Secondary leukemia or unfavorable leukemia?

Secondary leukemia usually describes a myeloid leukemia that follows treatment with chemotherapy or radiation or an exposure to any leukemogen. Smith and colleagues (page 43) describe a remarkable series, spanning 3 decades, of 306 consecutive patients with therapy-related myeloid leukemia. The majority of these patients had deletions in chromosomes 5 or 7 or t(11q23). These data have elegantly emphasized the predominant abnormalities in secondary leukemia and the expected latency period for each cytogenetic subtype. This paper confirms the poor prognosis for patients with secondary leukemia, revealing a dismal long-term outcome that is identical to acute myeloid leukemia (AML) patients with similar unfavorable cytogenetics who have not been previously exposed to leukemogens (Slovak et al, Blood. 2000;96:4075-4083). Other reports also indicate a similarity when known prognostic determinants such as morphology, immunophenotyping, and multidrug resistance (MDR) are considered (Leith et al, Blood. 1997;89:3323-3329).

This raises several issues. First, does the diagnosis of secondary leukemia confer a worse prognosis other than being a surrogate for a particular cytogenetic subtype? Despite a recent analysis from the Medical Research Council (MRC) in Britain suggesting an independent poor prognosis for patients with therapy-related leukemia (Goldstone et al, ASH 44th annual meeting, abstract 322, Blood. 2002;100:88a), there is no convincing evidence that this is a true biologic phenomenon other than a poorer prognosis related to antecedent disease and prior therapy. Furthermore, several reports indicate that patients with more favorable cytogenetics who have secondary leukemias have a prognosis that is similar to patients with de novo AML, with the same balanced translocations. Second, there are no data that secondary leukemias respond better to any combination, during induction or post-remission therapy, than to the standard therapy offered to AML patients with similar cytogenetic and other biologic prognostic factors. We may all be better served by focusing exclusively on the biology of a particular leukemia rather than a further subcategorization based on previous events in a patient's history.