Citations

Background: Anaplastic large cell lymphomas (ALCLs) are CD30-positive T-cell lymphomas that share pathologic features but differ in clinical presentation, outcome, and molecular features. The World Health Organization (WHO) and International Consensus Classification (ICC) classify ALCLs by presence or absence of ALK rearrangements (R) and clinical presentation (systemic, cutaneous [c], or breast implant-associated [BIA]). ICC, but not WHO, recognizes DUSP22-R as defining a new genetic subtype of ALK- ALCL. The classifications otherwise do not reflect additional molecular heterogeneity in genetics (e.g., TP63-R) or therapeutic vulnerabilities (e.g., JAK-STAT3 pathway activation). Methods: ALCLs (N=689) underwent expert consensus review (WHO/ICC) through the Lymphoma/Leukemia Molecular Profiling Project (LLMPP). All cases also underwent genetic subtyping (ALK, DUSP22, TP63, and triple-negative [TN]) using fluorescence in situ hybridization (FISH) and immunohistochemistry (IHC), as well as IHC for phospho-STAT3 Tyr705 (pSTAT3). RNAseq was performed and evaluable in 393 cases; the remaining cases had insufficient tissue, tumor content, RNA yield or quality, and/or sequencing data quality. Sequenced and non-sequenced sub-cohorts had similar demographics and subtype distribution. Results: Unsupervised gene expression profiling (GEP) identified 2 main molecular types of ALCL. Type I ALCLs predominantly included ALK+ ALCLs, BIA-ALCLs, and a subset of TN ALCLs (designated TN-I), whereas Type II ALCLs predominantly included ALCLs with DUSP22-R, TP63-R, or both (double-hit; DH), and the remaining TN ALCLs (TN-II). Type I ALCLs were strongly associated with pSTAT3 expression (74.2±26.4% positive malignant cells vs 9.9±22.7% for Type II; P<0.0001). An independently derived pSTAT3 staining threshold of 30% assigned Type I vs II with 91% accuracy. A third cluster of ALCLs, mostly cALCL, showed an epithelial GEP signature rather than a lymphoma signature; these cases were assigned to Types I or II based on pSTAT3 IHC. Distinct sub-signatures were identified for ALK+, DUSP22-R, TP63-R, and BIA ALCL (Fig. 1), but not ALK- ALCL or cALCL. Gene set enrichment analysis showed Type I ALCLs to be enriched for JAK-STAT3 signaling genes (normalized enrichment score [NES], 2.22; FDR<0.0001) and related pathways, such as TNFα-NFκB signaling (NES, 2.21; FDR<0.0001). In contrast, Type II ALCLs were enriched for cell cycle, DNA repair, epigenetic, and metabolic pathway genes, but not for major tyrosine kinase-mediated signaling pathway genes. Enriched epigenetic pathways included chromatin modifying enzymes (NES, -1.86; FDR=0.002) and histone methylation (NES, -1.71; FDR=0.002). EZH2 was the most overexpressed gene in Type II ALCLs (fold-change, 7.74; FDR=8.84×10 -306). At the protein level, EZH2 IHC H-scores were 275±45 in Type II and 168±70 in Type I ALCLs (P<0.0001); H3K27me3 H-scores were 170±81 and 76±63, respectively (P<0.001). The top metabolic gene set involved cholesterol biosynthesis (NES, -1.88; FDR=0.002). Overall survival (OS) data were available in 257 systemic ALCL patients (145 sequenced and 112 non-sequenced; cALCL and BIA-ALCL were excluded). Non-sequenced TN ALCLs were assigned to TN-I or TN-II using pSTAT3 IHC. The results supported earlier data indicating favorable prognosis of DUSP22-R ALCL (5 y OS, 95%) and ALK+ ALCL (87%), intermediate prognosis of TN ALCL (TN-I, 52% and TN-II, 38%; P=NS), and poor prognosis of TP63-R/DH ALCL (0%)(P<0.0001; Fig. 2). Conclusions: Two overarching molecular types of ALCL exist, predominantly associated with the presence (Type I) or absence (Type II) of the JAK-STAT3 signaling program. Distinct GEP signatures exist for ALK+ ALCL and BIA-ALCL (predominantly Type I), and DUSP22-R ALCL and TP63-R ALCL (predominantly Type II). TN ALCLs lacking ALK-R, DUSP22-R, and TP63-R can be stratified into TN-I and TN-II subtypes. pSTAT3 IHC has >90% accuracy as a surrogate for GEP-based subtyping. ALK- ALCL and cALCL cluster by molecular subtype rather than by defining GEP signatures. Type II ALCLs are enriched for targetable epigenetic and metabolic pathways, including EZH2/histone methylation and cholesterol biosynthesis. This molecular classification is diagnostically, prognostically, and potentially therapeutically relevant, and can be applied using FISH and IHC in routine practice and in the clinical trial setting.

Background: An interim analysis (data cutoff date April 18, 2022) of the international, open-label, phase 3 KarMMa-3 trial (NCT03651128) showed the chimeric antigen receptor T cell therapy idecabtagene vicleucel (ide-cel) significantly improved progression-free survival and treatment response rates, as well as HRQoL, as compared with standard regimens, in patients with triple-class exposed (TCE) RRMM who had received 2-4 prior regimens (Rodriguez-Otero P, et al. N Engl J Med 2023;388:1002-1014; Delforge M, et al., HemaSphere, 2023;7(S3):P905). The aim of the present study was to analyze the effect of ide-cel versus standard regimens (daratumumab [DARA], pomalidomide, and dexamethasone [DEX]; DARA, bortezomib, and DEX; ixazomib, lenalidomide, and DEX; carfilzomib and DEX; or elotuzumab, pomalidomide, and DEX) on HRQoL in KarMMa-3 trial participants with an extended follow-up of patient-reported outcomes (PRO) data collection (data cutoff date April 28, 2023). Methods: In KarMMa-3, HRQoL was assessed using the European Organisation for Research and Treatment of Cancer (EORTC) Quality-of-Life Questionnaire-Core 30 (QLQ-C30), EORTC Quality of Life Questionnaire Multiple Myeloma Module (QLQ-MY20), and EQ-5D-5L. HRQoL assessments were completed at screening (baseline); the day of ide-cel infusion or the first dose of standard treatment; and monthly from 2 to 28 months; and every 3 months thereafter. The prespecified primary outcomes of interest were EORTC QLQ-C30 global health status/QoL, physical functioning, cognitive functioning, fatigue, and pain; EORTC QLQ-MY20 disease symptoms and side effects of treatment; and the EQ-5D visual analog scale (EQ-VAS). Least squares (LS) mean changes from baseline through month 25 for ide-cel and standard regimens were compared by constrained longitudinal data analysis (cLDA). Effect sizes were estimated by Hedges' g. The effect of ide-cel versus standard regimens on time to confirmed improvement and deterioration in HRQoL (defined by prespecified change thresholds and sustained duration) were compared by stratified Cox proportional hazards regression. There was no adjustment for multiple testing, and all P values reported are nominal. Results: Overall, 254 patients were randomized to the ide-cel arm and 132 to the standard regimens arm. Baseline characteristics were well balanced across treatment arms; both arms had a median age of 63 years, a median of 3 prior anti-myeloma regimens, and median time from MM diagnosis to screening was 4 years. The baseline HRQoL scores were generally comparable between treatment arms. Overall LS mean changes from baseline to month 25 showed significant differences (nominal P value < 0.05) in favor of ide-cel for 18 out of 21 HRQoL domains, with effect sizes (Hedges' g) of 0.27 to 0.82 (Table). For 13 of these domains (including 6 primary domains of interest), the difference in overall LS mean change from baseline exceeded the prespecified minimum important difference (MID) threshold. Time to confirmed improvement was significantly shorter in the ide-cel arm than the standard regimens arm for 19 domains, including all primary domains of interest (Table). Time to confirmed deterioration showed trends of being longer in the ide-cel arm than the standard regimens arm for most domains and was significantly longer for emotional functioning, cognitive functioning, social functioning, dyspnea, and constipation as measured by the EORTC QLQ-C30. The time to confirmed deterioration in side effects of treatment (based on the EORTC QLQ-MY20) was significantly longer for patients in the ide-cel arm and time to confirmed improvement was significantly faster, as compared with the standard regimens arm (Table). Conclusions: In the KarMMa-3 trial, patients with TCE RRMM who had received 2-4 prior regimens, ide-cel significantly and meaningfully improved MM-relevant symptoms, functioning, and overall health status/HRQoL compared with standard regimens. Confirmed HRQoL improvements in the ide-cel arm occurred sooner than standard regimens. PRO improvements were sustained over more than 2 years, which is notable after a single infusion treatment with ide-cel. These patient-reported findings further support the benefit of ide-cel in this patient population.

Background: Waldenström Macroglobulinemia (WM) is an indolent lymphoma with heterogeneous outcomes. The recently developed, Revised International Prognostic Scoring System for WM (rIPSS-WM) did not examine the prognostic impact of MYD88L265P, a somatic alteration present in most patients with WM. Moreover, rIPSS-WM could only be partially validated in an external cohort. Here, we report the performance of rIPSS-WM and propose a refined prognostic model with external validation. Methods: We reviewed medical records of treatment-naïve patients with active WM consecutively seen at Mayo Clinic, Rochester (MCR) between 01/01/1996 and 12/31/2017. We applied the rIPSS-WM model in patients with data available for all required parameters, including age, serum beta-2 microglobulin (β2M), serum lactate dehydrogenase (LDH) and serum albumin at diagnosis to assess the prognostic utility of this model. We then identified the dichotomized clinical and laboratory parameters prognostic for overall survival (OS) by univariate analyses (UVA) in our cohort. The significant parameters were analyzed to independently predict OS in a multivariable analysis (MVA) and to create a prognostic model, which was subsequently validated in an independent multi-institutional cohort. Results: We identified 889 patients with active WM at MCR, with a median follow-up of 8.2 [95% confidence interval (CI): 7.5-9] years. The estimated median OS for this cohort was 10.6 (95%CI: 9.7-11.4) years. Patients with available data in this cohort for the validation of rIPSS-WM (n=241) were risk stratified into very low (n=46), low (n=64), intermediate (n=58), high (n=46) and very-high risk groups (n=27) per rIPSS-WM, with the respective 5-year OS rates of 96%, 76%, 72%, 77% and 32%, demonstrating overlapping survival curves for the low, intermediate and the high-risk groups. Because rIPSS-WM could not be validated, we used the data from the MCR cohort and identified age at diagnosis of active WM, LDH >upper limit of normal (ULN), β2M >3µg/mL and albumin <3.5g/dL as statistically significant prognostic markers for OS on UVA. Notably, the presence of MYD88 L265P did not impact OS [median 11.4 years (95%CI: 10.1-not reached (NR) for MYD88 L265P genotype (n=231) versus 10 years (95% CI: 8.1-NR) for MYD88 WT genotype (n=65); p=0.33]. On MVA, LDH >ULN (HR 2.34), serum albumin <3.5 g/dL (HR 1.5) and age groups 65-75 years (HR 1.4) and >75 years (HR 2.6) were independently prognostic. Based on comparable HRs, we assigned a score of 1 point each to serum albumin <3.5 g/dL and age group 66-75 years and similarly, 2 points each for age >75 years and serum LDH >ULN per their impact on OS. Patients with the composite scores of 0, 1 and 2 were categorized as low risk, low- intermediate risk and intermediate risk, respectively. Due to similar OS for the scores of 3 to 5, these were combined to form the high-risk group of our proposed model: Modified Staging System for WM (MSS-WM). The median OS was 14.6 years [95%CI: 9.1 years-NR] for low-risk (score 0, n=71, 21%), 11.2 years (95% CI: 9.1-15.2 years) for low-intermediate risk (score 1, n=110, 32%), 8.1 years (95% CI: 6.4-12.2) for intermediate risk (score 2, n=81, 24%) and 5.5 years (95 % CI: 3.9-11 years) for the high-risk group (score ≥3, n=79, 23%); p<0.0001. In this derivation cohort, the 5-year OS rate was 93%, 82%, 69% and 55% and 10-year OS rate was 60%, 53%, 45% and 30% for the low, low-intermediate, intermediate, and high-risk cohort, respectively ( Figure 1A). We also validated MSS-WM using competing risk analysis (p=0.001) and in the cohort of rituximab treated patients (p<0.0001). We then used data from an independent series of 335 treatment-naïve symptomatic patients from academic institutions (n=5) across the US and Europe for external validation of the MSS-WM model. The median follow-up for the validation cohort was 6.3 years (95% CI: 5.5-7.2 years). The 5-year OS was 93%, 90%, 75% and 57% and the estimated 10-year OS was 91%, 80%, 64% and 35% for the low risk (n= 86, 26%), low-intermediate (n=107, 32%), intermediate (n=86, 26%) and high-risk (n=59, 18%) cohorts, respectively (p<0.0001, Figure 1B). Conclusions: Our proposed model, MSS-WM, is a simple, clinically useful, externally validated model that reliably risk stratifies previously untreated patients with active WM into four groups that have distinct outcomes based on the composite scores derived from the patients' age, serum albumin and serum LDH at diagnosis.

Introduction Newly-diagnosed multiple myeloma (MM) patients often share tumor genetic abnormalities associated with a high risk of progression and poor survival. We hypothesized that single-cell transcriptomic characterization of the tumor immune microenvironment (TIME) may be combined with tumor-based risk stratification to uncover novel disease biology mechanisms associated with clinical outcomes and improve risk stratification. Methods We performed single-cell RNA sequencing of 361 CD138 negative sorted bone marrow mononuclear cells (BMMCs) samples from 263 MM patients enrolled at time of diagnosis in the Multiple Myeloma Research Foundation CoMMpass study (NCT01454297). Specimens were analyzed at baseline (n=263) across a network of five collaborating academic research centers following assay harmonization. Genetic factors that determine risk were defined as one or more of the following gain- or loss-of-function events derived from baseline genomic sequencing data: del17p14, translocations of the Ig locus with the MAF/A/B or WHSC1/MMSET/NSD2 loci, and gain of chromosome 1q21. Statistical modeling strategies were used to assess single-cell populations or RNA signatures associations with known risk factors or disease progression (FDR<0.05). Results The study generated high-quality single-cell profiles of 731,643 cells from 123 high-risk and 107 standard-risk MM patients based on genetics. Based on distinct transcriptome profiles, BMMCs formed 106 clusters corresponding to 5 major compartments: NK and T cells, B cells, erythrocytes, myeloid cells, and plasma cells. Our preliminary differential abundance analysis comparing high- vs standard-risk patients at baseline uncovered significant differences in the TIME associated with genetic risk. The high-risk group was identified as enriched in effector CD8+, and CD4+ T cells. In contrast, the high-risk group exhibited depletion of naive CD8+ and interferon-associated CD4+ and CD8+ T as compared to the standard-risk group. Within subclusters in the T and NK cells compartment, a higher cytotoxicity signature was observed in high-risk patients, characterized by the expression of granzymes and perforin. Finally, our preliminary survival and progression analysis associated a higher frequency of IDO1-expressing macrophages in worst progression patients, while better outcomes were associated with CD8+ T and B cells expressing androgen response and IL6/STAT3/EMT programs, respectively. Conclusions Current approaches for risk stratification are based on bulk measurements of tumor's genetic or epigenetics features, which do not capture the diversity and dynamics of the TIME. Through single-cell transcriptomics, we were able to capture, in high granularity, the TIME and correlate specific immune cell populations and phenotypes with relapse risk and poor prognostic outcomes. These results suggest that immune subpopulations may be an essential novel aspect for improving current risk stratification models.

Background: The anti-apoptotic protein Mcl-1 contributes to the pathophysiology of certain B-cell malignancies and AML, and dependence on Mcl-1 is associated with resistance to venetoclax. Efforts to target Mcl-1 through direct inhibition have faced challenges due to cardiovascular toxicities. Voruciclib, a potent oral CDK9 inhibitor (CDK9i), was shown to indirectly downregulate Mcl-1 protein levels and synergize with venetoclax in preclinical models in lymphoid and myeloid malignancies. In solid tumors, voruciclib maximum tolerated dose (MTD) was defined as 350 mg with continuous daily dosing, and 600 mg when used on days 1-14 on a 21-day cycle. This phase 1 study is the first to evaluate the safety, dose-limiting toxicities (DLT), preliminary efficacy, pharmacokinetics, and pharmacodynamics of voruciclib in R/R hematologic malignancies (NCT03547115). Methods: Patients (pts) with relapsed B-cell NHL, CLL, or AML aged ≥18 years who had ECOG performance status ≤1, disease progression after standard therapies, adequate hepatic and renal function, and no prior CDK9i were eligible. The study followed a 3+3 design and DLTs were assessed in Cycle 1. Voruciclib was administered daily continuously at a starting dose of 50 mg on a 28-day cycle (Cohort I). After 2 DLTs were observed at 100 mg in AML pts, dosing was changed to daily on days 1-14 on a 28-day cycle (Cohort II). Also, pts with prior allogeneic transplant were excluded and dose escalation (100, 150, 200 mg) proceeded separately in AML and B-cell malignancies. Disease response was assessed by investigators using the Lugano, 2008 iwCLL, and 2017 ELN criteria. Results: 40 pts were enrolled: 21 AML, 19 NHL (9 DLBCL, 3 CLL, 3 MCL, 3 FL and 1 MZL). Cohort I enrolled 16 pts, 8 each at 50 mg and 100 mg. Cohort II enrolled 24 pts (17 AML and 7 B-cell malignancies), including 6 at 100 mg, 7 at 150 mg, and 11 at 200 mg. Dose escalation in Cohort II did not reach the MTD and was stopped at 200 mg, a dose that achieved plasma concentrations sufficient for target inhibition. Median age was 70 years (range 40-90), 72.5% of pts were ≥65 years, and 77.5% had ECOG performance status = 1. Pts had a median of 3 prior therapies (range, 1-8). Median duration on drug was 5 weeks (range 1-22). Drug-related serious adverse events (AEs) were reported in Cohort I only. Two pts with AML had a DLT of Grade (Gr) 3 interstitial pneumonitis at 100 mg; both had had allogeneic transplant and active graft versus host disease, one of whom also had a concurrent Gr 3 differentiation syndrome. One pt with DLBCL at 50 mg had Gr 3 hypoxemic respiratory failure in Cycle 4. In Cohort II, no DLTs were observed and there were no non-hematologic Gr 3-5 drug-related AEs. In the overall population, the most common AEs were diarrhea (30%), nausea (25%), anemia (22.5%), fatigue (22.5%), constipation (17.5%), dizziness (15%) and dyspnea (15%) [Table 1). One pt with AML had Gr 3 hyperbilirubinemia. There was no evidence of drug-related neutropenia in pts with NHL. No significant cardiovascular AEs were observed. One pt with AML in Cohort I achieved a morphologic leukemia-free state. At 200 mg, 5 of 10 AML pts (50%) had stable disease. Voruciclib pharmacokinetics was dose proportional, mean accumulation ratio was 2.4, mean T max was 4 hours, and at 200 mg, steady state C max was 925 ng/mL and mean C trough was 442 ng/mL (~1 microM). Analysis of Mcl-1 protein expression in normal lymphocytes by flow cytometry showed a trend of decrease in on-treatment samples compared to pre-treatment. Consistent with this, RNA-Seq analysis of 3 paired CLL samples confirmed rapid (after 6 hours) downregulation of Mcl-1 mRNA transcript in the malignant B-cells. Furthermore, scRNA-Seq analysis of 2 longitudinal AML samples demonstrated downregulation of MYC transcriptional program and oxidative phosphorylation in circulating AML blasts, both evident 24 hours after start of voruciclib and ongoing at C1D15. Conclusions: Voruciclib at doses up to 200 mg administered on 14 consecutive days on a 28-day cycle was well tolerated, with no DLTs observed, and no cardiovascular toxicities. There was no significant myelosuppression in pts with B-cell malignancies. As the safety profile suggested non-overlapping toxicities with venetoclax, the study is now evaluating voruciclib in combination with venetoclax in pts with R/R AML to exploit dual inhibition of Bcl-2 and Mcl-1.

Background: Oncolytic virotherapy, an innovative immunomodulatory therapeutic approach, is currently being investigated for its potential in treating relapsed refractory (R/R) hematologic malignancies. This study focuses on the use of an engineered Vesicular Stomatitis Virus called VSV-IFNβ-NIS, a bullet-shaped rhabdovirus which encodes both interferon beta (IFNβ) and sodium iodine symporter (NIS) with particular tumor cell tropism. Viral IFNβ serves as an indicator of viral proliferation and boosts the host's anti-tumor immune response. This is a phase 1 clinical trial with the identifier NCT03017820, wherein patients (pts) with R/R multiple myeloma (MM), and non-Hodgkin lymphoma (NHL) are treated with a single systemic administration of VSV-IFNβ-NIS. Study Design and Methods: Patient Eligibility Criteria: The Mayo Clinic Institutional Review Board approved the protocol. The study includes adult patients with relapsed or refractory multiple myeloma (MM), or T-cell lymphoma (TCL). For MM or plasmacytoma, patients should have <15% plasma cells and a largest diameter of <5 cm. TCL patients should have a mass <5 cm. Study Design: This is a single-center phase 1 trial aiming to determine the maximum tolerated dose of VSV-IFNβ-NIS in patients with R/R MM, and TCL. Secondary objectives included assessing the safety profile and estimating the clinical response rate of VSV-IFNβ-NIS. Correlative objectives involved monitoring viral replication through plasma IFN-β transgene levels and using SPECT/CT imaging with the NIS gene. Viral genomes were measured through quantitative reverse transcription (qRT) polymerase chain reaction (PCR) of VSV RNA, and any virus shedding was examined in body fluids. The tested doses were 5 × 109 TCID50, 1.7 × 1010, 5 × 1010, and 1.7 × 1011 TCID50. The virus was administered via IV infusion over 30 minutes, and patients received 650 mg acetaminophen and 50 mg diphenhydramine hydrochloride before infusion and every 6 hours for 24 hours to manage fever and hypotension. Two expansion cohorts at DL4 have been added (20 additional PTCL, 10 B-cell lymphoma and histiocytic neoplasms) to obtain additional data on tolerability and efficacy and biomarkers of response in these histologies. Safety and Response Assessments: All patients undergo toxicity evaluation, and adverse events (AEs) assessment using the Common Terminology Criteria for Adverse Events (CTCAE) version 4 criteria, except for cytokine release syndrome (CRS) grading, which follows the 2014 Lee criteria. Dose escalation and tolerability are based on the standard 3+3 phase 1 trial design. Ruxolitinib is added from days 2-5 or day -1 to day 9 depending on the arm to mitigate IFNβ toxicity. Response assessment followed the International Myeloma Working Group criteria for MM and the Lugano Classification for NHL and histiocytic neoplasms. Pharmacodynamic and Pharmacokinetic Analyses: For viremia and viral shedding assessment, RNA is isolated from whole blood, mouth rinse, buccal swabs, and urine samples. Quantitative RT-PCR determined VSV-nucleocapsid (N) RNA copy number, and virus shedding is assessed through coculturing samples on Vero cells. Anti-VSV antibodies are evaluated using the standard plaque reduction neutralization assay. Plasma cytokine levels are measured with a 30-plex human cytokine kit, and human IFN-β levels is quantified using an enzyme-linked immunosorbent assay (ELISA) kit. Statistical Analysis: Incidence of dose-limiting toxicities and AEs are tabulated by dose level (DL). AEs are summarized by type, maximum grade, and duration across patients in each DL. Spearman rank correlation tests are used evaluated correlations between laboratory-based measures, and Wilcoxon rank sum tests compared continuous measures between response groups. Response rates will be summarized by DL and disease group, with estimated response rates and corresponding 95% binomial confidence intervals calculated.

Introduction: Immune-mediated graft failure (also termed graft rejection) is a feared complication of hematopoietic stem cell transplant (HSCT) and there is no effective intervention. Multiple pre-transplant risk factors for graft failure are known and recent studies from our groups identified interferon gamma (IFNγ) as a mechanistically important and pharmacologically targetable cytokine involved in graft failure pathophysiology. CXCL9, a downstream marker of IFNγ production, was also identified as a biomarker. We describe an international multicenter retrospective pooled analysis of emapalumab (EMA), an IFNγ neutralizing agent, for the prevention/treatment of immune-mediated graft failure after HSCT. Methods: The safety and preliminary efficacy of EMA were studied in predominantly pediatric patients who received EMA using two different strategies: 1) graft failure treatment and 2) graft failure prophylaxis. Patients at US centers were treated with EMA if they developed clinical and laboratory signs of graft failure after HSCT. All treatment patients met 2 or more of the following criteria: 1) HLA mismatched donor and/or prior graft failure, 2) unexplained fever >39°C (or >38°C if prior history of graft failure) after HSCT, 3) absolute neutrophil count decline after initial engraftment, 4) delayed neutrophil engraftment after HSCT, 5) real-time CXCL9 > 2.5x upper limit of normal (ULN). Temperature and CXCL9 cutoffs were based on a prior graft failure biomarker publication (Sabulski et al, Blood Advances 2021). Patients treated at European centers received prophylactic EMA to prevent graft failure if they were deemed at high-risk prior to transplant. High-risk patients had a prior history of graft failure after HSCT or had two or more of the following established risk factors: a) affected by a disease with high risk of graft failure, including acquired severe aplastic anemia, thalassemia, primary HLH; b) ex-vivo T cell-depleted stem cell product; c) transplant from unrelated cord blood unit or mismatched related/unrelated donor. Results: Thirty-six HSCTs used EMA for graft failure treatment or prophylaxis and were included in this study.EMA was used for graft failure treatment in 25 HSCTs wherein patients developed clinical and laboratory signs of graft failure (Table 1). Most HSCTs in the treatment cohort were performed for bone marrow failure syndromes (n=16). All patients tolerated EMA without complications and no side effects or infections were attributed to EMA. Fifty six percent (14/25) of patients treated with EMA for impending graft failure engrafted and had resolution of graft failure concerns, while the remaining 44% suffered graft failure. Responders had a median ANC of 0.29 x10 3 cells/μL (range 0.05-0.79) prior to EMA, which increased to a median of 0.6 x10 3cells/μL (range, 0-5.1) and 2.97 x10 3 cells/μL (range, 0.85-8.7) on days 3 and 7 after EMA dosing. Eighty percent (20/25) of CXCL9 levels were elevated in the treatment cohort prior to EMA dosing. The median CXCL9 level prior to EMA dosing was elevated 2.4 x ULN (range, 0.3-157.8). The median pre-EMA CXCL9 level in patients who ultimately suffered graft failure in the treatment cohort was 3.3x ULN (range, 0.7-157.8) compared to a median of 2.2x ULN (range, 0.3-30.4) in subjects who had signs of graft failure but did not suffer graft failure (p=0.24). Eleven patients received prophylactic EMA based on a prior history of graft failure or other pre-transplant risk factors. EMA was well tolerated in these patients and 91% (10/11) of patients engrafted while one patient with prior history of graft failure suffered recurrent graft failure. The pre-HSCT (day -2) CXCL9 level in the lone prophylactic patient who suffered graft failure was 4x the ULN. Conclusions: This study reports the largest cohort of HSCTs that used EMA therapy for the novel purpose of treating or preventing graft failure. EMA was well tolerated in all patients. Efficacy conclusions are preliminary and limited by the study design, however our treatment study successfully identified a cohort of patients with a markedly high graft failure rate (44%) and we show preliminary evidence to suggest emaplaumab prevented graft failure in patients who had signs of impending graft failure at the time of treatment. The safety and preliminary efficacy observations in this study support a larger prospective study of IFNγ blockade for graft failure prevention after HSCT.

Peripheral T-cell lymphoma (PTCL) encompasses a diverse group of post-thymic lymphomas, with ~40% of cases not further classifiable and designated as PTCL-not otherwise specified (PTCL-NOS). Two molecular prognostic subtypes of PTCL-NOS, PTCL-TBX21 and PTCL-GATA3, were identified through gene expression profiling (GEP). A subset of PTCL-NOS with T-follicular helper (TFH) differentiation was subsequently categorized as nodal PTCL with TFH phenotype (PTCL-TFH), also known in current classifications as nodal PTCL-TFH, NOS, or TFH lymphoma, NOS. However, the boundary between PTCL-NOS and PTCL-TFH remains poorly defined. To refine these subtypes, 101 PTCL-NOS with centralized pathology review and extensive immunophenotyping by immunohistochemistry (IHC) using 22 antibodies, digital GEP (nCounter, NanoString Inc) (n= 70), and RNA sequencing (n= 73) were included in this study. IHC was used to identify PTCL-TFH cases (those with strong expression of more than two TFH markers); the remaining patients were classified as PTCL-GATA3 and PTCL-TBX21 using pre-defined GEP signatures. Cases were reclassified into PTCL-NOS (n= 63) and PTCL-TFH (n= 38), with PTCL-NOS subclassified into PTCL-GATA3 (n= 22; 34%) and PTCL-TBX21 (n= 41; 66%) and showing significant differences in OS (p<0.001). PTCL-GATA3 was characterized by medium to large transformed cells (average= 90%, 70-100%) and had minimal tumor microenvironment (TME) (TME-poor: 100%). In contrast, PTCL-TBX21 was more heterogeneous, associated with pleomorphic cells in a polymorphous background (TME-rich: 78%), including Lennert lymphoma-like cases (22%). mRNA and CIBERSORT analysis substantiated the findings and identified a subset enriched in cytotoxic features in the PTCL-TBX21 subtype. IHC indicated that PTCL-GATA3 cases were CD4+/CD8- (83%) or CD4-/CD8- (17%) and lacked expression of CD8 or cytotoxic markers compared to PTCL-TBX21 (p<0.01). PTCL-GATA3 showed significantly higher expression of LEF1 (average= 80%), Ki67 (80%), and MYC (25%) than PTCL-TBX21 (25%, 30%, <5%, respectively, p<0.01). mRNA and protein expression of these biomarkers showed a significant positive correlation (r=0.5, p<0.001), and expectedly higher mRNA expression of LEF1 and MYC was observed in the PTCL-GATA3 versus PTCL-TBX21 (p<0.05). Strong expression of CD30 (>50% of cells) was only seen in PTCL-GATA3 cases. EBER positivity, found only in rare background cells, was not significantly different (18% of PTCL-GATA3 vs. 12% of PTCL-TBX21). Within PTCL-TBX21, we identified cytotoxic and non-cytotoxic subsets with divergent morphological, phenotypic, and clinical findings. The cytotoxic PTCL-TBX21 exhibited an activated cytotoxic phenotype (23/25, 96%), denoted by TIA1 and granzyme-B and/or perforin expression. Extranodal involvement and single-cell apoptosis were observed in 17% and 45% of the cytotoxic cases, respectively, but absent in non-cytotoxic PTCL-TBX21 cases. A trend towards higher Ki67 expression (average= 40% vs. 20%, p= 0.08) was seen in the cytotoxic subgroup. In contrast, the non-cytotoxic PTCL-TBX21 was associated with a CD4+/CD8- phenotype and higher ICOS expression (average= 30%) and CCR4 (60%) compared to the cytotoxic PTCL-TBX21 (p= 0.001). PTCL-TFH cases showed a CD4+/CD8- phenotype (90%) and rarely a CD4-/CD8- phenotype (10%). While a subset of PTCL-TFH cases had AITL-like features such as numerous clear cells and/or prominent vasculature (31% of cases), which were not seen in the other subgroups, the remaining cases exhibited morphology that was indistinguishable from other PTLC-NOS cases including sheets of transformed cells or pleomorphic cells in a polymorphous background. Morphologically, PTCL-TFH cases, like PTCL-TBX21 were associated with a rich TME (TME-rich: 75%). CIBERSORT analysis showed enrichment of plasma cells (p<0.01) in PTCL-TFH, compared to PTCL-TBX21, and validated by morphology (30% of cases vs. 5%). Expression of one TFH marker was frequent in some PTCL-NOS cases (PTCL-GATA3= 41% of cases, PTCL-TBX21-non cytotoxic= 47%, PTCL-TBX21-cytotoxic= 5%). Conclusion: Our comprehensive evaluation underscores the importance of integrating morphology, immunophenotyping, and GEP in achieving an accurate diagnosis, potentially leading to more tailored treatment strategies for PTCL. Correlation with pending whole-exome sequencing studies will be provided at the time of presentation.

Introduction: Axi-cel utilization expanded after demonstration of superior event-free survival in patients (pts) with early relapsed or primary refractory LBCL (Locke, et al. NEJM. 2022) and was reinforced by improved overall survival (OS) among these pts (Westin, et al. NEJM. 2023). Despite increased demand, limited data exist on the impact of bridging therapies (BT) on pt outcomes. Thus, BT use between leukapheresis and lymphodepletion (LD) is largely at the discretion of treating physicians. Here, we assessed the impact of BT on effectiveness and safety outcomes after axi-cel. Methods: Pts receiving axi-cel for 3L+ LBCL in the US between 2017-2020 were prospectively collected through the CIBMTR registry in a post-authorization safety study; data up to May 2023 were included. Effectiveness outcomes were overall response rate (ORR), complete response (CR) rate, duration of response (DOR), progression-free survival (PFS), and OS. Safety outcomes were cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) per ASTCT grading, non-relapse mortality (NRM), and other key adverse events. Propensity score-based inverse probability of treatment weighting (IPTW) was used in logistic and Cox regressions to account for potential confounding in BT use. For pts who had BT, effects of BT types and responses on outcomes were evaluated using conventional logistic and Cox regressions. Results: Of 1497 pts registered, 1307 had BT data available and were included (median follow-up, 36 mo). Overall, 430 pts (33%) received BT (median time from leukapheresis to LD, 23 days [IQR, 21-28]). Systemic therapy was used in 403 pts (31%), including 305 (23%) with chemotherapy (CT)-based BT (19% treated with CT + anti-CD20 monoclonal antibody [mAb; mainly rituximab] and 4% with CT only), and 98 (7%) with non-CT systemic therapy (Table). Radiotherapy (± systemic therapy) was used in 44 pts (3%). Pts with ≥3 lines of prior therapy (odds ratio [OR], 10.67 [95% CI, 7.17−15.87]) or mild hepatic impairment prior to infusion (OR, 1.64 [1.06−2.55]) were more likely to receive BT, while those with histologic transformation (OR, 0.69 [0.51−0.92]) or longer time from diagnosis to leukapheresis were less likely to receive BT. ORRs to axi-cel were 70% and 79% (CR rates 51% and 64%) for pts with or without BT, respectively. Unadjusted DOR, PFS, and OS at 3 years were 40%, 28%, and 37%, respectively, in pts with BT vs 49%, 39%, and 49%, respectively, in those without BT. After multivariate adjustment, no statistical significance was found between BT and any effectiveness outcome. Adjusted hazard ratios (HRs) for BT were 1.13 (95% CI, 0.94-1.36) for PFS and 1.11 (95% CI, 0.92-1.34) for OS, respectively (Figure). Most safety endpoints were consistent regardless of BT. Grade ≥3 CRS and ICANS occurred in 12% and 30% of pts, respectively, who received BT vs 10% and 24% of pts who did not. Use of BT was associated with more frequent prolonged neutropenia (OR, 2.24 [1.44-3.50]) and thrombocytopenia (OR, 2.33 [1.67-3.24]), but not associated with a higher rate of infections or NRM (Table). Among pts receiving BT, 74 (17%) had a CR or partial response to BT. Of them, median PFS and median OS post infusion were 12.7 mo (3.8-not estimable [NE]) and 48.7 mo (18.6-NE), respectively. Response to BT was associated with increased CR rate (OR, 2.00 [1.07-3.75]), PFS (HR, 0.69 [0.49-0.99]), and OS (HR, 0.59 [0.40-0.88]) and decreased any-grade ICANS (OR, 0.51 [0.29-0.92]). When different BT strategies were considered, effectiveness and most safety outcomes for CT + anti-CD20 mAb vs non-CT treatment were comparable without significant difference. CT + anti-CD20 mAb was associated with improved ORR (OR, 1.90 [1.00-3.60]), PFS (HR, 0.69 [0.49-0.98]), and OS (HR, 0.66 [0.46-0.95]) but increased any-grade CRS (OR, 2.16 [1.02-4.57]) vs CT only. CT use was associated with an increased risk of prolonged thrombocytopenia (41% vs 24% for pts receiving non-CT treatment; OR, 2.37 [1.39-4.02]). No significant association was found between use of radiotherapy as BT and any outcome. Conclusion: These US real-world findings suggest that the addition of BT prior to axi-cel is associated with more frequent prolonged cytopenias but does not impact effectiveness or other critical safety outcomes. Response to BT may prognosticate more favorable outcomes after axi-cel infusion. Further study is warranted to identify the optimal bridging regimen.

Introduction A single ide-cel infusion showed significantly longer median progression-free survival (PFS) vs std regimens (13.3 vs 4.4 months [mo], HR 0.49, 95% CI 0.38-0.65, P < 0.001) with deep, durable responses in heavily pretreated TCE RRMM at an interim analysis (IA) of KarMMa-3 (NCT03651128); safety data were consistent with prior studies (Rodríguez-Otero et al. NEJM 2023). Ide-cel benefits were consistent across pts in high-risk subgroups and across 2-4 prior lines of therapy (Tx). Results of the preplanned final PFS analysis of KarMMa-3 with 12.3 mo additional follow-up are reported. Methods In the phase 3 KarMMa-3 trial, pts with RRMM who received 2-4 prior regimens, including an immunomodulatory agent, proteasome inhibitor, and daratumumab, and were refractory to last regimen were randomized 2:1 to ide-cel or a std regimen (DPd, DVd, IRd, Kd, or EPd). In the ide-cel arm, pts could receive ≤ 1 cycle of optional bridging Tx for disease control. Pts in the std regimens arm could receive ide-cel after confirmed disease progression (PD). The primary endpoint was IRC-assessed PFS in the ITT population; final PFS was planned to be analyzed with ~289 events. Key secondary endpoints were IRC-assessed overall response rate (ORR) and overall survival (OS); other secondary endpoints included complete response rate (CRR), duration of response (DOR), minimal residual disease (MRD) status, time to next anti-myeloma Tx (TTNT; time from randomization to next anti-myeloma Tx [MTx]), event-free survival (EFS; time from randomization to first PD, next MTx or any-cause death, whichever is first), PFS2 (time from randomization to second objective PD or any-cause death, whichever is first), safety, and health-related quality of life (QOL). Results Of 386 randomized pts (ide-cel, n = 254; std regimens, n = 132), 225 received ide-cel and 126 received a std regimen. Baseline characteristics were generally balanced. Median follow-up from randomization to data cutoff (April 28, 2023) was 30.9 mo (range 12.7-47.8). Ide-cel significantly improved median PFS (95% CI) vs std regimens (13.8 [11.8-16.1] vs 4.4 [3.4-5.8] mo), representing a 51% reduced risk of PD or death (HR 0.49, 95% CI 0.38-0.63; Figure); 18 mo PFS rates were 41% vs 19%, respectively. Ide-cel significantly improved ORR vs std regimens (71% vs 42%) with deeper (CRR 44% vs 5%; ≥CR and MRD negative status [sensitivity level 10 -5], 22% vs 1%), more durable responses (median DOR 16.6 vs 9.7 mo; Table). PFS and ORR benefits of ide-cel vs std regimens were consistent with the IA. Interim OS will be included in the presentation. In pts who received ide-cel (n = 225) or a std regimen (n = 126), median PFS (95% CI) was 15.7 (12.5-18.9) vs 4.4 (3.4-5.8) mo, respectively. In the ITT population, median TTNT, EFS, and PFS2 were numerically longer with ide-cel vs std regimens. Median (range) TTNT was 20.9 (16.6-24.2) vs 7.0 (5.3-8.5) mo. Median (95% CI) EFS was 13.3 (11.3-15.7) vs 3.9 (3.0-5.3) mo. Median PFS2 (95% CI) was 23.5 (18.4-27.9) vs 16.7 (12.2-20.3) mo; ide-cel was next MTx in 70 (53%) pts in the std regimens arm. In the treated population, grade (gr) 3/4 infections occurred in 66/249 (27%) pts in the ide-cel arm vs 25/126 (20%) in the std regimens arm. In the ide-cel safety population, any gr cytokine release syndrome occurred in 197/225 (88%) pts, gr ≥ 3 in 11 (5%); median time to first onset was 1 d (1-14), median duration was 4 d (1-51). Any gr investigator-identified neurotoxicity occurred in 34/225 (15%) pts, gr ≥ 3 in 7 (3%); median time to onset was 3 d (range 1-317); median duration was 2.5 d (range 1-252). Ide-cel continued to demonstrate durable, clinically meaningful improvements in pt-reported outcomes, including symptoms, functioning, and QOL vs std regimens. Conclusions In this final PFS analysis of KarMMa-3, significantly longer PFS was maintained with ide-cel; PD or death risk reduced by 51%; responses were deeper and more durable vs std regimens. CRR with ide-cel increased since the IA, indicating a deepening response, but were unchanged with std regimens. A single ide-cel infusion vs continuous treatment with std regimens resulted in longer median TTNT and PFS2, indicating improved long-term disease control. The ide-cel safety profile was consistent with previous reports, with no parkinsonism or Guillain-Barré syndrome reported. These data continue to support use of ide-cel in pts with TCE RRMM. Study support 2seventy bio and Celgene, a Bristol-Myers Squibb Company.