Basophils and plasmacytoid dendritic cells are potential sources for error in flow cytometric monitoring of patients receiving anti‐CD22 therapies. AKA not all anti‐CD22 antibodies are created equal

We would like to share our laboratory's experience with flow cytometric monitoring of patients receiving anti-CD22 targeted therapies, to describe a potential pitfall and propose a simple solution for this type of analysis. This experience also highlights the importance of strengthening interactions between treating physicians and laboratory physicians in the rapidly evolving era of targeted and/or personalized medicine. Several pharmacological agents based on anti-CD22 antibodies are currently available in the clinic at various stages of development. These specifically target B lymphocytes that uniformly express high levels of CD22 molecule on their surface, including both benign conditions associated with aberrant B-cell function (e.g., systemic lupus erythematosus) and malignant B-cell non-Hodgkin lymphomas (B-NHL). Flow cytometry may be used to assess the effect of anti-CD22 therapeutic agents by following the level of CD22-positive B lymphocytes in the blood and/or bone marrow. However, we have observed potentially misleading flow cytometric results while monitoring “CD22-positive cell counts” ordered on patients receiving anti-CD22 therapies. Specifically, we have observed persistence of low levels of CD22-positive events in patients treated for B-NHL with inotuzumab ozogamicin (IO)—a targeted cytotoxic immunoconjugate composed of a humanized IgG4 anti-CD22 antibody covalently linked to N-acetyl-gamma-calicheamicin dimethyl hydrazide—who otherwise achieved a complete clinical and morphological remission, suggesting that there may be minimal residual disease. These events are not targeted neoplastic B lymphocytes, rather basophils that are also known to express CD22. The basophils in these patients were unaffected by the therapy, even in patients who achieved complete suppression of CD22-positive lymphocytes (Fig. 1). Thus, reporting only a “total CD22-postive cell count” as requested by the treating physicians would have led to a false-positive minimal residual disease results and possibly inappropriate treatment decisions. CD22 expression, as determined by reactivity to the S-HCL-1 antibody, has also been reported recently on normal plasmacytoid dendritic cells (PDCs) in the blood [1], leading to another source of false positive flow cytometry results. Example of serial CD22-positive cell counts in a patient treated with inotuzumab ozogamicin. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] Closer inspection of the biochemistry of the CD22 molecule and specific antibodies used in therapeutics and diagnostics explains the reactivity of IO with B lymphocytes but not with basophils. The antibody used in IO is humanized antibody derived from clone g5/44, which binds to the N-terminus of the CD22 molecule. In contrast, the most commonly used diagnostic anti-CD22 antibody, and the one used in our laboratory, is derived from the S-HCL-1 clone, a murine antibody that binds to the ligand binding interface between IgLD 1 and IgLD 2. Previous work has shown that several of the commercially available anti-CD22 monoclonal antibodies show differential affinities for the CD22 molecules on normal B lymphocytes versus basophils [2, 3]. Specifically, S-HCL-1-derived antibody binds well to CD22 on the surface of B lymphocytes, basophils, and PDCs; whereas g5/44-derived antibody binds CD22 on B lymphocytes but not on basophils or PDCs. The mechanism of this differential binding of anti-CD22 antibodies on B lymphocytes and basophils appears to be due to three-dimensional conformational differences in the CD22 molecule rather than alternative splicing, as the mRNA from B lymphocytes and basophils is identical. With this background, it is quite clear why IO effectively eliminates B lymphocytes but not basophils from the peripheral blood. Based on these observations and on the protocols for analyzing basophils suggested by Han et al. [4], Toba et al. [5], and Han [6], our laboratory has adopted a simple, cost efficient and effective four-color flow cytometric assay that readily separates basophils and PDCs from the targeted CD22-positive B lymphocytes for monitoring patients receiving anti-CD22 antibody therapies. We use the following combinations of antibodies in a single tube: CD45 PerCP/CD13+CD33 PE/CD22 FITC/CD19 APC (all from BD Biosciences, San Jose, CA). With this simple methodology, standard gating using CD45 versus side-scatter will provide fairly good discrimination between basophils (R1) and the mature lymphocytes (R2). Confirmatory analysis will separate basophils as CD22-positive, CD13+CD33-positive, and CD19-negative, and PDCs as CD22-positive, CD13+CD33-positive/negative [7], and CD19-negative; whereas B cells will be CD22-positive, CD13+CD33-negative, and CD19-positive (Fig. 2). Discrimination of CD22-positive basophils and PDCs from mature lymphocytes by flow cytometric analysis before and after treatment with anti-CD22 therapy. Basophils and PDCs (R1), which express CD13+CD33 and CD22, are contrasted with B lymphocytes (R2) expressing CD19 and CD22. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] As this example indicates, in the age of rapidly blossoming targeted therapies, disease monitoring in the clinical laboratory will need to expand to provide accurate results to guide therapies. The pivotal role of a professional laboratory physician to develop and interpret increasingly complex tests must expand to meet future needs of patients and their treating physicians. Malcom Acon-Laws*, Michael G. Bayerl*, W. Christopher Ehman , Jozef Malysz*, Cinda Boyer , * Penn State Milton S. Hershey Medical Center, Pathology and Laboratory Medicine, Penn State Hershey Cancer Institute, Hematology/Oncology, Penn State Milton S. Hershey Medical Center, Special Hematology Laboratory.