This retrospective study enrolled children with AML treated at our institution between March 2012 and March 2023. AML diagnoses were based on morphology, flow cytometry, immunohistochemistry, and genetic testing, following the latest WHO pediatric classification system [11]. The inclusion criteria were as follows: (1) aged < 18 years at diagnosis and (2) AML diagnosis confirmed at our institution. The exclusion criteria were as follows: (1) incomplete clinical data and (2) patients who were untreated after diagnosis. Follow-up was conducted via outpatient visits or phone consultations until January 2024.

A total of 111 patients were included and divided into NUP98-R-positive (n = 10) and NUP98-R-negative (n = 101) groups. Clinical characteristics, treatment responses, and prognoses were compared between the groups. The study was approved by the Ethics Committee of the Union Hospital of Tongji Medical College, Huazhong University of Science and Technology (No. 2024–0753). Written consent for the publication of the details of three cases was obtained from the patients' parents.

Bone marrow (BM) samples from pediatric patients were collected via PAXgene Blood RNA Tubes. Genomic DNA was extracted with the QIAamp DNA Blood Mini Kit. For targeted analysis of structural variants (SVs) and single-nucleotide variants (SNVs), a customized panel of biotinylated oligoprobes (Roche NimbleGen) was designed to capture likely gene breakpoints and hotspot mutations in hematological malignancies identified in prior leukemia studies.

BM evaluations were performed on day 28 of each chemotherapy cycle. CR was defined as the presence of less than 5% blasts in the BM and the regeneration of normal hematopoietic cells. OS was defined as the time from diagnosis to death or the last follow-up. Events included relapse, death, and loss to follow-up due to disease progression, abandonment of treatment, or unstable vital signs. Patients who completed most of the chemotherapy and achieved remission but were subsequently lost to follow-up were not considered events. Event-free survival (EFS) was defined as the time from treatment initiation to the first occurrence of any event or the last follow-up.

Baseline patient characteristics are presented as frequencies and percentages for categorical data and as medians with interquartile ranges for continuous variables. Categorical variables were analyzed with chi-square tests, whereas non-parametric tests were used for continuous variables. OS and EFS were estimated via the Kaplan‒Meier method. Continuous variables included age, BM and white blood cell (WBC) count, whereas categorical variables included sex, age group, French–American–British classification, chromosomal aberrations, molecular genetics, stem cell transplantation status, CR at the end of course 1, and minimal residual disease (MRD) positivity at the end of course 1. P values were calculated via two-tailed tests, with P < 0.05 indicating statistical significance. Analyses were conducted via SPSS (IBM, version 24.0), Prism (version 9), and R (version 4.3.3).

Among the 111 patients, 10 had NUP98 translocations, whereas the remaining 101 formed the reference cohort. Patients in the NUP98-R group had significantly greater WBC counts (134.09 vs. 16.02 × 109/L, P < 0.05) and BM blast percentages (78.25% vs. 57.5%, P < 0.05) than did those in the reference cohort (Table 1). The mutation burden was also greater in the NUP98-R group, with FLT3-ITD mutations found in 80% of patients compared with 14.8% in the reference group (P < 0.05). Similarly, IDH1/2 mutations were more common in the NUP98-R subgroup (50% vs. 11.8%, P < 0.05). Notably, 60% of NUP98-R patients (6/10) had a normal karyotype at diagnosis, which differed from that in the reference cohort. The most common NUP98 fusion subtype was NUP98-NSD1, which was present in 8 of 10 cases (80%). Other fusion partners included KDM5A and PRRX2 (Table 2, Fig. 1). The NUP98-R group also had a male predominance, with eight males (80%) and two females (20%) (Table 1).

Displaying co-occurring mutations and fusion partner genes of NUP98-R pediatric AML patients. a Chord diagram depicting commonly cooccurring mutations in NUP98-translocated pediatric AML patients. b Heatmap depicting the distribution of fusion and mutant genes associated with each patient who tested positive for NUP98-R

All 101 patients in the reference cohort received a standard chemotherapy regimen, including anthracyclines and nucleosides, during induction. Among these patients, 65 (64.3%) achieved CR after the first course (Table 1). In the NUP98-R cohort, two (20%) achieved CR after the first course, seven (70%) did not achieve remission, and one discontinued follow-up before assessment (Tables 1, 2). The CR rate after the first induction course was significantly lower in the NUP98-R group than in the reference cohort (20% vs. 64.3%, P < 0.05) (Table 1). The 3-year EFS rate was significantly greater in the reference cohort (55.3% vs. 30%, P < 0.05). Although the 3-year OS rate was greater in the reference cohort (67.5% vs. 42.9%, P = 0.055), this difference was not statistically significant (Fig. 2).

Survival of pediatric patients with NUP98-R-positive AML. a Kaplan–Meier estimates of OS. 3-year OS rate: 42.9% (NUP98 fusion) vs. 67.5% (No NUP98 fusion), P = 0.055. b Kaplan–Meier estimates of EFS. 3-year EFS rate: 30% (NUP98 fusion) vs. 55.3% (No NUP98 fusion), P < 0.05)

Patient 1 (N2): An 11-year-old boy presented with persistent cervical lymphadenopathy for over 20 days and hematological abnormalities for two days. The laboratory results revealed a WBC count of 76.87 × 109/L, a hemoglobin (Hb) level of 107 g/L, and a platelet (PLT) count of 203 × 109/L. BM examination revealed 75% blasts, which were morphologically classified as M5. Immunophenotyping revealed an aberrant myeloid progenitor phenotype, with strong expression of CD33, CD123, and HLA-DR and weak expression of CD14, CD15, and CD38. Chromosomal analysis revealed a karyotype of 46–47, XY, Inc. [3]/46, XY [17]. Genetic analysis confirmed the presence of the NUP98-NSD1 fusion gene and mutations in FLT3-ITD, IDH1 and MYC. The patient was diagnosed with high-risk AML (M5).

The initial DAH regimen (daunorubicin, cytarabine, and homoharringtonine [HHT]) combined with sorafenib and venetoclax achieved CR on day 28, with an MRD of 1%. At this point, NUP98-NSD1 fusion and IDH1 mutations were still detectable, while the FLT3-ITD mutation was negative. A second course of IAH chemotherapy (idarubicin, cytarabine, and HHT) combined with sorafenib and venetoclax reduced MRD to be less than 10–4, NUP98-NSD1 remained positive, and FLT3-ITD and IDH1 mutations were undetectable. The patient subsequently underwent hematopoietic stem cell transplantation (HSCT) but unfortunately succumbed to a severe infection 1 month posttransplantation.

Patient 2 (N6): A 5-year-old boy presented with a 1-day history of leukocytosis. Laboratory tests revealed a WBC count of 401.6 × 109/L, a Hb of 57 g/L, and a PLT of 52 × 109/L. BM examination revealed 93% myeloid blasts, which were morphologically classified as M1. Immunophenotyping confirmed the myeloid origin of the blasts, with strong expression of CD33, CD123, CD38, CD117, and HLA-DR. Chromosomal analysis revealed a normal karyotype (46, XY). Genetic testing revealed the NUP98-NSD1 fusion gene and mutations in FLT3-ITD and IDH1, confirming high-risk AML (M1).

The patient received induction therapy with the DAH regimen. On day 28, BM analysis revealed 8.5%, an MRD of 6.5%, and persistence of NUP98-NSD1 fusion and IDH1 mutations, while FLT3-ITD was undetectable. By day 46, disease progressed with myeloid blasts increasing to 28% and MRD also at 28%, along with continued positivity for NUP98-NSD1 and IDH1 mutations, indicating resistance to chemotherapy.

Treatment was modified to include gilteritinib, palbociclib, venetoclax, and azacitidine. By day 28 of the revised treatment, the patient achieved CR, with MRD reduced to less than 10–4. NUP98-NSD1 remained positive but was significantly reduced; FLT3-ITD and IDH1 mutations were undetectable. After HSCT, a 1-month follow-up confirmed CR, with an MRD less than 10–4 and no detectable NUP98-NSD1, FLT3-ITD, or IDH1 mutations. At the final follow-up, 243 days post-HSCT, the patient remained in CR, with no evidence of NUP98-NSD1.

Patient 3 (N9): A 5-year-old girl presented with a 1-day history of leukocytosis and bilateral eyelid petechiae. Laboratory tests revealed a WBC count of 482.79 × 109/L, a Hb of 73 g/L, and a PLT of 58 × 109/L. BM analysis revealed 90.5% blast cells. Immunophenotyping confirmed a myeloid blast phenotype with aberrant markers; strong expression of CD33, CD34, CD123, CD13, and HLA-DR; and weak expression of CD7 and CD71. Cytogenetic analysis revealed a normal karyotype (46, XY). The patient was diagnosed with unclassified AML, with the fusion gene NUP98-NSD1 identified, and with mutations in FLT3-ITD, WT1 and CEBPA.

After DAH induction therapy, day 28 evaluations revealed BM blast cells of 35%, an MRD of 25%, and persistent FLT3-ITD, WT1, and CEBPA mutations. A second course of IAH chemotherapy reduced the percentage of BM blasts to 15% and to 3.5%, indicating a limited response. Treatment was modified to the FLAG regimen (fludarabine, cytarabine, and granulocyte colony-stimulating factor). Unfortunately, the patient died of severe infection during this treatment. Posttreatment evaluations for fusion and mutation were not performed due to financial constraints.