Citations

Introduction Registrational trials are clinical trials which are used to justify regulatory approval of a new medication by the FDA. However, clinical trials may not be designed around patient-centred outcomes, like overall survival (OS) or quality of life. This disconnect raises concerns about whether current trial designs are truly aligned with the long-term needs and priorities of patients. A comprehensive analysis of trial designs and endpoints is necessary to ensure that future oncological therapies deliver meaningful clinical improvements. Our objective was to describe types of clinical trial design and primary endpoints used for registrational trials, and their trends over time. Research design and methods In this cohort study, the HemOnc database was used to identify registrational trials supporting cancer drug approval from 2004 through 2023. Clinical trial design (escalation, de-escalation, in-class switch, out-of-class switch) and primary endpoint (OS; time-to-event including progression-free survival; response-based endpoint) were characterized. Results Escalation trial designs were more common than in-class switches or out-of-class switches. De-escalation and mixed designs were exceedingly rare. The most common primary endpoint was a composite time-to-event endpoint. Conclusion Most clinical trials leading to FDA approval for cancer-directed therapies employ escalation designs with composite time-to-event endpoints. Designs reflecting paradigm shifts in therapy (in-class switches and out-of-class switches) are sparsely used. This approach raises the concern that trials may increase therapeutic toxicity without delivering meaningful improvements in OS or quality of life.

Objective To provide evidence‐based and expert guidance for the treatment and management of non‐renal systemic lupus erythematosus (SLE); treatment and management of lupus nephritis are addressed in a separate guideline. Methods Clinical questions for treatment and management of SLE were developed in the PICO format (population, intervention, comparator, and outcome). Systematic literature reviews were developed for each PICO question, and the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) methodology was used to assess evidence quality and formulate recommendations. The Voting Panel achieved a consensus of ≥70% agreement on the direction (for or against) and strength (strong or conditional) of each recommendation. Results We present recommendations and ungraded, consensus‐based good practice statements for the treatment and management of SLE that are applicable to pediatric and adult patients. Recommendations emphasize uniform treatment with hydroxychloroquine, limiting duration of glucocorticoid use, and early introduction of conventional and/or biologic immunosuppressive therapies to achieve and maintain control of SLE inflammation (remission or a low level of disease activity), reduce SLE‐related morbidity and mortality, and minimize medication‐related toxicities. Conclusion This guideline presents direction regarding treatment and management of SLE and provides a foundation for well‐informed, shared clinician–patient decision‐making. These recommendations should not be used to limit or deny access to therapies, as treatment decisions may vary due to the unique clinical situation and personal preferences of each person with SLE.

Objective To provide evidence‐based and expert guidance for the treatment and management of non‐renal systemic lupus erythematosus (SLE); treatment and management of lupus nephritis are addressed in a separate guideline. Methods Clinical questions for treatment and management of SLE were developed in the PICO format (population, intervention, comparator, and outcome). Systematic literature reviews were developed for each PICO question, and the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) methodology was used to assess evidence quality and formulate recommendations. The Voting Panel achieved a consensus of ≥70% agreement on the direction (for or against) and strength (strong or conditional) of each recommendation. Results We present recommendations and ungraded, consensus‐based good practice statements for the treatment and management of SLE that are applicable to pediatric and adult patients. Recommendations emphasize uniform treatment with hydroxychloroquine, limiting duration of glucocorticoid use, and early introduction of conventional and/or biologic immunosuppressive therapies to achieve and maintain control of SLE inflammation (remission or a low level of disease activity), reduce SLE‐related morbidity and mortality, and minimize medication‐related toxicities. Conclusion This guideline presents direction regarding treatment and management of SLE and provides a foundation for well‐informed, shared clinician–patient decision‐making. These recommendations should not be used to limit or deny access to therapies, as treatment decisions may vary due to the unique clinical situation and personal preferences of each person with SLE.

e13656 Background: Evidence-based medicine is crucial for clinical decision-making, yet studies suggest that a significant proportion of treatment decisions do not fully incorporate the latest evidence. Large Language Models (LLMs) show promise in bridging this gap, but their reliability for medical recommendations remains uncertain. Methods: We conducted an evaluation study comparing five LLMs’ recommendations across 50 clinical scenarios related to multiple myeloma diagnosis, staging, treatment, and management. The evaluation included three general-purpose LLMs (OpenAI o1-preview, Claude 3.5 Sonnet, Gemini 1.5 Pro), one retrieval-augmented generation (RAG) system (Myelo), and one agentic workflow-based system (HopeAI). General-purpose LLMs generated responses based solely on their internal knowledge, while the RAG system enhanced these capabilities by incorporating external knowledge retrieval. The agentic workflow system extended the RAG approach by implementing multi-step reasoning and coordinating with multiple tools and external systems for complex task execution. Three independent hematologist-oncologists evaluated the LLM-generated responses using standardized scoring criteria developed specifically for this study. Performance assessment encompassed five dimensions: accuracy, relevance, comprehensiveness, hallucination rate, and clinical use readiness. Results: HopeAI demonstrated superior performance across accuracy (82.0%), relevance (85.3%), and comprehensiveness (74.0%), compared to OpenAI o1-preview (64.7%, 57.3%, 36.0%), Claude 3.5 Sonnet (50.0%, 51.3%, 29.3%), Gemini 1.5 Pro (48.0%, 46.0%, 30.0%), and Myelo (58.7%, 56%, 32.7%). Hallucination rates were consistently low across all systems: HopeAI (5.3%), OpenAI o1-preview (3.3%), Claude 3.5 Sonnet (10.0%), Gemini 1.5 Pro (8.0%), and Myelo (5.3%). Clinical use readiness scores were relatively low for all systems: HopeAI (25.3%), OpenAI o1-preview (6.0%), Claude 3.5 Sonnet (2.7%), Gemini 1.5 Pro (4.0%), and Myelo (4.0%). Conclusions: This study demonstrates that while current LLMs show promise in medical decision support, their recommendations require careful clinical supervision to ensure patient safety and optimal care. Further research is needed to improve their clinical use readiness before integration into oncology workflows. These findings provide valuable insights into the capabilities and limitations of LLMs in oncology, guiding future research and development efforts toward integrating AI into clinical workflows.

Abstract Background Instruments to assess patient-reported outcomes (PRO) should generate high-quality evidence. Reliable PRO evidence is essential to policymakers, in conjunction with outcomes such as survival and radiological response, to understand the net clinical benefit of antitumor treatments. This study aimed to establish the content validity of 215 identified PRO measures used in patients with brain tumors. Methods A survey (n = 148 items) was developed reflecting aspects of the WHO International Classification of Functioning, Disability, and Health (ICF) framework. Patients with brain tumors, their proxies, and healthcare professionals (HCPs) were asked to rate each survey item on relevance. An item was considered a relevant issue if ≥25% of the patients or proxies or ≥50% of the HCPs considered that item to be an issue. Next, all items in the identified PRO measures were linked to ICF and relevant items in the survey, and the percentage of content coverage was calculated. Results In total, 114 patients, 71 proxies, and 65 HCPs from different countries completed the survey. Fifty-six of 148 (37.8%) items in the survey were considered relevant. The most important aspects mentioned by both patients and proxies were difficulty concentrating, difficulty remembering, multitasking, and handling stress. Depending on the definition, between 35% and 49% of PRO measures were considered to have sufficient content validity (≥80% coverage). Conclusion The content validity was insufficient in more than half of the identified PRO measures, particularly multidimensional measures. Future research should investigate whether different approaches to PRO assessment better meet the needs of all stakeholders.

Abstract Lineage switch (LS), defined as the immunophenotypic transformation of acute leukemia, has emerged as a mechanism of relapse after antigen-targeted immunotherapy, which is associated with dismal outcomes. Through an international collaborative effort, we identified cases of LS after a host of antigen-targeted therapies (eg, CD19, CD22, CD38, and CD7), described how LS was diagnosed, reviewed treatment approaches, and analyzed overall outcomes for this form of postimmunotherapy relapse. Collectively, 75 cases of LS were evaluated, including 53 (70.7%) cases of B-cell acute lymphoblastic leukemia (B-ALL) transforming to acute myeloid leukemia (AML), 17 (22.7%) cases of B-ALL transforming to mixed phenotypic acute leukemia (MPAL)/acute leukemias of ambiguous lineage (ALAL), and 5 (6.7%) cases of rare LS presentation (ie, T-cell ALL to AML). An additional 10 cases with incomplete changes in immunophenotype, referred to as “lineage drift” were also described. With a primary focus on the 70 cases of LS from B-ALL to AML or MPAL/ALAL, LS emerged at a median of 1.5 months (range, 0-36.5) after immunotherapy, with 81.4% presenting with LS within the first 6 months from the most proximal immunotherapy. Although most involved KMT2A rearrangements (n = 45, 64.3%), other rare cytogenetic and/or molecular alterations were uniquely observed. Treatment outcomes were generally poor, with remission rates of <40%. The median overall survival after LS diagnosis was 4.8 months. Outcomes were similarly poor for those with rare immunophenotypes of LS or lineage drift. This global initiative robustly categorizes lineage changes after immunotherapy and, through enhanced understanding, establishes a foundation for improving outcomes of LS.

ABSTRACTRubinstein‐Taybi syndrome (RTS) is a rare genetic disorder featuring craniofacial abnormalities, intellectual disability, and multi‐organ involvement, including increased risk of keloid formation, pilomatricomas, and hypertrichosis. Hidradenitis suppurativa (HS), a chronic inflammatory skin condition characterized by recurrent abscess formation and draining tunnels, has never been described in a patient affected by RTS. In this report, we detail the history of a 32‐year‐old male with RTS and refractory Hurley stage III HS and discuss the important considerations and risks of managing this unique patient population.

Background: Contemporary questions about treatment and outcomes of patients (pts) living with multiple myeloma (MM) demand efficient and timely evidence generation methods, where real-world/ everyday practice data can play an important role. Relevant potential applications include understanding and addressing disparities in treatment access, mitigating toxicities of novel therapies while maintaining quality of life, appropriately sequencing treatments within long-term management strategies, and disseminating and appropriately using surrogate endpoints in practice, such as minimal residual disease. However, existing real-world data (RWD) repositories are limited by an inability to implement interventions in response to issues identified through retrospective data analysis, which is pivotal for the use of newer therapies. The ASH Research Collaborative (ASH RC) MM Data Hub Program was developed to address these gaps. Methods: We developed a multicenter network of US-based tertiary academic medical centers, engaging site clinicians, patients, biostatisticians, informaticians, and data scientists. Initial activities included constructing a shared real-world MM data model comprised of consensus core data elements. Pt data from each site, dating back to at least 2019, is currently being entered through electronic case report forms, batch uploaded utilizing existing institutional data repositories, and soon direct electronic health record import in standard data acquisition formats. Data are aggregated in the Data Hub and shared in a confidential Site Portal with each site. The Portal allows sites to track their own pts, analyze outcomes, and identify ways to improve the quality of care provided to individuals living with MM. We also developed an innovative evidence generation learning community of site clinicians who meet monthly and collaborate on the “COVID-19 OutcomeS in Myeloma and the Impact of VaCcines (COSMIC)” study, a prospective, decentralized observational study that rapidly accrued pts and collects relevant practice information. The group also shares best practices in study conduct to accelerate study enrollment and execution. Results: The MM Data Hub Program now contains twelve centers: Memorial Sloan Kettering Cancer Center, Mayo Clinic, Icahn School of Medicine at Mount Sinai, University of California San Francisco, University of California Davis, Medical College of Wisconsin, Barbara Ann Karmanos Cancer Institute, University of North Carolina Medical Center at Chapel Hill, Emory University Hospital, Washington University School of Medicine in St. Louis, Dana-Farber Cancer Institute, and Atrium Health. From this Network, six centers have contributed data from 8,824 pts (2,825 with sufficiently complete data are included in this analysis). The median age at time of MM diagnosis is 63, and of the 627 deceased pts, the median age at death is 70.6 years, with median time to death from first MM diagnosis of 6 years. The proportions of pts by decade of diagnoses are 34.9% from 2020-2024; 57.6% from 2010-2019; 7.3% from 2000-2009; and 0.9% prior to 2000. 83.1% of pts had a diagnosis of plasma cell myeloma (non-light chain), and 14.4% of pts had a diagnosis of light chain myeloma. 42.3% of pts are female, and the self-reported patient races are: 85.0% white; 9.2% black or African; and 4.3% all other races (race was not documented in 1.5% of pts). 2.8% of pts are of Hispanic ethnicity. Participant data files are currently undergoing completion and cleaning procedures; the analysis will include updated results at the meeting, and a Network-wide analytic data set will be available for initial retrospective data analyses in early 2025. The COSMIC study opened for enrollment on January 22, 2024, and completed enrollment of 200 pts on May 17, 2024. Conclusions: The ASH RC Data Hub MM Program launched in 2023 and is comprised of 12 sites that collectively have contributed participant files for 8,824 pts thus far. Utilizing the structure of the Network, we successfully implemented and established the feasibility of collecting data directly from pts electronically to monitor SARS-CoV-2 vaccines and COVID-19 related outcomes. The Program serves as a successful proof-of-concept model, providing an enduring RWD resource and a prospective evidence generation platform for the myeloma community.

Background: The COVID-19 pandemic had an outsized impact on individuals with immune compromise, such as multiple myeloma, leading to increased rates of severe COVID, hospitalization, and death in these groups. The pandemic also demonstrated the need for rapid and efficient evidence generation to address rapidly changing medical needs and accelerated the use of innovative trial designs to accomplish this goal. Methods: We designed and established a prospective, decentralized study within a site-based provider network, COVID-19 OutcomeS in Myeloma and the Impact of VaCcines (COSMIC), for patients with multiple myeloma at ASH Research Collaborative sites to monitor outcomes related to COVID-19 infection and vaccination. The direct-to-patient approach was accomplished using a study application accessed with a QR code provided to patients during usual clinical care. All study activities occurred within the patient portal at Baseline, Day 30, and Day 180, minimizing burden on site personnel and staff. Study activities included eligibility confirmation, consenting, electronic health record (EHR) data permissions, and patient-reported outcome (PRO) surveys. PRO data (EORTC QLQ-C30, EORTC QLQ MY-20) were made available to clinicians through the portal to visualize results from their patients. Patients shared if they received a COVID-19 vaccination or booster and if they had a COVID-19 infection. Patients with a history of COVID-19 infection included method of confirmation and if hospitalization was required. Patients were provided compensation for participation. Automated text and email messages were sent at three separate intervals: 24 hours, Day 8, and Day 15 for participants who initiated but did not complete study activities. A personal email reminder per participant was sent for those whose second study visit window was approaching. ASH Research Collaborative clinicians met regularly to share best practices related to enrollment and study conduct, thereby creating an innovative evidence generation learning community. We studied the feasibility of data collection in a decentralized format by focusing on the number of participants who interacted with the system, those who consented, and those who completed baseline PRO assessments. Results: The COSMIC study rapidly accrued and consented 200 patients from eight participating ASH Research Collaborative sites over four months (1/22/2024 through 5/17/2024): 69% self-reported as non-Hispanic White (NHW). Of enrolled participants, 57.5% opted to permission sharing of their EHR data, which included disease and treatment-related clinical data, laboratory results, pathology reports, and imaging reports. A total of 196 patients (98%) provided demographic and COVID-related disease and vaccine data at baseline; nearly all of these patients (n=195) provided baseline PRO data, and 150 patients provided further information at Day 30 at the time of this analysis. Those who did not provide Day 30 data were not significantly different from those who did in terms of age (median: 68 [no Day 30 data] vs. 67 [Day 30 data] years; p=0.44), sex (41% [no Day 30 data] vs. 47% [Day 30 data] Female; p=0.47), or race (65% [no Day 30 data] vs. 71% [Day 30 data] NHW; p=0.60). A higher proportion of Hispanic patients were retained at Day 30 (13/15=87%) vs. 78% of non-Hispanic patients or those who declined reporting ethnicity (2/7=29%; p=0.007). COVID-19 vaccination data were submitted by 196 patients, 186 of whom had ever received a COVID-19 vaccination. Data on number, type, brand, and dates for these vaccinations were provided. Conclusions: The COSMIC study conducted by the ASH Research Collaborative was able to rapidly accrue a diverse cohort of patients using a decentralized workflow within a site-based provider network, with high consent and compliance rates. This platform, and the innovative evidence generation learning community of ASH Research Collaborative investigators and patients, sets a precedent to address a variety of real-world questions related to disease management, infections, vaccinations, and PROs in this population of patients with multiple myeloma.

Background: Data from the ASH Research Collaborative COVID-19 Registry for Hematology demonstrated that patients with multiple myeloma and other immune compromise had elevated risk for severe complications from COVID-19 infection. To address current evidence gaps and to demonstrate proof of concept for a pandemic preparedness and response network, the ASH Research Collaborative Multiple Myeloma Program developed a prospective, decentralized study within a site-based provider network to investigate COVID-19 infections, vaccinations, and quality of life in patients with multiple myeloma. Methods: We designed and established the COVID-19 OutcomeS in Myeloma and the Impact of VacCines (COSMIC) to study COVID-19 in patients with multiple myeloma at ASH Research Collaborative network sites. A study application was accessed with a QR code and shared with patients during usual clinical care. Patients provided consent, optionally gave permission for electronic health record data collection, and completed patient-reported outcomes (EORTC QLQ-C30 and MY20) at baseline, Day 30, and Day 180 through the study portal. Clinicians at participating sites facilitated study enrollment and shared best practices monthly to accelerate study execution. Study objectives related to feasibility of data collection in a decentralized format are reported in a separate abstract. Here, we report interim results related to SARS-CoV-2 vaccination history, COVID-19 infection, and how these relate to patient reported outcomes and clinical outcomes in enrolled participants. Results: The COSMIC study rapidly accrued and consented 200 patients from eight participating ASH Research Collaborative sites over four months (1/22/2024 through 5/17/2024). Consented participants' ages ranged from 29 to 93 years, with median of 67 years; 46% were female, 69% were non-Hispanic White (NHW), 17% were non-Hispanic Black (NHB), and 8% were Hispanic/Latino. Of the 196 patients who reported COVID-19 data, 111 (57%) reported at baseline that they had a history of COVID-19 infection, with 12 having required hospitalization for this (12/111=11%). Overall, 95% (186/196) reported a history of COVID-19 vaccination, where the median number of vaccinations was 4 (range: 1 to 6); 44 patients (23%) expressed willingness to discuss getting a yearly booster vaccine with their provider. At 1-month follow-up, there were 3 additional patients who reported a COVID-19 infection and 18 patients who received an additional COVID-19 vaccination. The baseline EORTC QLQ-C30 median Global Health Status/QOL score was 66.7 (mean=64.4, range: 8.3 - 100, n=196). This overall global QOL score was not significantly different between those who did vs. did not have a history of COVID-19 infection (mean [SD]: 63.7 [19.3] vs. 65.8 [17.6]; p=0.57). In analyses adjusted for age, biological sex, and race/ethnicity group, history of COVID-19 infection was not associated with HRQOL (odds ratio=0.93, 95% CI: 0.78 - 1.10; p=0.38). Data collection is ongoing and final study results will be reported at a later date. Conclusion: COSMIC is a prospective, decentralized study within a site-based provider network conducted by the ASH Research Collaborative Multiple Myeloma Program that demonstrates the feasibility of using an innovative evidence generation strategy to drive rapid accrual of patients to address a rapidly evolving area of need. The COSMIC study also provides proof of concept that a real-world evidence generation network can play an important role in national pandemic preparedness and response. The COSMIC experience reinforces the ability to accrue diverse patient populations representing the demographic of myeloma patients in the US.