Transient Abnormal Myelopoiesis in a Non‐Down Syndrome Infant With Subsequent Evolution to Acute Myeloid Leukemia

To the Editor: Transient abnormal myelopoiesis (TAM) occurs predominantly in neonates with Down syndrome and nearly always involves mutations in the GATA 1 gene 1, 2. These changes impair megakaryocytic maturation and lead to transient accumulation of megakaryoblasts. Approximately one-tenth of affected infants later develop acute myeloid leukemia (AML) 3. Here we describe the case of TAM in a non-Down syndrome infant with subsequent evolution to AML. A male full-term infant, born to a 28-year-old gravida 3 para 1 mother with a history of two miscarriages, was born with “blueberry muffin” lesions. There was no physical evidence of Down syndrome or hydrops fetalis on physical examination. CBC showed WBC 26 × 109/L with 27% blasts, hemoglobin 9 g/dL, and platelets 13 × 109/L. Flow cytometry of the peripheral blood demonstrated 57% myeloid blasts expressing CD36, CD33, CD117, and CD41. The patient had minimally elevated serum bilirubin, normal transaminases, and a normal coagulation profile. Additionally, infectious disease investigations for Cytomegalovirus, Toxoplasma, and Parvovirus B19 were negative in the mother and the child. Imaging studies revealed mild hepatosplenomegaly but was otherwise negative for a cardiac, pleural, or peritoneal effusion. Because of the characteristic findings, TAM associated with Down syndrome was suspected. Peripheral blood karyotype, however, was 46, XY. Bone marrow biopsy at 1 week of age revealed 20% CD117+ myeloid cells, with strong CD41 and CD61 expression on 41% of blasts, identifying a megakaryoblastic (M7) phenotype. Bone marrow karyotype was again 46, XY and fluorescence in situ hybridization (FISH) for trisomy 21 was negative. Chromosomal microarray did not detect any abnormalities. Next-generation sequencing (NGS) for myeloid neoplasms from the peripheral blood identified a KRAS c.179G>A (p.Gly60Asp) variant of potential significance (variant allele frequency VAF 12%) and an IKZF1 c.56C>T (p.Pro19Leu) variant of unknown significance (VAF 50%). No abnormalities of the GATA1 gene were noted. The same IKZF1 abnormality was present in the bone marrow at 52%. The buccal swab sample did not show any trisomy 13 or 21, indicating a lack of mosaicism for chromosome 21. The patient was given platelet transfusions and 1 g/kg of IVIG (intravenous immune globulin) and his platelet counts started rising after 1 week. Repeat flow cytometry at 5 weeks of age showed no blasts and the patient relocated abroad at that time. By 3 months of age, complete blood counts were normal with no evidence of persistent disease. At 6 months of age, the patient developed thrombocytopenia during routine follow-up in his home country. Bone marrow examination there showed 18% blasts and a normal 46, XY karyotype. NGS detected a pathogenic NRAS c.38G>A (p.Gly13Asp) mutation (VAF 12%) with no note of IKZF1. After a mild COVID-19 infection, a repeat bone marrow examination demonstrated 69% myeloid blasts. The patient was treated according to a Children's Oncology Group–based AML0531 protocol, receiving three cycles of induction chemotherapy with cytarabine, daunorubicin, and etoposide, followed by two intensification cycles, although the last and fifth cycle was not fully completed. As the patient was in remission and lacked high-risk features, bone marrow transplantation was deferred. Bone marrow examination after the fourth cycle was in remission with minimal residual disease below 0.02% by flow cytometry. Bone marrow karyotype was 46, XY and repeat molecular testing again identified the above-mentioned IKZF1 variant (VAF 50%) in the peripheral blood, the bone marrow and the buccal mucosa, consistent with a germline change. The previously detected KRAS and NRAS mutations were absent. At 14 months after completion of chemotherapy, the patient remained in remission with normal blood counts and normal physical examination. Subsequent genetic testing in the child and the parents confirmed the heterozygous maternal inheritance of the above-mentioned IKZF1 variant and a benign paternal HAX1 variant. Both parents were healthy with normal blood counts. TAM in non-Down syndrome infants is uncommon. Some cases result from undetected mosaic trisomy 21, but others show normal cytogenetics and lack GATA1 mutations 4. The current patient exhibited the typical biphasic pattern of Down-syndrome-related TAM yet consistently demonstrated a normal karyotype and no GATA1 abnormality. The persistent germline IKZF1 variant may have altered the same developmental pathway affected by GATA1 loss. IKZF1 encodes Ikaros, a transcription factor regulating early hematopoiesis. Experimental data suggest that Ikaros may influence GATA1 expression in fetal megakaryocytes 5. A constitutional IKZF1 alteration could therefore impair megakaryocytic differentiation and predispose to myeloproliferation. Transient KRAS and NRAS mutations were identified during disease evolution. These likely represented secondary RAS-pathway events that promoted blast proliferation. Their disappearance following therapy, while the IKZF1 variant persisted, supports a two-step model similar to that proposed for Down-syndrome-associated TAM 6. The authors declare no conflicts of interest.