Background: The lineage assignment in acute leukemias is critical in therapeutic decisions. Immunophenotyping by flow cytometry plays the main role in the lineage assignment; however, few studies have been done to determine the optimal set of markers. In this regard, we tried to find out the optimal first-line set of markers with a minimal compromise in its diagnostic sensitivity.
Materials and methods: We retrospectively analyzed 321 cases of acute leukemias whose diagnoses were based on the EGIL (European Group for Immunological Classification of Acute Leukemia) scores. At our institution, flow cytometic analyses included 15 first-line markers and 4 additional second-line markers as needed, along with immunohistochemical stains. We performed simulational studies for the expected EGIL scores involving every possible combination of markers and analyzing the overall diagnostic sensitivities in each combination.
Results: The cytoplasmic antigens including MPO stain and CD79a stain contributed greatly to the lineage assignment. For a sensitivity over 95%, there needed a combination of MPO stain with other 5 flow markers (CD33, CD13, CD14, CD15 and CD117) for myeloid lineage; CD79a stain with 3 flow markers [CD19, CD10, and CD20 (or TdT)] for B-lymphoid lineage; and 4 flow markers (CD2, CD3, CD5, and CD7) for T-lymphoid lineage.
Conclusions: To maintain diagnostic sensitivities over 95% for each lineage, at least 14 markers (including MPO stain and CD79a stain) were needed; while 16 markers were needed for a sensitivity of 100%. When combined with other important markers for specific aims such as CD45, the minimum number of markers needed for the accurate diagnosis of acute leukemias would be more than about 18 to 20.