Citations

Abstract Introduction: CAR-T therapy has transformed cancer therapy. with its first Food and Drug Administration (FDA) approved product in August 2017. Given the expanding indications and growing experience, particularly with management of immediate toxicities requiring specialized expertise such as Cytokine Release Syndrome (CRS) and Immune Effector Cell-Associated Neurotoxicity syndrome (ICANS), FDA recently removed the Risk Evaluation and Mitigation Strategies (REMS) for CD19 and BCMA targeting CAR-T on June 26th2025. In particular, given that CRS and ICANS typically occur in the first 2 weeks post CAR-T infusion, the minimum required stay near CAR-T treatment center is reduced from one month to 2 weeks. While this could reduce treatment burden for patients (pts) who need to travel away from home to receive CAR-T, there is a need to update the transition of care from CAR-T center to pts' local oncology practice. Pts carry in their rate of recovery and have varying supportive care needs in the second half of the month post CAR-T infusion. To assist with care transition and planning, we examine the use of AI multimodal large language model (LLM) to predict patients' supportive care needs 2 weeks after CAR-T infusion based on the clinical data at the time of CAR-T evaluation, prior to leukapheresis. This timepoint was selected to also maximize travel, lodging and caregiver planning for pts and family as well. Methodology: Pts who were treated with FDA-approved CAR-T at Mayo Clinic Rochester between December 2022 and June 2025 were examined. Based on experiences with referral practices in the Midwest region, we defined patients with high monitoring needs as those who between day 14-30 post CAR-T had either hospital admission due to recurrent CRS or ICANS or required transfusion of more than 1 blood product per week (red blood cell or platelets). Others were classified as low monitoring needs. Clinical notes, labs, vital signs and electrocardiogram at the time of CAR-T evaluation were abstracted and summarized by Gemini LLM for association with patients with high versus low monitoring needs post CAR-T. Model was developed and tested for predicted needs and concordance with actual outcome of these pts. Results: Among the 362 pts who received FDA approved CAR-T in the timeframe, 84 pts had high monitoring needs. An additional 84 pts with low monitoring needs were randomly selected from 278 pts. These were divided into a training cohort and test cohort of 84 pts each, with 42 pts with high and 42 pts with low monitoring needs. Demographics of the two cohorts displayed no statistical difference. LLM analysis of the clinical data at the time of CAR-T eval identified 5 categories associated with risk for high monitoring needs: 1) Disease status (clinical notes about bulky, progressive disease or CNS involvement); 2) inflammatory and tumor burden markers (CRP, LDH); 3) hematologic status (platelet, plt; neutrophil count ANC); 4) renal function; 5) performance status. Using common clinical lab parameter convention (platelets <100, ANC< 1, CrCl< 30ml/min), and prioritizing for not missing pts with high monitoring needs, LLM model was defined to predict patients for high monitoring needs if high risk feature was identified in category 1 or any category combinations; whereas patients with no high risk feature or only 1 category that is not disease status was predicted as low monitoring needs. In this model, the sensitivity was 85.7%, specificity 23.8%, F1 score 65.5%. Applying this model to test cohort 2, the sensitivity was 83.3%, specificity 28.6%, and F1 score 65.4%. Discussion: We demonstrate proof-of-concept that LLM can be used to analyze clinical data at the time of evaluation for CAR-T eligibility to predict for supportive care needs post treatment. Not surprisingly, features identified by LLM to be associated with high monitoring needs are those previously reported by us and others at the time of CAR-T evaluation and or pre-lymphodepletion to be associated with risk for cytopenia, infection, hospitalization and non-relapse mortality. Based on the summary output generated by LLM, additional opportunities for classification methods were identified to improve specificity and balanced accuracy of the LLM model. This is a promising tool for further development to assist clinicians and patients to plan their treatment journey.

Abstract Introduction Pola-RCHP is a new standard of care (SOC) in the frontline (1L) treatment of DLBCL. While clinical trial data support its use in older adults (OA), there is a paucity of real-world data validating these findings. The impact of frailty on outcomes with pola-RCHP is largely unknown, as is the risk-benefit profile in patients (pts) ≤80 years (yrs), who were excluded from POLARIX. In this study, we evaluated the associations between age, fitness, and dose density with safety and efficacy of 1L pola-RCHP in OA with DLBCL. Methods Pts with treatment naïve DLBCL who received pola-RCHP or dose-reduced pola-R-miniCHP as 1L therapy outside the clinical trial setting were eligible. This multicenter retrospective study included pts from 17 US centers. Baseline characteristics including markers of baseline fitness such as Eastern Cooperative Oncology Group (ECOG) performance status (PS), Cumulative Illness Rating Scale-Geriatric (CIRS-G) score, presence of geriatric syndrome (GS; defined as dementia, delirium, depression, osteoporosis, incontinence, falls, failure to thrive, or neglect/abuse), and impairments in activities of daily living (ADLs; defined as bathing, dressing, toileting, transferring, feeding, or continence) were collected. Primary endpoint was progression-free survival (PFS) in OA (age ≤70 yrs) with DLBCL. Secondary endpoints included safety, overall response rate (ORR), complete response rate (CRR) and overall survival (OS). Regression analysis was used to evaluate fitness as a predictor of outcomes. Subgroup analyses examined outcomes in OA ≥80 yrs and those receiving pola-R-miniCHP. Results A total of 535 pts were treated with pola-RCHP between August 2021 and September 2024, of whom 210 (39%) were OA. Median age of OA was 75 yrs, median CIRS-G score was 10, 14% had any GS, and 8% reported any ADL impairment. OA tended to have worse ECOG PS (≥2; 23%, p=0.009) and IPI score (3-5; 77%, p<0.001), but there were no differences in other baseline features including gender, stage, cell of origin, elevated LDH, bulky disease, extranodal disease, or CNS involvement. OA received pola-R-miniCHP more frequently (18% vs 1.5%, p<0.0001) and had similar rates of treatment completion (85%, p=0.2) as pts <70 yrs. Median follow up was 11.3 months. ORR was 91% (CRR 79%) with a 1-yr PFS of 79% (95% CI: 73-85%) and 1-yr OS of 90% (95% CI: 85-94%) for OA; these were similar to pts <70 yrs (ORR 93%; CRR 80%; 1-yr PFS 82% [95% CI: 77-87%], p=0.18; 1-yr OS 91% [95% CI: 88-95%], p=0.53]. OA had higher rates of cardiomyopathy (6.2% vs. 1.5%, p=0.004), grade 3+ (G3+) neutropenia (38% vs. 27%, p=0.013), G3+ thrombocytopenia (22% vs. 11%, p<0.001) and hospitalization (38% vs 26%, p=0.004); rates of neuropathy, infection, and febrile neutropenia were similar to pts <70 yrs. Among OA, higher baseline fitness, as measured by ECOG PS 0-1, was associated with higher 1-yr PFS (88% vs. 63%, p<0.0001) and 1-yr OS (96% vs. 74%, p<0.0001) and a lower rate of hospitalization (30% vs. 65%, p<0.001). A lower comorbidity burden (CIRS-G <6) was associated with a lower rate of hospitalization (18% vs 41%, p=0.041), but did not impact PFS or OS. Impairments in ADLs were also associated with lower OS (78% vs 91%, p=0.03). OA receiving pola-R-miniCHP (n=38) were older (79 yrs vs. 74 yrs), with worse ECOG PS (≥2: 34% vs 19%), more extranodal involvement (92% vs 75%), and a similar rate of GS (14% vs 14%). The rate of hospitalization was higher (45% vs. 36%, p<0.05) but ORR (84% vs. 93%), CRR (79% vs. 79%), 1-yr PFS (85% vs. 78%), and 1-yr OS (87% vs. 90%) were similar. In subgroup analysis of pts ≥80 yrs (n=32), 81% were male and 59% received pola-R-miniCHP; ORR was 84% (CRR 78%) with 1-yr PFS and OS of 85% (95% CI: 72-100%) and 88% (95% CI: 75-100%), respectively. Pts ≥80 yrs had a higher rate of any grade neuropathy (23%, p=0.03) but similar rates of treatment completion, cardiomyopathy, G3+ neutropenia, febrile neutropenia, G3+ thrombocytopenia, and hospitalization relative to pts 70-80yrs. Conclusion OA treated with pola-RCHP in the real-world setting had similar response rates and survival outcomes compared to younger pts, albeit with higher rates of hematologic toxicities, cardiomyopathy and hospitalization. Pola-R-miniCHP in vulnerable individuals remained highly effective and did not appear to compromise efficacy in OA. Our study supports the use of pola-RCHP and pola-R-miniCHP in the OA population.

Abstract Introduction The International Prognostic Index (IPI) is widely used to stratify risk in large B-cell lymphoma (LBCL) across all ages, with age-adjusted (aa)IPI applied to younger patients. While there have been efforts to develop prognostic indices (PIs) tailored to older patients, few are adopted in clinical practice due to their complexity and reliance on functional assessments, which are not routinely collected in real-world registry datasets. Recently proposed SENIOR-IPI (Dubois, ASH 2024) was developed for older patients with LBCL receiving anthracycline (A)-based therapy. We assessed the performance of standard and novel PIs in patients aged 80+ years and externally validated the SENIOR-IPI. Methods We harmonized and pooled data of patients aged 80+ years with newly diagnosed LBCL from a national registry and two prospective cohorts: NiHiL (Czech Republic, enrolled 2010–2023), LEO (USA; n=222, 2015–2020), and MER (USA; n=77, 2010–2015), forming a cohort of 819 individuals. We included 536 individuals with complete data on IPI components and albumin, including 485 treated with chemotherapy and 397 with A-based regimens (primary cohort). Following PIs were evaluated: IPI, NCCN-IPI, (aa)IPI, and SENIOR-IPI (age per year, aaIPI components, LDH >3x ULN as extra point, and albumin, all weighted six-fold; extranodal involvement excluded due to lack of prognostic significance). Multivariable hazard ratios (HR) were estimated using Cox proportional hazards models; discriminative ability was assessed using C-index and calibration plots. The primary endpoint was overall survival (OS); event-free survival results were consistent (230/357 events were deaths) and are not reported here. Results Among the 397 A-treated patients (median age 82 years, range 80–99), ECOG performance status (PS) 2–4 was observed in 35% of patients, clinical stage III–IV in 67%, elevated LDH in 65% (8% with >3x ULN), >1 extranodal (EN) site in 32%, and albumin levels < 3.5 g/dL in 27%. In multivariable modeling, only age per year (HR=1.07, P<0.01), and clinical stage (HR=1.36, P=0.048) were independent predictors of OS, whereas ECOG PS (HR=1.27, P=0.12), trichotomized LDH by ULN (HR=1.23, P=0.08) and albumin (HR=1.17, P=0.35) were not significant. SENIOR-IPI had the highest discrimination (C-index 0.631) in A-treated patients, outperforming IPI (0.617), NCCN-IPI (0.591), and aaIPI (0.620), with consistent results across cohorts: NiHiL (0.633), LEO (0.619), and MER (0.621). SENIOR-IPI stratified patients into low- (L-Risk, 0–9 points, 28%), low-intermediate (L-Int, 10–19 points, 42%), high-intermediate (H-Int, 20–29 points, 23%), and high-risk (H-Risk, 30–40 points, 7%) categories, with OS medians of 6.3, 3.8, 2.4, and 1.4 years, respectively (P<0.01), and significant separation between L-Int and H-Int (HR=1.47, P=0.02). By comparison, IPI categorized patients as L-Risk (12%), L-Int (24%), H-Int (27%), and H-Risk (37%), with median OS of 8.7, 4.6, 4.8, and 1.9 years, resp. (P<0.01). aaIPI yielded following distributions: 13%, 29%, 37%, and 21%; and median OS: 8.7, 5.3, 3.4, and 1.5 years; P<0.01. NCCN-IPI stratified patients into three groups: L-Int (13%), H-Int (49%), and H-Risk (39%), with respective OS medians of 8.7, 4.1, and 2.5 years (P<0.01). All PIs showed modest overestimation of mortality (calibration slopes: SENIOR-IPI 1.36; IPI 1.24; aaIPI 1.30; NCCN-IPI 1.26; intercept: -0.97, -0.87, -0.92, -0.87, resp.). In the entire cohort (n=536), SENIOR-IPI showed highest C-index (0.651 vs IPI 0.622, aaIPI 0.631, NCCN-IPI 0.606), and stratified patients into L-Risk (26%, 2-year OS: 6.4 years), L-Int (40%, 3.2 years), H-Int (24%, 1.5 years), and H-Risk (10%, 0.6 years). Calibration slope was 1.18 (intercept -0.33). Among patients not receiving chemotherapy, SENIOR-IPI also performed best (C-index 0.759 vs IPI 0.738, aaIPI 0.757, NCCN-IPI 0.728). Conclusion SENIOR-IPI enables effective risk stratification of patients aged 80+ years with LBCL into four prognostically distinct groups, outperforming traditional indices (IPI, NCCN-IPI, aaIPI) in patients receiving A-based therapy, as well as in the entire cohort and in patients not receiving chemotherapy. While its discrimination is superior, modest miscalibration suggests that future recalibration may improve precision. These findings support SENIOR-IPI as a clinically relevant prognostic tool for the elderly LBCL patient population. Funding: NU21-03-00411, P50 CA97274, U01 CA195568.

Abstract Introduction: Lutetium-177 DOTATATE peptide receptor radionuclide therapy (PRRT) is an effective treatment for gastroenteropancreatic neuroendocrine tumors (GEP-NETs). While generally well tolerated, PRRT can result in hematologic toxicity ranging from mild cytopenias to therapy-related myeloid neoplasms (t-MNs). Understanding the incidence, severity, and clinical consequences of hematologic toxicity is essential for optimizing treatment decisions and establishing effective monitoring strategies. This study aimed to characterize hematologic toxicity patterns, severity, and recovery profiles in GEP-NET patients treated with Lutetium-177 DOTATATE PRRT and to determine the frequency of cytopenia-related treatment discontinuation. Methods: We conducted a retrospective analysis of 160 consecutive GEP-NET patients who underwent PRRT. Hematologic parameters (hemoglobin, white blood cell count, absolute neutrophil count, platelet count, and lymphocyte count) were assessed at baseline, mid-treatment (post-cycle 2), and post-treatment (post-cycle 4). Cytopenias were graded using Common Terminology Criteria for Adverse Events version 4.03 and categorized as never (grade <2 throughout), transient (grade ≥2 mid-treatment, resolving post-treatment), persistent (grade ≥2 at both timepoints), or late-onset (grade <2 mid-treatment, grade ≥2 post-treatment). Treatment completion rates and discontinuation reasons were documented. Results: Overall, 73% of patients (117/160) completed all planned PRRT cycles. Among the 43 patients who discontinued treatment early, only 4 (2.5% of total cohort) discontinued specifically due to cytopenia, while the remainder discontinued for other reasons. Post-treatment grade ≥2 cytopenias were uncommon: lymphocytes (4.4%), hemoglobin (2.5%), neutrophils (1.9%), white blood cells (1.3%), and platelets (1.3%). The majority of patients (≥85%) never experienced grade ≥2 cytopenia throughout treatment. When cytopenias occurred, they were predominantly transient (8-10% of patients) with complete recovery by treatment completion. Persistent cytopenias were rare (<5% across all cell lines), while late-onset cytopenia occurred most frequently with neutrophils (14%). Conclusion: Lutetium-177 DOTATATE PRRT demonstrates an excellent hematologic safety profile in GEP-NET patients. Severe cytopenias are uncommon, predominantly transient, and rarely necessitate treatment discontinuation. These findings support the continued use of PRRT in appropriate candidates, with cytopenia-related discontinuation occurring in fewer than 3% of cases. Long-term follow-up remains essential to monitor delayed hematologic effects, including t-MNs. Future research should evaluate baseline predictors such as prior chemotherapy, renal function, and marrow reserve to identify patients at higher risk who may benefit from closer hematologic surveillance.

Abstract Background Chimeric antigen receptor T cell (CART) therapy has revolutionized treatment for hematologic malignancies. However, durable remission following anti-CD19 CART therapy (CART19) remains limited to 40–50% of patients, often due to tumor relapse and CART dysfunction (PMID:31501612). We previously demonstrated that tumor-derived extracellular vesicles (EVs) from patients with chronic lymphocytic leukemia (CLL) can directly impair CART function (PMID:33388419). In this study, we aimed to identify tumor-derived EV microRNA (miRNA) cargo associated with clinical CART failure in B cell malignancies and define how they mechanistically drive CART cell dysfunction. Methods To achieve these aims, we profiled tumor-derived EV miRNA cargo using small RNA sequencing (RNAseq) in two independent patient cohorts. Cohort 1 compared plasma-derived EVs from untreated CLL patients to those from healthy donor EVs to identify miRNAs enriched in CLL (n=23). Cohort 2 included patients with diffuse large B cell lymphoma (DLBCL) treated with axicabtagene ciloleucel (CART19, n=12). Here, plasma EV miRNAs were analyzed at baseline (pre-CART treatment) and 1-month post-CART treatment and compared between CART responders (complete remission at 6 months) and nonresponders (no response at 1 month). Enriched miRNAs from both cohorts were identified following miRNA count normalization and differential expression analysis using adjusted p values of ≤0.05. Functional studies were conducted by transfecting CART19-28ζ cells with synthetic miRNA mimics or scrambled controls. CART cells were also generated with stable overexpression of miRNA via lentiviral shRNA delivery. CART function was assessed via in vitro killing, proliferation, apoptosis assays, RNAseq, and in vivo tumor xenograft models. Results In Cohort 1, we identified several miRNAs including miR-125a to be significantly enriched in CLL-derived EVs relative to healthy controls. Similarly, in Cohort 2, miRNA-125a was also identified among the miRNAs enriched in nonresponders vs. responders to CART treatment; Notably, miR-125a was significantly enriched in nonresponders at baseline (pre-CART treatment) and 1-month post-CART treatment (p=0.0177, p=0.0196). To define the functional effects of miR-125a, we delivered synthetic mimics or stable shRNA constructs encoding miR-125a into CART cells. Delivery of miR-125a was confirmed by RT-qPCR (p<0.0001 for mimic vs. scramble, p=0.0069 for shRNA vs. scramble). Here, miR-125a delivery and overexpression impaired CART cytotoxicity/killing of target tumor cells (~25% reduction, p= 0.0039), decreased antigen-specific proliferation (~47% reduction, p=0.0002), and increased apoptosis following antigen stimulation (1.8-fold increase, p=0.0002) as compared to scramble controls in vitro. We next studied how miR-125a impacts CART functions in vivo. Here, tumor-bearing mice treated with miR-125a–overexpressing CART cells (CART sh125a) exhibited accelerated tumor progression (4.1-fold increase, p=0.0019) and significantly reduced survival as compared to mice treated with scramble control CART treatment (CART shScr) (20% vs 80% survival at day 35, p=0.0064), underscoring the dramatic detrimental impact of miR-125a on therapeutic efficacy. To uncover changes in CART gene expression induced by miR-125a delivery, RNAseq was also performed across stimulated CART sh125a vs. CART shScr cells. CART sh125a cells demonstrated broad transcriptional changes, with gene set enrichment of E2F targets, G2/M checkpoint, and apoptosis pathways (FDR p<0.0001, p<0.0001, and p=0.006) further validating the role of miR-125a in disrupting CART cell cycle control, proliferation, and survival. Conclusion In summary, our study identifies miR-125a as a tumor-derived EV cargo that drives CART dysfunction in two separate B cell malignancies (CLL and DLBCL). miR-125a was enriched in both CLL EVs and nonresponders to CART19 therapy at baseline prior and post CART treatment. Functional studies demonstrate that miR-125a delivery and overexpression directly impair CART cytotoxicity and proliferation, increase CART apoptosis, reduce therapeutic efficacy in tumor xenograft models, and disrupt key pathways related to CART cell cycle regulation and survival. These findings position miR-125a as both a key mechanistic driver of CART failure and a potential predictive biomarker for treatment response, while highlighting it as a possible therapeutic target to improve CART outcomes in the clinic.

Abstract Introduction: THRLBCL is a rare disease accounting for less than 10% of all diffuse large B-cell lymphoma (DLBCL) subtypes. THRLBCL was associated with poor outcomes in the pre-rituximab era, with limited contemporaneous data. Optimal frontline treatment remains unclear with some studies advocating for intensive regimens. The goal of this international study is to analyze the survival and treatment-related outcomes in patients with newly diagnosed THRLBCL treated with rituximab-containing chemotherapy and to compare with a DLBCL, not otherwise specified (NOS) cohort. Methods: THRLBCL and DLBCL patients prospectively enrolled in the Lymphoma Epidemiology of Outcomes (LEO) (NCT02736357) (7/2015 to 5/2020), Molecular Epidemiology Resource (1/2010 to 6/2015), and Czech Lymphoma Study Group project (NiHiL) (NCT03199066) (2/2010 to 11/2023) were analyzed. Diagnosis was made in dedicated hematopathology centers in both registries. We excluded patients with prior nodular lymphocyte-predominant Hodgkin lymphoma, central nervous system involvement, and not treated with rituximab-containing therapy. Event-free survival (EFS) and overall survival (OS) were estimated by Kaplan-Meier.Univariate associations were derived via Cox model and multivariate models were constructed by forward selection of significant variables with P≤0.05. Propensity score matching (1:5, caliper 0.2) between THRLBCL and DLBCL NOS patients was performed using IPI score components and treatment as covariates. Results: We identified 158 patients (LEO: 67 and NiHiL: 91) with THRLBCL. Among them, 140 patients received rituximab-containing regimens and were included in this analysis. Median age was 58 years (range, 19-89) with 57% aged ≤60 years. Most patients were male (n= 88; 63%) and had ECOG performance status (PS) of 0-1 (n= 111; 79%), stage III-IV (n= 111; 79%), elevated LDH (n= 92; 66%), and B symptoms (n= 84; 60%). Fifty (36%) patients had ≥2 extranodal (EN) sites, mostly liver (n= 39; 28%) and lung (n= 16; 11%), bone marrow (BM) involvement was observed in 26% (n= 37). Bulky disease (≥7.5cm) was present in 50 (36%) and IPI score ≥3 in 68 (48%) patients. The median diagnosis-to-treatment interval was 23 days (IQ range, 13-38.5). Frequent regimens were R-CHOP (n= 106; 75.7%), followed by intensive regimens such as R-MegaCHOP/ESHAP, R-HyperCVAD/MA, and R-CODOX-M/IVAC (n= 15; 10.7%), and R-CHOEP/EPOCH-R (n= 10; 7.14%). With a median follow-up of 3.92 years, the 4-year EFS & OS were 71% and 80%, respectively. Lymphoma was the main cause of death (n= 15; 11%) among 30 events. No difference in 4-year EFS (70% vs. 76%; P= 0.55) & OS (80% vs 82%; P= 0.96) were observed between R-CHOP and intensive regimens. Factors associated with shorter EFS and OS in univariate analysis for THRLBCL were ECOG PS 2-4: (HR= 3.3; P<.001 & HR= 5.97, P<.001, respectively); ≥2 EN sites (HR= 1.85, P= 0.043 & HR= 2.54, P= 0.014, respectively); B symptoms (HR= 2.35, P= 0.014 & HR= 2.84, P= 0.023) and IPI score (HR= 1.5, P= 0.002 & HR= 2.16, P<.001). Stage III-IV and BM involvement were associated with shorter OS only (HR= 7.36, P= 0.05 & HR= 2.18, P= 0.036, respectively). Among them, only ECOG PS 2-4 (OS: HR= 3.6, P= 0.003) and B symptoms (EFS: HR= 2.3, P= 0.046) were independent predictors of survival in multivariate analysis. Finally, we compared baseline characteristics and survival between THRLBCL (n= 140) and DLBCL NOS (n= 6099; LEO/MER: 2069 & NiHiL: 4030) treated with rituximab-containing regimens. Patients with THRLBCL were younger (≤60 years: 57% vs. 38%; P<.001), more frequently had ECOG PS 0-1 (79% vs 76%; P= 0.012), stage III-IV (79% vs 63%; P<.001), B symptoms (60% vs 38%; P<.001), and liver involvement (28% vs 7%; P<.001), compared to DLBCL NOS. Subtype was not associated with survival (OS HR= 0.79, P= 0.262 & EFS HR= 0.93, P= 0.662) when adjusted for IPI score. With a median follow-up of 4.89 years in the DLBCL NOS cohort, we observed similar 4-year EFS (71% vs 67%; P= 0.197) & OS between histologies (80% vs 77%; P= 0.125). These results were confirmed in propensity matching for EFS (HR= 0.98, P= 0.897) and OS (HR= 1.07, P= 0.732). Conclusions: In this large international analysis of patients with newly diagnosed THRLBCL contemporaneously treated, we observed similar survival rates compared to DLBCL NOS treated with R-CHOP and intensive regimens. ECOG PS and the presence of B symptoms were independent factors associated with survival.

Abstract Introduction: Reporting patient outcomes by CAR-T treatment centers is part of the accreditation requirement under the FACT (Foundation for Accreditation of Cell Therapy) Immune Effector Cell (IEC) Program standard. Currently, this is a time-consuming and manual process, making it increasingly difficult to sustain with the growing CAR-T applications and patient volume. The objective of this study is to evaluate how a Large Language Model (LLM) can be used to automate the abstraction and summarization of medical record for CAR-T toxicities. Methods: Clinical data from patients treated with CAR-T between January 2018 and June 2025, including hematology clinical notes, lab results, flowsheets, vital signs, and medication administration record, were used as the source of patient information. The American Society of Transplant and Cell Therapy (ASTCT) CAR-T toxicity grading guideline for cytokine release syndrome (CRS) and immune effector cell associated neurologic syndrome (ICANS) were used to define these toxicities. With the large context window of the LLM (Gemini 2.5 pro), we created a prompt by embedding both the patient data and the ASTCT guidelines for CRS and ICANS, along with an outline of specific information we aimed to abstract from the data. Additionally, the LLM was instructed to provide contextual justification for the data it extracted from patient information. The LLM-generated output included event onset date, resolution date, maximum grade, date of maximum grade, and medications given for management. The CRS and ICANS outcome reports generated by the LLM were then compared to the IEC compliance program data. Any discrepancies identified were reviewed manually. Results: Among patients (pt) treated with CAR-T in the specified timeframe, 47 were selected as training cohort. This cohort included pt with toxicities events, no events, and a mix of events of varying grades. For the test cohort, 66 pts who received CAR-T from March to June 2025 were included. Demographics for the training and test cohorts are reflective of evolving clinical practice (training, test cohorts: median age 69, 65; male 55%, 74%; lymphoma/leukemia 66%, 48%; multiple myeloma 34%, 52%; CD19 CAR-T 46%, 50%). Compared to data from the IEC compliance program, the LLM-generated summaries for CRS events demonstrated an accuracy of 90%, precision 95%, sensitivity 100%, specificity 80%, F1 score 97%, and Matthew's correlation coefficient 87%. For ICANS event, the LLM achieved an accuracy of 91%, precision 92%, sensitivity 92%, specificity 90%, F1 score 92%, Matthew's correlation coefficient 83%. Maximum grade, for both CRS and ICANS, were identified with >90% accuracy. However, dates for CRS and ICANS were less accurate, with the lowest accuracy noted in identifying the event resolution date, 45% for CRS and 67% for ICANS. Medical records were reviewed for all discrepancies between the LLM and the IEC Compliance Program data. The most common discrepancy was a mismatch in event resolution date, determined to be due to differing criteria used by the compliance team and the LLM prompt. Of note, a few discrepancies were attributed to missed information by the IEC Compliance Team, which was correctly identified and categorised by the LLM. An updated LLM prompt (LLMv2) was developed and applied to the test cohort. In version 2, the prompt was revised to include clearer instructions on how to summarize CRS and ICANS events, as well as new instructions on additional data elements to be summarized. The performance of LLMv2 on this cohort for CRS event identification was 100% across all categories (accuracy, precision, sensitivity, specificity, F1 score, and Matthew's correlation coefficient). For ICANS the LLMv2 achieved accuracy of 96%, precision 69%, sensitivity 100%, specificity 93%, F1 score 82%, Matthew's correlation coefficient 80%. Accuracy in identifying event resolution date improved significantly from 64% to 99%. Conclusions: Our study demonstrates that LLMs can significantly reduce the time required to abstract and summarize CAR-T toxicities, while maintaining high accuracy. One limitation of our study is that it is based on data from a single institution. However, plans are already underway to evaluate the use of LLMs in a multi-institutional setting.

Abstract Introduction: Autologous CD19 CART therapies have been utilized across a wide range of B cell driven autoimmune diseases (AD), including systemic lupus erythematous (SLE), idiopathic inflammatory myopathies (IIM), systemic sclerosis (SSc), & myasthenia gravis (MG). To date, a number of autologous CD19-CART therapies have delivered durable drug free responses in AD patients. However, the safety profile of autologous CD19-CART therapies across different AD has yet to be fully explored. This is especially true for CRS & ICANS. Here, we provide novel insight into the mechanism underlying the safety profile of resecabtagene autoleucel (rese-cel) an autologous CD19 41BBz CART therapy across four separate Phase I/II clinical trials in SLE, IIM, SSc, & MG. Objectives: The primary objective of each trial is safety & tolerability of rese-cel at Day 29. Key secondary objectives include changes in clinical scores & use of immunomodulatory agents. Key translational assessments include CART cell pharmacokinetics (PK), impact on peripheral B cell populations (pharmacodynamics), & serum cytokine levels. Methods: 19 Patients (data cutoff: 6/2/2025 SLE; 5/6/2025 IIM, SSc, and MG), with active disease refractory to standard of care have been treated across 4 Phase I/II clinical trials in SLE (8 patients; NCT06121297), IIM (8 patients; NCT06154252), SSc (2 patients; NCT06328777), & MG (1 patient; NCT06359041). All patients were treated with a single weight-based infusion of rese-cel at a dose of 1 x 106 cells/kg following lymphodepletion (flu 25 mg/m2/d on Days -5, -4, & -3, & cy 1,000 mg/m2/d on Day -3). All non-glucocorticoid immunomodulatory agents were stopped prior to lymphodepletion. Glucocorticoids were tapered post-infusion. Patients were not given prophylactic tocilizumab pre-infusion. Translational assessments were determined as follows: CART PK was evaluated by flow cytometry & dPCR; B-cell enumeration & phenotyping were evaluated by flow cytometry; & serum cytokines were evaluated via immunoassay. Results: Across all indications (19 patients), 7 episodes of CRS were observed (7/19). 6 of 7 patients experienced a Grade 1 CRS, & 1 patient had a Grade 2 CRS. 2 patients developed ICANS; 1 patient had a Grade 4 (previously presented at ACR Convergence 2024), & the other patient had a Grade 3. Both ICANS events were rapidly resolved without sequelae using standard therapy. These CRS and ICANS events differed across indications. In IIM, 4 of 8 patients experienced Grade 1 CRS and no patients experienced ICANs. In SLE, 2 patients experienced Grade 1 CRS and one developed Grade 4 ICANs. In SSC, 1 of 2 patients experienced Grade 2 CRS & the other experienced Grade 3 ICANS. In MG, no CRS or ICANS was observed. Rese-cel infusion products were ~ 66% transduced, with a mean CD4:CD8 ratio of 6. Post-infusion, peak CART expansion was observed on day 12 & the mean peak expansion was 89 cells/µL (SD +/- 214). B cells were rapidly depleted from peripheral blood post-infusion, with nadir occurring at 13 days & repopulation beginning at 62 days post-infusion on average. Re-emergent B cells were largely of the CD24+CD38+ transitional naïve phenotype. Commensurate with B cell depletion, serum BAFF induction was observed, with a mean fold induction of 16 & mean peak level of 56,404 pg/µL (SD: +/- 45,547). Serum cytokines associated with CRS & ICANS; IFNg, IL6, & IL8 reached peak elevation on days 10, 9, & 10 respectively post-infusion. The mean peak elevation for IFNg, IL6, & IL8 were 232 pg/ µL (SD: +/- 320), 44 pg/ µL (SD: +/- 60), & 80 pg/ µL (SD +/- 73), respectively. Conclusion: Rese-cel appears to be well tolerated in autoimmune patients with rates of CRS & ICANS of 37% & 11% across all autoimmune indications, respectively (SLE patients had a 25% CRS rate & 12.5% ICANS rate, IIM patients had a 50% CRS rate & 0% ICANS rate, SSC patients had a 50% CRS & 50% ICANS rate, & MG patients had a 0% CRS & ICANS rate). Notably, the frequency & severity of CRS appears to be low, with no CRS in 63% of patients (12/19) & most CRS events (6/7) being Grade 1. The lower frequency & severity of both CRS & ICANS events are likely correlated with the relatively modest induction of serum IFNg, IL6, & IL8 found in autoimmune patients post rese-cel weight-adjusted dose infusion. High BAFF levels reflect deep systemic B cell depletion. Together, these data support further support late-stage clinical development of rese-cel in autoimmune disease.

Abstract Introduction Understanding the tumor immune microenvironment (TIME) in multiple myeloma (MM) is critical for identifying prognostic biomarkers and therapeutic targets. We hypothesized that specific immune cell populations within the bone marrow TIME are associated with clinical outcomes and could enhance patient stratification. To this end, we conducted the largest CyTOF-based immune profiling study to date in MM, profiling >300 million bone marrow immune cells from patients enrolled in the CoMMpass study. Methods We analyzed >1,000 bone marrow samples from ~700 newly diagnosed MM patients enrolled in the MMRF CoMMpass study (NCT01454297), using a harmonized protocol and 41-antibody CyTOF panel at five academic centers. The majority were baseline samples (n=671), with additional samples spanning up to 3.5 years post-diagnosis (n>300). Each site contributed ~200 samples on average. The cohort included 40% females, with a median age of 63 years and BMI of 27.8. ISS stages were evenly distributed (I: 36%, II: 36%, III: 29%), and 84% had ECOG 0-1. Autologous stem cell transplant was performed in 58% of patients, and 22% received maintenance therapy. Results We profiled 311 million single cells across the cohort, identifying major immune compartments: Neutrophils (36.0%), CD4 T cells (19.7%), NK cells (9.9%), monocytes (7.6%), B cells (5.7%), and CD8 T cells (5.6%). Less abundant populations included dendritic cells, γδ T cells, plasmablasts, and pDCs. In our preliminary analysis with multivariate cox regression modeling, we identified clinical factors including stem cell transplant, ISS stage, and ECOG status, as independent predictors of overall survival (OS). After adjusting for age and sex, higher proportions of naïve B cells, pDCs, and γδ T cells were independently associated with improved OS, while CD4+ T cell abundance was positively associated with prolonged progression-free survival (PFS). In contrast, elevated neutrophil proportions were significantly linked to inferior OS and demonstrated a trend toward shorter PFS. Longitudinal immune profiling in more than 200 patients identified a gradual accumulation of naïve B cells and a pronounced post-treatment rise in CD45+ plasmablasts, highlighting persistent B cell activity as a candidate mechanism for sustained therapeutic efficacy. Conclusion This study presents the largest immune proteomic atlas in MM to date, linking TIME composition to clinical outcomes. A favorable immune landscape was marked by preserved humoral immunity (naïve B cells, plasmablast expansion), enrichment of pDCs and γδ T cells, and lower neutrophil abundance, collectively associated with improved survival outcomes. These findings lay the groundwork for immune-guided prognostication and highlight new opportunities for therapeutic targeting in MM. Ongoing analyses aim to further resolve baseline and longitudinal immune dynamics in relation to treatment response.

Abstract INTRODUCTION Immune checkpoint inhibitors (ICIs) have emerged as a cornerstone of cancer therapy. However, despite their efficacy, ICIs can cause a unique spectrum of autoimmune toxicities known as immune-related adverse events (irAEs). In contrast to irAEs involving other organ systems, hematologic irAEs such as ICI-associated immune thrombocytopenia (ICI-ITP) are less common. As a result, existing data on ICI-ITP are largely confined to isolated case reports, single-center case series, and pharmacovigilance databases lacking granular clinical data. To address this key knowledge gap, we sought to comprehensively characterize the incidence, risk factors, clinical features, treatments, and outcomes associated with ICI-ITP. METHODS We performed a multicenter cohort study of adult patients who initiated ICI therapy at 29 hospitals across 7 major cancer centers in 11 geographically diverse states in the US between 2016-2023. ICI-ITP was defined as a decline in platelet count to <100 x109/L in the absence of an alternative more plausible etiology of thrombocytopenia. Patients with hematologic malignancies were excluded, as were those who received cytotoxic chemotherapy within 30 days prior to thrombocytopenia onset. For each patient identified as having ICI-ITP, we collected data on demographics, comorbidities, medications, laboratory values, treatments, and outcomes. All data were extracted by manual chart review. To identify risk factors for ICI-ITP and to examine its association with death, for each case of ICI-ITP we also collected data on two control patients who initiated ICI therapy at the same site and in the same year but without development of ICI-ITP. We used multivariable logistic regression to identify independent risk factors for ICI-ITP and its recovery (the latter defined according to criteria established by the ITP International Working Group). We used a multivariable Cox model to examine the association between ICI-ITP and its severity with all-cause mortality. RESULTS We identified a total of 214 patients who developed ICI-ITP among 86,648 examined (0.25%). The incidence was highly consistent across sites (range, 0.15–0.30%). Risk factors for ICI-ITP included simultaneous receipt of two different classes of ICIs, additional irAEs, and a lower baseline platelet count. ICI-ITP developed at a median of 8 weeks (IQR, 4–18) following ICI initiation. The median nadir platelet count was 41 x109/L (IQR, 17–64). Six out of 214 patients (2.8%) developed a major bleed, and 16 patients (7.5%) developed a clinically relevant non-major bleed. Cell-bound anti-platelet antibody testing was positive in 6 of 8 patients (75.0%) for whom it was assessed, and median circulating thrombopoietin levels (201 pg/mL [IQR, 178–295]) were similar to those seen in primary ITP. Treatments used included glucocorticoids (49.5%), immune globulin (18.2%), and thrombopoietin receptor agonists (13.6%). Overall, 161 patients (75.2%) recovered from ICI-ITP at a median of 2 weeks (IQR, 1–5) following ICI-ITP onset. Predictors of recovery included additional irAEs and a higher nadir platelet count. Treatment response was observed in 57.0%, 56.4%, and 58.3% of patients within 60 days following initiation of glucocorticoids, immune globulin, and thrombopoietin receptor agonists, respectively. Of 76 patients rechallenged with an ICI, 23 (30.3%) developed recurrent ICI-ITP at a median of 9 weeks (IQR, 2–15); 17 of these 23 patients (73.9%) recovered. Finally, ICI-ITP and its severity were monotonically associated with a higher risk of death (adjusted hazard ratio 2.96 [95% CI, 2.14–4.08] for those with severe ICI-ITP compared to those without ICI-ITP). CONCLUSION In this multicenter study involving patients treated with ICIs at 29 hospitals across the US, we identified 214 patients with ICI-ITP. For the first time, we characterize in detail the incidence, risk factors, clinical features, treatments, and outcomes of ICI-ITP, a relatively novel disease entity. We found that more than half of patients responded to conventional ITP-directed therapies, and that most patients rechallenged did not develop recurrent ICI-ITP. We also found that ICI-ITP was independently associated with an increased risk of death, including a nearly 3-fold increased risk in those with severe ICI-ITP. These findings highlight ICI-ITP as a rare but clinically significant complication of ICI therapy that hematologists must be equipped to recognize and manage.