Detection of minimal residual disease (MRD) by multiparameter flow cytometry (MFC) has become an important clinical tool in the treatment of patients with acute leukemia. At our institution, we routinely employ Children's Oncology Group (COG) validated B-lymphoblastic leukemia (B-ALL) MFC-MRD analysis for the detection of MRD in pediatric patients after induction therapy for B-ALL. We have extended this COG-validated assay to the adult B-ALL population. This technique relies on the identification of aberrant immunophenotypes expressed by residual bone marrow B-lymphoblasts compared to the normal pattern of maturing hematogones using a 6-color, 3-tube MFC. This assay was initially designed and validated for the pediatric B-ALL population, but the majority of patients with B-ALL at our institution represent adults. Therefore, we characterized the frequency of abnormal cell surface antigen expression of B-ALL MRD with the COG-validated method in the adult B-ALL population and compared our findings with those of our pediatric population.
This study was approved by the Institutional Review Board (IRB) of Washington University School of Medicine in St Louis, MO. A total number of 289 bone marrow specimens were submitted for B-ALL MRD analysis over a time period of 14 months. Antibodies employed in the COG-validated method were CD19, CD20, CD10, CD38, CD9, CD45, CD58, CD13/CD33, CD34, CD71 and CD3. At least 1 million cell events per tube (or as many as possible) were acquired using Beckman Coulter Navios flow cytometers, and Kaluza software was used for analysis. Leukemia-associated immunophenotypic aberrancies were identified, and the proportion of aberrant leukemia cells was calculated as percentage of mononuclear cells. We defined any case with an aberrant immunophenotype that encompassed ≤5% of mononuclear cells as an MRD case, and it was included in subsequent analyses. For each case of B-ALL MRD, we examined side scatter along with the above-mentioned COG-validated markers to asses for aberrancies. Aberrant expression was defined as either over- or under-expression for an antigen when compared to the expression pattern of normal hematogones (CD45, CD19, CD10, CD38, CD58, CD9, CD34) or aberrant positivity/negativity (CD20, CD13/33). Side scatter was considered abnormal if higher or lower than normal hematogones. Fisher's Exact test to compare proportions was used for all statistical analyses.
Of the 289 cases, 64 cases (22%) from a total of 46 patients showed positivity for an aberrant B-lymphoblast population, of which 32 cases (50%) represented MRD ≤5%. Of these 32 cases, 22 were from adults and 10 from children. The MRD levels ranged from 0.010% to 3.8%. Next, we assessed the frequency at which each cell surface antigen was aberrantly expressed and compared results between adult and pediatric patients. For the majority of both adult and pediatric patients with B-ALL MRD, we observed concordant abnormal expression between adult and pediatric cases (listed respectively) of CD10 expression (95% vs. 90%, p=0.534), CD38 expression (68% vs. 70%, p=1.0), CD58 expression (64% vs. 60%, p=1.0), CD9 expression (91% vs. 90%, p=1.0), CD34 expression (86% vs. 70%, p= 0.346), and CD45 expression (55% vs. 80%, p= 0.248). Abnormal expression was less frequent for CD19 (27% vs. 40%, p= 0.683) and CD13/CD33 (18% vs. 20%, p=1.0), as well as side scatter changes (45% vs. 10%, p=0.106). Over-expression was the most common immunophenotypic aberrancy for CD10, CD58, and CD34; under-expression was more common for CD38, CD9, and CD45. Adult and pediatric patients had no significant differences in abnormal expression patterns, with the exception of CD20. In adult patients, CD20 expression was abnormal (negative) in 64% of cases, while only 10% of pediatric patients were negative (p=0.0052, Fisher's Exact test).
These results indicate that the range of aberrant immunophenotypes seen in adult and pediatric B-cell acute lymphoblastic leukemia minimal residual disease are similar in nature and suggest that the COG-validated assay designed for pediatric B-ALL can be successfully applied to the broader adult population. Because this assay allows for the detection of frank (as well as minimal residual) disease burden, we propose that this flow cytometry approach can be harmonized to accommodate monitoring of B-ALL by flow cytometry regardless of the extent of disease burden.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.