In this issue of Blood, Rubnitz and colleagues from St Jude Children's Research Hospital provide important information concerning the immunophenotypic, cytogenetic, gene expression, and response to therapy characteristics of 35 patients with so-called mixed lineage leukemia.
Identification of mixed lineage leukemia used to rest on the European Group for the Immunological Characterization of Leukemia (EGIL) criteria, which provided a lymphoid and a myeloid score based on immunophenotype; a score of 2 or more for both lymphoid and myeloid differentiation markers defined a mixed lineage leukemia.1 The definition of myeloid immunophenotype relied primarily on expression of CD13 and CD33. Recently, the World Health Organization (WHO) has revised the diagnostic criteria for mixed lineage leukemia.2 Criteria for myeloid differentiation include myeloperoxidase activity or monocytic differentiation (at least 2 of NSE, CD11c, CD14, or lysozyme). CD13 and CD33 are no longer considered relevant. T-cell criteria include cytoplasmic CD3—epsilon chain or surface CD3. B-lineage criteria include strong CD19 expression and coexpression of one of the following additional markers—CD79a, cytoplasmic CD22, or CD10. If CD19 expression is weak, coexpression of 2 additional markers is required.
In the current study, gene profiling was performed by the authors in 13 patients and revealed that only 5 cases clustered with known acute myeloid leukemia (AML) cases while the remaining 8 formed a unique cluster that had not previously been recognized.3 It is interesting that none of the cases resembled acute lymphoblastic leukemia (ALL), at least in terms of gene expression profiles.
Twenty-three patients initially received AML-type induction while 12 received ALL-type induction based on treatment protocols available over the two decade time course of patient presentation. Twelve of the 23 patients treated with AML induction and 10 of 12 treated with ALL induction achieved remission. Of note, 8 of the 10 patients who failed AML induction achieved remission after treatment with vincristine, prednisone, and L-asparaginase. Overall outcomes were similar for both the T/myeloid and the B myeloid subgroup (5-year survival 36% vs 54%, P = .67). Outcomes were clearly inferior to those seen for a contemporaneous sample of patients treated for standard ALL.
Now that the WHO criteria have been revised, we should be able to gather a clearer picture of this rare subgroup (3%-5%) of acute leukemia. At St Jude, patients with mixed lineage leukemia are now eligible for enrollment on the currently open AML trial. If patients fail initial therapy, they are then switched to lymphoid type therapy. I personally instead favor a combined lymphoid and myeloid induction for mixed lineage leukemia patients. One could consider a mixed lymphoid and myeloid induction with vincristine, prednisone, L-asparaginase with high-dose cytosine arabinoside, and etoposide on days 1 and 2, but this combination has not been formally studied.
Patients with mixed lineage leukemia by definition have unique leukemia-associated immunophenotypes that allow for successful detection of minimal residual disease. Patients with high-level minimal residual disease at the end of induction should be considered for allogeneic bone marrow transplantation in first remission. Clofarabine and cytoxan could be considered as a reinduction regimen for those patients who fail a combined AML/ALL induction. For patients who achieve remission with initial induction therapy, one could consider either augmented Berlin-Frankfurt-Munster (BFM)–type therapy or the BFM very high-risk therapy as continuation. For patients with morphologic remission after initial induction associated with intermediate levels of minimal residual disease (0.1%-1%), end consolidation minimal residual disease determinations could be used to determine subsequent therapy, either continuation chemotherapy or allogeneic stem cell transplantation.
No single pediatric leukemia study group will have enough patients with mixed lineage leukemia to do a study. Because many of the patients with mixed lineage leukemia are not eligible for either current ALL or AML trials, it would be very difficult to perform a meta-analysis concerning such patients. A consortium of national and international pediatric study groups, such as the Ponte Di Legno consortium, should consider initiating a prospective trial for such patients.
Conflict-of-interest disclosure: The author declares no competing financial interests. ■
