of CD123 and CD34 Is Discordant in Hematogones

In 75 (100%) of 75 bone marrow specimens, hematogones identified by their side scatter, variable CD45 (dim to moderate), and expression of both CD19 and CD10, were divided into 2 groups. The first group comprised less mature hematogones that expressed CD34 and had dim CD45 Image 1A (left panel). The second group was composed of more mature hematogones lacking CD34 but with moderate CD45 expression Image 1B (left panel). The expression of CD123 was found to be asynchronous in relation to CD34 in both groups of hematogones. Thus, the majority of the less mature hematogones expressed CD34 but did not express CD123. Only a small minority (range, 0%–30%; average, 9%) of CD34+ immature hematogones expressed CD123. On the other hand, the more mature hematogones, which did not express CD34, showed expression of CD123 on most cells (range, 70%–100%; average, 80%).

Image 1

Expression patterns of CD34 and CD123 on immature and mature hematogones. A, The less mature hematogones with dim CD45 expression are gated in the left dot plot (red dots). The middle histogram shows that a majority of these gated immature hematogones express a moderate level of CD34. The right histogram shows that the same population is negative for CD123 expression. B, The more mature hematogones with higher CD45 expression are gated (left panel) and show lack of CD34 expression (middle panel) but express a dim to moderate level of CD123 (right panel). FITC, fluorescein isothiocyanate; PE, phycoerythrin; PerCP, peridinin chlorophyll protein.

The expression of additional myeloid and lymphoid antigens was examined in 50 cases from this group. The hematogones did not show expression of the myeloid/monocytic markers CD13, CD14, CD33, or CD117 or the T-cell/NK-cell markers CD2, CD4, CD7, or CD56 in any case. Owing to the selection of these cases, additional B-cell markers were used in the original analysis only in 4 cases. All hematogones expressed CD22, whereas only the more mature hematogones expressed CD20.

CD123 Expression on Mature B Cells

Mature B lymphocytes identified by their specific side scatter, bright CD45, expression of CD19, and lack of CD10 expression were examined in 100 additional bone marrow specimens from patients who did not have a diagnosis of B-ALL. These specimens were divided into 2 equal groups depending on the presence or absence of detectable hematogones in the specimen. We found that in each case with a detectable hematogone population, the mature B lymphocytes also expressed CD123 Image 2A , whereas in every case without detectable hematogones, the mature B cells lacked the expression of CD123 Image 2B .

Image 2

CD123 expression on normal mature B lymphocytes. A, A representative case in which CD19+/CD10+ hematogones were identified and divided into the less mature (blue dots) and more mature (green dots) subgroups as before. The mature B cells (CD19+/CD10−) are represented by red dots and show expression of CD123 along with the more mature hematogones. B, Another representative case in which a distinct CD19+/CD10+ hematogone population was not identified. The mature B cells (CD19+/CD10−) lack expression of CD123 in this case. APC, allophycocyanin; FITC, fluorescein isothiocyanate; PE, phycoerythrin, PerCP, peridinin chlorophyll protein.

B-ALL Blasts Show Concordant Expression of CD34 and CD123

Next, we examined CD123 and CD34 expression in 45 newly diagnosed cases of B-ALL. We found that in 36 cases (80%), leukemic blasts expressed CD34 and CD123 Image 3A . Conversely, in 5 (11%) of 45 cases, neither antigen was expressed Image 3B . Thus, the expression of CD34 and CD123 was found to be concordant (both positive or both negative) in most B-ALL cases. Only in 4 additional cases in which the blasts did not express CD34, a variable proportion of blasts (7%–96%) expressed CD123 Image 3C . Examination of diagnostic flow cytometric analyses in these 45 cases of B-ALL showed that the leukemic blasts expressed CD33 in 23 cases (51%), CD117 in 3 cases (7%), CD7 in 5 cases (11%), CD56 in 6 cases (13%), and TdT in 39 cases (87%). The B-lineage markers CD20 and CD22 were expressed in 20 cases (44%) and 42 cases (93%), respectively.

Expression Pattern of CD34 and CD123 on B-ALL Blasts Is Stable

To determine if therapy alters the concordant expression pattern of CD34 and CD123, we examined the results of bone marrow analysis of 45 cases of B-ALL after chemotherapy. In 28 cases, residual leukemic blasts were detected. Of these, blasts from 27 cases were positive for CD34 and CD123 at initial diagnosis, and 1 was negative for CD34 and CD123. In each of these cases, the expression pattern of these 2 antigens remained constant after chemotherapy. Image 4A shows the expression pattern of these antigens at diagnosis in a representative case. After chemotherapy, the identity of the immature B cells was uncertain using conventional criteria. However, these cells were identified as hematogones by the discordant expression of CD34 and CD123 Image 4B . In 1 case, residual leukemic blasts and hematogones were present. The 2 cell types were correctly identified by using the concordant and discordant expression of CD34 and CD123 on leukemic blasts and hematogones, respectively Image 4C .

Image 3

Expression patterns of CD34 and CD123 on B-acute lymphoblastic leukemia (B-ALL) blasts. Three representative cases of B-ALL with the most common pattern in which blasts express CD34 and CD123 antigens (A), a less common pattern in which blasts lack both antigens (B), and the least observed pattern in which only CD123 is expressed by some blasts without expression of CD34 (C). FITC, fluorescein isothiocyanate; PE, phycoerythrin, PerCP, peridinin chlorophyll protein.


Following treatment for B-ALL, hematogones are often expanded in regenerating marrow and can potentially be mistaken for residual disease because of their similarities to neoplastic lymphoblasts.10 Hematogones range in size from 10 to 20 μm, but most of them tend to be small. The cytoplasm is usually scanty, and sometimes it is a thin rim of deep basophilic cytoplasm without granules or vacuoles. The nucleus is oval to round with homogeneous condensed chromatin. The nucleoli are usually absent or indistinct.11

Because of these morphologic similarities to B-ALL blasts, flow cytometric analysis is required to reliably distinguish the 2 populations. By flow cytometric analysis, hematogones have low side scatter and are CD19+/CD10+ and lack surface immunoglobulin expression. They can be divided into 2 or 3 stages of maturation based on expression level of CD45, CD34, and CD20.2,12 Neoplastic lymphoblasts may exhibit 1 or more aberrancies relative to normal B-lymphocyte precursors such as uniform expression of TdT and CD34; negative or underexpression of CD45, CD20, HLA-DR, and CD38; overexpression of CD10; an abnormal spectrum of CD22; and coexpression of CD34 and CD20.13

Image 4

CD34 and CD123 expression patterns in treated B-acute lymphoblastic leukemia (B-ALL). A, Representative histograms of bone marrow from a patient with B-ALL at diagnosis. The leukemic blasts express CD34 and CD123. B, Histogram of bone marrow analysis from the same patient after chemotherapy. The CD34+ B cells do not express CD123 and vice versa, indicating that these are hematogones. C, Bone marrow from another case of B-ALL following chemotherapy shows residual blasts (red dots) that express CD34 and CD123. The 2 populations of hematogones (immature population in blue and mature population in green) show discordant patterns of CD34 and CD123 expression (blue dots, CD34+/CD123−; green dots, CD34−/CD123+). FITC, fluorescein isothiocyanate; PE, phycoerythrin, PerCP, peridinin chlorophyll protein.

Our study provides another useful parameter to distinguish B-ALL blasts from hematogones by showing that the concordant expression pattern of CD34 and CD123 in B-ALL blasts differs from the discordant expression of these antigens in hematogones. We have successfully tested this strategy to differentiate between hematogones and residual or early recurrent leukemic blasts present in early regenerating or postchemotherapy B-ALL.

Other investigators have also reported methods for discriminating between normal B-cell precursors and neoplastic lymphoblasts. Farahat and associates,14 using quantitative double-labeling flow cytometry, found B-lineage ALL lymphoblasts to express fewer TdT and CD19 and more CD10 molecules than did hematogones. Weir and colleagues15 demonstrated quantitative differences in light scatter and intensity of antigen expression between these 2 populations. Rimsza and colleagues16 found a predominance of more mature B-cell precursors in hematogone-rich specimens relative to the least mature (CD34+/TdT+) cells that predominated in cases of ALL. In addition, the adhesion molecules CD44 and CD54 were expressed more heterogeneously on hematogones than on neoplastic lymphoblasts.17 However, these methods depend on evaluating subtle differences in the degree of antigen expression pattern that are subject to daily variation in staining and instrument setup. In contrast, the expression patterns of CD34 and CD123 provide a qualitative difference that is easy to define in 2-dimensional histograms (Image 4).

The stages of cell differentiation and lineage commitment in normal hematopoiesis are characterized by sequential expression of surface antigens.18 IL-3 stimulates cell cycle progression in early hematopoietic progenitors19 and promotes differentiation in a broad spectrum of hematopoietic cells, including pre-B and pro-B cells, in concert with other growth factors,8,9,20,21 while inhibiting apoptosis of hematopoietic cells.18 IL-3 acts through a heterodimeric cell membrane receptor comprising α and β chains.22 CD123, the α chain of IL-3 receptor, is a transmembrane, 378-amino-acid glycoprotein of 70 kDa that binds IL-3 with high specificity but low affinity. The β chain is involved in signal transduction, and the αβ heterodimer is a high-affinity, high-specificity functional receptor complex.7

The expression patterns of the IL-3 receptor on normal B cells and precursor B-cell neoplasms are not fully characterized. To our knowledge, this is the first study to demonstrate that there are distinct expression patterns of CD123 on subgroups of hematogones. This may provide a basis for studying normal B-cell maturation events under varying conditions.

We also noted that normal mature B-lymphocytes, defined by their CD19+/CD10− phenotype and CD45/side scatter features, express CD123 only if they are present in continuation with mature hematogones (Image 2). If there are no hematogones in the marrow, the mature B cells do not express CD123. This finding suggests that CD123 expression on normal mature B cells is probably lost in a time-dependent manner. Previous in vitro studies have shown that IL-3 functions predominantly as a differentiation factor on B-cell precursors,8 and short-term exposure to IL-3 at each stage of early B-cell development leads to stage-specific effects.23 It is intriguing to ask if a B-cell lymphoma such as hairy cell leukemia that constantly expresses CD123 arises from these “young” but fully mature B cells.

We have shown that hematogones display surface staining for CD123 only in the more mature fraction that lacks CD34 expression. This “discordant” pattern is in contrast with the almost invariable “concordant” expression of these 2 antigens in B-ALL blasts. This distinction is useful in correctly classifying immature B cells as residual leukemic blasts or hematogones in the bone marrow of patients treated for B-ALL.


Upon completion of this activity you will be able to:

  • discuss the molecular and functional properties of IL-3 receptor and the role of IL-3 in normal maturation of B cells.

  • define the morphologic and immunophenotypic characteristics of normal bone marrow B-cell precursors (hematogones).

  • describe the differences in the pattern of CD34 and CD123 expression on hematogones and on B-acute lymphoblastic leukemia (B-ALL) blasts.

  • analyze the flow cytometry data from bone marrow of a patient treated for B-ALL and distinguish residual leukemia cells and hematogones.

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 642. Exam is located at www.ascp.org/ajcpcme.


  • During this work, Dr Hassanein was supported as a member of the postdoctoral scientific mission by the Cultural and Educational Bureau of the Embassy of the Arab Republic of Egypt.


  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.
View Abstract