Запоздалое переливание после лечения CART связано с стойкой гематотоксичностью и безвозвратной смертностью во многих миеломах: постспециальный анализ.

Background

Chimeric antigen receptor T-cell (CAR-T cell) therapy has shown revolutionary efficacy in relapsed/refractory multiple myeloma (R/R MM) [ 1 – 3 ]. However, its safety can be compromised by unique adverse effects, including cytokine release syndrome (CRS), immune effector cell-associated neurotoxicity syndrome (ICANS), and hemophagocytic lymphohistiocytosis (HLH)-like syndrome [ 4 – 6 ]. Real-world data show that hematologic toxicities (HTs) are the most common grade 3 or higher adverse events in patients receiving CAR-T cell therapy. In studies of B-cell mature antigen (BCMA) CAR-T cell therapy, up to 95% of patients experienced prolonged cytopenia (low blood cell counts), which persisted for at least 30 days in more than half of these patients [ 7 , 8 ]. Such effects lead to prolonged hospital stays, elevated rates of infection, and increased non-relapse mortality (NRM) [ 9 ].

Comparing HT rates across trials is difficult due to differences in disease types and cell targets for CAR T-cell therapy. To address this, the European Hematology Association (EHA)/EBMT group proposed the concept of immune effector cell-associated hematotoxicity (ICAHT), categorized as early and late phases [ 10 ]. A study applying a neutrophil-based scoring system (N-ICAHT) found that approximately 15% of MM patients develop severe early N-ICAHT, which is linked to prolonged neutropenia, severe thrombocytopenia, and anemia [ 11 ]. Recently, Rejeski et al. [ 12 ] introduced the thrombocytopenia-based T-ICAHT grading system. Studies of both systems confirmed that patients with higher grades of ICAHT experience a higher burden of platelet and red blood cell (RBC) transfusions and shorter survival.

Clinical management of ICAHT is challenging. Our center has shown that treatment with granulocyte colony-stimulating factor (G-CSF) could reduce the risk of neutropenia after CAR-T cell therapy [ 13 ]. In addition to G-CSF, blood transfusion is the main supportive treatment. However, true transfusion needs depend on the primary disease and the CAR-T cell type. Statistics show that 23.1%–66% of patients need RBC or platelet transfusions after CAR-T cell therapy [ 14 – 16 ]. Such requirements are indicative of a patient’s bone marrow reserve. Long-term transfusion can cause iron overload as well as circulatory overload, reducing quality of life. Also, transfusion-related immunomodulation (TRIM) can affect the efficacy of CAR-T cell therapy [ 17 ]. Early studies in patients with solid tumors showed that allogeneic blood transfusion increases mortality and tumor recurrence [ 18 – 20 ]. Recent research also suggests that transfusions may reduce response rates to immunotherapy [ 21 ]. In large B-cell lymphoma, both early transfusion and late platelet transfusion were found to be associated with shorter progression-free survival (PFS) and overall survival (OS), as well as higher mortality [ 22 ]. Although high rates of cytopenia were previously attributed to bone marrow infiltration and the effects of multiple prior lines of therapy, later findings suggest that disease-specific differences are more important [ 23 ].

Currently, research on the transfusion needs of R/R MM patients after CAR-T cell therapy remains limited. Because understanding these needs and their impact on survival is crucial for optimizing treatment, the present study examined the patterns of transfusion needs among R/R MM patients who received CAR-T cell therapy, analyzed risk factors for transfusion need, and evaluated the effects of transfusion events on clinical outcomes.

Methods

Patients and data collection

We conducted a retrospective study on patients with R/R MM who received CAR-T cell therapy at the Affiliated Hospital of Xuzhou Medical University between May 2018 and December 2023 (clinical trial numbers: ChiCTR1900026219, ChiCTR2000033194, ChiCTR2100048888). We have previously described the methods applied for lymphodepletion chemotherapy and CAR-T cell preparation [ 1 , 2 ].

We collected baseline data, including age, sex, MM type, International Staging System (ISS) stage, Eastern Cooperative Oncology Group Performance Status (ECOG PS), tumor burden, cytogenetics, extramedullary disease (EMD), and prior treatments. Laboratory data were retrieved from electronic medical records and included blood counts, lactate dehydrogenase (LDH) concentration, and β2-microglobulin level before lymphodepletion. We also tracked absolute neutrophil counts (ANCs), hemoglobin (Hb) levels, and platelet counts from Day 0 to Day 30, and peak concentrations of inflammatory cytokines, including interleukin (IL)-2, IL-6, and IL-10; tumor necrosis factor-α (TNF-α); and interferon-α (IFN-α) and IFN-γ within the first month after infusion.

We recorded the number, frequency, timing, and type of all blood product transfusions from 6 months before to 6 months after CAR-T cell infusion (D-6 M to D6M). Transfusion of packed RBCs (PRBCs) or single-donor platelets (SDPs) was counted as one event. Administration of other products such as cryoprecipitate or small volumes of fresh frozen plasma was not counted. Blood products were prepared by fresh apheresis, leukoreduction, and irradiation. Transfusion decisions were based on clinical judgment and standard guidelines.

The patient inclusion criteria were adapted from Vic et al. [ 22 ] to include R/R MM patients treated with at least one CAR-T cell infusion (targets: G protein-coupled receptor class C group 5 member D [GPRC5D], BCMA, CD19) at our center, with at least 6 months of follow-up and detailed transfusion data. The early transfusion phase was defined as Day 0 to Month 1 (1 M), and the late phase was defined as Month 1 to Month 6 (1 M-6 M).

The study was approved by the Ethics Committee of the Affiliated Hospital of Xuzhou Medical University (approval number XYFY2021-KL154-01). Written informed consent was obtained from each patient before study entry in accordance with the Declaration of Helsinki.

Hematologic toxicity

CRS and ICANS were diagnosed according to official American Society for Transplantation and Cellular Therapy (ASTCT) criteria [ 24 ]. Neutropenia was evaluated based on the 2023 EHA/EBMT consensus guidelines for ICAHT [ 10 ]. According to the guidelines, early N-ICAHT is characterized by neutropenia (ANC ≤ 500/µL) from the day of CAR-T cell infusion (D0) to D28, whereas late N-ICAHT is defined as neutropenia beyond D28. Severe N-ICAHT involves persistence of ANC ≤ 500/µL for at least 14 days or ANC ≤ 100/µL for at least 7 days during the first 28 days after CAR-T cells infusion (early N-ICAHT), or ANC ≤ 500/µL at two or more time points or nontransient neutropenia after D28 (late N-ICAHT). Early T-ICAHT was graded based on the depth and duration of thrombocytopenia in the first 30 days. Late T-ICAHT represents the thrombocytopenia severity between Day 31 and Day 100 [ 12 ].

Severe infection was defined as an infectious event requiring hospitalization, intravenous antimicrobial therapy, ICU admission, or resulting in infection-related death.

Clinical outcomes

Treatment response in the included R/R MM patients was assessed using the International Myeloma Working Group (IMWG) criteria. PFS was defined as the time from the date of CAR-T cell infusion to disease progression or all-cause death. OS was defined as the time from the date of CAR-T cell infusion to death due to any cause.

Statistical analysis

Intergroup differences in continuous variables were analyzed using the Mann–Whitney U test, and percentage differences were identified using Fisher’s exact test. For univariate analysis, P < 0.05 was considered statistically significant, and factors with significant P values were retained in the multivariate analysis. P -values for the overall model significance and for non-linearity were derived from the likelihood ratio test. A linear Cox model was employed when RCS modeling was inapplicable for the threshold analysis, and optimal cutoffs were determined by the maximally selected rank statistic method. PFS and OS survival curves were estimated using the Kaplan–Meier method. The following software programs were used for statistical analysis and data visualization: R Resources (version 4.4.2).

Results

Basic and clinical characteristics of R/R MM patients

This retrospective study included 155 patients with R/R MM who received CAR-T cell therapy during the study period of May 2018 to December 2023 according to the inclusion criteria. The types of CAR-T cell therapy administered to the study population included BCMA CAR-T cell therapy for 38 patients, GPRC5D CAR-T cell therapy for 38 patients, BCMA+CD19 CAR-T cell therapy for 33 patients, bispecific BCMA/CD19 CAR-T cell therapy for 31 patients, and bispecific BCMA/GPRC5D CAR-T cell therapy for 15 patients. The patient characteristics are presented in Table 1 (and Additional file 1: Tables S1A - S1F ). The study design flow is depicted in Fig. 1 . At the time of CAR-T cell infusion, the median patient age was 58 years (range, 53–63 years), and the median number of prior lines of therapy was 4 (range, 3–6). Of the 155 included patients, 23 (14.1%) had a ECOG PS > 0, 67 (43.2%) had extramedullary disease, 62 (40.0%) had high-risk genetic factors, 42 (27.1%) had prior hematopoietic stem cell transplantation, and 32 (20.6%) had received bridging therapy. Thirty-nine patients (25.2%) experienced infections within 3 months prior to lymphodepletion.

Table 1 Patient baseline characteristics between patients receiving ≥ 1 transfusion after CAR T-cell infusion vs. no transfusion Characteristics No Transfusion ( N = 81) ≥ 1 transfusion ( N = 74) Total ( N = 155) P -value Age , median (IQR) 57.0 (53.0, 64.0) 58.00 (54.0, 63.0) 58.0 (53.0, 63.0) 0.907 Gender , no. (%) 0.320 Male 48 (59.3) 37 (50.0) 85 (54.8) Female 33 (40.7) 37 (50.0) 70 (45.2) ECOG PS , no. (%) 0.006 2 0 (0.0) 9 (12.2) 9 (4.7) 1 6 (7.4) 10 (13.5) 14 (9.4) 0 75 (92.6) 55 (74.3) 130 (85.9) Types , no. (%) 0.613 IgA 14 (17.3) 12 (16.2) 26 (16.8) IgD 6 (7.4) 5 (6.8) 11 (7.1) IgG 48 (59.3) 40 (54.1) 88 (56.8) k 7 (8.6) 5 (6.8) 12 (7.7) l 4 (4.9) 10 (13.5) 14 (9.0) Non-secretory 2 (2.5) 2 (2.7) 4 (2.6) ISS stage , no. (%) 0.685 I 20 (24.7) 15 (20.3) 35 (22.6) II 27 (33.3) 26 (35.1) 53 (34.2) III 28 (34.6) 30 (40.5) 58 (37.4) Missing 6 (7.4) 3 (4.1) 9 (5.8) CAR-T targets a , no. (%) 0.182 BCMA 18 (22.2) 20 (27.0) 38 (24.5) BCMA/CD19 14 (17.3) 17 (23.0) 31 (20.0) BCMA/GPRC5D 11 (13.6) 4 (5.4) 15 (9.7) BCMA+CD19 21 (25.9) 12 (16.2) 33 (21.3) GPRC5D 17 (21.0) 21 (28.4) 38 (24.5) EMD b , no. (%) 27 (33.3) 40 (54.1) 67 (43.2) 0.015 High risk cytogenetic profile c , no. (%) 23 (28.4) 39 (52.7) 62 (40.0) 0.003 High tumor burden d , no. (%) 4 (4.9) 20 (27.0) 24 (15.5) < 0.001 Triple-class refractory e , no. (%) 17 (21.0) 22 (29.7) 39 (25.2) 0.286 Penta-drug refractory f , no. (%) 7 (8.6) 9 (12.2) 16 (10.3) 0.649 Bridging therapy g (%) 11 (13.6) 21 (28.4) 32 (20.6) 0.038 Previous therapy lines , median (IQR) 3.0 (2.0, 5.0) 5.0 (3.0, 7.0) 4.0 (3.0, 6.0) < 0.001 Previous BCMA CAR-T therapy , no. (%) 9 (11.1) 17 (23.0) 26 (16.8) 0.055 Previous infection , no. (%) 9 (11.1) 30 (40.5) 39 (25.2) < 0.001 Previous HSCT , no. (%) < 0.001 Autologous 11 (13.6) 30 (40.5) 41 (26.5) Allogeneic 0 1 (1.4) 1 (0.6) Severe infection after infusion , no. (%) 24 (29.6) 36 (48.6) 60 (38.7) 0.024 MK counts h , median (IQR) 5.00 (2.00, 11.00) 2.00 (1.00, 4.00) 3.00 (1.00, 9.00) < 0.001 CRP , median (IQR) 1.25 (0.65, 3.15) 2.75 (0.98, 6.65) 1.60 (0.80, 4.75) 0.053 D-dimer , median (IQR) 0.32 (0.22, 0.55) 0.59 (0.40, 0.97) 0.47 (0.30, 0.81) 0.005 a. The BCMA/CD19 and BCMA/GPRC5D constructs were designed as bispecific targets. The BCMA+CD19 configurations represent a dual-target approach. b. Extramedullary disease (EMD) was defined as tissue masses in extraosseous locations. c. High-risk cytogenetics were defined as del(17p), t(4;14), t(14;16), t(14;20) or 1q21 gain or amplification by fluorescence in situ hybridization. d. High tumor burden was defined as at least 50% plasma cells in bone marrow. e. Refractory to an immunomodulatory agent, a proteasome inhibitor and an anti-CD38 monoclonal antibody. f. Refractory to lenalidomide, pomalidomide, bortezomib, carfilzomib and daratumumab. g. Bridging therapy was administered after leukapheresis and before lymphodepletion. h. the megakaryocyte count found within a 4.5 square centimeter area of the bone marrow tissue section on a glass slide.

Fig. 1 Patient stratification flowchart based on post-CAR-T transfusion requirements. Flowchart illustrating the inclusion of 155 patients with relapsed/refractory multiple myeloma treated with CAR-T cell therapy. Patients were stratified based on the timing (early phase < 1 month vs. late phase > 1 month) and type (both type transfusion vs. single type) of transfusion requirements.

Transfusion requirements among patients receiving CAR-T cell therapy

We recorded all transfusion events that occurred during the month preceding CAR-T cell reinfusion, during the early transfusion phase following CAR-T cell reinfusion, and in the late transfusion phase. Within 1 month prior to CAR-T cell infusion, 34 (21.9%) patients received transfusions, which included RBC transfusions for 30 (19.4%) patients and platelet transfusions for 17 (11.0%) patients. Among these 34 patients, 13 (8.4%) required both RBC and platelet transfusions (Fig. 2 A).

Fig. 2 Transfusion patterns after CAR-T cell infusion and clinical correlates. A-C. Evolution of transfusion requirements following CAR T-cell therapy. A Number of patients who received transfusions over time; B mean number of transfused red blood cell (RBC) units over time; and C mean number of transfused platelets units over time. SEM, standard error of the mean. D-E. Distribution of transfusion patterns across ICAHT grades. D Distribution of transfusion patterns across N-ICAHT grades; and E distribution of transfusion patterns across T-ICAHT grades. F-G. Severe infection and supportive care rates after CAR-T cell infusion. F Supportive care rate after CAR-T cell therapy; and G severe infection within 180 days after CAR T-cell infusion. H-I. Association of different transfusion events with inflammation ( H ) and clinical toxicities ( I ) after CAR-T cell therapy. J Impact of transfusion on CAR-T treatment efficacy in the early and late phases. *** P < 0.001; Continuous variables were compared between responders (≥ PR) and non-responders (< PR) using the Mann–Whitney U test. K-L Non-linear relationship between post-CAR T-cell transfusion volume and survival outcomes. K Association between total RBC transfusion and overall survival. L Association between total platelet transfusion and overall survival. Note: In Fig. 2G, severe infection is defined as requiring hospitalization, intravenous antimicrobial treatment, ICU admission, or resulting in infection-related death; In Fig. 2K–L, 0 units of post-CAR-T transfusion was used as the reference value for hazard ratio estimation. The displayed curves represent the full spline results across the observed range of transfusion exposure, without truncation or restriction

During the 6-month follow-up period after CAR-T cell infusion, transfusion requirements persisted (Fig. 2 A). A total of 74 (47.7%) patients received transfusions, which included RBC transfusions for 59 (38.1%) patients and platelet transfusions for 58 (37.4%) patients. Among the 74 patients, 30 (40.5%) needed both RBC and platelet transfusions.

Transfusion requirements were concentrated in the first month post-reinfusion (i.e., the early transfusion phase), during which 60 (38.7%) patients required transfusions [RBCs for 46 (29.7%) patients and platelets for 45 (29.0%) patients]. The mean RBC transfusion volume during this phase was 1.8 units (range, 0–30 units), and the mean platelet transfusion volume was 1.3 units (range, 0–5 units). Thereafter, transfusion requirements showed a declining trend (Fig. 2 B and C). During the 1- to 6-month post-transfusion period (i.e., the late transfusion phase), 55 (35.5%) patients received transfusions, with 36 (23.2%) receiving RBC transfusions and 47 (30.3%) receiving platelet transfusions. Although the average transfusion volume did not decrease further between 5 and 6 months post-CAR-T cell infusion, the proportion of patients requiring transfusions did drop to 9.7%, with only 15 patients needing at least one transfusion in this period (Fig. 2 A).

Early ICAHT grading and transfusion requirements

In our patient cohort, transfusion requirements increased with higher early N-ICAHT grades. Among patients with Grade 0 N-ICAHT ( n = 83), 69.9% required no transfusion, whereas 10.8% received RBC transfusion alone, 7.2% received platelet transfusion alone, and 12.0% required both RBC and platelet transfusion (Fig. 2 D and Additional file 1: Fig. S2C ). Patients with Grade 1 N-ICAHT ( n = 36) showed greater transfusion needs: 55.6% required no transfusion, while 2.8% received RBC alone, 13.9% received platelets alone, and 27.8% required both products (Fig. 2 D and Additional file 1: Fig. S2D ). In Grade 2 N-ICAHT ( n = 18), transfusion was required in 77.8% of patients, including 11.1% with RBC alone, 22.2% with platelets alone, and 44.4% with both RBC and platelets, whereas 22.2% required no transfusion (Fig. 2 D and Additional file 1: Fig. S2E ). Notably, among patients with severe early N-ICAHT (Grade 3–4, n = 18), 88.9% required combined RBC and platelet transfusion and 11.1% received RBC alone, with no patients requiring platelet-only transfusion or no transfusion (Fig. 2 D and Additional file 1: Fig. S2F ).

Similar to the pattern observed with N-ICAHT grading, the frequency of transfusion needs increased with higher T-ICAHT grades (Fig. 2 E and Additional file 1: Fig. S2G–J ). All 36 patients with early severe (Grade 3–4) T-ICAHT received at least one transfusion of RBCs and/or platelets, with 83.3% requiring transfusion of both blood products.

Early transfusion mirrors inflammatory toxicity and intensive supportive care

Our analysis revealed that early transfusion events were strongly associated with heightened inflammatory responses and clinical toxicity (Fig. 2 F-I and Additional file 1: Table S2 ). Early RBC transfusion was significantly associated with elevated levels of multiple proinflammatory cytokines, including IL-2, IL-6, IL-10, and IFN-α. Concurrently, patients in this group exhibited higher CRS grades ( P = 0.040) and a longer CRS duration ( P = 0.036). Early platelet transfusion was also significantly associated with elevated levels of cytokines, including IL-2, IL-6, IL-8, and IFN-γ. Regarding clinical toxicity, patients in this group had higher ICANS grades ( P = 0.001) and CRS grades ( P = 0.026), with a significantly extended CRS duration ( P = 0.019) and ICANS duration ( P = 0.030) (Fig. 2 H, I and Additional file 1: Table S2 ).

Regarding adverse events and supportive care (Figs. 2 F, G), the rates of severe infection in the groups requiring early RBC and early platelet transfusion were 57% and 58%, respectively. These patients also required frequent supportive interventions, particularly corticosteroids (72% and 82% of those requiring early RBC or platelet transfusion, respectively) and tocilizumab (41% and 47% of those requiring early RBC or platelet transfusion, respectively) (Fig. 2 F). The rates of severe infection among the patients who received late RBC and late platelet transfusions were 58% and 52%, respectively. Among the patients who required late transfusion of RBCs or platelets, the percentages who were treated with corticosteroids were 67% and 70%, respectively, and the percentages who received tocilizumab were 31% and 33%, respectively.

Late transfusion identifies poor response depth and excess nonrelapse mortality

Transfusion events were associated with the depth of therapeutic remission, particularly those occurring in the late phase. Early RBC and platelet transfusions showed no significant correlation with final efficacy (Fig. 2 J). Late RBC transfusion (overall response [ORR]: 61.1% vs. 86.6%; P < 0.001) and late platelet transfusion (ORR: 63.0% vs. 88.1%; P < 0.001) were both significantly associated with poorer treatment response depth, which manifested as a significantly lower response depth in patients who received transfusions than in those who did not.

In the analysis of causes of death, disease-related mortality (DRM) showed no significant difference across all transfusion subgroups compared with that among non-transfused patients (Table 2 A). However, NRM was significantly elevated in multiple transfusion subgroups compared with the non-transfusion group, including in the groups that received early RBC transfusion (26.1% vs. 10.1%; P = 0.021), late RBC transfusion (27.8% vs. 10.9%; P = 0.026), and late platelet transfusion (26.1% vs. 10.1%; P = 0.021) (Table 2 B).

Table 2 DRM and NRM by transfusion status A Transfusion category Transfused (%) Not transfused (%) P -value Early RBC (< 1 M) 30.4 30.3 1.000 Late RBC (≥ 1 M) 41.7 26.9 0.138 Early platelets (< 1 M) 31.1 30.0 1.000 Late platelets (≥ 1 M) 37.0 27.5 0.329 B Early RBC (< 1 M) 26.1 10.1 0.021 Late RBC (≥ 1 M) 27.8 10.9 0.026 Early platelets (< 1 M) 22.2 11.8 0.160 Late platelets (≥ 1 M) 26.1 10.1 0.021 NRM, nonrelapse mortality; DRM, disease related mortality; RBC, red blood cells; M, month. (A) Disease related mortality in patients with early blood transfusion and late blood transfusion. (B) Nonrelapse mortality in patients with early blood transfusion and late blood transfusion.

Risk factors associated with the need for transfusion

During the early phase (< 1 month), requirements for RBC and platelet transfusions shared multiple risk factors, including Grade ≥ 2 early ICAHT, history of transfusion within 1 month preceding CAR-T cell infusion, history of hematopoietic stem cell transplantation (HSCT), and tocilizumab use. Additionally, EMD was a risk factor specific to early platelet transfusion (Table 3 A).

Table 3 Multivariable analyses of early and late transfusions after CAR T-cell therapy A RBC Platelets Characteristic aOR (95% CI) P -value aOR (95% CI) P -value Early ICAHT grade ≥ 2 5.2 (1.9, 15) 0.002 2.8 (1.0, 7.9) 0.041 Any transfusion with 1 month before CAR T-cell infusion 6.1 (1.6, 27) 0.010 4.3 (1.1, 19) 0.042 Previous HSCT 3.4 (1.2, 10) 0.023 7.3 (2.6, 23) < 0.001 Tocilizumab use 3.6 (1.1, 12) 0.035 8.9 (2.7, 33) < 0.001 EMD NS 3.5 (1.3, 9.8) 0.015 B G-CSF use NS 5.5 (1.7, 22) 0.008 Early ICAHT grade ≥ 2 NS 3.4 (1.3, 9.0) 0.014 Previous HSCT 2.9 (1.1, 7.9) 0.035 4.9 (1.9, 13) < 0.001 High tumor burden a 3.8 (1.2, 12) 0.023 NS Previous BCMA CAR-T therapy 4.1 (1.3, 14) 0.019 NS Any transfusion with 1 month before CAR T-cell infusion 3.2 (1.0, 9.8) 0.041 NS EMD b 3.3 (1.3, 8.8) 0.013 NS a. High tumor burden was defined as at least 50% plasma cells in bone marrow. b. Extramedullary disease (EMD) was defined as tissue masses in extraosseous locations. ICAHT, immune effector cell-associated hematotoxicity; HSCT, hematopoietic stem cell transplantation; G-CSF, granulocyte colony-stimulating factor; EMD, extramedullary disease; BCMA, B-cell maturation antigen; NS, not significant. (A) Predictive factors for transfusion in the early phase (< 1 month). (B) Predictive factors for transfusion in the late phase (≥ 1 month). Factors tested in the univariate analysis index age > 60, ECOG PS > 1, ISS, stage III, high tumor burden, EMD at infusion, previous HSCT, any transfusion with 1month before CAR T-cell infusion, bridging therapy, tocilizumab use after CAR T-cell infusion, GM-CSF use after CAR T-cell infusion, G-CSF use after CAR T-cell infusion, CRS grade ≥ 3, previous BCMA CAR-T therapy, severe infection after CAR T-cell infusion and early ICAHT grade ≥ 2. Variables significant in univariate analysis ( P < 0.05) were included in the multivariate model.

Late-stage RBC transfusion (≥ 1 month) was independently associated with a history of HSCT, a high tumor burden, prior BCMA CAR-T cell therapy, history of transfusion within 1 month prior to reinfusion, and EMD. Independent predictors for late-stage platelet transfusion were G-CSF use, Grade ≥ 2 early ICAHT, and history of HSCT (Table 3 B).

Transfusion and survival outcomes

The last follow-up date was August 15, 2025, and the median time duration among all 155 patients was 41.7 months (95% confidence interval [CI], 35.1–48.3 months). The need for transfusion, particularly RBC transfusion, was a significant risk factor for shortened PFS (Fig. 3 B). Patients receiving late RBC transfusions demonstrated a significantly worse median PFS than those who did not require transfusion (4.93 vs. 17.17 months, P < 0.0001). Early RBC transfusion (median PFS: 8.17 months vs. 17.17 months, P = 0.0172) and late platelet transfusion (median PFS: 5.25 months vs. 17.53 months, P = 0.0038) also were significantly associated with poorer PFS compared with no transfusion. However, the effect of early platelet transfusion on PFS did not reach statistical significance ( P = 0.0972) (Fig. 3 B and Table S3 ).

Fig. 3 Overall and progression-free survival stratified by transfusion type and timing. A Overall survival comparing non-transfused vs. transfused patients, with further stratification of transfused patients by transfusion timing (early vs. late). B Progression-free survival comparing non-transfused vs. transfused patients, with further stratification of transfused patients by transfusion timing (early vs. late). C Survival outcomes stratified by transfusion types

Similar and more pronounced associations were observed for OS (Fig. 3 A). Patients who received late RBC transfusions had the shortest median OS (10.87 months vs. 37.87 months for patients who did not receive transfusions, P < 0.0001). Early RBC transfusion (median OS: 18.63 months vs. 34.97 months, P = 0.0003) and late platelet transfusion (median OS: 19.20 months vs. 38.77 months, P = 0.0003) were similarly significantly associated with reduced OS. Consistent with the result for PFS, early platelet transfusion had no significant impact on OS ( P = 0.1146). Overall, the need for transfusion of RBCs and platelets was significantly associated with poor survival outcomes.

In the group that received both types of transfusions in the early phase, the 5-year OS was 16.3% (95% CI: 5.2%–51.5%) (Fig. 3 C), which was significantly lower than that in the single type-transfusion group (26.5%) and in the no-transfusion group (43.5%; P = 0.0126). PFS also was significantly worse for patients who received both types of transfusions in the early phase ( P = 0.0287). These differences between the groups that received both types of transfusions versus a single type of transfusion or no transfusions became even more pronounced in the late phase (OS, P < 0.0001; PFS, P = 0.0002), with 5-year OS dropping to 6.8% (95% CI: 1.1%–42.1%) for the group that received both types of transfusions, which was far below the OS for both the single type-transfusion group (25.4%) and no-transfusion group (46.0%).

A nonlinear, dose- and time-dependent association was observed between transfusion volume post-CAR-T cell therapy and survival outcomes (Figs. 2 K, L and Additional file 1: Fig. S3 ). A total RBC transfusion volume exceeding 1 unit (hazard ratio [HR] 1.19; 95% CI, 1.06–1.34; P < 0.001) and a total platelet transfusion volume exceeding 10.1 units (HR, 1.76; 95% CI, 1.02–3.05; P < 0.001) were significantly associated with increased patient mortality risk (Figs. 2 K, L). Stratified analysis by transfusion timing showed that both early and late RBC and platelet transfusions were significantly associated with reduced OS (Additional file 1: Fig. S3A-D and Additional file 1: Table S4 ), but only late transfusions were associated with shortened PFS (Additional file 1: Fig. S3E-H ).

Time-dependent Cox proportional hazards regression showed that in multivariable time-dependent models early RBC transfusion remained independently associated with worse PFS (HR 1.677, 95% CI 1.034–2.719; P = 0.036) and OS (HR 2.307, 95% CI 1.208–4.407; P = 0.011). Late transfusion of any type was significantly associated with inferior OS (HR 2.534, 95% CI 1.303–4.927; P = 0.006), but showed no significant association with PFS (HR 1.546, 95% CI 0.905–2.641; P = 0.111). Landmark analyses yielded consistent findings: in the 30-day landmark analysis, early RBC transfusion was associated with significantly poorer PFS ( P = 0.040) and OS ( P < 0.001); in the 90-day landmark analysis, late transfusion was linked to inferior PFS and OS ( P = 0.020 and P < 0.001, respectively) (Additional file 1: Table S4 ; Additional file 1: Fig. S4–S5 ).

Discussion

In this large retrospective study of patients with R/R MM who were treated with CAR T-cell therapy based on five different cell targets, we comprehensively characterized the landscape of transfusion requirements. Within our cohort, 21.9% of patients required transfusion support in the month preceding CAR-T cell infusion, reflecting widespread hematopoietic impairment driven by prior multi-line therapies and inherent bone marrow involvement. Following CAR-T cell infusion, 47.7% of patients required transfusion within the first 6 months. The demand peaked in the first month (38.7%) and gradually declined, with only 9.7% of patients requiring transfusion by month 6. This trajectory aligns with the known kinetics of lymphodepletion-induced myelosuppression and ICAHT [ 10 , 25 ]. Our findings parallel those of Jin et al., 16 who reported that 30 days after BCMA CAR-T cell reinfusion, 23.1% of patients received RBC transfusions and 26.9% received platelet transfusions. Notably, we observed a distinct temporal shift in the type of blood product required. While RBC and platelet needs were comparable in the first month (RBC needs vs. platelet needs: 29.7% vs. 29.0%), platelet transfusion became the dominant requirement in the late phase (RBC needs vs. platelet needs: 23.2% vs. 30.3%). This supports the concept that early anemia is primarily associated with the lymphodepletion regimen, whereas persistent thrombocytopenia is a hallmark of late-phase toxicity. The occurrence of long-term thrombocytopenia after CAR-T cell therapy can be caused by multiple factors including CAR target structure, prior lines of therapy, bridging regimens, baseline intramedullary tumor burden, and levels of endothelial inflammation markers such as baseline D-dimer, and the precise pathophysiology remains to be fully elucidated [ 1 , 7 , 15 , 26 , 27 ].

Our data indicate a robust correlation between high-grade early N-ICAHT and increased transfusion requirements. The neutrophil count–based N-ICAHT grading likely captures the impact of the CAR-T-driven systemic inflammatory storm on the functional reserve of the bone marrow microenvironment [ 10 ]. Clinically, this manifests as a higher transfusion burden, particularly in the early post-infusion period, consistent with previous findings [ 28 ]. Patients with severe (Grade 3–4) early N-ICAHT in our study exhibited significantly higher transfusion needs. Regarding the relationship between inflammation and cytopenia, Jain et al. [ 15 ] and Rejeski et al. [ 29 ] found no association between the serum level of IL-6 and prolonged cytopenias. However, Juluri et al. [ 30 ] reported conflicting conclusions. Our results support the former conclusion, with early inflammatory adverse events not significantly predicting late transfusion requirements. This discrepancy demonstrates that the mechanisms underlying late cytopenias are complex, potentially involving persistent immune dysregulation, stem cell exhaustion from prolonged therapy, or sustained suppression of hematopoietic progenitor cells by CAR-T cells, rather than merely the initial cytokine storm [ 25 , 31 , 32 ]. Additionally, neutropenia and lymphopenia heighten the risk of infections. In patients with high N-ICAHT grades, lower baseline granulocyte levels and greater antibiotic exposure may exacerbate post-CAR-T hematologic toxicity, negatively impacting long-term survival and indirectly increasing transfusion needs [ 28 , 33 – 37 ]. Recently, Rejeski et al. [ 12 ] established the T-ICAHT grading system based on platelet counts and showed that higher grades correlate with increased transfusion burden and inferior survival, conclusions corroborated by the results of the present study. Moreover, predictive models such as Early ICAHT Prediction Models (eIPMs) [ 38 ] have been developed to forecast severe early N-ICAHT, and thus, could facilitate risk stratification to guide interventions including transfusion and infection prophylaxis.

Multivariate analysis identified that, during the early phase, a history of transfusion within the preceding month was a significant predictor, as was a history of HSCT, which likely reflects cumulative damage to hematopoietic progenitor cells [ 27 , 39 ]. The roles of G-CSF and granulocyte–macrophage colony-stimulating factor (GM-CSF) in modulating CAR-T cell therapy efficacy and toxicity remain topics of debate [ 40 – 42 ]. Our data indicate that neither factor predicted early transfusion requirements, although G-CSF use was associated with late-phase platelet transfusion needs. In the late phase, risk factors diverged. Late platelet transfusion was linked to markers of marrow exhaustion such as G-CSF use, whereas late RBC transfusion was primarily determined by baseline marrow reserve. Lower baseline MK counts in transfused patients may reflect pre-existing megakaryopoietic vulnerability, potentially contributing to persistent platelet transfusion needs after CAR-T cell infusion. As mentioned previously, older age and/or poorer performance status may be associated with the duration of hematologic toxicity. The MM patient population is predominantly middle-aged and elderly, often presenting with comorbidities such as cardiac and pulmonary diseases or reduced organ function. These factors impose limitations on anemia tolerance. Studies on large B-cell lymphoma have demonstrated that advanced age and/or poor performance status do not elevate transfusion risk. While our conclusions are consistent with these findings, this may be partly attributable to the relatively low proportion of elderly patients in our transfusion cohort [ 22 , 43 ]. We recommend assessment of transfusion volume tolerance in these patients prior to treatment. CAR-T target was reviewed as a baseline treatment characteristic. However, given the limited number of patients in each target subgroup, it was not incorporated into the final multivariable survival models to avoid unstable estimates and model overfitting.

Transfusion requirements were strongly associated with inferior clinical outcomes. The persistent need for late RBC transfusion was closely linked to reduced survival. This contrasts with lymphoma studies showing that early transfusion or late platelet dependence is associated with poor prognosis [ 22 ]. More recently, late hematologic toxicity after CAR-T has been systematically characterized in large B-cell lymphoma. In a cohort of 290 LBCL patients, Corona et al. [ 44 ] reported that late ICAHT remained frequent by day 100 (45% for any grade; 22% for grade ≥ 3), and late severe ICAHT was associated with a higher hazard of late infections. These data highlight that late cytopenias after CAR-T are clinically meaningful rather than isolated laboratory abnormalities. Rejeski et al. [ 25 , 45 ] defined persistent severe neutropenia (> 14 days) following CAR-T cell infusion as an aplastic phenotype and found that such patients experience poorer PFS and OS. Combined with our study results, this indicates that some aplastic phenotype patients overlap with thrombocytopenia, leading to similar outcomes. Furthermore, we observed that patients who required late-stage transfusions exhibited significantly poorer disease remission depth and bore a higher risk of NRM. For patients who have undergone multiple lines of therapy and suffered severe bone marrow dysfunction, the relatively slow-recovering erythroid hematopoietic function remains unreconstructed after cell reinfusion. This may indicate irreversible depletion of bone marrow reserve. In vitro studies have indicated that activated CAR-T cells secrete pro-inflammatory mediators, impair hematopoietic differentiation, and remodel transcriptional programs, contributing to persistent hematopoietic toxicity [ 25 , 44 , 46 ]. Previous animal and clinical studies also have revealed the involvement of transfusion-related immune modulation (TRIM) effects after allogeneic transfusion, which in multiple scenarios lead to significant immune downregulation and may further promote tumor relapse [ 21 , 47 ]. The hypothesis that blood product processing methods such as pre-storage leukoreduction can mitigate TRIM effects requires further validation. Furthermore, restricted cubic spline analysis confirmed a dose-dependent mortality risk with increasing transfusion volume. Although sparse data at the upper extremes limited precision, patients accumulating substantial transfusions represent a distinct subgroup with persistent marrow failure warranting further study.

Given that transfusion exposure after CAR-T therapy is inherently time-dependent, we performed additional time-dependent Cox and landmark analyses to further examine the robustness of the survival findings. These analyses indicate that the prognostic impact of transfusion is not constant over time: early RBC transfusion appears to identify an early post-infusion vulnerable period, whereas late transfusion exhibits a broader and more sustained adverse association, particularly with overall survival. Considered alongside its associations with poorer response depth and higher non-relapse mortality in the primary analyses, late transfusion appears to represent a clinically meaningful marker of persistent hematologic vulnerability.

To our knowledge, this investigation represents the largest international retrospective cohort study to date examining transfusion requirements following CAR-T cell therapy in patients with R/R MM. However, our study has limitations. First, its retrospective design may introduce selection bias. Second, heterogeneity among CAR-T cell products may contribute to variability in associations and outcomes. For example, clinical meta-analyses indicate that CD19-targeted CAR-T cell therapy is associated with a higher incidence of complications compared with BCMA-targeted therapy [ 48 ]. While the use of different targets enhances the universality of our core findings across therapeutic products, it may also influence transfusion requirements. Furthermore, the potential impacts of blood product processing methods and storage duration on hematologic toxicity and prognosis after CAR-T cell therapy remain unknown. In addition, differences in transfusion thresholds and outpatient follow-up patterns across centers may influence the incidence and timing of transfusion events, which should be considered when interpreting the external generalizability of our findings.

Conclusions

In summary, this study identifies post-CAR-T transfusion dependency as a key prognostic factor in R/R MM. Early transfusion indicates severe inflammation, whereas late dependence predicts inferior survival and increased NRM. Future research should focus on mechanisms and predictive models to optimize supportive care and improve long-term outcomes.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1: Additional file 1: Tables S1A–S5 and Figs. S1–S5.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Fan Xia and Bin Lv contributed equally to this work.

Acknowledgements

We thank Dr. Alex H Chang and his team for their assistance of this study. We also thank Medjaden Inc. for scientific editing of this manuscript.

Author contributions

F.X. and B.L. contributed to the design and conduct of the study. F.Z. and D.L. designed the study and supervised the project. D.Z., Q.S., J.X., Y.Q., H.Z., K.Q., Y.L., Y.W., W.C., J.C., Z.Y., M.S., D.L., and F.Z. recruited and followed up the patients. B.L., F.X., X.L., M.T., and X.G. collected the data. B.L. and F.X. performed the statistical analysis and drafted the manuscript. F.Z. revised the manuscript. All authors contributed to the article and read and approved the final version of the manuscript.

Funding

This study was supported by grants of National Key Research and Development Program of China (Grant Number: 2023YFC2508900 and 2023YFC2507800) and National Natural Science Foundation of China (No. 82170187).

Data availability

The datasets generated and analyzed in the present study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The study was approved by the Ethics Committee of the Affiliated Hospital of Xuzhou Medical University (approval number XYFY2021-KL154-01). Written informed consent was obtained from each patient before study entry in accordance with the Declaration of Helsinki.

Consent for publication

Not applicable.

Conflict of interests

The authors declare no competing interests.

Abbreviations

absolute neutrophil count

adjusted odds ratio

B-cell maturation antigen

beta-2 microglobulin

chimeric antigen receptor T cell

confidence interval

C-reactive protein

cytokine release syndrome

Eastern Cooperative Oncology Group performance status

extramedullary disease

granulocyte colony-stimulating factor

granulocyte–macrophage colony-stimulating factor

G-protein coupled receptor family C group 5 member D

hemoglobin

hazard ratio

hematopoietic stem cell transplantation

immune effector cell–associated hematotoxicity

immune effector cell–associated neurotoxicity syndrome

interferon

interleukin

International Myeloma Working Group

interquartile range

International Staging System

lactate dehydrogenase

multiple myeloma

neutropenia component of ICAHT

non-relapse mortality

overall response rate

overall survival

progression-free survival

platelets/platelet transfusion

packed red blood cells

red blood cells

restricted cubic spline

Revised International Staging System

relapsed/refractory

thrombocytopenia component of ICAHT

tumor necrosis factor alpha

transfusion-related immunomodulation

References

1. Zhou D Sun Q Xia J Gu W Qian J Zhuang W Anti-BCMA/GPRC5D bispecific CAR T cells in patients with relapsed or refractory multiple myeloma: a single-arm, single-centre, phase 1 trial Lancet Haematol 2024 11 10 e751 60 10.1016/S2352-3026(24)00176-5 39059405 Zhou D, Sun Q, Xia J, Gu W, Qian J, Zhuang W, et al. Anti-BCMA/GPRC5D bispecific CAR T cells in patients with relapsed or refractory multiple myeloma: a single-arm, single-centre, phase 1 trial. Lancet Haematol. 2024;11(10):e751–60. 39059405 10.1016/S2352-3026(24)00176-5

2. Wang Y Cao J Gu W Shi M Lan J Yan Z Long-Term Follow-Up of Combination of B-Cell Maturation Antigen and CD19 Chimeric Antigen Receptor T Cells in Multiple Myeloma J Clin Oncol 2022 40 20 2246 56 10.1200/JCO.21.01676 35333600 Wang Y, Cao J, Gu W, Shi M, Lan J, Yan Z, et al. Long-Term Follow-Up of Combination of B-Cell Maturation Antigen and CD19 Chimeric Antigen Receptor T Cells in Multiple Myeloma. J Clin Oncol. 2022;40(20):2246–56. 35333600 10.1200/JCO.21.01676

3. Gagelmann N Dima D Merz M Hashmi H Ahmed N Tovar N Development and Validation of a Prediction Model of Outcome After B-Cell Maturation Antigen-Directed Chimeric Antigen Receptor T-Cell Therapy in Relapsed/Refractory Multiple Myeloma J Clin Oncol 2024 42 14 1665 75 10.1200/JCO.23.02232 38358946 PMC11095856 Gagelmann N, Dima D, Merz M, Hashmi H, Ahmed N, Tovar N, et al. Development and Validation of a Prediction Model of Outcome After B-Cell Maturation Antigen-Directed Chimeric Antigen Receptor T-Cell Therapy in Relapsed/Refractory Multiple Myeloma. J Clin Oncol. 2024;42(14):1665–75. 38358946 10.1200/JCO.23.02232 PMC11095856

4. Karschnia P Jordan JT Forst DA Arrillaga-Romany IC Batchelor TT Baehring JM Clinical presentation, management, and biomarkers of neurotoxicity after adoptive immunotherapy with CAR T cells Blood 2019 133 20 2212 21 10.1182/blood-2018-12-893396 30808634 Karschnia P, Jordan JT, Forst DA, Arrillaga-Romany IC, Batchelor TT, Baehring JM, et al. Clinical presentation, management, and biomarkers of neurotoxicity after adoptive immunotherapy with CAR T cells. Blood. 2019;133(20):2212–21. 30808634 10.1182/blood-2018-12-893396

5. Hines MR Knight TE McNerney KO Leick MB Jain T Ahmed S Immune Effector Cell-Associated Hemophagocytic Lymphohistiocytosis-Like Syndrome Transpl Cell Ther 2023 29 7 438.e1 438.e16 10.1016/j.jtct.2023.03.006 PMC10330221 36906275 Hines MR, Knight TE, McNerney KO, Leick MB, Jain T, Ahmed S, et al. Immune Effector Cell-Associated Hemophagocytic Lymphohistiocytosis-Like Syndrome. Transpl Cell Ther. 2023;29(7):438.e1-438.e16 10.1016/j.jtct.2023.03.006 PMC10330221 36906275

6. Freyer CW Porter DL Cytokine release syndrome and neurotoxicity following CAR T-cell therapy for hematologic malignancies J Allergy Clin Immunol 2020 146 5 940 8 10.1016/j.jaci.2020.07.025 32771558 Freyer CW, Porter DL. Cytokine release syndrome and neurotoxicity following CAR T-cell therapy for hematologic malignancies. J Allergy Clin Immunol. 2020;146(5):940–8. 32771558 10.1016/j.jaci.2020.07.025

7. Berdeja JG Madduri D Usmani SZ Jakubowiak A Agha M Cohen AD Ciltacabtagene autoleucel, a B-cell maturation antigen-directed chimeric antigen receptor T-cell therapy in patients with relapsed or refractory multiple myeloma (CARTITUDE-1): a phase 1b/2 open-label study Lancet 2021 398 10297 314 24 10.1016/S0140-6736(21)00933-8 34175021 Berdeja JG, Madduri D, Usmani SZ, Jakubowiak A, Agha M, Cohen AD, et al. Ciltacabtagene autoleucel, a B-cell maturation antigen-directed chimeric antigen receptor T-cell therapy in patients with relapsed or refractory multiple myeloma (CARTITUDE-1): a phase 1b/2 open-label study. Lancet. 2021;398(10297):314–24. 34175021 10.1016/S0140-6736(21)00933-8

8. Munshi NC Anderson LD Jr Shah N Madduri D Berdeja J Lonial S Idecabtagene Vicleucel in Relapsed and Refractory Multiple Myeloma N Engl J Med 2021 384 8 705 16 10.1056/NEJMoa2024850 33626253 Munshi NC, Anderson LD Jr, Shah N, Madduri D, Berdeja J, Lonial S, et al. Idecabtagene Vicleucel in Relapsed and Refractory Multiple Myeloma. N Engl J Med. 2021;384(8):705–16. 33626253 10.1056/NEJMoa2024850

9. Cordas Dos Santos DM Tix T Shouval R Gafter-Gvili A Alberge JB Cliff ERS A systematic review and meta-analysis of nonrelapse mortality after CAR T cell therapy Nat Med 2024 30 9 2667 78 10.1038/s41591-024-03084-6 38977912 PMC11765209 Cordas Dos Santos DM, Tix T, Shouval R, Gafter-Gvili A, Alberge JB, Cliff ERS, et al. A systematic review and meta-analysis of nonrelapse mortality after CAR T cell therapy. Nat Med. 2024;30(9):2667–78. 38977912 10.1038/s41591-024-03084-6 PMC11765209

10. Rejeski K Subklewe M Aljurf M Bachy E Balduzzi A Barba P Immune effector cell-associated hematotoxicity: EHA/EBMT consensus grading and best practice recommendations Blood 2023 142 10 865 77 10.1182/blood.2023020578 37300386 Rejeski K, Subklewe M, Aljurf M, Bachy E, Balduzzi A, Barba P, et al. Immune effector cell-associated hematotoxicity: EHA/EBMT consensus grading and best practice recommendations. Blood. 2023;142(10):865–77. 37300386 10.1182/blood.2023020578

11. Rejeski K Greco R Onida F Sánchez-Ortega I Bonini C Sureda A An International Survey on Grading, Diagnosis, and Management of Immune Effector Cell-Associated Hematotoxicity (ICAHT) Following CAR T-cell Therapy on Behalf of the EBMT and EHA Hemasphere 2023 7 5 e889 10.1097/HS9.0000000000000889 37125259 PMC10145722 Rejeski K, Greco R, Onida F, Sánchez-Ortega I, Bonini C, Sureda A, et al. An International Survey on Grading, Diagnosis, and Management of Immune Effector Cell-Associated Hematotoxicity (ICAHT) Following CAR T-cell Therapy on Behalf of the EBMT and EHA. Hemasphere. 2023;7(5):e889. 37125259 10.1097/HS9.0000000000000889 PMC10145722

12. Rejeski K Sanz J Fei T Nair MS Hashmi H Avigdor A T-ICAHT: grading and prognostic impact of thrombocytopenia after CAR T-cell therapy Blood 2025 146 7 834 46 10.1182/blood.2025028833 40258181 PMC12783514 Rejeski K, Sanz J, Fei T, Nair MS, Hashmi H, Avigdor A, et al. T-ICAHT: grading and prognostic impact of thrombocytopenia after CAR T-cell therapy. Blood. 2025;146(7):834–46. 40258181 10.1182/blood.2025028833 PMC12783514

13. Ma S Wang Y Qi K Lu W Qi Y Cao J Associations of granulocyte colony-stimulating factor with toxicities and efficacy of chimeric antigen receptor T-cell therapy in relapsed or refractory B-cell acute lymphoblastic leukemia Cancer Immunol Immunother 2024 73 6 104 10.1007/s00262-024-03661-1 38630258 PMC11024067 Ma S, Wang Y, Qi K, Lu W, Qi Y, Cao J, et al. Associations of granulocyte colony-stimulating factor with toxicities and efficacy of chimeric antigen receptor T-cell therapy in relapsed or refractory B-cell acute lymphoblastic leukemia. Cancer Immunol Immunother. 2024;73(6):104. 38630258 10.1007/s00262-024-03661-1 PMC11024067

14. Fried S Avigdor A Bielorai B Meir A Besser MJ Schachter J Early and late hematologic toxicity following CD19 CAR-T cells Bone Marrow Transpl 2019 54 10 1643 50 10.1038/s41409-019-0487-3 30809033 Fried S, Avigdor A, Bielorai B, Meir A, Besser MJ, Schachter J, et al. Early and late hematologic toxicity following CD19 CAR-T cells. Bone Marrow Transpl. 2019;54(10):1643–50. 10.1038/s41409-019-0487-3 30809033

15. Jain T Knezevic A Pennisi M Chen Y Ruiz JD Purdon TJ Hematopoietic recovery in patients receiving chimeric antigen receptor T-cell therapy for hematologic malignancies Blood Adv 2020 4 15 3776 87 10.1182/bloodadvances.2020002509 32780846 PMC7422135 Jain T, Knezevic A, Pennisi M, Chen Y, Ruiz JD, Purdon TJ, et al. Hematopoietic recovery in patients receiving chimeric antigen receptor T-cell therapy for hematologic malignancies. Blood Adv. 2020;4(15):3776–87. 32780846 10.1182/bloodadvances.2020002509 PMC7422135

16. Jin C Chen R Fu S Zhang M Teng Y Yang T Long-term follow-up of BCMA CAR-T cell therapy in patients with relapsed/refractory multiple myeloma J Immunother Cancer 2025 13 3 e010687 10.1136/jitc-2024-010687 40154960 PMC11956354 Jin C, Chen R, Fu S, Zhang M, Teng Y, Yang T, et al. Long-term follow-up of BCMA CAR-T cell therapy in patients with relapsed/refractory multiple myeloma. J Immunother Cancer. 2025;13(3):e010687. 40154960 10.1136/jitc-2024-010687 PMC11956354

17. Remy KE Hall MW Cholette J Juffermans NP Nicol K Doctor A Mechanisms of red blood cell transfusion-related immunomodulation Transfusion 2018 58 3 804 15 10.1111/trf.14488 29383722 PMC6592041 Remy KE, Hall MW, Cholette J, Juffermans NP, Nicol K, Doctor A, et al. Mechanisms of red blood cell transfusion-related immunomodulation. Transfusion. 2018;58(3):804–15. 29383722 10.1111/trf.14488 PMC6592041

18. Shen JG Cheong JH Hyung WJ Kim J Choi SH Noh SH Pretreatment anemia is associated with poorer survival in patients with stage I and II gastric cancer J Surg Oncol 2005 91 2 126 30 10.1002/jso.20272 16028285 Shen JG, Cheong JH, Hyung WJ, Kim J, Choi SH, Noh SH. Pretreatment anemia is associated with poorer survival in patients with stage I and II gastric cancer. J Surg Oncol. 2005;91(2):126–30. 16028285 10.1002/jso.20272

19. Kim SH Lee SI Noh SM Prognostic significance of preoperative blood transfusion in stomach cancer J Gastric Cancer 2010 10 4 196 205 10.5230/jgc.2010.10.4.196 22076186 PMC3204508 Kim SH, Lee SI, Noh SM. Prognostic significance of preoperative blood transfusion in stomach cancer. J Gastric Cancer. 2010;10(4):196–205. 22076186 10.5230/jgc.2010.10.4.196 PMC3204508

20. Sun C Wang Y Yao HS Hu ZQ Allogeneic blood transfusion and the prognosis of gastric cancer patients: systematic review and meta-analysis Int J Surg 2015 13 102 10 10.1016/j.ijsu.2014.11.044 25486261 Sun C, Wang Y, Yao HS, Hu ZQ. Allogeneic blood transfusion and the prognosis of gastric cancer patients: systematic review and meta-analysis. Int J Surg. 2015;13:102–10. 25486261 10.1016/j.ijsu.2014.11.044

21. Mispelbaum R Hattenhauer ST Brossart P Heine A Red blood cell transfusions impact response rates to immunotherapy in patients with solid malignant tumors Front Immunol 2022 13 976011 10.3389/fimmu.2022.976011 36159812 PMC9492841 Mispelbaum R, Hattenhauer ST, Brossart P, Heine A. Red blood cell transfusions impact response rates to immunotherapy in patients with solid malignant tumors. Front Immunol. 2022;13:976011. 36159812 10.3389/fimmu.2022.976011 PMC9492841

22. Vic S Thibert JB Bachy E Cartron G Gastinne T Morschhauser F Transfusion needs after CAR T-cell therapy for large B-cell lymphoma: predictive factors and outcome (a DESCAR-T study) Blood Adv 2024 8 6 1573 85 10.1182/bloodadvances.2023011727 38181767 PMC10982963 Vic S, Thibert JB, Bachy E, Cartron G, Gastinne T, Morschhauser F, et al. Transfusion needs after CAR T-cell therapy for large B-cell lymphoma: predictive factors and outcome (a DESCAR-T study). Blood Adv. 2024;8(6):1573–85. 38181767 10.1182/bloodadvances.2023011727 PMC10982963

23. Xia Y Zhang J Li J Zhang L Li J Fan L Cytopenias following anti-CD19 chimeric antigen receptor (CAR) T cell therapy: a systematic analysis for contributing factors Ann Med 2022 54 1 2951 65 10.1080/07853890.2022.2136748 36382675 PMC9673810 Xia Y, Zhang J, Li J, Zhang L, Li J, Fan L, et al. Cytopenias following anti-CD19 chimeric antigen receptor (CAR) T cell therapy: a systematic analysis for contributing factors. Ann Med. 2022;54(1):2951–65. 36382675 10.1080/07853890.2022.2136748 PMC9673810

24. Lee DW Santomasso BD Locke FL Ghobadi A Turtle CJ Brudno JN ASTCT Consensus Grading for Cytokine Release Syndrome and Neurologic Toxicity Associated with Immune Effector Cells Biol Blood Marrow Transpl 2019 25 4 625 38 10.1016/j.bbmt.2018.12.758 PMC12180426 30592986 Lee DW, Santomasso BD, Locke FL, Ghobadi A, Turtle CJ, Brudno JN, et al. ASTCT Consensus Grading for Cytokine Release Syndrome and Neurologic Toxicity Associated with Immune Effector Cells. Biol Blood Marrow Transpl. 2019;25(4):625–38. 10.1016/j.bbmt.2018.12.758 PMC12180426 30592986

25. Rejeski K Hansen DK Bansal R Sesques P Ailawadhi S Logue JM The CAR-HEMATOTOX score as a prognostic model of toxicity and response in patients receiving BCMA-directed CAR-T for relapsed/refractory multiple myeloma J Hematol Oncol 2023 16 1 88 10.1186/s13045-023-01465-x 37525244 PMC10391746 Rejeski K, Hansen DK, Bansal R, Sesques P, Ailawadhi S, Logue JM, et al. The CAR-HEMATOTOX score as a prognostic model of toxicity and response in patients receiving BCMA-directed CAR-T for relapsed/refractory multiple myeloma. J Hematol Oncol. 2023;16(1):88. 37525244 10.1186/s13045-023-01465-x PMC10391746

26. Cordeiro A Bezerra ED Hirayama AV Hill JA Wu QV Voutsinas J Late Events after Treatment with CD19-Targeted Chimeric Antigen Receptor Modified T Cells Biol Blood Marrow Transpl 2020 26 1 26 33 10.1016/j.bbmt.2019.08.003 PMC6953906 31419568 Cordeiro A, Bezerra ED, Hirayama AV, Hill JA, Wu QV, Voutsinas J, et al. Late Events after Treatment with CD19-Targeted Chimeric Antigen Receptor Modified T Cells. Biol Blood Marrow Transpl. 2020;26(1):26–33. 10.1016/j.bbmt.2019.08.003 PMC6953906 31419568

27. Frenking JH Zhou X Rejeski K Wagner V Costello P Hielscher T Bridging intensity is associated with impaired hematopoietic recovery after BCMA CAR-T therapy for multiple myeloma Blood Adv 2025 9 16 4151 66 10.1182/bloodadvances.2024015732 40267180 PMC12361772 Frenking JH, Zhou X, Rejeski K, Wagner V, Costello P, Hielscher T, et al. Bridging intensity is associated with impaired hematopoietic recovery after BCMA CAR-T therapy for multiple myeloma. Blood Adv. 2025;9(16):4151–66. 40267180 10.1182/bloodadvances.2024015732 PMC12361772

28. Rejeski K Wang Y Hansen DK Iacoboni G Bachy E Bansal R Applying the EHA/EBMT grading for ICAHT after CAR-T: comparative incidence and association with infections and mortality Blood Adv 2024 8 8 1857 68 10.1182/bloodadvances.2023011767 38181508 PMC11007437 Rejeski K, Wang Y, Hansen DK, Iacoboni G, Bachy E, Bansal R, et al. Applying the EHA/EBMT grading for ICAHT after CAR-T: comparative incidence and association with infections and mortality. Blood Adv. 2024;8(8):1857–68. 38181508 10.1182/bloodadvances.2023011767 PMC11007437

29. Rejeski K Perez A Sesques P Hoster E Berger C Jentzsch L CAR-HEMATOTOX: a model for CAR T-cell-related hematologic toxicity in relapsed/refractory large B-cell lymphoma Blood 2021 138 24 2499 513 10.1182/blood.2020010543 34166502 PMC8893508 Rejeski K, Perez A, Sesques P, Hoster E, Berger C, Jentzsch L, et al. CAR-HEMATOTOX: a model for CAR T-cell-related hematologic toxicity in relapsed/refractory large B-cell lymphoma. Blood. 2021;138(24):2499–513. 34166502 10.1182/blood.2020010543 PMC8893508

30. Juluri KR Wu QV Voutsinas J Hou J Hirayama AV Mullane E Severe cytokine release syndrome is associated with hematologic toxicity following CD19 CAR T-cell therapy Blood Adv 2022 6 7 2055 68 10.1182/bloodadvances.2020004142 34666344 PMC9006285 Juluri KR, Wu QV, Voutsinas J, Hou J, Hirayama AV, Mullane E, et al. Severe cytokine release syndrome is associated with hematologic toxicity following CD19 CAR T-cell therapy. Blood Adv. 2022;6(7):2055–68. 34666344 10.1182/bloodadvances.2020004142 PMC9006285

31. Davis JA McGann M Marini J Hashmi H How to manage prolonged immune effector cell-associated hematotoxicity (ICAHT) related to BCMA-directed myeloma therapy Expert Rev Anticancer Ther 2024 24 6 335 8 10.1080/14737140.2024.2344650 38626305 Davis JA, McGann M, Marini J, Hashmi H. How to manage prolonged immune effector cell-associated hematotoxicity (ICAHT) related to BCMA-directed myeloma therapy. Expert Rev Anticancer Ther. 2024;24(6):335–8. 38626305 10.1080/14737140.2024.2344650

32. Davis JA Dima D Ahmed N DeJarnette S McGuirk J Jia X Impact of Frailty on Outcomes after Chimeric Antigen Receptor T Cell Therapy for Patients with Relapsed/Refractory Multiple Myeloma Transpl Cell Ther 2024 30 3 298 305 10.1016/j.jtct.2023.12.015 38142943 Davis JA, Dima D, Ahmed N, DeJarnette S, McGuirk J, Jia X, et al. Impact of Frailty on Outcomes after Chimeric Antigen Receptor T Cell Therapy for Patients with Relapsed/Refractory Multiple Myeloma. Transpl Cell Ther. 2024;30(3):298–305. 10.1016/j.jtct.2023.12.015 38142943

33. Kampouri E Little JS Rejeski K Manuel O Hammond SP Hill JA Infections after chimeric antigen receptor (CAR)-T-cell therapy for hematologic malignancies Transpl Infect Dis 2023 25 Suppl 1 e14157 10.1111/tid.14157 37787373 Kampouri E, Little JS, Rejeski K, Manuel O, Hammond SP, Hill JA. Infections after chimeric antigen receptor (CAR)-T-cell therapy for hematologic malignancies. Transpl Infect Dis. 2023;25(Suppl 1):e14157. 37787373 10.1111/tid.14157

34. Liu J Li Y Zhu F Lv B Xu J Du M Prior antibiotics exposure predicts early and prolonged CD19 CAR T-cell-related hematologic toxicity and prognosis in acute B-cell leukemia Bone Marrow Transpl 2025 60 6 902 5 10.1038/s41409-025-02578-2 PMC12151845 40195552 Liu J, Li Y, Zhu F, Lv B, Xu J, Du M, et al. Prior antibiotics exposure predicts early and prolonged CD19 CAR T-cell-related hematologic toxicity and prognosis in acute B-cell leukemia. Bone Marrow Transpl. 2025;60(6):902–5. 10.1038/s41409-025-02578-2 PMC12151845 40195552

35. Yin L Lv B Ge J Qi Y Xia J Ma S The impact of antibiotic use on outcomes of relapsed/refractory multiple myeloma patients treated with CAR-T therapy Front Immunol 2025 16 1566016 10.3389/fimmu.2025.1566016 40313953 PMC12043596 Yin L, Lv B, Ge J, Qi Y, Xia J, Ma S, et al. The impact of antibiotic use on outcomes of relapsed/refractory multiple myeloma patients treated with CAR-T therapy. Front Immunol. 2025;16:1566016. 40313953 10.3389/fimmu.2025.1566016 PMC12043596

36. Smith M Dai A Ghilardi G Amelsberg KV Devlin SM Pajarillo R Gut microbiome correlates of response and toxicity following anti-CD19 CAR T cell therapy Nat Med 2022 28 4 713 23 10.1038/s41591-022-01702-9 35288695 PMC9434490 Smith M, Dai A, Ghilardi G, Amelsberg KV, Devlin SM, Pajarillo R, et al. Gut microbiome correlates of response and toxicity following anti-CD19 CAR T cell therapy. Nat Med. 2022;28(4):713–23. 35288695 10.1038/s41591-022-01702-9 PMC9434490

37. Stein-Thoeringer CK Saini NY Zamir E Blumenberg V Schubert ML Mor U A non-antibiotic-disrupted gut microbiome is associated with clinical responses to CD19-CAR-T cell cancer immunotherapy Nat Med 2023 29 4 906 16 10.1038/s41591-023-02234-6 36914893 PMC10121864 Stein-Thoeringer CK, Saini NY, Zamir E, Blumenberg V, Schubert ML, Mor U, et al. A non-antibiotic-disrupted gut microbiome is associated with clinical responses to CD19-CAR-T cell cancer immunotherapy. Nat Med. 2023;29(4):906–16. 36914893 10.1038/s41591-023-02234-6 PMC10121864

38. Liang EC Rejeski K Fei T Albittar A Huang JJ Portuguese AJ Development and validation of an automated computational approach to grade immune effector cell-associated hematotoxicity Bone Marrow Transpl 2024 59 7 910 7 10.1038/s41409-024-02278-3 PMC11380036 38627450 Liang EC, Rejeski K, Fei T, Albittar A, Huang JJ, Portuguese AJ, et al. Development and validation of an automated computational approach to grade immune effector cell-associated hematotoxicity. Bone Marrow Transpl. 2024;59(7):910–7. 10.1038/s41409-024-02278-3 PMC11380036 38627450

39. Baumhardt TM Amoah A Hoenicka M Liebold A Sakk V Soller K Functional and molecular analyses reveal impaired HSPCs in Multiple Myeloma patients post-induction Stem Cells Transl Med 2025 14 11 szaf061 10.1093/stcltm/szaf061 41259784 PMC12629531 Baumhardt TM, Amoah A, Hoenicka M, Liebold A, Sakk V, Soller K, et al. Functional and molecular analyses reveal impaired HSPCs in Multiple Myeloma patients post-induction. Stem Cells Transl Med. 2025;14(11):szaf061. 41259784 10.1093/stcltm/szaf061 PMC12629531

40. Chung K Ng LS Yee CI Bansal R Kharfan-Dabaja MA Nowakowski GS Impact of granulocyte colony-stimulating factor on safety of chimeric antigen receptor-T-cell therapy in non-Hodgkin lymphoma Br J Haematol 2025 207 2 606 11 10.1111/bjh.20207 40524548 Chung K, Ng LS, Yee CI, Bansal R, Kharfan-Dabaja MA, Nowakowski GS, et al. Impact of granulocyte colony-stimulating factor on safety of chimeric antigen receptor-T-cell therapy in non-Hodgkin lymphoma. Br J Haematol. 2025;207(2):606–11. 10.1111/bjh.20207 40524548

41. Sterner RM Sakemura R Cox MJ Yang N Khadka RH Forsman CL GM-CSF inhibition reduces cytokine release syndrome and neuroinflammation but enhances CAR-T cell function in xenografts Blood 2019 133 7 697 709 10.1182/blood-2018-10-881722 30463995 PMC6376281 Sterner RM, Sakemura R, Cox MJ, Yang N, Khadka RH, Forsman CL, et al. GM-CSF inhibition reduces cytokine release syndrome and neuroinflammation but enhances CAR-T cell function in xenografts. Blood. 2019;133(7):697–709. 30463995 10.1182/blood-2018-10-881722 PMC6376281

42. Winget MD Sowell H Shultes K Sawyer J Jayani-Kosarzycki RV Biltibo E Impact of Granulocyte Colony Stimulating Factor Use Following CD-19 Chimeric Antigen Receptor T-Cell Therapy Transpl Cell Ther 2026 32 1 48 58 10.1016/j.jtct.2025.09.017 40953732 Winget MD, Sowell H, Shultes K, Sawyer J, Jayani-Kosarzycki RV, Biltibo E, et al. Impact of Granulocyte Colony Stimulating Factor Use Following CD-19 Chimeric Antigen Receptor T-Cell Therapy. Transpl Cell Ther. 2026;32(1):48–58. 10.1016/j.jtct.2025.09.017 40953732

43. Simon GI Craswell A Thom O Chew MS Anstey CM Fung YL Impacts of Aging on Anemia Tolerance, Transfusion Thresholds, and Patient Blood Management Transfus Med Rev 2019 33 3 154 61 10.1016/j.tmrv.2019.03.001 31129009 Simon GI, Craswell A, Thom O, Chew MS, Anstey CM, Fung YL. Impacts of Aging on Anemia Tolerance, Transfusion Thresholds, and Patient Blood Management. Transfus Med Rev. 2019;33(3):154–61. 31129009 10.1016/j.tmrv.2019.03.001

44. Corona De Lapuerta M Shouval R Alarcon Tomas A Flynn JR Devlin SM Batlevi CL Severe Persistent Cytopenias Following CAR T-Cell Therapy in Patients with Large B-Cell Lymphoma Blood 2022 140 Supplement 1 12770 2 10.1182/blood-2022-163457 Corona De Lapuerta M, Shouval R, Alarcon Tomas A, Flynn JR, Devlin SM, Batlevi CL, et al. Severe Persistent Cytopenias Following CAR T-Cell Therapy in Patients with Large B-Cell Lymphoma. Blood. 2022;140(Supplement 1):12770–2.

45. Rejeski K Perez A Iacoboni G Blumenberg V Bücklein VL Völkl S Severe hematotoxicity after CD19 CAR-T therapy is associated with suppressive immune dysregulation and limited CAR-T expansion Sci Adv 2023 9 38 eadg3919 10.1126/sciadv.adg3919 37738350 PMC10516499 Rejeski K, Perez A, Iacoboni G, Blumenberg V, Bücklein VL, Völkl S, et al. Severe hematotoxicity after CD19 CAR-T therapy is associated with suppressive immune dysregulation and limited CAR-T expansion. Sci Adv. 2023;9(38):eadg3919. 37738350 10.1126/sciadv.adg3919 PMC10516499

46. Palacios-Berraquero ML Rodriguez-Marquez P Calleja-Cervantes ME Berastegui N Zabaleta A Burgos L Molecular mechanisms promoting long-term cytopenia after BCMA CAR-T therapy in multiple myeloma Blood Adv 2024 8 21 5479 92 10.1182/bloodadvances.2023012522 39058976 PMC11532743 Palacios-Berraquero ML, Rodriguez-Marquez P, Calleja-Cervantes ME, Berastegui N, Zabaleta A, Burgos L, et al. Molecular mechanisms promoting long-term cytopenia after BCMA CAR-T therapy in multiple myeloma. Blood Adv. 2024;8(21):5479–92. 39058976 10.1182/bloodadvances.2023012522 PMC11532743

47. Blumberg N Triulzi DJ Heal JM Transfusion-induced immunomodulation and its clinical consequences Transfus Med Rev 1990 4 4 Suppl 1 24 35 10.1016/S0887-7963(90)70239-8 2134638 Blumberg N, Triulzi DJ, Heal JM. Transfusion-induced immunomodulation and its clinical consequences. Transfus Med Rev. 1990;4(4 Suppl 1):24–35. 2134638 10.1016/s0887-7963(90)70239-8

48. Zhu Y Liu K Rosen ST Liu W Zhu H Treatment-related adverse events of chimeric antigen receptor-T therapies for cancers in clinical trials: a systematic review and meta-analysis EClinicalMedicine 2025 84 103267 10.1016/j.eclinm.2025.103267 40687742 PMC12273842 Zhu Y, Liu K, Rosen ST, Liu W, Zhu H. Treatment-related adverse events of chimeric antigen receptor-T therapies for cancers in clinical trials: a systematic review and meta-analysis. EClinicalMedicine. 2025;84:103267. 40687742 10.1016/j.eclinm.2025.103267 PMC12273842