This multicenter retrospective study was approved by the institutional review boards (No. 2025R022-E01) following discussion in a multicenter advisory panel across four pediatric medical centers: Shanghai Children’s Hospital, Shanghai Children’s Medical Center, Anhui Children’s Hospital, and Shandong Provincial Hospital. Patients ≤18 years old with R/R B-ALL and MRD positivity at any time who received blinatumomab therapy were enrolled. Patients with chemotherapy intolerance or severe infection who received blinatumomab as bridging therapy were enrolled as the control group. The exclusion criteria were as follows: (i) patients with severe infection and cardiac, liver, or kidney insufficiency who had an expected survival time of less than 3 months; and (ii) those who received blinatumomab for fewer than 7 days. Patient enrollment lasted from September 2021 to June 2024, with follow-up through March 2025. A total of 105 patients were enrolled.

Blinatumomab was administered via a stepwise dose-escalation protocol during the initial cycle: 5 μg/m²/day as continuous intravenous infusion on days 2–7, followed by escalation to 15 μg/m²/day for a total cycle duration of 14–28 days. The infusion duration depended on family financial conditions and physician discretion. In some cases, BM aspiration was performed on day 15, and treatment was discontinued upon achieving BM remission. Each treatment cycle was separated by a 14-day treatment break.

Dexamethasone prophylaxis (5 mg/m²/day for 1 day) was routinely administered. Subsequent treatment cycles commenced directly at 15 μg/m²/day. Adverse events (AEs) were managed according to the manufacturer’s instructions. BM assessment was performed upon completion of the infusion cycle. Intrathecal injections of methotrexate, cytarabine, and dexamethasone were administered before, during, or after blinatumomab cycles. Upon achieving BM remission, patients proceeded to hematopoietic stem cell transplantation (HSCT), continued the original protocol, or received alternative treatment. Non-responders were transitioned to salvage protocols, chimeric antigen receptor T-cell (CAR-T) therapy, or palliative HSCT, as clinically indicated.

Some patients received reinduction therapy to reduce tumor burden. The reinduction therapy followed the initial induction regimen: dexamethasone 6 mg/m² on days 1–4; vincristine 1.5 mg/m² on days 5, 12, 19, and 26; prednisone 45 mg/m² on days 5–28; daunorubicin 25 mg/m² on days 5 and 12; and peg-asparaginase 2,000 U/m² on days 6 and 26.

For patients with Philadelphia chromosome–positive (Ph+) disease, a tyrosine kinase inhibitor (TKI) was added, with dasatinib 80 mg/m² preferred. Bridging chemotherapy prior to blinatumomab included induction chemotherapy or continued consolidation chemotherapy consisting of cyclophosphamide 1,000 mg/m² on day 1, cytarabine 50 mg/m² on days 1–7, and mercaptopurine 40 mg/m² on days 1–7.

Patients were divided into three groups: CR-MRD^neg, CR-MRD^pos, and R/R. Response was categorized as either cytological CR or MRD CR. Cytological CR was defined as <5% BM blasts in patients with R/R status. MRD CR, detected by flow cytometry (FCM), was defined as a reduction in MRD to <0.01% or maintenance of MRD negativity in patients with CR-MRD^pos. No response was defined as partial remission (PR) or no remission (NR) in R/R patients, and persistent MRD ≥0.01% in patients with CR-MRD^pos.

Poor cytogenetics were defined as KMT2Ar, BCR-ABL1, and TCF3-HLF, according to the Chinese Children’s Cancer Group ALL (CCCG-ALL) 2015 protocol. Event-free survival (EFS) was defined as the time from diagnosis to relapse, death, secondary cancer, or last contact for those who were event-free. Overall survival (OS) was defined as the time from diagnosis to death from any cause or last contact if alive.

Serum concentrations of target cytokines [IL-1β, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12p70, IL-17, interferon (IFN)-γ, tumor necrosis factor (TNF)-α, and IFN-α] were measured using a multiplex microsphere-based flow immunofluorescence assay (12-cytokine kits, Raisecare, China) according to the manufacturer’s instructions. Cytokines were assessed before blinatumomab infusion, at the onset of cytokine release syndrome (CRS) or immune effector cell–associated neurotoxicity syndrome (ICANS), and at the end of blinatumomab treatment.

Basic lymphocyte subpopulations were analyzed using a FACSCalibur flow cytometer (BD Biosciences) and reported as both percentages and absolute counts. FCM with CellQuest software (BD Biosciences) was used for analysis of lymphocyte subsets (CD3/CD45/CD4/CD8/CD16CD56/CD19, BD Biosciences), including T cells (CD3+CD45+), cytotoxic T cells (CD3+CD8+CD45+), helper T cells (CD3+CD4+CD45+), NK cells (CD16+CD56+CD3−CD45+), and B cells (CD19+CD45+). A total of 15,000 lymphocytes were acquired for analysis. Data were collected at baseline and on days 14, 21, and 28 (end of treatment). T-cell activation magnitude was defined as the difference between post-blinatumomab and baseline (pre-treatment) measurements.

Quantitative data with a Gaussian distribution were presented as mean ± standard deviation (SD) and compared using the t-test. Non-normally distributed data were presented as medians with full ranges. Comparisons between two groups were performed using the Mann–Whitney U test or Wilcoxon matched-pairs test. Comparisons involving two or more factors were conducted using one-way or two-way analysis of variance (ANOVA). Categorical variables were presented as percentages and compared using Fisher’s exact test or the chi-square (χ²) test. Patients lost to follow-up were censored at the last date they were known to be alive. OS and EFS were estimated by the Kaplan–Meier method, and curves were compared using the log-rank test. The Cox proportional hazards model was used for univariate and multivariate analyses. All statistical analyses were performed using GraphPad Prism version 9 (GraphPad Software Inc., La Jolla, CA, USA) and SPSS version 19.0 (SPSS Inc., Chicago, IL, USA). All tests were two-sided, and P < 0.05 was considered statistically significant.

A total of 105 patients with B-ALL received blinatumomab across 125 cycles, following standardized clinical practice. The median age was 72 months (range, 5–210 months). Sixty-four patients (61%) were male and 41 (39%) were female.

Thirty cases were diagnosed with R/R ALL, including 23 with relapse and seven with induction failure. Eight patients received blinatumomab directly as reinduction therapy, and 22 patients received burden-reduction chemotherapy prior to blinatumomab. Among these, three patients failed to reduce tumor burden, with blast levels remaining above 5%.

Ten patients achieved CR-MRD^pos, and nine patients achieved CR-MRD^neg. An additional 21 patients had CR-MRD^pos following induction or developed MRD positivity during treatment. Fifty-four patients received blinatumomab due to chemotherapy delay caused by chemotherapy intolerance or severe AEs, among whom nine patients had CR-MRD^pos.

At the initial cycle of blinatumomab, 40 patients had CR-MRD^pos, 11 patients had R/R status, and 54 patients had MRD negativity. Among the latter, four cases had BM blasts ranging from 5%–9.5%, while seven had blasts ≥20%. For these 51 patients, the median BM blast percentage was 2% and the median MRD percentage was 0.52% ( Table 1 ). In patients achieving MRD negativity, a total of 74 cycles were administered. The demographic and clinical characteristics of the R/R, CR-MRD^pos, and CR-MRD^neg groups are summarized in Table 1 .

The CR rate of R/R patients was 81.8% (9/11), while the MRD-negative CR rate was 72.7% (8/11). MRD CR was achieved in 33 of 40 cases (82.5%) with CR-MRD^pos. The overall CR response rate was 82.4%.

Among the R/R and CR-MRD^pos patients, 31 received a 14-day infusion and 20 received a 3–4-week infusion. The response rate was 83.9% for the 2-week regimen and 80% for the 3–4-week regimen.

Of the 74 MRD-negative cycles, 73 patients remained MRD negative until the follow-up day. Only one patient experienced central nervous system (CNS) relapse 14 months after blinatumomab.

In nine cases with NR after the first blinatumomab cycle, three received a second cycle of blinatumomab, two patients underwent HSCT (achieving CR), three patients with persistent MRD continued chemotherapy (achieving CR), and one patient discontinued treatment and subsequently died. Among the three patients receiving a second cycle of blinatumomab, one achieved CR and bridged to HSCT. The remaining two patients failed to achieve remission and continued treatment with CAR-T. All three patients remained alive.

When patients with R/R and CR-MRD^pos status were compared with those with CR-MRD^neg, notable differences in the immune cell repertoire were observed. The data exhibited a non-Gaussian distribution and were expressed as median (range), with comparisons performed using the Mann–Whitney U test.

The increase in CD3+ T cells was significantly greater in the R/R + CR-MRDpos group than in the CR-MRD^neg group [0.57 (−1.08 to 6.07) vs. 0.12 (-1.83 to 3.34)×10⁹/L; P = 0.047; Figure 2A ]. CD4+ cells expanded more significantly in the R/R + CR-MRDpos group than in the CR-MRD^neg group [0.24 (−0.55 to 1.81) vs. 0.03 (−0.73 to 1.19) × 10^9/L; P = 0.039; Figure 2B ]. In contrast, no significant difference in CD8+ T-cell expansion was observed between the two cohorts [0.16 (−0.48 to 5.21) vs. 0.11 (−0.58 to 1.33) × 10^9/L; P = 0.174; Figure 2C ].

Enhanced B-cell eradication was observed in the R/R + CR-MRD^pos group compared with the CR-MRD^neg group [−0.03 (−1.12 to 0.00) vs. −0.004 (−0.35 to 0.08)×10⁹/L, P = 0.017; Figure 2G ]. No significant intergroup differences were detected in NK cells, total lymphocytes, CD4+/CD8+ ratio, neutrophils, or platelets ( Figures 2D–F, H, I ; Supplementary Table S2 ).

At the onset of CRS, significant elevations in serum levels of IL-2, IL-5, IL-10, and IFN-γ were observed. The data were expressed as median (range), and comparisons were performed using the Wilcoxon matched-pairs test.

IL-10 increased from 2.4 (0.3–8.3) to 2.4 (0.3–34.3) pg/mL (P < 0.0001; Figure 3A ). IL-5 rose from 2.7 (0.3–9.6) to 2.7 (0.5–205.6) pg/mL (P = 0.0006; Figure 3B ). IFN-γ increased from 4.6 (1.0–84.5) to 12.3 (1.3–328.3) pg/mL (P = 0.003; Figure 3C ). IL-2 rose from 2.4 (0.5–17.3) to 2.4 (0.7–22.8) pg/mL (P = 0.001; Figure 3D ).

No significant changes were observed in IL-6, IL-8, IL-4, IL-1β, IL-12p70, IL-17, TNF-α, or IFN-α between pre-blinatumomab and CRS onset ( Figures 3E–L ).

In total, nine patients had NR after the first cycle of blinatumomab. During follow-up, four relapses and one death were documented. Among the relapses, three occurred in the R/R group and one in a CR-MRD^pos patient with subsequent MRD reversion to positivity. Of these, three were BM recurrences and one was a CNS relapse. Additionally, one CR-MRD^neg patient died from CNS-invasive aspergillosis.

In patients with and without R/R status, the 1-year EFS rates were 73.3% ± 8.1% and 93.3% ± 2.9%, respectively (P = 0.0004; Figure 4A ). The 1-year OS rates were 93.3% ± 4.6% and 98.6% ± 1.3%, respectively (P = 0.009; Figure 4B ).

In patients with CR-MRD^neg and CR-MRD^pos status, the 1-year EFS rates were 97.8% ± 2.2% and 86.7% ± 6.2%, respectively (P = 0.001; Figure 4C ). The 1-year OS rates were 97.8% ± 2.2% and 100%, respectively (P = 0.029; Figure 4D ).

In patients with R/R and CR-MRD^pos status at initial classification, the 1-year EFS was 94.0% ± 3.4% for those achieving MRD negativity versus 10.0% ± 9.5% for those not achieving MRD negativity after blinatumomab (P < 0.0001). The corresponding 1-year OS rates were 100% and 78.8% ± 13.4% (P = 0.001).

In the R/R ALL subgroup, the 1-year EFS was 87.5% ± 6.8% for patients who achieved MRD negativity after blinatumomab therapy, compared with 16.7% ± 15.2% for those who remained MRD positive (P < 0.0001). The 1-year OS rates were 100% and 66.7% ± 19.2%, respectively (P = 0.002).

In the CR-MRDpos cohort, the 1-year EFS was 100% for patients who achieved MRD negativity versus 0% for those who remained MRD positive (P < 0.0001). The 1-year OS was 100% in both groups (P = 1.000).

Among the 51 cases with CR-MRD^pos or R/R status, nine cases failed to respond. Univariable analysis revealed poor cytogenetics, BCR-ABL1 fusion, and low absolute B-cell count as risk factors for treatment failure ( Table 2 ). Multivariable analysis using a Cox regression model further demonstrated that high MRD level (P = 0.014), BCR-ABL1 fusion (P = 0.065), and poor cytogenetics (P = 0.025) were independent risk factors.

To evaluate the impact on MRD negativization, univariable analysis revealed poor cytogenetics (P = 0.0003), BCR-ABL1 fusion (P = 0.004), and low absolute B-cell count (P = 0.065) as risk factors ( Supplementary Table S3 ). The Cox regression model showed that high MRD level (P = 0.014) and poor cytogenetics (P = 0.009) were independent risk factors.

In the initial cohort of 30 patients with R/R disease, 22 patients receiving bridging therapy demonstrated a CR rate of 86.4% (19/22), while the CR rate was 75.0% (6/8) in patients without bridging therapy (P = 0.589). Details of bridging therapy and response for R/R patients are presented in Supplementary Table S4 . A subset of eight patients received bridging chemotherapy following induction therapy, with cyclophosphamide administered at doses of either 1,000 mg/m² or 300 mg/m². No statistically significant difference in CR rate was observed between patients receiving cyclophosphamide-containing bridging chemotherapy (87.5%, 7/8) and those who did not undergo lymphodepleting chemotherapy (81.8%, 18/22; P = 1.000; Supplementary Table S4 ).

Each patient received one to four courses of blinatumomab, with a total of 125 cycles across the entire cohort. The most common AEs were CRS and hematologic toxicity. The incidence of severe CRS and ICANS in the R/R and CR-MRD^pos groups was comparable to that in the CR-MRD^neg group (3.9% vs. 0% and 0% vs. 5.4%, respectively; P = 0.146 and P = 0.399; Table 3 ).

Among the six patients who developed ICANS, the median age was 160 months, significantly older than that of patients without ICANS [160.5 (69–210) vs. 73 (5–212) months; P = 0.014]. Four patients underwent serum cytokine profiling both before and at ICANS onset, while three patients additionally received cerebrospinal fluid (CSF) cytokine profiling. During ICANS onset, CSF showed a white blood cell count of 0–40 cells/μL, albumin levels between 300–783 mg/L, and cytokine profiling in three patients revealed elevations of IL-5, IL-6, and IL-8, while IL-2, IL-10, and IFN-γ remained within reference ranges. Meanwhile, most serum cytokines showed no abnormal elevations; however, elevated IL-8 was observed. Detailed data are shown in Table 4 .

Five of six patients with ICANS underwent T-cell subset analysis. A more pronounced inversion of the CD4+/CD8+ ratio—particularly with higher CD8+ proportions, together with lower absolute B-cell counts and reduced B-cell percentages—was implicated in a higher risk of ICANS ( Supplementary Table S5 ).

Severe neutropenia occurred more frequently in the CR-MRD^neg group compared with the high-MRD group (32.4% vs. 17.4%; P = 0.003). Thrombocytopenia was more common in the R/R plus MRD^pos group than in the CR-MRD^neg group (16.4% vs. 1.4%; P = 0.003).