Abstract
Chimeric antigen receptor (CAR) T‐cell therapy has significantly improved outcomes in patients with relapsed/refractory aggressive large B‐cell lymphoma (R/R LBCL). Although thrombosis and bleeding complications have been increasingly reported, their clinical relevance and risk factors remain unclear, also due to heterogeneity between reporting studies. We investigated the incidence of thrombosis and bleeding in a more homogeneous real‐world multicenter cohort by analyzing 250 adults with R/R LBCL who were included in the Dutch “Follow that CAR!” multicenter registry, and all received axicabtagene ciloleucel (axi‐cel) CAR T‐cell therapy. We assessed the impact of thrombosis and bleeding, incorporated as time‐dependent covariates on overall survival and explored associations with cytokine release syndrome (CRS), immune effector cell‐associated neurotoxicity syndrome (ICANS), anticoagulant therapy, and laboratory parameters. One‐year cumulative incidences of thrombosis and bleeding were 6.3% and 11.0%, respectively. Median time to thrombosis or bleeding was 28 days after infusion for both events. Bleeding was associated with poorer overall survival (hazard ratio [HR] 3.77, 95% CI 1.96–7.27). Patients with a CAR‐HEMATOTOX score ≥2 before CAR T‐cell therapy (HR 3.62, 95% CI 1.23–10.69), therapeutic anticoagulation at time of infusion (HR 3.74, 95% CI 1.33–10.49), or thrombocytopenia grade ≥3 (platelet count < 50 × 109/L) at time of infusion (HR 3.55, 95% CI 1.12–11.26, P = 0.03) had a significantly higher risk of bleeding. Our study confirms the high risk and clinical relevance of bleeding in patients receiving CAR T‐cell therapy. Prospective studies are necessary to confirm these results and to guide recommendations regarding modifiable risk factors such as the use of anticoagulation or transfusion thresholds during CAR T‐cell therapy.
BACKGROUND
Chimeric antigen receptor (CAR) T‐cell therapy has revolutionary changed treatment and outcomes of patients with hematological malignancies, particularly those with refractory or relapsed aggressive large B‐cell lymphoma (R/R LBCL). 1 , 2 CAR T‐cell therapy has improved 2‐year progression‐free survival from 16% to 41% compared with second line standard‐care high‐dose (myeloablative) chemotherapy followed by autologous hematopoietic stem cell transplantation (aHSCT). 3 In the third line setting, CD19‐directed CAR T‐cell therapy has also demonstrated high response and improved survival rates in patients with R/R LBCL. 4 Long‐term follow‐up showed sustained and durable responses, 5 , 6 leading to the approval of several CAR T‐cell products for the treatment of R/R LBCL in second‐ and/or third‐line: axicabtagene ciloleucel (axi‐cel, Yescarta®), tisagenlecleucel (tisa‐cel, Kymriah®), and lisocabtagene maraleucel (liso‐cel, Breyanzi®). Clinical trials for other products and targets, as well as for other hematological malignancies and non‐malignant disorders, are ongoing. 7
Cytokine release syndrome (CRS) and immune effector cell‐associated neurotoxicity syndrome (ICANS) are well‐known and frequently reported complications of CAR T‐cell therapy. These are characterized by the release of proinflammatory cytokines, resulting in amplified inflammatory responses. 8 Immune effector cell‐associated hematotoxicity (ICAHT) and infections are other commonly observed complications, and patients at risk of neutropenia or infections can be identified before CAR T‐cell therapy with the CAR‐HEMATOTOX (CAR‐HT) score. 9 Recently, reports on thrombosis after CAR T‐cell therapy have also emerged. Thrombosis has been described in 7%–11% of patients undergoing CAR T‐cell therapy, especially in the first 30 days after infusion. 10 , 11 Another study reported thrombosis (6.7%) as well as bleeding (11.2%) in the first 3 months after CAR T‐cell infusion in LBCL patients. 12 A systematic review and meta‐analysis estimated thrombosis and bleeding incidence rates at 2.6% and 1.9% per person‐month, respectively, in studies with follow‐up durations up to 6 months. 13 Thrombosis rates seem higher than in reports on patients undergoing myeloablative chemotherapy and aHSCT for R/R LBCL (3%–4%), while data on bleeding have not been reported in this setting. 14 , 15
However, most studies involved heterogeneous populations regarding underlying disease or CAR T‐cell product, were of variable quality, and not primarily designed to study thrombosis and bleeding as outcomes in patients receiving CAR T‐cell therapy. Moreover, they applied different event definitions for thrombosis or bleeding and used variable follow‐up durations, which complicates comparisons and reproducibility. Therefore, we investigated the incidence of thrombosis and bleeding in a more homogeneous Dutch multicenter cohort of LBCL patients undergoing CAR T‐cell therapy, assessed their impact on overall survival (OS), and explored associations with anticoagulant therapy, CRS/ICANS, and laboratory parameters.
METHODS
Study design
For this observational study, we used data from the Dutch “Follow that CAR!” registry, a real‐world large multicenter cohort of patients with R/R LBCL undergoing CAR T‐cell therapy in the Netherlands. 16 We included adults aged ≥18 years at the time of CAR T‐cell infusion with R/R LBCL after ≥2 lines of systemic therapy. All patients were treated with standard lymphodepleting chemotherapy (i.e., fludarabine and cyclophosphamide, Day −5 until −3) before infusion of the commercially available CAR T‐cell product axi‐cel (on Day 0) between May 2020 and December 2023. For this analysis, we excluded patients who received CAR T‐cell therapy as part of a clinical trial or for other indications than LBCL. Collaboration was approved by the steering committee of the Dutch National Lymphoma CAR T‐cell Therapy Tumorboard, and data exchange was facilitated according to the existing Data User Agreement of the registry. The study was conducted in accordance with the Declaration of Helsinki and approved by the medical ethical committee of the Erasmus University Medical Center (NL76835.018.21, MEC‐2024‐0283).
Data collection and outcomes
We used available registry data on patient characteristics, medical history, LBCL diagnosis, previous treatment lines, and disease status before lymphodepleting chemotherapy (assessed according to the Lugano classification 17 ). Maximum CRS and ICANS grades (according to the American Society for Transplantation and Cellular Therapy [ASTCT] criteria 18 ), laboratory results (complete blood count and, if available, coagulation parameters) up to 90 days after CAR T‐cell infusion, supportive care measures (blood transfusions, use of granulocyte‐colony stimulating factor [G‐CSF], and/or thrombopoietin receptor [TPO‐R] agonists), disease response after CAR T‐cell therapy, and all‐cause death were also documented. CAR‐HT scores before CAR T‐cell therapy were calculated as previously described 19 , 20 (Table S1), and CAR‐HT scores for the majority of our cohort have been previously reported. 21 CAR‐HT scores for the remaining patients (n = 5) were calculated using the same methodology. High risk HT was defined as CAR‐HT score ≥2, and CAR‐HT score <2 was considered low risk HT. Additionally, we retrospectively extracted data from medical records systematically, using electronic data capture (Castor EDC), on thrombosis (objectively confirmed by imaging) and bleeding events, classified as major bleeding and clinically relevant non‐major bleeding (CRNMB) by the investigators according to the International Society on Thrombosis and Haemostasis (ISTH) criteria. 22 , 23 Presence of a central line (central venous catheter [CVC] or peripherally inserted central catheter [PICC]) and antithrombotic therapy (any type of anticoagulant or antiplatelet therapy) at the time of CAR T‐cell infusion were also noted. Our primary outcomes were thrombosis (venous or arterial) and bleeding (composite of major bleeding and CRNMB). Secondary outcomes were OS and CRS/ICANS. Finally, we explored the impact of the use of anticoagulant therapy and laboratory parameters on the risk of thrombosis and bleeding. Follow‐up duration for all outcomes was up to 1 year after CAR T‐cell infusion.
Statistical analysis
Descriptive statistics were used to summarize patient characteristics at baseline, using frequencies (%) for categorical and medians with range or interquartile range (IQR) for continuous variables. We estimated 1‐year cumulative incidences of thrombosis or bleeding event using first events only, by applying Fine and Gray's subdistribution hazard models treating death as a competing risk. Patients were censored after progression or relapse requiring the start of additional anti‐cancer therapy, or if lost to follow‐up. Furthermore, 1‐year OS, defined as the time from CAR T‐cell infusion until death from any cause, was calculated using the Kaplan–Meier method. Associations between thrombosis, bleeding, or progressive disease (PD) and OS were assessed using extended Cox proportional hazard models with thrombosis, bleeding, or PD incorporated as time‐dependent covariates, and adjusted survival curves were visualized using Simon–Makuch plots. 24 We also applied Fine and Gray's subdistribution hazard models to evaluate the impact of anticoagulant therapy on the incidence of thrombosis or bleeding, using anticoagulant therapy (at time of CAR T‐cell infusion) as a time‐fixed covariate. Associations between laboratory values and thrombosis or bleeding were assessed in four ways: (1) CAR‐HT score before CAR T‐cell therapy; (2) median levels of hemoglobin and platelets at time of CAR T‐cell infusion; (3) presence of anemia grade ≥3 (i.e., hemoglobin level < 4.9 mmol/L) or thrombocytopenia grade ≥3 (i.e., platelet count < 50 × 109/L) at time of CAR T‐cell infusion; and (4) the lowest (nadir) values reached during the first 30 days after infusion. Continuous and categorical variables were evaluated using Mann–Whitney U and chi‐squared tests, respectively. Laboratory measurements after thrombosis or bleeding were excluded to avoid event‐driven bias. Finally, we performed multivariable Cox proportional hazard regression analyses to evaluate associations between clinical or laboratory parameters and thrombosis, bleeding, or all‐cause death. As covariates, we included established risk factors for thrombosis (age, sex, and previous thrombosis), bleeding (anemia grade ≥3, thrombocytopenia grade ≥3, and use of therapeutic anticoagulant therapy; all at time of CAR T‐cell infusion), refractory disease status (defined as stable disease [SD] or PD) and high CAR‐HT score (≥2), both at initiation of lymphodepleting therapy before CAR T‐cell infusion. Results were expressed as adjusted hazard ratios (HRs) with 95% confidence intervals (95% CIs). Statistical tests and computed P values were two‐sided, with P < 0.05 considered statistically significant. All data were analyzed and visualized using R statistical software version 4.5.2 or higher (R Foundation for Statistical Computing).
RESULTS
Baseline patient and treatment characteristics
We included 250 consecutive patients in our study with R/R LBCL who all received axi‐cel therapy (Table 1). Median age was 61 years (IQR 53–67); 83 (33.2%) were female. Most patients were diagnosed with diffuse large B‐cell lymphoma (DLBCL; n = 130, 52.0%), followed by transformed follicular lymphoma (tFL; n = 81, 32.4%), high‐grade B‐cell lymphoma (HGBCL; n = 28, 11.2%), primary mediastinal B‐cell lymphoma (PMBCL; n = 5, 2.0%), and other types of B‐cell lymphoma (n = 6, 2.4%). Median number of prior lines of systemic chemotherapy before CAR T‐cell therapy was 2, and 198 (79.2%) patients had received bridging therapy before CAR T‐cell therapy. Disease status before lymphodepleting chemotherapy was complete or partial remission (CR or PR) in 65 patients (26.0%), SD in 32 (12.8%), and PD in 153 (61.2%). All patients received the same lymphodepleting regimen except for two (bendamustine instead of fludarabine/cyclophosphamide, for toxicity considerations).
Table 1.
Baseline patient and treatment characteristics.
| All patients (n = 250) | |
|---|---|
| Demographics | |
| Median age at CAR T‐cell infusion, years (IQR) | 61 (53–67) |
| Female sex, n (%) | 83 (33.2%) |
| History of thrombosis, n (%) | 79 (31.6%) |
| Venous, n | 62 |
| Arterial, n | 16 |
| Both, n | 1 |
| History of cardiovascular disease, n (%) | 53 (21.2%) |
| Atrial fibrillation, n | 23 |
| Disease characteristics | |
| Lymphoma diagnosis | |
| DLBCL, n (%) | 130 (52.0%) |
| tFL, n (%) | 81 (32.4%) |
| HGBCL, n (%) | 28 (11.2%) |
| PMBCL, n (%) | 5 (2.0%) |
| Other types of BCL, n (%) | 6 (2.4%) |
| Median lines of prior chemotherapy, n (IQR) | 2 (2–2) |
| Bridging therapy before CAR T‐cell therapy, n (%) | 198 (79.2%) |
| Disease status before lymphodepleting therapy | |
| CR/PR, n (%) | 65 (26.0%) |
| SD, n (%) | 32 (12.8%) |
| PD, n (%) | 153 (61.2%) |
| CAR T‐cell therapy | |
| Central line inserted, n (%) | 241 (96.4%) |
| CVC, n (%) | 146 (60.6%) |
| Tunneled, n | 33 |
| Non‐tunneled, n | 111 |
| PICC, n (%) | 95 (39.4%) |
| Lymphodepleting chemotherapy | |
| Fludarabine/cyclophosphamide, n (%) | 248 (99.2%) |
| Bendamustine, n (%) | 2 (0.8%) |
| Antithrombotic therapy | |
| Total, n (%) | 115 (46.0%) |
| Therapeutic anticoagulation, n (%) | 63 (25.2%) |
| Heparin (LMWH or UFH), n | 52 |
| DOAC, n | 10 |
| VKA, n | 1 |
| Prophylactic LMWH, n (%) | 37 (14.8%) |
| Antiplatelet therapy, n (%) | 15 (6.0%) |
Abbreviations: BCL, B‐cell lymphoma; CAR, chimeric antigen receptor; CR, complete remission; CVC, central venous catheter; DLBCL, diffuse large B‐cell lymphoma; DOAC, direct oral anticoagulant; HGBCL, high‐grade B‐cell lymphoma; IQR, interquartile range; LMWH, low‐molecular‐weight heparin; PD, progressive disease; PICC, peripherally inserted central catheter; PMBCL, primary mediastinal B‐cell lymphoma; PR, partial remission; SD, stable disease; tFL, transformed follicular lymphoma; UFH, unfractionated heparin; VKA, vitamin K antagonist.
A history of thrombosis was present in 79 (31.6%) patients (62 venous, 16 arterial, and 1 both), while 53 (21.2%) had a history of cardiovascular disease (23 atrial fibrillation). A central line was present at CAR T‐cell infusion in 241 (96.4%) patients; 146 CVCs (of which 33 tunneled) and 95 PICCs. Antithrombotic treatment was used in 115 (46.0%) patients at the time of CAR T‐cell infusion; 100 received anticoagulation, and 15 antiplatelet therapy. In total, 63 (25.2%) patients received therapeutic anticoagulation (51 with low‐molecular‐weight heparin and 1 with unfractionated heparin; 10 direct oral anticoagulants; and 1 vitamin K antagonists). Indications for therapeutic anticoagulation were a recent thrombosis (n = 46), atrial fibrillation (n = 13), other cardiovascular disease (n = 2), thrombophlebitis (n = 1), or unknown (n = 1). In total, 37 (14.8%) patients were treated with prophylactic dose LMWH, of whom 7 with previous thrombosis. In total, 15 (6.0%) patients received antiplatelet therapy, for a previous thrombosis (n = 7), cardiovascular disease (n = 7), or essential thrombocythemia (n = 1).
Clinical follow‐up characteristics and survival
Median duration of hospital admission after CAR T‐cell infusion was 14 days (range 4–78; Table 2). In total, 237 (94.8%) patients developed CRS grade ≥1, with grade ≥2 in 124 (49.6%). ICANS grade ≥1 occurred in 160 (64.0%) patients, with grade ≥2 in 99 (39.6%). In total, 113 (45.2%) and 81 (32.4%) patients received at least one red blood cell or platelet transfusion after CAR T‐cell infusion, respectively. G‐CSF was prescribed to 112 (44.8%) patients, with a median time to initiation of 36 days (IQR 28–64) following CAR T‐cell infusion and for a median duration of 57 days (IQR 28–152). TPO‐R agonists were started in 25 (10.0%) patients, with a median time to initiation of 28 days (IQR 18–37) following CAR T‐cell infusion and for a median duration of 178 days (IQR 77–240).
Table 2.
Follow‐up characteristics of chimeric antigen receptor (CAR) T‐cell therapy.
| All patients (n = 250) | |
|---|---|
| Median duration of hospitalization after CAR T‐cell infusion, days (range) | 14 (4–78) |
| Toxicities | |
| CRS | |
| Any, n (%) | 237 (94.8%) |
| Grade ≥2, n (%) | 124 (49.6%) |
| Median time to CRS after infusion, days (IQR) | 2 (1–3) |
| Median duration of symptoms, days (IQR) | 5 (3–7) |
| ICANS | |
| Any, n (%) | 160 (64.0%) |
| Grade ≥2, n (%) | 99 (39.6%) |
| Median time to ICANS after infusion, days (IQR) | 6 (4–8) |
| Median duration of symptoms, days (IQR) | 5 (2–9) |
| Supportive care | |
| Patients with ≥1 red blood cell transfusions, n (%) | 113 (45.2%) |
| Patients with ≥1 platelet transfusions, n (%) | 81 (32.4%) |
| Patients receiving G‐CSF, n (%) | 112 (44.8%) |
| Median time to initiation, days (IQR) | 36 (28–64) |
| Median duration, days (IQR) | 57 (28–152) |
| Patients receiving TPO‐R agonist, n (%) | 25 (10.0%) |
| Median time to initiation, days (IQR) | 28 (18–37) |
| Median duration, days (IQR) | 178 (77–240) |
Abbreviations: CRS, cytokine release syndrome; G‐CSF, granulocyte‐colony stimulating factor; ICANS, immune effector cell‐associated neurotoxicity syndrome; IQR, interquartile range; TPO‐R, thrombopoietin receptor.
Median follow‐up duration after CAR T‐cell infusion was 305 days (range 14–366). After a 1‐year follow‐up, 111 (44.4%) patients had PD, of whom 75 received additional anti‐cancer therapy. In total, 91 (36.4%) patients had deceased, translating into a 1‐year OS of 75.3% (95% CI 69.5–81.5; Figure S1). Most patients (65.6%) deceased of PD after CAR T‐cell therapy, with a statistically significant association between PD after CAR T‐cell therapy and OS (8.8% vs. 75.8%; HR 7.85, 95% CI 4.92–12.50, P < 0.001).
Incidence of thrombosis and bleeding
During 1‐year follow‐up after CAR T‐cell infusion, 17 patients experienced a first thrombosis; 11 venous thromboembolism (VTE, including 4 catheter‐associated thrombosis) and 6 arterial thrombosis (Table 3A), translating into a 1‐year cumulative incidence of 6.3% (95% CI 3.7–10.0; Figure 1). Median time to first thrombosis was 28 days (IQR 9–71) after CAR T‐cell infusion. Of the patients with thrombosis, 15 (88%) experienced CRS grade ≥1, with grade ≥2 in 9 (60%), while 11 patients (65%) experienced ICANS grade ≥1, with grade ≥2 in 9 (82%); all preceding the thrombosis events. In patients without thrombosis, 52% had CRS grade ≥2, while 60% had ICANS grade ≥2. In total, 29 patients experienced at least one bleeding event during 1‐year follow‐up; 12 major bleedings of which 1 was a fatal retroperitoneal bleed, and 17 CRNMB (Table 3B), translating into a 1‐year cumulative incidence of 11.0% (95% CI 7.4–15.5; Figure 1). Median time to first bleed (major or CRNMB) was 28 days (IQR 14–99) after CAR T‐cell infusion. Most bleeding events occurred in the gastrointestinal tract (34.5%), and 5 (17.2%) were lymphoma‐related, all with gastrointestinal localization. Of the patients with bleeding, 27 (93%) experienced CRS grade ≥1, with grade ≥2 in 16 (59%), while 20 (69%) experienced ICANS grade ≥1, with grade ≥2 in 16 (80%); all preceding the bleeding events. In patients without bleeding, 51% had CRS grade ≥2 and 59% ICANS grade ≥2.
Table 3A.
Overview of patients with a first thrombosis (n = 17).
| Age | Sex | Previous thrombosis | Diagnosis | Antithrombotic therapy | Time to event (days) | Type of thrombosis | Location | Maximum CRS grade | Maximum ICANS grade |
|---|---|---|---|---|---|---|---|---|---|
| 64 | M | None | tFL | None | 2 | VTE | lDVT | 1 | 1 |
| 71 | M | None | HGBCL | ASA | 3 | VTE | PE | 2 | 3 |
| 68 | M | VTE | DLBCL | pLMWH | 8 | VTE | PE | 2 | 2 |
| 56 | M | VTE | tFL | tLMWH | 9 | AT | TIA | 1 | 2 |
| 63 | F | None | DLBCL | pLMWH | 9 | VTE | uDVT | 2 | 3 |
| 64 | M | None | DLBCL | None | 14 | AT | TIA | 2 | 2 |
| 61 | F | None | tFL | None | 14 | VTE | uDVT | 2 | 1 |
| 69 | M | None | tFL | None | 28 | VTE | uDVT | 1 | 0 |
| 46 | M | None | tFL | None | 28 | VTE | PE | 1 | 0 |
| 70 | M | None | DLBCL | None | 30 | VTE | PE | 2 | 3 |
| 65 | M | VTE | tFL | None | 45 | AT | TIA | 2 | 2 |
| 51 | M | None | HGBCL | None | 71 | VTE | uDVT | 1 | 2 |
| 72 | M | None | DLBCL | None | 108 | AT | MCI | 2 | 0 |
| 50 | M | VTE | tFL | None | 111 | AT | CVA | 1 | 2 |
| 61 | M | None | tFL | None | 114 a | VTE | Other | 0 | 0 |
| 68 | M | AT | DLBCL | ASA | 182a | VTE | PE | 0 | 0 |
| 69 | M | AT | tFL | tLMWH | 246 | AT | PAD | 2 | 0 |
Abbreviations: ASA, acetylsalicylic acid; AT, arterial thrombosis; CRS, cytokine release syndrome; CVA, cerebrovascular accident; DLBCL, diffuse large B‐cell lymphoma; F, female; HGBCL, high‐grade B‐cell lymphoma; ICANS, immune effector cell‐associated neurotoxicity syndrome; lDVT, lower extremity deep venous thrombosis; M, male; MCI, myocardial infarction; PAD, peripheral arterial disease; PE, pulmonary embolism; pLMWH, prophylactic dose low‐molecular‐weight heparin; tFL, transformed follicular lymphoma; TIA, transient ischemic attack; tLMWH, therapeutic dose low‐molecular‐weight heparin; uDVT, upper extremity deep venous thrombosis (note: all catheter‐related); VTE, venous thromboembolism.
Patients with events after progression/relapse requiring the start of additional anti‐cancer therapy.
Figure 1.
One‐year cumulative incidences of thrombosis and bleeding. +, censored.
Table 3B.
Overview of patients with bleeding (n = 29).
| Age | Sex | Previous thrombosis | Diagnosis | Antithrombotic therapy | Time to event (days) | Type of bleeding (ISTH criteria) | Location | Maximum CRS grade | Maximum ICANS grade |
|---|---|---|---|---|---|---|---|---|---|
| 20 | F | VTE | PMBCL | tLMWH | 1 | CRNMB | Genitourinary | 2 | 0 |
| 51 | F | VTE | DLBCL | tLMWH | 5 | CRNMB | Genitourinary | 2 | 3 |
| 51 | M | None | HGBCL | None | 8 | CRNMB | Cerebral | 1 | 2 |
| 50 | M | None | HGBCL | None | 10 | MB (non‐fatal) | Gastrointestinal | 1 | 0 |
| 63 | M | VTE | DLBCL | tLMWH | 11 | MB (non‐fatal) | Gastrointestinal | 3 | 3 |
| 66 | M | None | DLBCL | pLMWH | 11 | CRNMB | Gastrointestinala | 3 | 4 |
| 71 | M | VTE | DLBCL | tLMWH | 12 | MB (non‐fatal) | Gastrointestinal a | 2 | 4 |
| 71 | M | None | HGBCL | ASA | 14 | MB (fatal) | Retroperitoneal | 2 | 3 |
| 73 | M | None | HGBCL | None | 14 | CRNMB | Genitourinary | 2 | 2 |
| 67 | M | None | DLBCL | tLMWH | 16 | CRNMB | Gastrointestinal | 2 | 1 |
| 48 | M | VTE | tFL | DOAC | 20 | CRNMB | Genitourinary | 0 | 0 |
| 65 | M | AT | tFL | tLMWH | 21 | MB (non‐fatal) | Epistaxis | 1 | 3 |
| 61 | F | None | DLBCL | None | 27 | CRNMB | Subcutaneous | 2 | 3 |
| 68 | F | None | tFL | pLMWH/UFHb | 28 | CRNMB | Subcutaneous | 2 | 4 |
| 73 | F | None | DLBCL | None | 28 | CRNMB | PICC bleeding | 1 | 2 |
| 46 | M | None | DLBCL | pLMWH | 33 | MB (non‐fatal) | Gastrointestinal | 1 | 3 |
| 76 | M | VTE | DLBCL | tLMWH | 43 | MB (non‐fatal) | Genitourinary | 2 | 1 |
| 63 | M | None | tFL | None | 45 | MB (non‐fatal) | Gastrointestinal a | 0 | 0 |
| 53 | F | VTE | DLBCL | pLMWH | 53 | CRNMB | Genitourinary | 1 | 0 |
| 69 | F | None | DLBCL | None | 57 | MB (non‐fatal) | Subcutaneous | 1 | 1 |
| 41 | F | VTE | PMBCL | None | 73 | CRNMB | Mucosal | 2 | 2 |
| 49 | F | None | DLBCL | tLMWH | 99 | MB (non‐fatal) | Cerebral | 1 | 2 |
| 70 | M | None | DLBCL | None | 148 | MB (non‐fatal) | Gastrointestinal | 2 | 3 |
| 73 | M | None | DLBCL | tLMWH/DOACc | 184 | CRNMB | Subungual | 1 | 0 |
| 35 | F | None | DLBCL | None | 194 | CRNMB | Subcutaneous | 2 | 3 |
| 36 | M | VTE | DLBCL | None | 205d | CRNMB | Epistaxis | 2 | 0 |
| 60 | M | AT | DLBCL | tLMWH | 248 | CRNMB | Gastrointestinal a | 1 | 0 |
| 21 | M | None | HGBCL | None | 250 d | CRNMB | Genitourinary | 1 | 1 |
| 49 | M | None | DLBCL | None | 250 d | MB (non‐fatal) | Gastrointestinal a | 2 | 0 |
Abbreviations: ASA, acetylsalicylic acid; AT, arterial thrombosis; CRNMB, clinically relevant non‐major bleeding; CRS, cytokine release syndrome; DLBCL, diffuse large B‐cell lymphoma; DOAC, direct oral anticoagulant; F, female; HGBCL, high‐grade B‐cell lymphoma; ICANS, immune effector cell‐associated neurotoxicity syndrome; ISTH, International Society of Thrombosis and Haemostasis; M, male; MB, major bleeding; PICC, peripheral inserted central venous catheter; pLMWH, prophylactic dose low‐molecular‐weight heparin; PMBCL, primary mediastinal B‐cell lymphoma; tFL, transformed follicular lymphoma; tLMWH, therapeutic dose low‐molecular‐weight heparin; UFH, unfractionated heparin; VTE, venous thromboembolism.
Intratumoral bleeding events.
Patient switched from pLMWH to UFH after infusion (Day +14) because of atrial fibrillation and was on UFH during bleeding event.
Patient started tLMWH/DOAC after infusion (Day +5) because of atrial fibrillation, treated with DOAC during bleeding event.
Patients with events after progression/relapse requiring start of additional anti‐cancer therapy.
Thrombosis or bleeding and overall survival
Patients who experienced thrombosis had an OS of 52.2% versus 74.6% for those without thrombosis; this difference was not statistically significant (HR 1.82, 95% CI 0.73–4.58, P = 0.20; Figure 2A). Patients with bleeding had a statistically significant poorer OS compared to patients without bleeding (34.5% vs. 74.8%; HR 3.77, 95% CI 1.96–7.27, P < 0.001; Figure 2B). Subgroup analysis by bleeding type showed that this association was mainly driven by major bleeding (10.7% vs. 75.2%; HR 6.90, 95% CI 2.95–16.16, P < 0.001; Figure S2A), whereas no statistically significant difference was observed for patients with CRNMB only (50.8% vs. 74.9%; HR 2.37, 95% CI 0.94–5.95, P = 0.07; Figure S2B). 13/29 (44.8%) patients with bleeding also had PD, versus 98/221 (44.3%) without bleeding.
Figure 2.
One‐year overall survival of patients with thrombosis (A) or bleeding (B). 95% CI, 95% confidence interval; +, censored; HR, hazard ratio.
Anticoagulant therapy and thrombosis or bleeding
Thrombosis incidence was lower among anticoagulated patients, but there was no statistically significant association between use of any anticoagulation (prophylactic or therapeutic) at the time of CAR T‐cell infusion and cumulative incidence of thrombosis (4.4% vs. 7.6%; HR 0.54, 95% CI 0.17–1.70, P = 0.29). Although bleeding incidence was higher among anticoagulated patients (15.9% vs. 7.9%), there was no statistically significant association between use of any anticoagulation at the time of CAR T‐cell therapy and cumulative incidence of bleeding (HR 2.15, 95% CI 0.97–4.75, P = 0.06). When analyzing prophylactic anticoagulation at the time of CAR T‐cell infusion separately, there were also no statistically significant associations with cumulative incidence of thrombosis (5.4% vs. 7.6%; HR 0.73, 95% CI 0.16–3.41, P = 0.69) or bleeding (8.1% vs. 7.9%; HR 1.08, 95% CI 0.30–3.93, P = 0.90). For patients using therapeutic anticoagulation at time of CAR T‐cell infusion, there was no statistically significant association with cumulative incidence of thrombosis (4.0% vs. 7.6%; HR 0.43, 95% CI 0.10–1.93, P = 0.27; Figure 3A), but bleeding incidence was significantly higher (21.1% vs. 7.9%; HR 2.87, 95% CI 1.25–6.60, P = 0.01; Figure 3B). Subgroup analyses for therapeutic anticoagulation and major bleeding (8.4% vs. 3.2%; HR 2.89, 95% CI 0.78–10.64, P = 0.11; Figure S3A), or CRNMB (12.7% vs. 4.7%; HR 2.67, 95% CI 0.91–7.87, P = 0.08; Figure S3B) individually did not reach statistical significance.
Figure 3.
One‐year cumulative incidences of thrombosis (A) or bleeding (B), by use of antithrombotic therapy at time of chimeric antigen receptor (CAR) T‐cell infusion. 95% CI, 95% confidence interval; +, censored; HR, hazard ratio.
Laboratory parameters and thrombosis or bleeding
Table 4 summarizes continuous and binary CAR‐HT scores before CAR T‐cell therapy, hemoglobin and platelet levels, anemia grade ≥3, and thrombocytopenia grade ≥3 at the time of CAR T‐cell infusion, and nadirs of hemoglobin and platelet levels reached during the first 30 days after CAR T‐cell infusion. CAR‐HT scores could be calculated for 216/250 patients (86.4%). Median CAR‐HT score was 1 (IQR 0–2), and CAR‐HT score was ≥2 in 98 (39.2%) patients. There was no difference in CAR‐HT scores for patients with and without thrombosis, but a CAR‐HT score ≥ 2 was more common in patients with subsequent bleeding (55.2% vs. 37.1%; P = 0.04). Median hemoglobin level at infusion was 6.2 mmol/L (IQR 5.5–6.9); levels were not significantly different between patients with and without thrombosis, but patients with subsequent bleeding had lower median hemoglobin levels (5.8 mmol/L [IQR 5.4–6.3] vs. 6.2 mmol/L [IQR 5.5–6.9]; P = 0.03). Anemia grade ≥3 was present in 18 (7.2%) patients at CAR T‐cell infusion, and nadir hemoglobin level reached during the first 30 days after infusion was 5.5 mmol/L (IQR 4.8–6.1); for both, no significant differences were found between patients with thrombosis or bleeding and those without. Median platelet count at CAR T‐cell infusion was 144 × 109/L (IQR 97–187), and thrombocytopenia grade ≥3 was observed in 21 (8.4%) patients; there were no significant differences between patients with thrombosis or bleeding and those without. Nadir platelet count reached during the first 30 days was 63 × 109/L (IQR 26–115). Levels did not differ significantly between patients with and without thrombosis, but nadir platelet count was lower in patients with bleeding in comparison to those without (27 × 109/L [IQR 11–102] vs. 61 × 109/L [IQR 25–113]; P = 0.02). Coagulation parameters were only available in 83 (33.6%) patients, which was insufficient to evaluate associations with thrombosis or bleeding.
Table 4.
Laboratory parameters.
| Total | Thrombosis | Bleeding | |||||
|---|---|---|---|---|---|---|---|
| Yes (n = 17) | No (n = 233) | P‐value | Yes (n = 29) | No (n = 221) | P‐value | ||
| Before CAR T‐cell infusion | |||||||
| CAR‐HEMATOTOX (CAR‐HT) score | |||||||
| Continuous, median (IQR) | 1 (0–2) | 1 (1–3) | 1 (0–2) | 0.54 | 2 (1–3) | 1 (0–2) | 0.07 |
| CAR‐HT score ≥ 2, n (%) | 98 (39.2%) | 9 (52.9%) | 109 (46.8%) | 1 | 16 (55.2%) | 82 (37.1%) | 0.04 |
| At time of CAR T‐cell infusion | |||||||
| Hb, mmol/L (IQR) | 6.2 (5.5–6.9) | 6.4 (5.7–6.7) | 6.2 (5.5–6.9) | 0.86 | 5.8 (5.4–6.3) | 6.2 (5.5–6.9) | 0.03 |
| Anemia grade ≥3 (%)a | 18 (7.2%) | 1 (5.9%) | 17 (7.3%) | 1 | 4 (13.8%) | 14 (6.3%) | 0.30 |
| Platelets, ×109/L (IQR) | 144 (97–187) | 145 (103–172) | 140 (90–186) | 0.87 | 130 (69–186) | 139 (93–186) | 0.37 |
| Thrombocytopenia grade ≥3 (%)b | 21 (8.4%) | 2 (11.8%) | 19 (8.2%) | 0.91 | 5 (17.2%) | 16 (7.2%) | 0.16 |
| Nadir within first 30 days after CAR T‐cell infusion | |||||||
| Hb, mmol/L (IQR) | 5.5 (4.8–6.1) | 5.5 (5.1–6.5) | 5.5 (4.8–6.1) | 0.66 | 5.5 (4.6–5.7) | 5.5 (4.8–6.1) | 0.14 |
| Median time to nadir, days (IQR) | 5 (1–10) | 1 (0–5) | 5 (1–10) | 0.11 | 1 (0–9) | 5 (1–9) | 0.29 |
| Platelets, ×109/L (IQR) | 63 (26–115) | 93 (54–138) | 61 (24–115) | 0.11 | 27 (11–102) | 61 (25–113) | 0.02 |
| Median time to nadir, days (IQR) | 10 (5–28) | 5 (1–5) | 10 (5–28) | <0.001 | 5 (5–26) | 10 (5–28) | 0.64 |
Abbreviations: CAR, chimeric antigen receptor; Hb, hemoglobin level; IQR, interquartile range.
Hb level < 4.9 mmol/L.
Platelet count < 50 × 109/L.
Multivariable analyses for thrombosis, bleeding, or all‐cause death
None of the selected variables were significantly associated with thrombosis using multivariable analysis (Figure 4A). However, CAR‐HT score ≥ 2 before CAR T‐cell therapy (HR 3.62, 95% CI 1.23–10.69, P = 0.02), use of therapeutic anticoagulation at time of CAR T‐cell infusion (HR 3.74, 95% CI 1.33–10.49, P = 0.01) and thrombocytopenia grade ≥3 at time of CAR T‐cell infusion (HR 3.55, 95% CI 1.12–11.26, P = 0.03) were significantly associated with bleeding (Figure 4B). Finally, CAR‐HT score ≥2 before CAR T‐cell therapy (HR 8.40, 95% CI 3.31–21.32, P < 0.001) was also significantly associated with all‐cause death (Figure 4C).
Figure 4.
Multivariable Cox proportional hazards regression analysis for thrombosis (A), bleeding (B), or all‐cause death (C). 95% CI, 95% confidence interval; CAR‐HT, CAR‐HEMATOTOX; HR, hazard ratio.
DISCUSSION
In this observational study, we investigated the incidence of thrombosis and bleeding in a large real‐world multicenter cohort of 250 patients with R/R LBCL receiving treatment with commercially available CAR T‐cell therapy after standard lymphodepleting chemotherapy in the Netherlands. We found a relatively high 1‐year cumulative incidence of thrombosis (6.3%) and especially bleeding (11.0%) in our cohort, with events occurring predominantly in the first 30 days after CAR T‐cell infusion. Bleeding was associated with poorer OS. CAR‐HT score ≥2 before CAR T‐cell therapy, use of therapeutic anticoagulation at the time of CAR T‐cell infusion, and thrombocytopenia grade ≥3 at the time of CAR T‐cell infusion were associated with a higher risk of bleeding. Our study confirms the high risk and clinical relevance of especially bleeding during CAR T‐cell therapy, and the importance of evaluating potentially modifiable risk factors, such as anticoagulant therapy, thrombocytopenia, or anemia, in patients undergoing CAR T‐cell therapy.
Coagulopathies including thrombosis and bleeding are increasingly recognized as complications after CAR T‐cell therapy. It is likely that previous CAR T‐cell studies have underreported these complications as they were not predefined as primary or secondary study outcomes. 13 The incidence of thrombosis in our cohort is in line with previous studies involving patients receiving axi‐cel therapy for LBCL. 10 , 11 , 12 Some studies describing lower thrombosis rates excluded catheter‐related or arterial events, 25 or investigated patients treated with tisa‐cel instead of axi‐cel. 26 , 27 Although direct comparative studies are lacking, the incidence of thrombosis in patients with LBCL treated with CAR T‐cell therapy seems higher than in historical cohorts treated with myeloablative chemotherapy followed by aHSCT (~1%). 14 , 15 , 28
Bleeding has been reported less than thrombosis in patients receiving CAR T‐cell therapy, but our study highlights bleeding as a more frequent and clinically relevant complication, systematically assessed using the ISTH criteria. Bleeding incidence in our cohort was in line with one previously published report, 12 but higher than in other studies reporting on bleeding in CAR T‐cell patients. 26 , 27 These studies generally lacked standardized bleeding definitions, potentially causing underreporting of events and explaining lower reported incidences. Importantly, we showed that bleeding, incorporated as a time‐dependent covariate, was significantly associated with poorer OS (HR 3.77, 95% CI 1.96–7.27). An increased risk of death following bleeding has previously been reported in the CAT‐BLED study, demonstrating a case fatality rate of 5% across bleeding events in patients with cancer. 29 We observed only one fatal bleed in our cohort, suggesting that bleeding itself was not the direct cause of death in our population. Bleeding could be an indicator of poor treatment response after CAR T‐cell therapy, although we did not observe higher rates of PD in patients with bleeding. Notably, a CAR‐HT score ≥2 was associated with bleeding as well as all‐cause death in our study. This is in line with previous studies demonstrating an association between high CAR‐HT score and poorer OS, largely attributed to prolonged neutropenia and infectious complications. 19 , 20 Although the CAR‐HT score was only validated for severe neutropenia, our findings suggest that it could also identify patients at risk of bleeding. This may be related to the contribution of severe (grade ≥3) thrombocytopenia or anemia, which could be a relevant indicator of limited bone marrow recovery or treatment response, predisposing to bleeding after lymphodepleting chemotherapy and CAR T‐cell therapy. A pathophysiological link between hematotoxicity and bleeding in LBCL patients treated with CAR T‐cell therapy, and the predictive value of the CAR‐HT score for bleeding, should be confirmed in future studies.
We have identified several potentially modifiable risk factors associated with bleeding in patients receiving CAR T‐cell therapy. Patients using therapeutic anticoagulation at the time of CAR T‐cell infusion had a significantly higher risk of bleeding (21.1% vs. 7.9%; HR 2.87, 95% CI 1.25–6.60). This risk was comparable to a previous study in LBCL patients receiving anticoagulation during myeloablative chemotherapy and aHSCT. 30 Our results should be interpreted with caution though, as use of therapeutic anticoagulation was assessed as a fixed‐time variable at the time of CAR T‐cell infusion and thus, did not account for changes over time (discontinuation or dose adjustments). This is relevant as therapeutic anticoagulation is often adjusted if severe thrombocytopenia (platelet count < 30 × 109/L) occurs, in which case platelet transfusions, anticoagulation dose reduction, or temporary cessation are recommended according to guidelines. Advanced time‐dependent or multistate modeling could possibly better estimate bleeding risk with anticoagulation, but require more complete longitudinal data than available in our cohort. We also found that thrombocytopenia grade ≥3 and CAR‐HT score ≥2 were significantly associated with bleeding. Hence, careful consideration of continuing therapeutic anticoagulation or introducing higher transfusion thresholds to prevent grade ≥3 thrombocytopenia or anemia at the start of lymphodepletion or CAR T‐cell infusion may help in reducing the bleeding risk in the setting of CAR T‐cell therapy and deserves further investigation.
We did not observe statistically significant associations between thrombosis and OS (52.2% vs. 74.6%; HR 1.82, 95% CI 0.73–4.58, P = 0.20), or use of therapeutic anticoagulation and thrombosis (4.0% vs. 7.6%; HR 0.43, 95% CI 0.10–1.93, P = 0.27). However, confidence intervals were wide, and the shortcoming of statistically significant associations likely reflects the limited number of thrombosis events in the therapeutic anticoagulation group and the lack of statistical power of our study, rather than the absence of a clinically relevant effect. Similarly, our data also do not allow us to provide any recommendations regarding the benefit or risk of prophylactic anticoagulation. These findings should therefore be interpreted with caution, and future larger studies are needed for verification.
The relatively high risk of thrombosis and especially bleeding in our cohort may reflect the higher lymphoma activity in patients eligible for CAR T‐cell therapy, but potentially also the mechanisms of CAR T‐cell therapy itself. CRS and ICANS are characterized by substantial cytokine release, which can promote endothelial activation and injury, exposing subendothelial collagen layers and increasing tissue factor release. 31 This contributes to a procoagulant state through platelet and coagulation activation, leading to fibrin deposition. 32 As CRS resolves, the procoagulant phase may shift towards hyperfibrinolysis, underscoring the dynamic interplay between coagulopathy and inflammation. 33 In parallel, systemic inflammation can suppress hematopoiesis, leading to prolonged cytopenias in CAR T‐cell patients and increasing the risk of bleeding in patients with prolonged and/or severe thrombocytopenia. 34 Coagulation abnormalities consistent with disseminated intravascular coagulation (DIC) have been reported in patients with severe CRS and were closely correlated with elevated cytokine levels, 35 suggesting this may contribute to CAR T‐cell–related coagulopathy. DIC may also contribute to platelet consumption and thereby impact the severity or duration of thrombocytopenia. Due to the limited availability of coagulation markers in our cohort, we were unable to evaluate associations between DIC parameters and thrombosis or bleeding.
Our study has important limitations. The “Follow that CAR!” registry contains a comprehensive dataset but did not include systematic prospective registration of thrombosis and bleeding events. Consequently, we retrospectively collected data on these outcomes manually by reviewing medical records, which could have potentially led to underreporting of events and surveillance bias. We also encountered missing data regarding laboratory results (particularly coagulation markers) and transfusions or other supportive care. Because anticoagulation was assessed as a time‐fixed variable at the time of CAR T‐cell infusion, subsequent interruptions or dose adjustments during follow‐up were not captured in our analyses, which may have confounded the observed statistically significant association with bleeding. We also acknowledge that there was considerable heterogeneity in anticoagulation use in the cohort, which may have impacted more outcomes than thrombosis or bleeding only, and limits generalizability. Finally, potential other confounders in the associations between thrombosis or bleeding with OS could not be ruled out despite our adjusted analyses, which may limit the robustness of our findings.
Nonetheless, the strengths of our study are that we assessed the incidence of thrombosis and bleeding in a large and well‐characterized cohort of patients, relatively homogeneous as all were diagnosed with R/R LBCL and received axi‐cel, using standardized and reproducible evaluation of thrombosis and bleeding events, with a 1‐year follow‐up after CAR T‐cell therapy, while also incorporating the role of anticoagulant therapy and laboratory parameters. As we focused on patients with R/R LBCL receiving axi‐cel in the Netherlands only, this limits the generalizability of our results, and incidence or risk factors for thrombosis and bleeding may differ for other CAR T‐cell products, indications, or countries. Nonetheless, we hope that our findings increase awareness of thrombosis and bleeding risks in patients undergoing CAR T‐cell therapy and stimulate future prospective international multicenter cohort studies to systematically evaluate the incidence of thrombosis and bleeding, the impact of anticoagulant therapy, and, ideally, the value of monitoring of coagulation markers (such as DIC parameters) in assessing thrombosis or bleeding risk.
In summary, we showed a 1‐year cumulative incidence of 11.0% for bleeding in a large cohort of 250 patients with R/R LBCL who were all treated with commercially available axi‐cel. Bleeding incidence was higher than thrombosis (6.3%) and associated with poorer OS. CAR‐HT score ≥2 before CAR T‐cell therapy, and therapeutic anticoagulation and grade ≥3 thrombocytopenia at time of CAR T‐cell infusion were associated with a higher bleeding risk. Future studies are needed to confirm these results and to guide recommendations regarding modifiable risk factors such as anticoagulation or transfusion thresholds in patients receiving CAR T‐cell therapy.
AUTHOR CONTRIBUTIONS
Amica K. Ko: Conceptualization; investigation; writing—original draft; writing—review and editing; resources; methodology; validation; visualization; project administration; formal analysis; data curation. Pim G. N. J. Mutsaers: Conceptualization; writing—review and editing; resources. Lukas M. van Kleij: Writing—review and editing; formal analysis. Amária Vledder: Data curation; writing—review and editing; resources. Elise R. A. Pennings: Writing—review and editing; resources. Anne M. Spanjaart: Writing—review and editing; resources. Janneke W. de Boer: Writing—review and editing; resources. Kylie Keijzer: Writing—review and editing; resources. Astrid M. P. Demandt: Writing—review and editing; resources. Jaap A. van Doesum: Writing—review and editing; resources. Suzanne van Dorp: Writing—review and editing; resources. Margot Jak: Writing—review and editing; resources. Maria T. Kuipers: Writing—review and editing; resources. Tom van Meerten: Writing—review and editing. Marjolein W. M. van der Poel: Writing—review and editing; resources. Joost S. P. Vermaat: Writing—review and editing; resources. Frank W. G. Leebeek: Writing—review and editing; conceptualization; supervision. Marie J. Kersten: Writing—review and editing; resources. Mandy N. Lauw: Conceptualization; writing—original draft; writing—review and editing; visualization; validation; supervision; formal analysis; project administration; resources; methodology; investigation.
CONFLICT OF INTEREST STATEMENT
P.G.N.J.M. has received funding from AstraZeneca. S.v.D. has received honoraria from Novartis, Gilead, Kite (a Gilead Company), and Johnson & Johnson. M.J. has received honoraria from BMS/Celgene and Kite (a Gilead company); has received funding from Novartis; and has served in an advisory role for Janssen. M.T.K. has received honoraria from Galapagos. T.v.M. has received honoraria from BMS/Celgene, Siemens, and Kite (a Gilead company); has received funding from BMS/Celgene, Kite (a Gilead company), and Genentech; and has served in an advisory role for BMS/Celgene, Eli Lilly, Kite (a Gilead company), and Janssen. M.W.M.v.d.P. received honoraria from Kite (a Gilead company) and Takeda. F.W.G.L. has received funding from Takeda, CSL Behring, and Sobi; and has served in an advisory role for Takeda, Biomarin, and CSL Behring. M.J.K. has received honoraria from Janssen, Mustang Bio, Roche, Miltenyi Biotech, Bristol Myers Squibb/Celgene, Adicet Bio, BeiGene, Novartis, BMS, Galapagos, Kite (a Gilead company), and AbbVie; and has served in an advisory role for Janssen, Mustang Bio, Roche, Bristol Myers Squibb/Celgene, Adicet Bio, BeiGene, Novartis, BMS, Galapagos, and Kite (a Gilead company). M.N.L. has received funding from LeoPharma and has served in an advisory role for Amgen, AbbVie, Inari Stryker, and Viatris. For the remaining authors, no relevant conflicts of interest were declared.
FUNDING
This research received no funding.
Supporting information
Supporting Information.
ACKNOWLEDGMENTS
The authors would like to thank all data managers and other coinvestigators of the “Follow that CAR!” registry, as well as the patients who participated in this study.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
REFERENCES
- 1. Kalos M, Levine BL, Porter DL, et al. T cells with chimeric antigen receptors have potent antitumor effects and can establish memory in patients with advanced leukemia. Sci Transl Med. 2011;3(95):95ra73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Mitra A, Barua A, Huang L, Ganguly S, Feng Q, He B. From bench to bedside: the history and progress of CAR T cell therapy. Front Immunol. 2023;14:1188049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Locke FL, Miklos DB, Jacobson CA, et al. Axicabtagene ciloleucel as second‐line therapy for large B‐cell lymphoma. N Engl J Med. 2022;386(7):640‐654. [DOI] [PubMed] [Google Scholar]
- 4. Neelapu SS, Locke FL, Bartlett NL, et al. Axicabtagene ciloleucel CAR T‐cell therapy in refractory large B‐cell lymphoma. N Engl J Med. 2017;377(26):2531‐2544. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Neelapu SS, Jacobson CA, Ghobadi A, et al. Five‐year follow‐up of ZUMA‐1 supports the curative potential of axicabtagene ciloleucel in refractory large B‐cell lymphoma. Blood. 2023;141(19):2307‐2315. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Westin JR, Oluwole OO, Kersten MJ, et al. Survival with axicabtagene ciloleucel in large B‐cell lymphoma. N Engl J Med. 2023;389(2):148‐157. [DOI] [PubMed] [Google Scholar]
- 7. Goyco Vera D, Waghela H, Nuh M, Pan J, Lulla P. Approved CAR‐T therapies have reproducible efficacy and safety in clinical practice. Hum Vaccines Immunother. 2024;20(1):2378543. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Bonifant CL, Jackson HJ, Brentjens RJ, Curran KJ. Toxicity and management in CAR T‐cell therapy. Mol Ther Oncolytics. 2016;3:16011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Rejeski K, Subklewe M, Aljurf M, et al. Immune effector cell‐associated hematotoxicity: EHA/EBMT consensus grading and best practice recommendations. Blood. 2023;142(10):865‐877. [DOI] [PubMed] [Google Scholar]
- 10. Hashmi H, Mirza AS, Darwin A, et al. Venous thromboembolism associated with CD19‐directed CAR T‐cell therapy in large B‐cell lymphoma. Blood Adv. 2020;4(17):4086‐4090. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Schorr C, Forindez J, Espinoza‐Gutarra M, Mehta R, Grover N, Perna F. Thrombotic events are unusual toxicities of chimeric antigen receptor T‐cell therapies. Int J Mol Sci. 2023;24(9):8349. 10.3390/ijms24098349 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Johnsrud A, Craig J, Baird J, et al. Incidence and risk factors associated with bleeding and thrombosis following chimeric antigen receptor T‐cell therapy. Blood Adv. 2021;5(21):4465‐4475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Bindal P, Patell R, Chiasakul T, et al. A meta‐analysis to assess the risk of bleeding and thrombosis following chimeric antigen receptor T‐cell therapy: communication from the ISTH SSC Subcommittee on Hemostasis and Malignancy. J Thromb Haemost. 2024;22(7):2071‐2080. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Gonsalves A, Carrier M, Wells PS, McDiarmid SA, Huebsch LB, Allan DS. Incidence of symptomatic venous thromboembolism following hematopoietic stem cell transplantation. J Thromb Haemost. 2008;6(9):1468‐1473. [DOI] [PubMed] [Google Scholar]
- 15. Benfield M, He J, Arnall J, et al. Venous thromboembolism incidence and risk factors in patients undergoing hematopoietic stem cell transplantation. Transplant Cell Ther. 2025;31(2):111.e1‐111.e10. [DOI] [PubMed] [Google Scholar]
- 16. Spanjaart AM, Pennings ERA, Mutsaers PGNJ, et al. Dutch CAR‐T Tumorboard Consortium. The Dutch CAR‐T tumorboard experience: population‐based real‐world data on patients with relapsed or refractory large B‐cell lymphoma referred for CD19‐directed CAR T‐cell therapy in the Netherlands. Cancers (Basel). 2023;15(17):4334. 10.3390/cancers15174334 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Cheson BD, Fisher RI, Barrington SF, et al. Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non‐Hodgkin lymphoma: the Lugano classification. J Clin Oncol. 2014;32(27):3059‐3068. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Lee DW, Santomasso BD, Locke FL, et al. ASTCT consensus grading for cytokine release syndrome and neurologic toxicity associated with immune effector cells. Biol Blood Marrow Transplant. 2019;25(4):625‐638. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Rejeski K, Perez A, Sesques P, et al. CAR‐HEMATOTOX: a model for CAR T‐cell‐related hematologic toxicity in relapsed/refractory large B‐cell lymphoma. Blood. 2021;138(24):2499‐2513. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Rejeski K, Perez A, Iacoboni G, et al. The CAR‐HEMATOTOX risk‐stratifies patients for severe infections and disease progression after CD19 CAR‐T in R/R LBCL. J Immunother Cancer. 2022;10(5):e004475. 10.1136/jitc-2021-004475 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. de Boer JW, Keijzer K, van Dorp S, et al. Population‐based validation of the CAR‐HEMATOTOX for hematotoxicity, infections, and survival after CART in R/R LBCL. Blood Adv. 2025;9(21):5641‐5650. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Schulman S, Kearon C, Subcommittee on Control of Anticoagulation of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis . Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non‐surgical patients. J Thromb Haemost. 2005;3(4):692‐694. [DOI] [PubMed] [Google Scholar]
- 23. Kaatz S, Ahmad D, Spyropoulos AC, Schulman S. Definition of clinically relevant non‐major bleeding in studies of anticoagulants in atrial fibrillation and venous thromboembolic disease in non‐surgical patients: communication from the SSC of the ISTH. J Thromb Haemost. 2015;13(11):2119‐2126. [DOI] [PubMed] [Google Scholar]
- 24. Simon R, Makuch RW. A non‐parametric graphical representation of the relationship between survival and the occurrence of an event: application to responder versus non‐responder bias. Stat Med. 1984;3(1):35‐44. [DOI] [PubMed] [Google Scholar]
- 25. Melody M, Gandhi S, Saunders H, et al. Incidence of thrombosis in relapsed/refractory B‐cell lymphoma treated with axicabtagene ciloleucel: Mayo Clinic experience. Leuk Lymphoma. 2022;63(6):1363‐1368. [DOI] [PubMed] [Google Scholar]
- 26. Schuster SJ, Tam CS, Borchmann P, et al. Long‐term clinical outcomes of tisagenlecleucel in patients with relapsed or refractory aggressive B‐cell lymphomas (JULIET): a multicentre, open‐label, single‐arm, phase 2 study. Lancet Oncol. 2021;22(10):1403‐1415. [DOI] [PubMed] [Google Scholar]
- 27. Goldman A, Maor E, Bomze D, et al. Adverse cardiovascular and pulmonary events associated with chimeric antigen receptor T‐cell therapy. J Am Coll Cardiol. 2021;78(18):1800‐1813. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Wadhwa PD, Fu P, Koc ON, et al. High‐dose carmustine, etoposide, and cisplatin for autologous stem cell transplantation with or without involved‐field radiation for relapsed/refractory lymphoma: an effective regimen with low morbidity and mortality. Biol Blood Marrow Transplant. 2005;11(1):13‐22. [DOI] [PubMed] [Google Scholar]
- 29. Englisch C, Moik F, Steiner D, et al. Bleeding events in patients with cancer: incidence, risk factors, and impact on prognosis in a prospective cohort study. Blood. 2024;144(22):2349‐2359. [DOI] [PubMed] [Google Scholar]
- 30. Jalaber E, Orvain C, Papadopoulou V, et al. Management of thrombocytopenia and anticoagulant therapy in patients with hematological malignancy on chemotherapy: a binational prospective study (TAT study). J Thromb Thrombolysis. 2025;58(5):646‐656. [DOI] [PubMed] [Google Scholar]
- 31. Li Y, Jiang H, Luo L, Mei H. Immunotherapy‐associated hemostatic abnormalities: bleeding and thrombotic complications. Ann Hematol. 2025;104(7):3537‐3551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Wang J, Doran J. The many faces of cytokine release syndrome‐related coagulopathy. Clin Hematol Int. 2021;3(1):3‐12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Peng X, Zhang X, Zhao M, et al. Coagulation abnormalities associated with CAR‐T‐cell therapy in haematological malignancies: a review. Br J Haematol. 2024;205(2):420‐428. [DOI] [PubMed] [Google Scholar]
- 34. Dahunsi D, Eleanya C, Akintunde A, Oluwole O. Prolonged cytopenia with CAR‐T cell therapy and management recommendations. Clin Hematol Int. 2025;7(1):47‐54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Wang Y, Qi K, Cheng H, et al. Coagulation disorders after chimeric antigen receptor T cell therapy: analysis of 100 patients with relapsed and refractory hematologic malignancies. Biol Blood Marrow Transplant. 2020;26(5):865‐875. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.