Answers are grounded in this paper and cite their sources.
Case report reveals peripheral blast clearance after G-CSF withdrawal in severe congenital neutropenia, indicating drug cessation may stabilize patients prior to stem cell transplantation.
To the Editor: Severe congenital neutropenia (SCN) is a rare hematologic disorder characterized by severe neutropenia (absolute neutrophil count, ANC, < 500/µL) and severe bacterial infections since early infancy. The availability of granulocyte colony-stimulating factor (G-CSF) has dramatically improved the survival and quality of life of patients with SCN due to an increase in granulocyte counts and a significant reduction in infections [1]. However, patients with SCN have an intrinsic predisposition to develop myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) [2]. G-CSF therapy may contribute to leukemogenesis, especially in patients with acquired G-CSF receptor (CSF3 R) mutations [3], although its role remains poorly understood. We herein report the case of a female patient diagnosed with SCN at the age of three, who harbored an ELANE mutation (c.1851 G>A, p.Val16Met). Since diagnosis, she received G-CSF with doses of 1 µg/kg/day subcutaneously, with adequate neutrophil response (median values > 1.5 × 103/µL). Serial bone marrow aspirates showed partial myeloid maturational arrest without cytogenetic abnormalities or somatic RUNX1 mutations. However, during follow-up, fluctuating variant allele frequencies of multiple CSF3 R clones (c.2384 C>T, c.2390 C>T, c.2215 C>T) were identified by next-generation amplicon sequencing of Exon 7 in bone marrow polymorphonuclear cells. At the age of 16, she developed pharyngitis and oral stomatitis associated with severe neutropenia (30 cells/µL). The bone marrow aspirate ruled out the presence of MDS and AML (Figure 1A), and the G-CSF dose was increased to 4.45 µg/kg/day, obtaining a transient ANC improvement. Four months later, she developed progressive neutropenia associated with monocytosis (10.2 × 103/µL) despite G-CSF therapy. The morphological bone marrow evaluation showed a complete maturational arrest at the promyelocyte/myelocyte stage with monocytic hyperplasia; histopathological examination and immunophenotyping ruled out evident signs of progression to MDS and AML. Karyotype analysis showed a pseudodiploid clone 46, XX with 7q partial deletion (7q22-7qter), associated, in one of the metaphases, with a balanced translocation t(4p;21q) involving RUNX1, confirmed by fluorescent in-situ hybridization in 2/50 metaphases and 8/200 interphase nuclei. A rise of a previously detectable CSF3 R mutated clone (c.2215 C>T) up to 16% was detected with amplicon sequencing. Furthermore, a reverse transcription polymerase chain reaction-based panel for screening common molecular abnormalities associated with malignant transformation excluded chronic and acute features of myeloproliferative disorder. In a following cytogenetic analysis, the presence of a partial deletion of Chromosome 7 was confirmed on 9/20 metaphases, but the rearrangement involving RUNX1 was no longer detectable. Three weeks later, the patient developed splenomegaly and cervical lymphadenopathy. A complete blood count showed trilineage cytopenia, with 6% myeloblasts at the peripheral blood smear. BMA showed 50% myeloperoxidase-positive blasts SSClow, CD45+dim, CD34±, CD117+, HLA-DR+dim, CD13+dim, CD33+/++, and CD15+, confirming the progression to acute myeloproliferative disease (Figure 1B,C). Over the following days, the peripheral blast count rapidly increased up to 45% (Figure 1D), and G-CSF therapy was then promptly withdrawn. After G-CSF discontinuation, the bone marrow myeloblast rate decreased to 20% in 3 weeks, without reaching morphologic or cytogenetic remission. Conversely, peripheral blasts rapidly decreased to 15% shortly after and were no longer detectable two weeks later. Considering the clinical stability of the patient and the peripheral blood blast clearance, the patient was promptly referred to the transplant unit to receive T–cell-depleted haploidentical allogeneic hematopoietic stem cell transplantation (HSCT) from her brother. The conditioning regimen included thiotepa, busulfan, and fludarabine. Post-transplant complications included graft-versus-host disease with gastrointestinal tract and skin involvement, responsive to corticosteroids, and polyomavirus-induced hemorrhagic cystitis, responsive to cidofovir. Four years after HCST, the patient remains clinically stable, with a normal neutrophil count and full donor chimerism. The risk of malignant myeloid transformation in SCN patients varies according to the genetic cause of the neutropenia [2] and is estimated to be around 13% in patients with ELANE mutations [4]. Leukemogenesis in SCN is widely considered a multi-step process, in which several acquired mutations occur sequentially [2]. It has been proposed that cell clones harboring CSF3 R mutations have a growth advantage and that the acquisition of cooperating mutations [2, 5]–such as RUNX1–together with the activation of a chronic pro-inflammatory signaling, contribute synergistically to disease evolution [6, 7]. G-CSF exposure is likely to mainly contribute to leukemic progression in patients who require higher doses of G-CSF to achieve an adequate neutrophil response [8, 9]. However, it is not clear if it directly contributes to the acquisition of Chromosome 7 alterations [10] and growth advantage on the myeloid cells, especially after the acquisition of CSF3 R mutations, or if it just improves survival, revealing the leukemic propensity of SCN [3]. Notably, studies using patient-derived induced pluripotent stem cells (iPSCs) models shown that G-CSF exposure in CSF3 R-mutant hematopoietic progenitor cells from ELANE-mutated SCN patients exacerbates the inflammatory state initiated by the germline mutation, thereby contributing to leukemic transformation [11]. However, AML has also been reported in untreated SCN patients [12], highlighting the biological complexity of leukemogenesis in SCN. Jeha et al. described an SCN patient who developed AML associated with a CSF3 R mutation during prolonged G-CSF therapy and achieved complete morphologic and cytogenetic remission following G-CSF discontinuation, although a CSF3 R mutation was still detectable during the remission period [13]. Conversely, our patient developed genomic instability, with increasing CSF3 R mutant clones and chromosomal alterations involving RUNX1 and Chromosome 7, during long-term treatment with G-CSF for SCN. These alterations were followed by loss of G-CSF efficacy and progression to MDS/AML. G-CSF discontinuation induced a clearance of the peripheral blastosis, but no morphologic or cytogenetic bone marrow remission, requiring prompt referral for HSCT. The loss of G-CSF response might represent an early clinical marker of incoming leukemogenesis, biologically reflected by the increase of the CSF3 R clone and chromosomal alterations. Therefore, G-CSF therapy should be promptly discontinued in case of progression to myeloproliferative disease in patients with CSF3 R mutations, in order to reduce the proinflammatory response and the selective advantage for the mutant clones. However, discontinuation of G-CSF therapy alone is not always sufficient to achieve complete bone marrow remission. Regular monitoring of patients with SCN is therefore crucial to identify those with expanding CSF3 R-mutated clones and concurrent chromosomal alterations, who should be promptly referred to a transplant unit. A.R. and I.T. drafted the paper, and F.P. and M.V. substantially revised it. M.T., F.F., V.T., A.T., and M.V. followed the patient. L.G. performed laboratory analysis. All authors read and approved the final version of the manuscript. The authors have nothing to report. The authors have nothing to report. This manuscript adheres to the ethical principles outlined in the Declaration of Helsinki. Informed consent was obtained for the publication of the clinical data and diagnostic images included in the manuscript. The authors declare no conflicts of interest. Data sharing is not applicable to this article, as no datasets were generated or analyzed during the current study.
No takes yet. Share an insight, caveat, or question.
Roveta et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: