Advancing Cancer Diagnostics
Improving Lives
Warning! You won't be able to use the quotation basket until you enable cookies in your Web browser.
Warning! Your Web browser is no longer supported. Please upgrade to a modern browser.

First CD33 Antibody Effective in FFPE Tissue - New reagent to detect acute myeloid leukemias

Detection of CD33 using monoclonal antibodies has been critical in immunophenotyping acute leukemias, particularly acute myeloid leukemias. To date this has only been possible by either flow cytometry or by frozen section immunohistochemistry. Now the first CD33 antibody has been developed that is effective on formalin-fixed, paraffin-embedded (FFPE) tissue.


CD33 is a 67 kD glycosylated transmembrane protein that is a member of the sialic acid-binding immunoglobulin-like lectin (Siglec) family1,2. This antigen is expressed in the earliest myeloid progenitor cells, but not in hematopoietic stem cells and is present during myelomonocytic differentiation as well as in granulocytes and resident histiocytes at low levels. It is an antigen retained on monocytes, expressed in dendritic cells and mast cells. Anti-CD33 antibodies have in the main been utilised to phenotype acute myelogenous leukemias 3,4. CD33 is also described as the target for Myelotarg® (Gemtuzumab Ozogamicin), a therapy used in the treatment of patients with acute myeloid leukemias, so immunophenotyping these and related tumors has also assumed greater therapeutic importance 5,6.

In April of 2007, Novocastra™ scientists developed the first ever antibody to CD33 antigen, (NCL-L-CD33, clone PWS44), effective on formalin-fixed, paraffin-embedded tissue. So now pathologists can view the morphology together with the immunophenotyping of CD33 positive cells in routinely fixed and immunohistochemically stained tissues without dependence on flow cytometry or on frozen section immunohistochemistry.

Fig. 1: Acute myeloid leukemia (M3) in bone marrow: Hematoxylin & Eosin stained section (left) and cytoplasmic and membrane immunohistochemical staining for CD33 antigen using NCL-L-CD33 (right). Paraffin sections.

Evaluated by pathologists

After “in-house” evaluation on a range of various human tissue types, both normal and neoplastic, the new reagent was evaluated in the field by pathologists to maximise the understanding of its specificity, sensitivity and utility. The first published work with the new Novocastra CD33 (clone PWS44) antibody was a poster presented by Dr Alan Ramsay, University College London Hospital, at the 191st Pathological Society meeting of Great Britain and Ireland. Subsequent presentations were made by Ahmet Dogan and colleagues from the Department of Laboratory Medicine and Pathology, Mayo Clinic Rochester, MN USA, and Elizabeth Hyjek and her team from Weill Cornell Medical College, New York at the 2007 USCAP meeting.

Fig. 2: Granulocytic Sarcoma of the tongue: Hematoxylin & Eosin stained section (left) and cytoplasmic and membrane immunohistochemical staining for CD33 antigen using NCL-L-CD33 (right). Paraffin sections.

Dr. Alan Ramsey

Dr. Alan Ramsay evaluated the CD33 antibody (clone PWS44) against a range of tumors that included acute myeloid leukemia (Fig. 1) , granulocytic sarcoma (Fig. 2), T-ALL and B-ALL cases and also cases of Burkitt lymphoma, classical Hodgkins lymphoma and plasma cell myeloma using Leica Microsystems’ Bond-max™ automated immunostainer together with the Bond Polymer Refine Detection System (Leica Microsystems, Biosystems Division) and ER2 (pH9.0) unmasking solution for 30 minutes. Dr Ramsay observed the following:

  • Clone PWS44 stains myeloid and monocytic cells in bone marrow and elsewhere, and is positive in most cases of acute myeloid and myelomonocytic leukemia.
  • Clone PWS44 shows crisp staining with a broad a range of myeloid and myelomonocytic leukemias (Fig. 3).
  • Clone PWS44 stains mature histiocytes and macrophages in all tissues.
  • Erythroid cells, megakaryocyte and lymphoid cells do not stain with Clone PWS44.
  • Clone PWS44 stains myelo-monocytic cells reactive in bone marrows and in myelodysplastic/myeloproliferative conditions but is not useful for assessing cellular maturity.
Fig. 3: Plasma cell leukemia in bone marrow: Hematoxylin & Eosin stained section (left) and cytoplasmic and membrane immunohistochemical staining for CD33 antigen using NCL-L-CD33 (right). Paraffin sections.

Dr. Elizabeth Hyjek

Dr. Elizabeth Hyjek’s evaluation was based on the comparison of clone PWS44 using a Leica Microsystems’ Bond-max™ and the Bond Polymer Refine Detection System (Leica Microsystems, Biosystems Division) with different CD33 antibodies bound to different fluorochromes. Furthermore, clone PWS44 was evaluated to characterize its pattern of staining with a range of other hematopoietic tumors. Her team made the following conclusions:

  • Clone PWS44 can reliably detect CD33 expression in paraffin tissue sections by IHC on normal/neoplastic myelomonocytic lineage cells, mast cells and dendritic cells. The pattern of PWS44 reactivity parallels that of other anti-CD33 monoclonal antibodies by flow cytometry and frozen section immunohistochemistry.
  • Clone PWS44 can identify low expression of CD33 in AML cases not detected by some FITC-conjugated anti-CD33 monoclonal antibodies (clone MY9-FITC used for immunophenotyping by flow cytometry).
  • Clone PWS44 is useful in differentiating diagnosis of myeloid tumors from their lymphoid mimics and other non-hematopoietic tumors.
  • Clone PWS44 can be reliably used for identifying patients with CD33 (positive) Acute Myeloid Leukemia and other CD33 (positive) hematopoietic tumors that may benefit from anti-CD33 therapy when viable cells or frozen sections are not available for immunophenotypic analysis.

Dr. Ahmet Dogan

Further works by Dr. Ahmet Dogan and colleagues at the Mayo Clinic, provided results that agreed with the above findings where immunohistochemistry results correlated well with flow cytometry results showing similar sensitivity as well as defining other potential uses for this product that included cases of minimally differentiated acute myeloid leukemia (AML-M0), and acute monocytic leukemia (AML-M5) where other markers including myeloperoxidase may be negative. Recommendations in this communication7 concluded that clone PWS44 should be used with other myeloid and lymphoid markers, as there are cases of myeloid-antigen-positive acute lymphoblastic leukemias that show genuine expression of the CD33 antigen.

Many scientific and utility-based questions can be asked about any new product, and for CD33 (PWS44), many of the answers are as yet unknown, or are just appearing on the horizon through continuous work at various centers. This information will become increasingly available via peer-reviewed publications.

My sincere thanks to Dr. Elizabeth Hyjek, Dr. Alan Ramsay and Dr. Ahmet Dogan for communication of their results.