The detection of minimal residual disease status in chronic lymphocytic leukemia

Abstract

In this issue of Clinical Cytometry, Durrieu et al. on behalf of GEIL-GOELAMS present evidence that circulating levels of CD19+ CD5+ chronic lymphocytic leukemia (CLL)-like cells can be used to define a threshold for the detection of minimal residual disease (MRD) in CLL (this issue: page 346). This study was performed in 16 different hematology laboratories or services, and the data is analyzed centrally. The panel consisted of five tubes of four reagents each with CD19 and CD5 being present in each tube. They determined that the level of two CLL-like immunophenotypes CD19+/CD5+/CD43int/CD79blo and CD19+/ CD5+/CD81lo/CD22lo were less than 4 × 10−4 per WBC in unwashed, lyzed normal whole blood. Levels greater than this found in post treatment CLL samples were taken to signify the presence of MRD. The author further state that they are unable to determine monoclonality using light chain restriction due to their polyclonal nature in normal blood donors and the presence of cytophilic antibodies being present in the unwashed, lyzed method. And that bright CD5+ B cells are often lambda positive, and polyclonal. I have several preferences (biases) in this matter, and they include the use of CD20 rather than CD19 or preferably both CD20 and CD19 in combination to define B cells and the clone, demonstration of monoclonality or monotypia, and that reports be based on the absolute cell-count (WBC or absolute lymphocyte count ALC) and not just the percentage < or > 10−4. If the WBC is used, it should be used in conjunction with the ALC and even a B cell absolute count. I sympathize with the problem of cytophilic antibodies but in monoclonal B cell lymphocytosis studies, values of 5–50 light chain restricted B cells per microliter can be routinely detected (1). I was also surprised to see that the lower range of 1–3% CD5+ B cells seems to be the norm, whereas values of 10–25% seem to be more the rule. The European Research Initiative on CLL (ERIC) has published a standardized four-color approach to MRD detection in CLL (2). Both the authors of that study and this study find a WBC threshold of 10−4 below which is considered MRD negativity. The ERIC study allows for the primary role of kappa lambda tube for historical purposes but favors the combination of CD20/CD38/CD19/CD5 over the combinations of CD43/CD79b and CD81/CD22. They further point out the need for the evaluation of gate contamination, number of total events collected (500,000), and operator training. Certain precautions are required for differentiating between blood and marrow involvement in the setting of antibody therapies. However, due to the correlation of MRD between blood and marrow, blood should be the first site tested for MRD. The complete absence of B cells (B cell aplasia) might also be taken as a clinical measure of MRD. If ascertainment of marrow involvement in this setting were considered important, a bone marrow study would be indicated. Conversely, the timing of the blood MRD analysis could be postponed. The German CLL Study Group have also standardized three four tube panels of four-color reagents using a larger volume of blood, and where appropriate 2 million leucocytes were collected (3). This paper also shows that the overall ability of MRD flow is to detect and quantify residual CLL cells and is not affected by the addition of rituximab to FC chemotherapy (3). In a 2005 editorial, Montserrat (4) called for standardization and revision of 1996 NCI response guidelines to include MRD negative status as a goal. In 2008, Hallek et al. (5) revised the NCI guidelines to include not only a definition of MRD but also encouraged that the MRD detection become part of future CLL clinical trials because MRD negative status has prognostic impact. The GCLLSC, ERIC, and GEIL studies have all demonstrated that multicolor flow cytometry can meet the needs of a robust MRD assay. All three groups used a four-color, five tube panel and a sequential gating strategy. Newer clinical flow cytometers now allow for the use of 6–8 colors. The Salamanca group has developed an automated method for doing MRD detection by merging files on the same patient with the addition normal donor virtual 1:1 dilution as a control (6). Knowledge of the original CLL immunophenotype would be useful. One could envision an integration of the GCLLSC, ERIC, and GEIL methods combining CD19/CD20/CD5/CD38/CD79b/ CD22/CD41/CD81 and CD19/CD20/CD5/CD45/CD3/CD14/kappa/lambda. Del Poeta et al. (7) concluded that the addition of rituximab consolidation and maintenance therapy prolonged response duration significantly in patients with MRD + CLL. MRD detection is indicated in the investigation, care, and treatment of the CLL patient. Clinical trials will be required to answer the question of overall survival in CLL. Regardless, CLL remains a model disease for the appropriate treatment of the elderly with progressive disease.