Two tales of infant leukemia

Comment on Nagayama et al, page 4663, and comment on Massey et al, page 4606

Nagayama and colleagues show that infants with B-cell–precursor acute lymphoblastic leukemia and a germ-line MLL gene rearrangement are highly curable with relatively short-term intensive chemotherapy, while Massey and colleagues describe the natural history of transient leukemia in neonates with Down syndrome, suggesting a potential role for low-dose cytarabine in those at increased risk for early death or conversion to frank leukemia.

Infant leukemias are fascinating and challenging to oncologists and molecular biologists alike because of their marked heterogeneity in presentation and outcome, which is almost certainly linked to differences in genetic abnormalities, host susceptibilities, and the target cells of malignant transformation. MLL rearrangements, for example, occur in both acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML), but are associated with a poor prognosis only in ALL.¹ 

Two articles addressing the important topic of infant leukemias appear in this issue of Blood. Nagayama and coworkers showed that infant B-cell–precursor ALL with a germ-line MLL gene is highly curable with chemotherapy alone. Of the 22 infants studied, 21 remained in first complete remission for 3.5 to 8.8 years; the sole relapse occurred in the only patient with T-cell ALL. Notably, the total duration of treatment did not exceed 20 months. Like virtually all infants with or without an MLL rearrangement, the patients in this study received an intensive regimen of chemotherapy that included 3 high doses of methotrexate (3 g/m²), 19 doses of triple intrathecal therapy, cyclophosphamide at a cumulative dose of 3.9 g/m², anthracyclines at 250 mg/m², and 16 doses of etoposide. The regimen also specified 15 doses of asparaginase administered at 10 000 U/m², which would be considered quite intensive, given the use of less-potent product at the same dose in the United States.²  With the exception of 3 seizures during remission induction, this treatment was relatively well tolerated, and the patients grew normally in weight and height over 3.5 to 9 years of follow-up. However, because of a potential sex bias (20 of the 22 patients were boys) and the relatively small patient sample size, this excellent result will require confirmation, not only to establish the therapeutic efficacy of the regimen, but also to gauge the extent of late effects, such as second cancers and neurocognitive deficits. It is also possible that infants with germ-line MLL rearrangement and a good early treatment response could be treated less intensively without any substantial loss of antileukemic potency.

A prospective Children's Oncology Group (COG) study by Massey and colleagues of 47 patients with Down syndrome with a diagnosis of transient leukemia (more commonly known as transient myeloproliferative disorder) revealed early deaths in 8 patients due to liver failure and disseminated intravascular coagulation, and the subsequent development of acute megakaryoblastic leukemia in 8 patients and B-cell–precursor ALL in another patient at a mean of 20 months from diagnosis. As one might expect from past reports,³  all but one of these patients, a child with acute megakaryoblastic leukemia, responded well to chemotherapy and remained in continuous complete remission. Early death was associated with a high presenting leukocyte count, increased levels of bilirubin and hepatic enzymes, and failure to regain normal blood counts. By contrast, the presenting leukocyte count, blast cell percentage, and time to clearance of blasts or normalization of blood counts did not predict the emergence of frank leukemia. The presence of karyotypic abnormalities in addition to trisomy 21 was only weakly correlated with leukemia development; that is, 4 of 7 such patients, compared with 5 of 35 lacking additional karyotypic abnormalities, subsequently developed leukemia. Interestingly, patients with overt leukemia were less likely than the unaffected group to be symptomatic at diagnosis of transient leukemia (one of 9 patients vs 26 of 39 patients).

The COG study was not intended to evaluate therapy, so patients were treated according to the discretion of their physicians. Only one child received standard AML therapy, and 2 patients got low doses of cytarabine shortly after diagnosis. All three of these patients remained well. Two additional patients received low-dose cytarabine for progressive symptoms and died 1 or 2 days later. On the strength of those observations and limited experience by others,⁴  the authors recommend early use of low-dose cytarabine as prophylactic chemotherapy in patients with progressive organomegaly and evidence of liver dysfunction or in boys with cytogenetic abnormalities other than trisomy 21. However, I would argue for a prospective trial of low-dose cytarabine in all Down syndrome patients with transient leukemia because of the lack of factors that can be used with certainty to predict leukemia development or early death. Indeed, as mentioned, asymptomatic patients were more likely than symptomatic patients to develop leukemia. While 6 of 7 boys with additional cytogenetic abnormalities had a subsequent adverse event (4 leukemias and 2 early deaths), 3 of the remaining 22 boys also converted to overt leukemia, and 4 died early. Finally, low-dose cytarabine appears to be an effective and relatively nontoxic agent in infants with Down syndrome and transient leukemia.⁴ 

It should be emphasized that the true incidence of transient leukemia is currently unknown, since blood counts are not routinely performed in Down syndrome patients. Conceivably, some of these leukemias resolve permanently without being diagnosed in the first place. Thus, an immediate challenge is to learn the exact prevalence and history of this disorder by prospectively studying all neonates with Down syndrome. ▪