OUP user menu

Immunohistochemical Analysis of Monocytic Leukemias
Usefulness of CD14 and Krüppel-like Factor 4, a Novel Monocyte Marker

Jeffery M. Klco MD, PhD, Shashikant Kulkarni PhD, Friederike H. Kreisel MD, Tu-Dung T. Nguyen MD, PhD, Anjum Hassan MD, John L. Frater MD
DOI: http://dx.doi.org/10.1309/AJCPZ46PMMAWJROT 720-730 First published online: 1 May 2011


Detection of monocytic differentiation in myeloid neoplasms by immunohistochemical analysis is challenging owing to a lack of sensitive and/or specific antibodies. We tested the usefulness of immunohistochemical analysis for CD14, an antigen commonly detected by flow cytometry, and Krüppel-like factor 4 (KLF4), a potentially novel marker of monocytic differentiation, in a series of myeloid leukemias, including 53 acute myeloid leukemias with monocytic differentiation. These findings were compared with immunohistochemical findings for CD68 (KP-1), CD34, and CD163 and were also correlated with flow cytometric and enzyme cytochemical results. CD163 and CD14 are the most specific markers of monocytic differentiation, followed by KLF4. CD68, in contrast, is the most sensitive monocytic marker, and KLF4 is also significantly more sensitive than CD14 and CD163. These studies show that KLF4 is another marker of monocytic differentiation and that the combination of CD14 and CD163 can increase the diagnostic sensitivity for monocytic neoplasms.

Key Words:
  • Monocytic leukemia
  • Immunohistochemistry
  • CD14
  • Krüppel-like factor 4
  • Monocytic differentiation markers

The characterization of monocytic proliferations in the bone marrow requires a combination of morphologic examination complemented by immunophenotypic, enzyme cytochemical, cytogenetic, and molecular genetic studies. Positivity for nonspecific esterases, such as α-naphthyl butyrate esterase (ANBE) or α-naphthyl acetate esterase, by enzyme cytochemical studies and expression of CD14 and/or CD64 antigens by flow cytometric analysis are convincing evidence of monocytic differentiation.14 In addition, the presence of a translocation involving the mixed lineage leukemia (MLL) locus, especially t(9;11), or a translocation involving the core binding factor β at 16q22 is commonly seen in monocytic neoplasms. However, these translocations are not absolutely specific for monocytic lineage and identify only a limited number of acute myeloid leukemias (AMLs) with monocytic differentiation.57

Evidence of monocytic lineage can be established by immunohistochemical analysis, although this method is limited by the lack of highly sensitive and specific markers. For example, CD68 antigen, which can be detected by KP1 and PG-M1 antibodies, is not only expressed by monocytes/macrophages but is also detected in neutrophil-lineage primary granules8 and a number of nonmonocytic AMLs.9 In addition, CD68 expression can be observed in a wide range of neoplasms, including melanomas, carcinomas, and schwannomas.10 The hemoglobin scavenger receptor CD163,11 which is also frequently used as a monocytic marker, is more specific than CD68 for monocytic leukemias12,13 and has an expression profile limited to cells of the monocyte/macrophage lineage.14 However, detection of CD163 by immunohistochemical approaches has a low sensitivity for monocytic neoplasms, ranging from 6% to 49% for acute monocytic leukemias12,14,15 and chronic myelomonocytic leukemias (CMMLs).16,17

Thus, additional immunohistochemical markers of monocytic differentiation are needed. One intriguing possibility is KLF4, a member of the Krüppel-like factor (KLF) family of transcription factors, which has been shown to have a critical role in monocytic differentiation.1820 However, its expression in monocytic neoplasms has not been investigated. Furthermore, although CD14 is frequently assessed by flow cytometric analysis and is thought to be a specific marker for monocytes, its immunohistochemical usefulness and the correlation of flow cytometric and immunohistochemical studies have not been thoroughly investigated.

Herein we report the immunohistochemical expression profile of acute and chronic monocytic leukemias in a large retrospective study. We compared the expression patterns of KLF4 and CD14 with those of other commonly used markers in the workup of monocytic proliferations, including CD34, CD68 (KP-1), and CD163. These findings were also correlated with enzyme cytochemical and flow cytometric data. We show that KLF4 is a novel monocyte marker useful in the workup of potential monocytic neoplasms that is more specific for monocytic differentiation than CD68 (KP-1) and more sensitive than CD163 and CD14. CD14 and CD163 are the most specific markers of monocytic differentiation and can be used together to increase the diagnostic sensitivity for the detection of monocytic differentiation.

Materials and Methods

Case Selection

The Washington University School of Medicine (St Louis, MO) Institutional Review Board approved this study. A search of the database of the Division of Anatomic and Molecular Pathology, Washington University School of Medicine, was conducted for adults with a bone marrow biopsy with a diagnosis of a de novo monocytic neoplasm, including acute myelomonocytic leukemia, acute monocytic/monoblastic leukemia, and CMML from 2000 to 2008. All but 2 cases had flow cytometric studies performed. Cases were excluded from analysis if the core biopsy material was not sufficient for additional studies. Diagnoses were established according to World Health Organization Classification of Tumours of the Haematopoietic and Lymphoid Tissues 3 and were also classified according to the French-American-British (FAB) criteria.21 This identified 17 cases of acute monocytic/monoblastic leukemia (FAB M5), 26 of acute myelomonocytic leukemia (FAB M4), 22 of CMML, 6 of AML with MLL abnormalities, and 4 of AML with inv(16)(p13.1q22). In addition, 14 cases of nonmonocytic AML and 7 negative bone marrow biopsy specimens from patients undergoing staging for lymphoma were randomly selected.

Immunohistochemical Analysis

All tissue samples were fixed in 10% neutral buffered formalin and decalcified in EDTA before paraffin embedding. Immunohistochemical analysis was performed on 4-μm-thick sections using the antibodies listed in Table 1. Antigen retrieval was performed using citrate buffer, pH 6.0 (KLF4 and CD14) or EDTA buffer, pH 8.0 (CD34, CD68, and CD163). A Decloaking Chamber (Biocare Medical, Concord, CA) and heat-induced epitope retrieval (125°C for 30 seconds) were also used for CD14 immunohistochemical analysis. Staining was performed using the BenchMark XT (Ventana, Tucson, AZ) and antibody detection with UltraView Universal 3,3′-diaminobenzidine detection system (Ventana). Staining was considered positive for all antibodies tested if present in more than 10% of the mononuclear cells, regardless of the staining intensity. All immunohistochemical stains were blinded and independently reviewed by 2 hematopathologists (J.M.K. and J.L.F.), and discrepancies were settled over a multihead microscope.

View this table:
Table 1

Flow Cytometry

Immunophenotypic markers were scored as positive if expressed (dim or bright) in more than 20% of the analyzed cells. Analysis was performed using whole bone marrow and/or blood following RBC lysis (0.15 mol/L ammonium chloride, 10 mmol/L potassium bicarbonate, and 0.05 mmol/L EDTA) using instruments manufactured by Becton Dickinson (Franklin Lakes, NJ) and commercially available antibodies (Table 1), along with isotype controls for each assay.


Normal Bone Marrow

KLF4 expression was limited to occasional interstitial mononuclear cells, consistent with monocytes, along with dim reactivity in a subset of neutrophils in normal bone marrow. The staining was exclusively nuclear. These cells typically accounted for less than 10% of the overall cellularity with focal clusters. Subsets of endothelial cells also expressed KLF4, although not all vascular structures, such as those seen with CD34 immunohistochemical analysis, showed immunoreactivity. No KLF4 expression was present in the erythroid, megakaryocytic, or lymphoid series, similar to CD14 Image 1. CD14 expression was limited to scattered mononuclear cells with cytomorphologic features of monocytes. CD163 immunohistochemical analysis highlighted scattered monocytes and macrophages, but was largely negative, whereas CD34 immunohistochemical analysis was positive in only rare, singly dispersed mononuclear cells and the vascular endothelium. CD68 (KP-1) was positive in more than 10% of the cells in the normal bone marrow and highlighted mononuclear cells and scattered maturing myeloid elements.

Image 1

Expression of Krüppel-like factor 4 (A and B) and CD14 (C) in normal bone marrow (AC, ×1,000).

Acute Leukemias With Monocytic Differentiation

We studied 53 cases of AMLs with monocytic differentiation, including acute monocytic/monoblastic leukemias (FAB M5, n = 17), acute myelomonocytic leukemia (FAB M4, n = 26), and acute leukemias classified by the presence of recurring cytogenetic abnormalities such as inv(16)(p13.1q22) (n = 4) or MLL rearrangements (n = 6). Of the 17 cases of acute monocytic leukemia, 15 were best classified as acute monoblastic leukemia (FAB M5a). The 6 MLL-rearranged cases included 2 monoblastic leukemias with t(9;11), 1 monoblastic leukemia with t(6;11), 2 acute myelomonocytic leukemias with t(11;19), and 1 monoblastic leukemia with a cryptic MLL rearrangement detected only by fluorescence in situ hybridization break-apart probes. All 4 cases of AML with inv(16) (p13.1q22) demonstrated myelomonocytic differentiation.

Nuclear KLF4 expression was present in 39 of 53 cases of AML with monocytic differentiation Table 2. In contrast, CD68 immunohistochemical analysis was positive in all cases, whereas immunoreactivity for an antibody to CD163 was present in only a subset (8/53). Expression of CD14 was noted in 9 of the 52 cases. Only 2 cases demonstrated positivity for CD14 and CD163. CD34 expression was absent in all cases of acute monocytic leukemia but was positive in a portion of the cases with myelomonocytic features. Representative cases of acute monoblastic leukemia and acute myelomonocytic leukemia are shown in Image 2 and Image 3, respectively.

Acute Myeloid Leukemia

We evaluated 14 cases of AML, not otherwise specified, for expression of CD14 and KLF4, including 3 cases of minimally differentiated AML (FAB M0), 3 AMLs without maturation (FAB M1), and 8 AMLs with maturation (FAB M2). CD34 expression was present in 12 of the 14 cases. The 2 cases without CD34 expression by immunohistochemical analysis also demonstrated no expression by flow cytometric studies. CD68 was also expressed in 12 of 14 cases, in contrast with 8 of 14 cases expressing KLF4. Expression of KLF4 was frequently dim and in a subset of the blasts. Immunohistochemical studies for CD14 and CD163 were negative in the blasts in all cases. CD163 was expressed by numerous activated macrophages, which were not highlighted by CD14 immunohistochemical analysis. A representative case is shown in Image 4.

Chronic Myelomonocytic Leukemia

Expression of CD34 was increased in only 2 cases of CMML, both diagnosed as CMML-2, whereas immunohistochemical analysis for CD68 was positive in all cases. In contrast, 13 (59%) of 22 cases expressed KLF4. CD14 and CD163 were present in only 3 (14%) of 22 cases. A representative case is shown in Image 5.

View this table:
Table 2
Image 2

Histologic features and expression pattern in acute monoblastic leukemia. A, H&E. B, CD34. C, Krüppel-like factor 4 (KLF4). D, CD68. E, CD163. F, CD14. Case demonstrates positivity for CD68, KLF4, CD14, and CD163 with no expression of CD34 (AF, ×1,000).

Image 3

Histologic features and expression pattern in acute myelomonocytic leukemia. A, H&E. B, CD34. C, Krüppel-like factor 4 (KLF4). D, CD68. E, CD163. F, CD14. Case demonstrates positivity for CD68 and KLF4 with no expression of CD34, CD163, or CD14 (AF, ×1,000).

Image 4

Histologic features and expression pattern in acute myeloid leukemia without maturation (French-American-British class M1). A, H&E. B, CD34. C, Krüppel-like factor 4 (KLF4). D, CD68. E, CD163. F, CD14. Blasts express CD34 and a subset are positive for CD68 with no expression of KLF4, CD163, or CD14 (AF, ×1,000).

Ancillary Studies

Enzyme Cytochemistry

Of the 51 cases of acute leukemia with monocytic differentiation, 49 (96%) demonstrated positivity for ANBE. The 2 negative cases, both acute myelomonocytic leukemias, demonstrated morphologic features of immature monocytes with distinct and prominent monocytic components by flow cytometry. None of the cases without monocytic features had positivity for ANBE. In contrast, enzyme cytochemical studies for myeloperoxidase were negative in all cases with pure monocytic features. In contrast, 30 of the 32 myelomonocytic leukemias and 11 of the 14 AMLs without monocytic differentiation demonstrated positivity for myeloperoxidase.

Flow Cytometry

The flow cytometric findings for all cases of AML are reported in Table 2. Strong concordance was present between the immunohistochemical and flow cytometric results for CD34 expression, with 63 of the 65 cases demonstrating similar results. Positivity for CD14, in contrast, was more variable with concordance between the two modalities in 55 of the 63 cases analyzed. In 5 cases, positivity was demonstrated by flow cytometry with no immunohistochemical reactivity, whereas 3 were positive by immunohistochemical analysis but not by flow cytometric studies. Expression of CD64 by flow cytometry was a more reliable marker of monocytic differentiation because it was negative in all 14 AMLs without monocytic differentiation.


Myeloid leukemias with monocytic differentiation represent a heterogeneous group of bone marrow disorders. Acute monocytic leukemias frequently lack expression of CD34 or CD117, which are more often seen in other AMLs, but commonly express HLA-DR, CD33, and CD64 with variable positivity for CD14.22,23 Enzyme cytochemical stains for nonspecific esterases, such as ANBE, are perhaps the most sensitive and specific markers of monocytic differentiation,1,24 which is further illustrated by the results of this study. Unfortunately, the current battery of immunohistochemical stains for monocytic lineage, including CD68 and CD163, have low specificity and low sensitivity, respectively. It is clear that additional markers of monocytic differentiation that are suitable for immunohistochemical analysis are needed. In the present study, we investigated the usefulness of immunohistochemical analysis for KLF4, a potentially novel marker of monocytic lineage, and the toll-like receptor coreceptor CD14.

The KLF family of transcription factors, which all share a common 2 cysteine and 2 histidine zinc finger DNA binding domain, are important in cellular differentiation and proliferation, especially of epithelial cells.25 KLF4 was originally termed G-KLF for its expression in the digestive tract, where it is most abundant in terminally differentiated epithelium,26 but it is also found in the reproductive tract,27 lung,28 and vascular endothelium.29 KLF4 is also important in somatic cell reprogramming and embryonic cell self-renewal.30,31 Several studies have also demonstrated a role in monocytic differentiation.1820 The study by Feinberg et al20 implied that expression of KLF4 can induce monocyte differentiation and that KLF4 functions downstream of PU.1 by binding to and activating the CD14 promoter, suggesting that KLF4 may be a useful marker of monocytic differentiation.

In the present study, we demonstrated that in normal bone marrow, KLF4 expression is present in a subset of endothelial cells and mononuclear cells consistent with monocytes, along with dim expression in a subset of neutrophils. In addition, KLF4 was expressed in 39 of 53 AMLs with monocytic features and in 13 of 22 cases of CMML. Of the 14 cases of AML, 8 dimly expressed KLF4. Expression, when present, was similar in immature and mature monocytes. Compared with the other common markers used in the evaluation of monocytic leukemias, expression of KLF4 is more sensitive for monocytic differentiation than CD163 and CD14 (69.3% vs 14.7% and 16.0%, respectively) and more specific than the KP-1 antibody for CD68 (42.3% vs 14.3%). A polyclonal KLF4 antibody was used in this study, and it is unclear if the dim expression seen in normal neutrophils and in many of the nonmonocytic blasts represents normal physiology or low-level nonspecific binding. Unfortunately, no monoclonal antibodies that have been validated for immunohistochemical analysis in formalin-fixed, paraffin-embedded tissue samples are currently available.

CD14, along with CD163, is the most specific marker for monocytic differentiation. However, as shown by the present study and others,4,17 CD14 expression is not sensitive by immunohistochemical or flow cytometric analyses. Although CD14 and CD163 have low sensitivity, there is little overlap in positivity for these markers. Of the 53 cases of AMLs with monocytic differentiation, 15 (28%) were positive for CD14 or CD163, and both of these markers were negative in all cases of AMLs without monocytic differentiation. This finding suggests that the combination of CD163 and CD14 can have added value. Unlike CD163, CD14 immunohistochemical analysis does not highlight the activated macrophages, which are commonly present in the background of high-grade neoplasms. In addition, although other studies have suggested down-regulation of CD14 during monocyte maturation,22 in the present study, we showed CD14 expression in 17% (9/53) of acute monocytic leukemias and in 14% (3/22) cases of CMML. Furthermore, there is some variation in the pattern of CD14 expression between flow cytometric and immunohistochemical analyses because 8 of the 63 cases showed discrepant findings. The reason for this is unclear and is most likely multifactorial, including the increased sensitivity of flow cytometry, the focal nature of monocytic aggregates frequently seen in bone marrow core biopsy specimens, and the variation between the epitopes recognized by the different antibodies. It has been previously shown that different CD14 antibodies can recognize different monocytic populations during monocyte maturation.32

Image 5

Histologic features and expression pattern in chronic myelomonocytic leukemia. A, H&E. B, CD34. C, Krüppel-like factor 4 (KLF4). D, CD68. E, CD163. F, CD14. Case illustrates an increase in monocytes that express KLF4, CD68, and CD14 with no expression of CD34 or CD163 (AF, ×1,000).

The results of this large study on the immunohistochemical profile of monocytic leukemias show that KLF4 is an additional marker of monocytic differentiation that can be a useful adjunct to the current panel of antibodies for identifying monocytic differentiation. KLF4 is a more sensitive marker than CD14 and CD163. Although it seems to be more specific than CD68, like CD68, it is still expressed in a significant percentage of nonmonocytic AMLs. These overall findings suggest that a study of the expression profile of KLF4 in a more diverse series of monocytic and histiocytic neoplasms is warranted. CD14, in contrast, is not expressed in AMLs that lack monocytic differentiation, and the combination of CD14 and CD163 immunohistochemical analysis can have added diagnostic sensitivity. Unfortunately, the current panel of antibodies for immunohistochemical analysis, now including CD14 and KLF4, still has limited ability to differentiate immature from mature monocytes.


We thank Erica Crouch, MD, PhD, and Jinglan Liu for technical help.


  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.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
View Abstract