Lymphoma Clinical Trials Research Articles

Preventing Cardiac Damage in Patients Treated for Breast Cancer and Lymphoma: The PROACT Clinical Trial

BackgroundCardiotoxicity is a concern for cancer survivors undergoing anthracycline chemotherapy. Enalapril has been explored for its potential to mitigate cardiotoxicity in cancer patients. The dose-dependent cardiotoxicity effects of anthracyclines can be detected early through the biomarker cardiac troponin. ObjectivesThe PROACT (Preventing Cardiac Damage in Patients Treated for Breast Cancer and Lymphoma) clinical trial assessed the effectiveness of enalapril in preventing cardiotoxicity, manifesting as myocardial injury and cardiac function impairment, in patients undergoing high-dose anthracycline-based chemotherapy for breast cancer or non-Hodgkin lymphoma. MethodsThis prospective, multicenter, open-label, randomized controlled trial employed a superiority design with observer-blinded endpoints. A total of 111 participants, scheduled for 6 cycles of chemotherapy with a planned dose of ≥300 mg/m2 doxorubicin equivalents, were randomized to receive either enalapril (titrated up to 20 mg daily) or standard care without enalapril. ResultsMyocardial injury, indicated by cardiac troponin T (≥14 ng/L), during and 1 month after chemotherapy, was observed in 42 (77.8%) of 54 patients in the enalapril group vs 45 (83.3%) of 54 patients in the standard care group (OR: 0.65; 95% CI: 0.23-1.78). Injury detected by cardiac troponin I (>26.2 ng/L) occurred in 25 (47.2%) of 53 patients on enalapril compared with 24 (45.3%) of 53 in standard care (OR: 1.10; 95% CI: 0.50-2.38). A relative decline of more than 15% from baseline in left ventricular global longitudinal strain was observed in 10 (21.3%) of 47 patients on enalapril and 9 (21.9%) of 41 in standard care (OR: 0.95; 95% CI: 0.33-2.74). An absolute decline of >10% to <50% in left ventricular ejection fraction was seen in 2 (4.1%) of 49 patients on enalapril vs none in patients in standard care. ConclusionsAdding enalapril to standard care during chemotherapy did not prevent cardiotoxicity in patients receiving high-dose anthracycline-based chemotherapy. (PROACT: Can we prevent Chemotherapy-related Heart Damage in Patients With Breast Cancer and Lymphoma?; NCT03265574)

Utilization of Real-World Data to Facilitate Clinical Trials for Patients with Lymphoma

The future directions in leveraging real-world evidence (RWE) and real-world data (RWD) in the field of lymphoma, as compared to traditional experimental clinical trials, are poised to significantly impact research methodologies, treatment strategies, and patient care. Current methods of clinical trials involve a well-controlled design and patient selection bias. Integrating RWE and RWD with experimental clinical trials offers a multifaceted approach to understanding lymphoma and enhancing patient outcomes. In this review, we discuss how RWE has helped shape lymphoma clinical trials, and we compare and evaluate evidence obtained from real-world lymphoma studies/databases with that obtained from clinical trials. We also discuss methods for utilizing surrogate endpoints to facilitate clinical trials and expedite drug development. RWE can be leveraged to bridge the gap between data obtained from clinical trial populations and the broader patient population encountered in clinical practice, by highlighting differences in outcomes and the need for effective treatment strategies across diverse patient groups.

DeepAFT: A nonlinear accelerated failure time model with artificial neural network.

The Cox regression model or accelerated failure time regression models are often used for describing the relationship between survival outcomes and potential explanatory variables. These models assume the studied covariates are connected to the survival time or its distribution or their transformations through a function of a linear regression form. In this article, we propose nonparametric, nonlinear algorithms (deepAFT methods) based on deep artificial neural networks to model survival outcome data in the broad distribution family of accelerated failure time models. The proposed methods predict survival outcomes directly and tackle the problem of censoring via an imputation algorithm as well as re-weighting and transformation techniques based on the inverse probabilities of censoring. Through extensive simulation studies, we confirm that the proposed deepAFT methods achieve accurate predictions. They outperform the existing regression models in prediction accuracy, while being flexible and robust in modeling covariate effects of various nonlinear forms. Their prediction performance is comparable to other established deep learning methods such as deepSurv and random survival forest methods. Even though the direct output is the expected survival time, the proposed AFT methods also provide predictions for distributional functions such as the cumulative hazard and survival functions without additional learning efforts. For situations where the popular Cox regression model may not be appropriate, the deepAFT methods provide useful and effective alternatives, as shown in simulations, and demonstrated in applications to a lymphoma clinical trial study.

Why We Fail: An Analysis of Terminated Interventional Clinical Trials in Lymphoma

Introduction: Clinical trials serve as the cornerstone for advancing therapeutic and clinical strategies in Lymphoma. Learning from prematurely terminated trials provides valuable insights for researchers and drug developers. This study analyses interventional clinical trials in lymphoma terminated between 2010 and 2023, aiming to identify common factors contributing to their failure and improve future research in Lymphoma. Methods The study identified clinical trials in Lymphoma registered in the US between 01/01/2010, and 01/07/2023, using the InovaSight platform. The analysis included trials investigating drugs in phases I-III with at least one recruited patient. Phase 4 studies, early access programs, and trials examining medical devices were excluded, as were trials terminated early due to positive efficacy signals. Natural language processing (NLP) &amp; classification algorithms were used to categorize trials into distinct phases of clinical development and were also applied to identify reasons for trial termination from official registrations and relevant publications. These reasons were then characterized using the InovaSight trial ontology for thematic analysis. Enriching features, including mechanism of action information, design characteristics, target enrolment, and inclusion criteria, were generated and linked for each eligible trial. The analysis was conducted using the InovaSight drug development platform, which employs responsible artificial intelligence to expedite the development of novel therapeutics for haematological malignancies. Results: A total of 306 trials met the inclusion criteria, with phase II trials being most frequently withdrawn (49.3%), followed by phase I (42.5%) and phase III (8.2%). The trials spanned from 2010-01-31 to 2021-12-15. Among terminated studies, single-arm trials were most frequent (64%), followed by two-armed designs (21.2%) and three or more arms (14.7%). A concentration of mechanisms of action were observed across all phases (further described in poster). Reasons for early termination included enrolment issues, sponsor decisions, lack of efficacy, safety concerns, funding withdrawal, investigator-related issues, study irrelevance, drug supply problems, COVID-19, and unspecified reasons. The most common reason for early termination was enrolment issues in Phase I and Phase II studies, and sponsor decisions in Phase III studies. Detailed results are presented in Figure 1. Discussion: Our analysis identified a total of 306 trials that met the inclusion criteria, encompassing different phases of clinical development. Interestingly, phase II trials were the most frequently withdrawn, followed by phase I and phase III trials. This observation suggests that a considerable number of trials face challenges during the intermediate stages of development, which could be attributed to various factors such as safety concerns, lack of efficacy, or financial constraints. Our analysis revealed a concentration of similar mechanisms of action across all trial phases. This finding suggests that certain therapeutic areas or drug targets may be more prone to early termination. Exploring the underlying reasons for this pattern could provide valuable insights for future trial planning and design (This is further elucidated in the full presentation). Among the various reasons for early termination, enrolment issues emerged as the most common factor in both Phase I and Phase II studies. This highlights the critical need for in depth and principled approaches to study feasibility assessment routed in an evidence-based methodology. In contrast, sponsor decisions were the leading cause of termination in Phase III trials. This finding suggests that late-stage trials are more susceptible to decisions made by the trial sponsors, which could be influenced by factors such as financial considerations or market dynamics. The results underscore the importance of careful trial planning, patient recruitment strategies, and effective collaboration between sponsors, investigators, and regulatory bodies. Understanding the challenges and factors contributing to trial termination can aid in optimizing future trial designs, mitigating risks, and improving the efficiency of drug development processes.

Progression of Disease within 2 Years Predicts Poor Survival in Patients with Classic Hodgkin Lymphoma

Background: Early progression within 24 months (POD24) of initial immunochemotherapy was repeatedly confirmed to be associated with poor survival in follicular, diffuse large B-cell and mantle-cell lymphomas, identifying its high-risk subgroup with different disease biology (Casulo C, JCO 2015; Maurer JM, JCO 2015; Bond DA, Blood Adv 2021). Little is known about the POD24 incidence and clinical impact in Hodgkin lymphoma (HL) patients, and current prognostic schemes have been designed with longer survival endpoints (Rodday AM, JCO 2022). Methods: This single-center study comprised patients (pts) with classic HL (cHL) treated with frontline chemotherapy prospectively enrolled in the Czech Hodgkin Lymphoma Study Group database in 1993-2020 (database lock was done on 15. June 2023). Early event was defined as progression, relapse, or death related to progressive HL within 24 months after the date of diagnosis. Overall (OS) and progression-free survival (PFS) were calculated from the date of diagnosis. To evaluate the association between early POD and OS from a risk-defining event, that is, survival from the time of POD for early progressors (POD24 group) or from 2 years after diagnosis for the reference group (noPOD24 group), was calculated. In total, we have identified 417 pts with cHL diagnosed from 1993 to the end of 2020. Ten pts have been excluded from the analysis because of early death (&amp;lt;24 months) with no recorded disease progression. Results: The analyzed cohort consisted of 407 pts: 205 females and 202 males with a median age of 35 years (min 16, max 78) with limited disease (stage I-II) in 58% and advanced disease (stage III-IV) in 42%. Histological subtypes were as follows: nodular sclerosis in 53.1%, mixed cellularity in 35.6%, lymphocyte-rich in 3.7%, and lymphocyte-depleted in 0.3%. Subtype was not available in 7.3% of the pts. Extranodal disease and B symptoms were present in 25% and 49%, respectively. 42% of patients had high-risk IPS of 3 or more points. Patients received BEACOPP in 55.5%, Stanford V in 23%, ABVD in 12%, and COPP in 7%. Other treatments were used in 2,5% of the cases. The complete and partial remission rates after frontline treatment were 88.7% and 8.6%, respectively, and only 3% of the pts did not respond to therapy. After a median follow-up of 12.1 years, a cumulative incidence of progression/relapse is 0.14 (95%CI 0,10-0.18). The POD24-event occurred in 36 pts (8.8%) and 38 (9.3%) patients relapsed after 2 years. The entire group's 5-y PFS and 5-y OS reached 81.0% (95% CI 0.77-0.85) and 91.4% (95% CI 88.6-0.94), respectively. The 5-y OS since the risk-defining event was 48.1% (95% CI 31.4-64.8) and 94.2% (95% CI 0.92-96.7) in the POD24 vs noPOD24 group (p&amp;lt;0.001, HR=5.35), Fig 1. When comparing 12 baseline clinical and laboratory characteristics between POD24 (n=35) with noPOD24 controls (n=371), we have found no major differences (Tab 1). There was a trend in more frequent POD24 in the patients with massive mediastinal tumor (p=0.07) and those with advanced clinical stage (p=0.047). The spectrum of the induction regimes was distributed equally (p=0.26). Conclusions: Early progression of the disease is rare; however, a catastrophic event in HL resulting in more than five times higher risk of death. Baseline disease characteristics - besides the clinical stage - do not predict the risk of later POD24 events. Large independent validation may enhance the validity of these data and help create a prognostic tool for POD24 prediction. This event should be considered an important endpoint in Hodgkin lymphoma clinical trials in the future. Acknowledgements: Supported by MH CZ - DRO (FNOL, 00098892), IGA_LF_2023_005 and AZV NU22-03-00182.

Geographic Disparities in Access to Cancer Clinical Trials in Canada.

This study aims to evaluate geographic disparities in access to cancer clinical trials across Canada. Cancer clinical trial data recorded within the clinicaltrials.gov and reporting the conduct of any of these trials in Canada, 2005 to 2023 were reviewed. Frequency analyses of the number of clinical trials that were registered on clinicaltrials.gov for Canada, individual Canadian provinces, main Canadian urban centers, and different cancer types, according to the funding source (industry versus non-industry), as well as according to different periods (using 3-y intervals) were conducted. Moreover, a comparison of cancer clinical trials per 10,000 persons was done between Canada and the United States. The number of cancer clinical trials per 10,000 individuals (according to the 2021 census) in each province/territory varied between 6.79 (New Brunswick) to 0 (the 3 territories). The number of cancer clinical trials in relation to 1000 projected cancer cases for some of the common tumor types in Canada was then reviewed. The highest number was for lymphoma clinical trials (32.85), whereas the lowest number was for bladder cancer clinical trials (7.06). Most of the trials have industry funding (69%). Using 3-year intervals, the highest number of cancer clinical trials was observed from 2014 to 2016 (778 trials), and the lowest number was observed from 2020 to 2022 (633 trials). Access to clinical trials in Canada is not equitably distributed, with geographical and primary tumor site disparities. Moreover, access to cancer clinical trials has been negatively impacted during the time of the COVID-19 pandemic.

COMPARATIVE EFFECTIVENESS OF SALVAGE CHEMOTHERAPY AND CHIMERIC ANTIGEN T‐CELL RECEPTOR THERAPIES IN R/R DLBCL: A SYSTEMATIC REVIEW AND NETWORK META‐ANALYSIS

David Hodgson and John Kuruvilla are co-senior investigators Introduction: The optimal salvage chemotherapy (SC) regimen for relapsed or refractory (R/R) diffuse large B-cell lymphoma (DLBCL) prior to autologous stem cell transplant (ASCT) remains unclear. Head-to-head comparison of chimeric antigen receptor T cell (CAR-T) therapies for second line use in primary refractory DLBCL is lacking. Given these limitations, we conducted a systematic review and network meta-analysis (NMA) of randomized clinical trials (RCTs) to assess the efficacy and safety of SC and CAR-T therapies for R/R DLBCL. Methods: We searched MEDLINE, EMBASE and CENTRAL from Jan 1 2006 to July 7 2022 for RCTs enrolling patients diagnosed with R/R DLBCL following treatment (trt) with at least 1 prior regimen. The primary outcome was overall survival (OS), and secondary outcomes were progression free survival (PFS) and grade ≥ 3 adverse events (AEs). NMA was conducted using a fixed-effect model to synthesize evidence. Surface under the cumulative ranking curve was used to rank the relative effect of compared trt. Results: A total of 22 RCTs (n=4505) met eligibility criteria, 13 included for qualitative synthesis only. NMA of SC (6 RCTs, n=1831) compared 7 trt: GDP (gemcitabine, dexamethasone, cisplatin), rituximab (R)-GDP, ICE (ifosfamide, mesna, carboplatin, etoposide), R-ICE, DHAP (dexamethasone, cytarabine, cisplatin), R-DHAP, ofatumumab (O)-DHAP, ESHAP (etoposide, cytarabine, cisplatin, methylprednisolone). Results indicated that R-GDP had improved OS and PFS over other regimens (Table), and ranked highest for OS and PFS. NMA of AEs showed no differences for infection. NMA of 3 CAR-T RCTs (axi-cel, liso-cel, tisa-cel; n=733) showed that axi-cel and liso-cel had improved PFS over SC and ASCT, while no OS difference was observed. Comparing between CAR-T products, both axi-cel and liso-cel had improved OS and PFS over tisa-cel. Liso-cel ranked highest for PFS. NMA of AEs showed that tisa-cel and axi-cel had higher risk of cytokine release syndrome over SC, while axi-cel had higher risk of neurologic toxicity over SC. Conclusions: Our results suggest that R-GDP may be the optimal regimen for R/R DLBCL prior to ASCT. Both axi-cel and liso-cel had improved PFS over SC and ASCT, while tisa-cel did not. Although liso-cel ranked highest for PFS, longer follow-up is required for comparative survival analysis as data matures. Keywords: Aggressive B-cell non-Hodgkin lymphoma, Therapeutics and Clinical Trials in Lymphoma - Other No conflicts of interests pertinent to the abstract.

RANDOMIZATION IS AN INFREQUENTLY USED METHOD IN INTERVENTIONAL PHASE II CLINICAL TRIALS OF ADULT PATIENTS WITH LYMPHOMA

Introduction: Clinical trials are important to inform decision making and to improve outcomes for patients with lymphoma. Phase II trials are designed to determine if a new treatment has enough promising efficacy and safety to warrant further investigation in a phase III trial. Randomized controlled trials have historically received the greatest consideration in providing reliable evidence of a drug's safety and efficacy. However, Single Arm Non-Randomized Studies (SANRS) have been considered as a valid option by regulators. It has been reported that 66% (49/74) of approvals granted between 1999 and 2014 from the European Medicines Agency (EMA) and U.S. Food and Drug Administration (FDA) were based on nonrandomized experiments in oncology including hematological malignancies, and lymphoma is one of the most common cancer types that has single arm FDA drug approval. The aim of this report is to evaluate the proportion and evolution in time of the use of randomization in interventional phase II clinical trials in adult patients with lymphoma, using a sample of the ClinicalTrials.govdatabase. Methods: We conducted a predefined search in the United States National Library of Medicine clinical trials database (ClinicalTrials.gov) in the field “condition or disease” for the term "lymphoma" with filters for "interventional studies", "phase 2 studies” and “adults = >18–64 years old” plus “older adults = 65 years and older”. We examined all clinical trials that met the inclusion criteria from January 1st, 2012, through December 31st, 2022. We excluded trials with overlapping terminology that did not enroll patients with mature lymphoid neoplasms, such as precursor b-lymphoblastic lymphoma/leukaemia, anaplastic lymphoma kinase (ALK) non-small-cell lung cancer and trials involving very broad disease categories such as “advanced cancer”, “hematological malignancies” or “high risk populations” with the exception when a significant proportion of patients with lymphoma was expected to be enrolled. Difficult cases were discussed and decided at the discretion of the authors. Results: Our search returned 2085 clinical trials, of those we included 1376 in the analysis according to our inclusion/exclusion criteria. We found that between 2012 and 2022 randomization was performed in 12.8% (177/1376) of interventional phase II clinical trials of adult patients with lymphoma. The histogram (Figure 1) shows that the number of trials increased during the analysis period, but the number of randomized clinical trials remained proportionally low. Keywords: therapeutics and clinical trials in lymphoma, other No conflicts of interests pertinent to the abstract.

17th International Conference on Malignant Lymphoma, Palazzo dei Congressi, Lugano, Switzerland, 13 - 17 June, 2023.

Introduction: Patients with B-cell lymphoma have poorer antibody responses to COVID-19 vaccines compared to patients with other malignancies. However, the immune responses are heterogeneous and its association with clinical outcomes are poorly understood. Here, we study the association between antibody and cellular responses with the outcomes from COVID-19 infection from the PROSECO study. Methods: THE UK PROSECO study is a prospective observational cohort study evaluating COVID-19 vaccine response in people with lymphoma. Peripheral blood antibody titres (anti-spike (S) IgG by Mesoscale Discovery), antibody avidity (surface plasmon resonance) and IFNγ T-cell responses (ELISpot) against spike protein after two to four COVID-19 vaccines, and clinical outcomes data were collected. Results: 524 patients (92 Hodgkin lymphoma, 180 aggressive B non-Hodgkin lymphoma (B-NHL), 234 indolent B-NHL and 18 peripheral T-cell lymphoma) were included: 334 (84·3%), 315 (79·5%) and 266 (67·1%) participants were eligible for post-two, three and four vaccine dose analysis, respectively. Breakthrough infections (BTI) occurred in 20 (5.9%), 40 (12.7%) and 36 (13.5%) participants after two, three and four vaccine doses. Of 96 participants with BTI, 12 (12.5%) needed hospitalisation. No deaths were observed due to COVID-19 infection. No differences were observed in T-cell responses between BTI and non-BTI groups, but 4/9 (44.5%) hospitalised BTI cases had undetectable cellular and antibody responses. There was no association between antibody titres and infection episodes after two vaccine doses, but antibody titres were associated with lower BTI risk after three or four vaccine doses on multivariable analysis. Intriguingly, the antibody threshold associated with infection risk was lower after four than three vaccine doses (41 vs. 820 BAU/mL). We also observed stronger antibody binding avidity to SARS-CoV-2 spike receptor-binding domain (RBD) proteins from Wuhan and Omicron BQ.1 variants, with increasing vaccine doses, i.e., fourth versus third dose: 10-fold increase (p: <0.01); fourth versus second dose:100-fold-increase than the second dose (p: <0.0001) (see Figure). The research was funded by: the Blood Cancer UK Vaccine Research Collaborative, which is led by Blood Cancer UK in partnership with Myeloma UK, Anthony Nolan and the British Society for Haematology; a Cancer Research UK Advanced Clinician Scientist Fellowship to SH Lim; Cancer Research UK/National Institute for Health Research (NIHR) Southampton Experimental Cancer Medicine Center; NIHR Southampton Clinical Research Facility (Southampton Research Biorepository) and NIHR Southampton Biomedical Research Center; NIHR Great Ormond Street Biomedical Research Center and the NIHR Oxford Biomedical Research Center and CRUK Experimental Cancer Medicines Center. Keywords: Lymphoid Cancers, Other: Covid-19, Therapeutics and Clinical Trials in Lymphoma, vaccination Conflicts of interests pertinent to the abstract. M. Ahearne Research funding: Pfizer C. P. Fox Consultant or advisory role: Astra Zeneca G. P. Collins Consultant or advisory role: Astra Zeneca and Pfizer Research funding: Pfizer A. J. Davies Honoraria: Astra Zeneca and Janssen Research funding: Astra Zeneca and Janssen S. H. Lim Honoraria: Astra Zeneca

Rural communities and health professional shortage areas (HPSA) and access to immune effector cellular therapy and lymphoma/myeloma clinical trials.

e18633 Background: The National Health Service Corps (NHSC) has created a shortage designation to identify medically underserved communities. HPSA (Health Professional Shortage Areas) designation identifies an area, population, or facility with inadequate access to healthcare resources such as primary care. Rural-urban continuum codes (RUCC) classify U.S. census tracts using urbanization, population density, and daily commuting. Additionally, there are racial, ethnic, and geographic disparities in enrollment on therapeutic cancer clinical trials. Chimeric antigen receptor T cell (CAR T) therapy has been at the forefront of immune effector cell therapy (IEC) for lymphoma and myeloma. However, it is delivered at specialized centers, and access to specialty centers is known to be a challenge due to financial and geographic barriers. At our institution, trials for CAR-T cell therapy in lymphoma and myeloma are conducted by the IEC disease working group (DWG), while non-IEC trials are conducted by the lymphoma/myeloma DWG. We hypothesized that patients from HPSA and rural communities and HPSA had lower access to IEC and non-IEC trials for lymphoma and myeloma . Methods: Patients enrolled between 1/1/2015 – 2/6/2023 on IEC and non-IEC interventional trials with treatment were included. Disease types included lymphoma (Follicular lymphoma, high-grade B cell lymphoma, diffuse large B cell lymphoma) and myeloma were included. The hypothesis was tested through a t-test. Results: There were 667 total subjects. 503 (75.4%) were non-IEC and 96 (14.3%) were IEC clinical trial participants. The majority in both groups were male (52.7%), White (74.8%), non-Hispanic (86%) .Only 69 (10.3%) pts were from rural areas and 223 (38.9%) were from HPSA communities When stratifying based on RUCC (Table), there is significantly higher rural accrual for IEC trials compared to non-IEC trials. Hispanics from rural areas were lower. There were no African Americans (AA) from rural areas. Conclusions: There is low representation of patients from HPSA and rural communities for IEC and non-IEC lymphoma and myeloma clinical trials. The ethnic and racial inequity in access and enrollment to lymphoma/myeloma trials are pronounced in rural and HPSA communities. Strategies to improve clinical trial enrollment of the underserved rural and HPSA populations is an area of unmet need and will benefit minorities and patients from lower socioeconomic status. [Table: see text]

Targeting tumour-associated macrophages in hodgkin lymphoma using engineered extracellular matrix-mimicking cryogels

Tumour-associated macrophages are linked with poor prognosis and resistance to therapy in Hodgkin lymphoma; however, there are no suitable preclinical models to identify macrophage-targeting therapeutics. We used primary human tumours to guide the development of a mimetic cryogel, wherein Hodgkin (but not Non-Hodgkin) lymphoma cells promoted primary human macrophage invasion. In an invasion inhibitor screen, we identified five drug hits that significantly reduced tumour-associated macrophage invasion: marimastat, batimastat, AS1517499, ruxolitinib, and PD-169316. Importantly, ruxolitinib has demonstrated recent success in Hodgkin lymphoma clinical trials. Both ruxolitinib and PD-169316 (a p38 mitogen-activated protein kinase (p38 MAPK) inhibitor) decreased the percent of M2-like macrophages; however, only PD-169316 enhanced the percentage of M1-like macrophages. We validated p38 MAPK as an anti-invasion drug target with five additional drugs using a high-content imaging platform. With our biomimetic cryogel, we modeled macrophage invasion in Hodgkin lymphoma and then used it for target discovery and drug screening, ultimately identifying potential future therapeutics.

Abstract 5743: Reprogramming myeloid cells by JAK inhibition to enhance checkpoint blockade immunotherapy

Abstract Background: Checkpoint inhibitors have achieved a wide adoption for treatment of advanced cancer, but the majority of patients fail to respond. Among the underlying causes are redundant mechanisms driving cytotoxic lymphocyte dysfunction. Motivated by the discovery of JAK inhibitors as hit compounds in a screen for the reversal of T cell exhaustion, this study investigates the hypothesis that JAK inhibition could enhance the efficacy of checkpoint inhibitors. Methods: Murine immunocompetent models of cancer were utilized (A20, EL4, LLC1, MC38). The combination therapy of systemic treatment with the JAK inhibitor ruxolitinib with anti-PD1 + anti-CTLA4 was evaluated for efficacy and biomarkers compared to checkpoint inhibitors (ICI) alone. Tumor-infiltrating, blood and lymphoid organ immune cells were phenotyped using single-cell transcriptomics, functional assays and flow cytometry. The combination therapy was clinically tested in an investigator-initiated Phase I/II clinical trial of ruxolitinib with nivolumab in relapsed or refractory Hodgkin lymphoma (NCT03681561). Patients who previously failed to respond to ICI received ruxolitinib for 1 week then nivolumab every 4 weeks concurrent with ruxolitinib twice daily. Peripheral blood was collected at baseline, 1 week after ruxolitinib and 4 weeks after 1st nivolumab dose. Hematologic and transcriptomic analyses were performed on peripheral blood samples. Results: The ruxolitinib + ICI combination was superior to ICI in 3/4 of the tumor models examined in controlling tumor growth. Tumor sizes were reduced by &amp;gt;50% in the MC38 (mean volume 123.1 vs 283.2 mm3, ICI + rux vs ICI, respectively, n=9 per group, ANOVA p=0.0094), LLC1 and A20 models (survival hazard ratio 0.24, ICI + rux vs ICI, n=8 per group, p=0.048). Remarkably, we observed a broad shift of tumor monocytes and granulocytes from a suppressive into an immunostimulatory state characterized by the expression of MHC molecules and the ability to stimulate T cell proliferation. Hodgkin lymphoma patients in the ruxolitinib with nivolumab trial exhibited an interim disease control rate of 76% (13/17). Ruxolitinib treatment in these patients did not impair T cell cytokine production but significantly reduced the neutrophil-to-lymphocyte ratio (NLR, mean difference -0.82, n=14, p=0.0023) and the expression of myeloid derived suppressor cell markers in monocytes (normalized enrichment score -1.74, n=14, q&amp;lt;0.001) compared to pre-treatment. The reduction in NLR was significantly greater in complete responders than in progressive disease patients (mean -2.6 vs -0.58, respectively, p=0.023). Conclusions: The combination of ruxolitinib with ICI was effective in preclinical models of cancer and in a Phase I/II Hodgkin lymphoma clinical trial. Unexpectedly, JAK inhibitor-mediated reprogramming of myeloid cells and the associated enhanced T cell activity may be important for the observed efficacy. Citation Format: Jaroslav Zak, Isaraphorn Pratumchai, Brett S. Marro, Kristi L. Marquardt, Michael B. Oldstone, Judith A. Varner, Veronika Bachanova, John R. Teijaro. Reprogramming myeloid cells by JAK inhibition to enhance checkpoint blockade immunotherapy. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5743.

Underrepresented Groups Often Excluded From Lymphoma Clinical Trials

Underrepresented Groups Often Excluded From Lymphoma Clinical Trials

Is local review of positron emission tomography scans sufficient in diffuse large B-cell lymphoma clinical trials? A CALGB 50303 analysis.

Quantitative methods of Fluorodeoxyglucose Positron Emission Tomography (FDG-PET) interpretation, including the percent change in FDG uptake from baseline (ΔSUV), are under investigation in lymphoma to overcome challenges associated with visual scoring systems (VSS) such as the Deauville 5-point scale (5-PS). In CALGB 50303, patients with DLBCL received frontline R-CHOP or DA-EPOCH-R, and although there were no significant associations between interim PET responses assessed centrally after cycle 2 (iPET) using 5-PS with progression-free survival (PFS) or overall survival (OS), there were significant associations between central determinations of iPET ∆SUV with PFS/OS. In this patient cohort, we retrospectively compared local vs central iPET readings and evaluated associations between local imaging data and survival outcomes. Agreement between local and central review was moderate (kappa=0.53) for VSS and high (kappa=0.81) for ∆SUV categories (<66% vs. ≥66%). ∆SUV ≥66% at iPET was significantly associated with PFS (p=0.03) and OS (p=0.002), but VSS was not. Associations with PFS/OS when applying local review vs central review were comparable. These data suggest that local PET interpretation for response determination may be acceptable in clinical trials. Our findings also highlight limitations of VSS and call for incorporation of more objective measures of response assessment in clinical trials.

The DLBCL Axis of Low TNFRSF10B Expression, TP53 Loss, and a Cold Immune Microenvironment May be Addressed By the Synergistic Combination of Eprenetapopt and Idasnutlin

Background: Genomically-targeted Diffuse Large B-cell (DLBCL) treatments for refractory or relapsing patients remain a clinical need alongside the growing success of Chimeric Antigen T-cells (CAR-T), which have ushered in durable responses in 40-70% of patients. However, recent studies have identified factors indicative of an inferior CAR-T response. These include a "cold" immune microenvironment, downregulation of extrinsic cell death factors such as TNFRSF10B, and alterations to the TP53 tumor suppressor. Few precision treatments exist capable of addressing the consequences, although two avenues are being pursued in clinical trials for lymphoma: inhibition of the TP53 negative regulator MDM2 via Idasnutlin and TP53 reactivation via the unique molecule Eprenetapopt (APR-246). Herein, we report the first observations of synergy between Idasnutlin and Eprenetapopt within DLBCL cell lines and data supporting that TP53 restoration may revive extrinsic apoptosis and immunological pathways capable of modulating CAR-T success. Methods: We analyzed 2 de-novo DLBCL patient cohorts (Xu-Monette et. al. 2020 and the TCGA) for differential expression and pathway analysis between TP53-mutant and Low TNFRSF10B (-1 Standard Deviation) patients versus normal counterparts. We integrated data from 3 additional sources to uncover actionable vulnerabilities within patients faced with dismal CAR-T response: differential gene expression between TP53-mutant and WT rrDLBCL patients (Shouval et. al. 2022), in-vitro CAR-T sgRNA-cytotoxicity scores (Dufva et. al. 2020), and gene expression following MDM2 inhibition (Louise et. al. 2021). We next assayed Idasnutlin and Eprenetapopt vs. 7 DLBCL cell lines, 5 of which harbor TP53 alterations, across a series of 9 concentrations (78nM-20µM) across 3-9 replicates. We followed these results by combining treatments in the TP53-impaired DHL6 and U2932 cell lines, analyzing results using the HSA synergy model. Lastly, we performed Western blot analysis to measure protein expression of TNFRSF10B following 24-hour single-agent or combination treatment. Results: Integrative analyses support that DNA alterations to TP53 are associated with significantly reduced T-cell activation and extrinsic apoptosis genes. TNFSRF10B was significantly reduced in these cases as well (FDR = 0.0034), and genes with significantly reduced expression in both TP53-mut and Low TNFRSF10B expression were associated with the loss of Immune System Development genes (FDR = 3.66E-09). The Staudt lab IFNG-Up and STAT-Up gene signature pathways were significantly reduced in TP53-mut patients (P < 0.0001, P = 0.483) and in Low TNFRSF10B Patients (P < 0.0001, P = 0.488). TNFRSF10B was also the only gene with significantly reduced expression in TP53-mut rrDLBCL patients, importance for CAR-T cytotoxicity, and increased expression following MDM2 inhibition. Our Western blot results support these observations, showcasing upregulation of TNFRS10B following treatment. Idasnutlin and Eprenetapopt assays vs. DLBCL cell lines demonstrated dose and TP53-dependent cell inhibition, with ED50 values ranging between 880nM to 19.6µM, independent of COO. The homozygous-impaired TP53 cell line Karpas-422 was the least sensitive. The molecules were next combined vs. DHL6 and U2932 cell lines, resulting in synergistic reduction of cell viability and HSA synergy scores of 10.94, and 5.32, respectively. Combined treatments of 400 nM Idasnutlin and Eprenetapopt reduced cell viability significantly greater than single agent treatments at the same concentration in DHL6 (P = 0.0001; P = 0.0176) and U2932 (P = 0.0012; P = 0.0322). Conclusions: Our results suggest the presence of an interconnected genomic DLBCL phenotype characterized by TP53 loss, reduced TNFRSF10B expression, and a debilitated immune presence, collectively predictive of an inferior CD19 CAR-T response. We applied the targeted therapies Eprenetapopt and Idasnutlin vs. cell line models, with both displaying anti-tumor activity as single agents. Most importantly, we observed synergistic activity in TP53-imparied DLBCL cell lines when the molecules were combined. These results are buoyed by integrative and Western blot analyses that highlight the promise of TNFRSF10B revival afforded by these agents, revealing a potential avenue to address inferior CAR-T response rates faced by patients harboring TP53 alterations. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Transcriptomic Comparison of Non-Hodgkin Lymphomas in Relapsed/Refractory Versus Newly Diagnosed Patients Using Single FFPE Slides

Background Patients newly diagnosed with non-Hodgkin Lymphomas (NHLs) have been historically treated with B-cell depleting antibodies and chemotherapy. Standard of care (SoC) for patients with refractory or relapsed (R/R) disease has rapidly evolved over the past decade to incorporate chimeric antigen receptor T-cell therapy (CAR T), lenalidomide, antibody-directed chemotherapies and stem cell transplantation. This diverse treatment landscape can transform the tumor microenvironment and drive heterogenous tumor escape mechanisms. Understanding this varied tumor contexture will help inform efficacy results from NHL therapeutic studies. Core needle biopsies are used to sample tumor from R/R NHL patients; the bulk of tissue may be consumed for histopathology testing, leaving insufficient sample for transcriptomic profiling. We have shown that HTG EdgeSeq is useful for single-slide analysis of formalin fixed, paraffin embedded (FFPE) tissues (Loya et. al., JCOASCO Suppl. 2022). Here, we use this technique to compare tumors of R/R and newly diagnosed patients with Diffuse Large B-cell Lymphoma (DLBCL) and Follicular Lymphoma (FL). MethodsSamples and Transcriptomics: NHL tumor samples from newly diagnosed, treatment naïve patients (DLBCL n = 99, FL n = 19) and lymph node controls (n = 18) were sourced commercially. Tumor samples from a Phase 1/2a trial of epcoritamab in R/R patients (NCT03625037) were collected prior to treatment (DLBCL n = 38, FL n = 5). Single FFPE slides were processed using the EdgeSeq platform (19,000 gene HT panel, HTG Molecular). Tumors originated from lymphoid organs, gastrointestinal tract, testes, and the pleural cavity. Immunohistochemistry imaging for CD3, PAX5, CD20 was also performed (CellCarta, Belgium). Bioinformatics: Differential expression analysis was performed using linear models in R. Gene set enrichment analysis (GSEA) was performed with GSVA in R (Hänzelmann et. al., BMC Bioinformatics. 2013) using published signature sets (mSigDB: Liberzon et al., Cell Syst. 2015, XCell: Aran, et. al., Genome Biol. 2017). Results Single 5μm slides were analyzed with an average tissue size of 8mm2 for core needle biopsies and 40mm2 for resections. The number of detected genes was comparable to RNAseq, suggesting the HTG platform provides comprehensive data using core needle biopsies obtained during routine histopathology. We compared DLBCL tumor to normal lymph nodes and derived a gene set that was highly correlated to pathologist derived tumor purity. This quantitative proxy allows automated identification of samples that may contain source organ tissue. We applied differential expression analysis to identify gene, pathway and immune cell signatures contrasting newly diagnosed and R/R DLBCL. Top genes upregulated in R/R DLBCL tumors included B-cell markers as expected, T-cell function (CD25/IL2RA), and an intriguing epigenetic mechanism (YTHDC1; a N6-methyladenosine reader with a role in lymphoma proliferation [Liu et al., Front Onco 2022]). MYC pathway signaling was enriched in R/R tumors; consistent with known mutational risk factors for progression (Bisso et. al., Immunol Rev. 2019). Th2 polarized CD4+ T-cell signature was higher in treatment naïve DLBCL compared to R/R and normal lymph nodes. Fibroblast gene signatures were higher in R/R DLBCL, potentially a result of microenvironment remodeling due to chemotherapy or tumor driven adaptation (Dumontet et. al., Front Immunol. 2021). Comparison of DLBCL to FL was also performed in both R/R and untreated patient cohorts. Immune cell signatures for Tregs and CD4+ naïve T-cells were enriched in FL, whereas M1 macrophage signatures were enriched in DLBCL (findings were similar in both R/R and treatment naïve groups). Conclusions The EdgeSeq platform enabled whole-transcriptome profiling from very limited tissue samples in lymphoma clinical trials. This study highlights the importance of characterizing the transcriptional heterogeneity of NHL in relation to genetic alterations and T-cell immunity. Ongoing analysis with large patient cohorts in epcoritamab clinical trials will add statistical power to uncover associations with International Prognostic Score, Cell of Origin, effects of different treatment combinations, and genomic alterations on clinical outcomes.

Evaluating the Impact of Lab-Based Eligibility Criteria By Race/Ethnicity in Frontline Clinical Trials for Diffuse Large B-Cell Lymphoma (DLBCL): A LEO Cohort Analysis

Introduction: Trial design and modern inclusion/exclusion criteria have impacted the patients enrolling in frontline trials for DLBCL. Our previous analysis from the University of Iowa/Mayo Clinic SPORE Molecular Epidemiology Resource (Khurana et al. JCO 2021) identified that up to 24% of patients treated with standard immunochemotherapy (IC) were excluded based on 5 lab-based criteria alone. Additionally, ineligible patients had worse clinical outcomes and increased deaths related to lymphoma progression suggesting the potential exclusion of patients who could have benefited most from novel therapies being evaluated in these trials. We sought to evaluate the impact of laboratory eligibility criteria from recent first-line DLBCL trials in patients from the Lymphoma Epidemiology Outcomes (LEO) Cohort shown to have demographics similar to the US SEER population (Flowers et al. ASH 2018) and evaluated the impact of trial eligibility criteria by race/ethnicity. Methods: Newly diagnosed patients with DLBCL were enrolled from 2015-2020 into the LEO Cohort from 8 large academic centers across the US. Clinical data, including baseline laboratory values, were abstracted from the medical record using a standard protocol. Patients were prospectively followed for outcomes. All patients for this analysis received anthracycline plus CD-20 antibody-based IC to standardize treatment. Organ function parameters were identified from the exclusion criteria for hemoglobin, absolute neutrophil count, platelet count, creatinine, and bilirubin, as reported in recent frontline trials (Table 1). Associations between trial eligibility and event-free (EFS) and overall survival (OS) were evaluated using Kaplan Meier curves. The POLARIX trial was used as an example here for outcomes as it is the most recently published frontline DLBCL trial showing a benefit in clinical outcomes over the standard of care R-CHOP. Results: A total of 2510 DLBCL patients received frontline IC. Of these 2330 had 3 or more of the 5 lab-based values available in LEO and comprised the analysis set. The group included 1972 non-Hispanic whites (NHW, 85%) and 272 Hispanic and/or non-whites (H/NW, 12%) participants with the rest unknown race/ethnicity. Compared to NHW, H/NWs were younger at diagnosis (median 54 vs. 64 years, p <0.001) but were comparable for elevated LDH (56% vs. 63%), B-symptoms (32% vs. 36%), bone marrow involvement (16% vs 18%), extranodal disease (26% vs 32%) and high-risk IPI (38% vs. 37%). Hemoglobin was significantly lower (mean + SD) in the H/NW (11.7±2.3) vs. NHW (12.5±2.2, p<0.001), while other labs values were similar between the two groups. Treatment regimens varied between the 2 groups with more NHW receiving R-CHOP (53% vs.38%, p <0.001), and more H/NWs receiving R-EPOCH (38% vs. 25%, p <0.001). Within the LEO cohort 9-26% of patients would have been excluded from recent frontline DLBCL clinical trials based on the 5 lab-based criteria alone (Table 1). Of these, ENGINE (26%), POLARIX (17%), and GOYA (17%) trials were the most restrictive. Trials with a higher cut-off value of hemoglobin (9gm/dL or more) had a larger impact on H/NW population's ineligibility. OS was significantly inferior for patients ineligible for all trials listed in table 1 based on lab criteria for both H/NW and NHW patients [POLARIX as an example - Figure 1]. Associations between ineligibility and poor outcomes were consistent across trial criteria and endpoints (OS and EFS). Conclusions: This study confirms our previously reported association between trial lab-based eligibility criteria and outcomes in newly diagnosed DLBCL in a larger, contemporary, diverse cohort and showed some criteria disproportionately limited eligibility for H/NW patients. Further studies to evaluate the impact of each lab-based criteria on eligibility in different racial/ethnic populations are required. Future trial designs need to tailor eligibility criteria to be more inclusive of the H/NW population. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Dose Intensity and Reasons for Dose Alterations in Patients Excluded from Frontline Diffuse Large B-Cell Lymphoma Clinical Trials Based on Eligibility Criteria: A Mayo Clinic Cohort Study

Background: Frontline clinical trials in DLBCL can exclude up to 24% of the real-world population based on 5 lab-based eligibility criteria (Khurana et al. JCO 2021). Furthermore, these patients have worse clinical outcomes and increased deaths due to lymphoma progression suggesting, that the patients that could benefit the most from novel therapies in clinical trials are left behind. However, the details of chemotherapy dose intensity and its contribution to poor outcomes remain unknown. We sought to evaluate the chemotherapy dose intensity and reasons for dose alterations in these clinical trial ineligible patients and to identify a subset of these patients that can potentially be rescued back into the trials. Methods: Patients with newly diagnosed lymphoma were enrolled from 2002 to 2015 in the Molecular Epidemiology Resource (MER) of the University of Iowa and Mayo Clinic Lymphoma SPORE and prospectively followed. Clinical data, treatment, and outcomes were abstracted from medical records using a standard protocol. Eligible patients had DLBCL, initiated frontline immunochemotherapy (IC), and were ineligible for the ENGINE trial (an ongoing phase III multicenter trial - most restrictive per our prior analysis) based on ENGINE trial's lab criteria. Average relative dose intensity (ADI) was calculated for doxorubicin and cyclophosphamide based on both the actual dose and time to deliver the doses (Bataillard et al. Blood Advances 2021). To identify patients that could be rescued back into trials we divided the patients into 4 categories (Table 1) based on disease risk and Charlson comorbidity index (CCI). High-risk disease was defined as IPI >2 and/or diagnosis to treatment interval (DTI) < 14 days; high-risk comorbidity score was CCI > 2. Baseline patient characteristics and chemotherapy details including reasons for dose alterations were compared between groups. Results: Of the 1265 IC treated DLBCL patients in MER, 305 (24%) patients were deemed ineligible for the ENGINE trial based on 5 lab criteria. Ineligible patients were significantly older (median 68 vs. 61 yrs), with worse performance status (PS >2, 29.8 vs. 12.6%), had elevated LDH (70 vs 52%), high-risk IPI (57 vs. 32%), shorter DTI (median 13 vs. 16 days), and more B-symptoms (43 vs 23%). Of the 305, 144 (47%) patients were treated at Mayo and had complete details on chemo dosing. The baseline characteristics of this cohort and different subgroups are shown in table 1. Over 95% of patients were planned for 6 or more chemo cycles but only 73% completed planned cycles. 56% of patients were admitted to the hospital for cycle 1. The most common reasons for inpatient chemo delivery were symptomatic disease (35%), and lab derangements other than tumor lysis or cytopenia requiring intervention (21%). 19% (28/144) of patients had planned reduction in cycle 1 for both doxorubicin and cyclophosphamide with the most common reasons of older age and comorbidities (57%). Hospitalization/ER visits occurred in 38% (55/144) of the patients after cycle 1, with 22% (32/144) febrile neutropenia. 51% of patients had 1 or more chemo cycles delayed with the most common reason of febrile neutropenia (24%), followed by count recovery (10%). The median age was greater for patients with high CCI compared to low CCI for both high risk (77 vs 57) and low risk (78 vs. 50) disease groups. Patients with high-risk disease more commonly required inpatient admission for cycle 1 and pre-chemo steroids and rituximab optimization. All patients maintained >60% ADI in the low CCI groups (range 60-100%) compared to high CCI (range 20-100%). Clinical outcomes (overall survival) in the high-risk disease and low CCI group were significantly better than high-risk disease and high CCI (Figure 1). Conclusions: Patients who were ineligible for clinical trials based on lab-based criteria had disproportionately more high-risk disease features. Patients with high-risk disease more commonly required pre-chemo steroids and rituximab optimization. While small numbers and unbalanced subgroups limit interpretation, low CCI patients despite high- risk disease were able to maintain ADI and attempts should be made to include these patients in clinical trials. High CCI and high-risk disease subgroup had the worst outcomes and most dose alterations suggesting an unmet need with alternative less toxic treatment strategies and comorbidity management. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Are pivotal clinical trials for lymphomas that led to drug approval representative of the population affected by these diseases?

6531 Background: There are significant inequities in cancer care and outcomes in the United States (US) and worldwide. Race, ethnicity, and sex all define sub-groups affected by these inequities. Randomized controlled trials (RCTs) provide new therapies in cancer care. Many new drugs have been recently approved for the management of lymphoma. We examined whether pivotal clinical trials included populations representative of those affected. Methods: Lymphoma clinical trials were collected from the US Food and Drug Administration (FDA) Databases ‘Oncology (cancer)/Hematologic Malignancies approval notifications’ and ‘Novel Drug Approvals’ between 2011 and 2021. Trials were also reviewed for demographic data on ClinicalTrials.gov (CT.gov) and Primary Literature related to the drug approval. Only studies for adult patients were included; one study was excluded due to inconsistencies in the total patients enrolled and the totals by race. Chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL) RCTs were not included. The burden of diseases based on Surveillance, Epidemiology, and End Results (SEER) data was used as a comparison for US participants. Results: Thirty-three pivotal trials were identified from the FDA databases (2011-2021); 18 (54.5%) had available race data on CT.gov or in Primary Literature. Three trials focused on classical Hodgkin’s Lymphoma (cHL) and 15 on Non-Hodgkin’s Lymphoma (NHL). Two of the three trials for cHL with available data for race reported ethnicity data. Only 10 of 15 NHL trials that reported race demographics had available ethnicity data. Table 1 shows trials that had available data for race and ethnicity along with proportions from the SEER data. There was underrepresentation of Blacks and to some extent Hispanics in both cHL and NHL trials. Females were also underrepresented, particularly in NHL trials. Conclusions: There is racial, ethnic, and sex misrepresentation within clinical trials that led to approval of drugs for lymphomas in the US between 2011 and 2021. In addition, there was significant underreporting of racial and ethnic subgroups noted in the clinical trials.[Table: see text]

Enrollment disparities in non-Hodgkin lymphoma clinical trials.

e18594 Background: Disparities in cancer clinical trials limit the real-world reproducibility of their results while remaining inapplicable to excluded patients (pts) like those living with chronic viral infections. Non-Hodgkin lymphoma (NHL) has an incidence distributed heterogeneously among races and is associated with chronic viral infections such as HIV, Hepatitis B (Hep B) and Hepatitis C (Hep C). Given these epidemiological characteristics, we assessed demographic representation in NHL clinical trials to identify areas for improvement. Methods: Clinicaltrials.gov was queried for NHL trials and specified for completed, interventional, phase 2, 3 &amp; 4 trials, with results and with recruitment in the US. Trials including malignancies other than NHL, with enrollment outside of the US or lacking reported demographics on clinicaltrials.gov or trial publications were excluded. SEER 21 and the 2020 US census were used to calculate incident cases for each specified subgroup. Enrollment fraction (EF) was defined as the number of trial enrollees divided by the estimated US cases in each subgroup. We compared EF between subgroups utilizing odds ratios (OR) with p-value &lt; 0.05 deemed significant. Results: Of 176 trials meeting criteria, 64 (36.4%) reported ethnicity and 79 (44.9%) reported race. EF of Black, Asian/Pacific Islander, American Indian and Hispanic (HSP) pts were significantly lower compared to White and non-HSP pts (Table). Racial minorities were less likely to be enrolled in relapsed/refractory trials (OR 0.31 - 0.57 p &lt; 0.008) compared to White pts but all were equally likely to be enrolled in front-line trials. HSP pts had a lower EF compared to non-HSP pts in both relapsed/refractory and front-line trials (OR 0.57, 0.46 p &lt; 0.0001). Black and HSP pts were also less likely to be enrolled in hematopoietic stem cell transplant (HSCT) trials (OR 0.27, 0.74, p &lt; 0.003). Women had a significantly lower EF compared to men (Table). 150 trials had inclusion/exclusion criteria specific for chronic viral infections: 141 (94%) excluded pts with HIV, 108 (72%) with chronic Hep B and 103 (69%) with Hep C. Conclusions: Our study demonstrates underrepresentation of not only women, but also of Black, Asian/Pacific Islanders, American Indian and HSP pts in NHL trials, especially in relapsing/refractory and HSCT trials. Moreover, pts living with chronic viral infections are systematically excluded by trial criteria. It is essential that trials reporting on safety, toxicity and efficacy represent the true patient population to limit ascertainment bias and improve external validity of results.[Table: see text]