Predictive Factors of the Development of Leukemia in Patients with Transient Abnormal Myelopoiesis and Down Syndrome: The Jccg Study JPLSG TAM-10

Introduction: Transient abnormal myelopoiesis (TAM) in neonates with Down syndrome (DS) is characterized by the transient appearance of blast cells that harbor somatic GATA1 gene mutation. Although most patients show spontaneously resolution without therapeutic interventions, approximately 20% of TAM cases result in early deaths within 9 months and 20% of survivors develop acute megakaryoblastic leukemia (AMKL) within 4 years. Although the risk factors associated with early deaths are known, the definite clinical predictive indicators of AMKL onset in patients with TAM remain unclear. Therefore, we analyzed 167 TAM patients with DS enrolled in the TAM-10 prospective observational study conducted by the Japan Pediatric Leukemia/Lymphoma Study Group (JPLSG) to determine the clinical characteristics of TAM and predictive factors of leukemia development.

Patients and Methods: Between May 2011 and February 2014, 167 neonates (89 boys and 78 girls) diagnosed with TAM were prospectively registered in the TAM-10 study. Somatic GATA1 gene mutations were confirmed in 163 (98%) patients using Sanger and/or next-generation sequencing. Minimal residual disease using flow cytometry (FCM-MRD; cut-off level, ≥0.1%) was monitored at 1 (n = 133) and 3 months (n = 104).

Results: Median (range) gestational age, birth body weight, white blood cell (WBC) count, and percentage of blasts at diagnosis were 37 (29-40) weeks, 2,612 (1,066-3714) g, 38.3 (2.4-478.7) × 10⁹ cells/L, and 37% (0.5%-95.5%), respectively. Systemic edema and organ hemorrhage was observed in 31/167 (19%) and 14/167 (8%) patients, respectively; 68/167 (41%) patients received some therapeutic interventions, including low-dose cytarabine (LDCA; n = 52), exchange blood transfusion (n = 20), and systemic steroid therapy (n = 31).

Early death (<9 months of age) occurred in 22/167 (13%) patients. In multivariate analysis, early death was significantly associated with a high WBC count [≥100 × 10⁹ cells/L; HR (95% CI) = 5.329 (2.194-12.945), P < 0.001] and systemic edema [HR (95% CI) = 8.073 (3.130-20.823), P < 0.001]. Subgroup analysis in patients with such high WBC count (n = 36) showed that LDCA therapy significantly improved survival [1-year OS (95% CI) = 78.3% (55.4-90.3; n = 23) vs. 38.5% (14.1-62.8; n = 13); P = 0.009].

Among 145/167 patients without early death, 28 (19%) developed AMKL. FCM-MRD positivity at 1 month [positive, n = 107; negative, n = 26; cumulative incidence ratio (CIR) (95% CI) = 25.2% (17.3-33.9%) vs 3.8% (0.3%-16.8%), P = 0.022] and 3 months (positive, n = 20; negative, n = 84; CIR (95% CI), 45.0% (22.3%-65.4%) vs. 16.0% (9.0%-24.8%), P = 0.002] was significantly associated with leukemia development. However, other clinical covariates, including sex, birth weight, gestational age, WBC count, blast percentage, and GATA1 gene mutational types, could not predict AMKL development.

Considering their severe clinical conditions, 13/31 (42%) patients who received systemic steroid therapy died before AMKL development; interestingly, none of the remaining 18 patients developed AMKL but they showed significantly lower CIR than those who did not receive this therapy [CIR (95% CI), 0% vs. 19.4% (10.9%-29.6%), P = 0.010]. Other therapeutic interventions, including LDCA and exchange blood transfusion, were not associated with AMKL development.

Conclusion: FCM-MRD positivity at 1 month and 3 months might be a useful marker to predict leukemia development in patients with TAM. Although LDCA therapy significantly decreased the rate of early deaths, it did not suppress leukemia development. Interestingly, systemic steroid therapy might suppress leukemia development. These results pave the way to design clinical trials for developing MRD-directed leukemia prevention therapy for patients with TAM.