2026 ASH®
Annual Meeting & Exposition December 12-15, 2026 ▪ New Orleans, Louisiana

Paper Details

Abstracts, authors, affiliations, and disclosures.

Saturday, December 12, 2026
05:30 p.m. - 07:30 p.m.
MCCNO - Hall F
Access, Representation, Population Health, Technology and Procedures, Bioinformatics, Emerging Technologies, Treatment Considerations, Combination Therapy, Supportive Care, Therapy Sequence, Biological Therapies, Antibody Therapy, Transplantation, Bispecific Antibody Therapy, (CAR)-T Cell Therapies, Immunotherapy, Adverse Events, Artificial Intelligence (AI), Generative AI, Clinical Practice (Health Services and Quality), Clinical Guidelines, Health Economics, Patient Safety, Practice Models, Quality Improvement, Quality of Care, Value, Diseases, Lymphoid Malignancies, Lymphoid Leukemias, Lymphomas, Plasma Cell Disorders, Myeloid Malignancies, CLL, Hodgkin Lymphoma, B Cell Lymphoma, Indolent Lymphoma, Aggressive Lymphoma, Research, Clinical Research, Clinical Trials, Epidemiology, Health Outcomes Research, Health Disparities Research, Patient-Reported Outcomes, Real-World Evidence, Registries, Study Population, Human, Adult
PAPER

2646 An accessible AI-based clinical trial matching application for patients with multiple myeloma demonstrates performance comparable with manual trial matching

Ori Grossberger, MD, Tamir Shragai, MD2, Yossi Bengigi, N/A3, Irit Avivi Mazza, MD4, Yair Herishanu, Prof5, Yotam Bronstein, M.D6, Maria Pavlov, N/A7, Ricky Avissara, N/A8, Audelia Eshel Fuhrer, MD9, Yael C. Cohen, MD10
1 Faculty of Medical and Health Sciences, Tel Aviv University, Tel Aviv, Israel, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel
3 OITO Ltd., Tel Aviv, Israel
4 Faculty of Medical and Health Sciences, Tel Aviv University, Tel Aviv, Israel, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel
5 Faculty of Medical and Health Sciences, Tel Aviv University, Tel Aviv, Israel, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel
6 Faculty of Medical and Health Sciences, Tel Aviv University, Tel Aviv, Israel, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel
7 Tel Aviv Sourasky Medical Center, Tel Aviv, Israel
8 Tel Aviv Sourasky Medical Center, Tel Aviv, Israel
9 Tel Aviv Sourasky Medical Center, Tel Aviv, Israel
10 Faculty of Medical and Health Sciences, Tel Aviv University, Tel Aviv, Israel, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel

Introduction

Participation in clinical trials provides Multiple Myeloma (MM) patients (pts) with early access to innovative therapies, while driving the development and approval of novel drugs. Nowadays, pts often seek to take an active role in managing their disease and independently explore available clinical trials, yet matching trial eligibility criteria with pts features is a complex and time-consuming task and existing interfaces are challenging to navigate. To overcome this challenge, we developed and validated “MMimy”, an AI-powered web application, aimed to suggest potential clinical trials to MM pts, which match their disease features.

Methods

MMimy was developed in collaboration between MM physicians and AI experts. Based on extensive experience in MM trials, key drivers for MM trial eligibility were identified, primarily: prior exposure and refractoriness to MM drugs, washout period, lines of therapy, key laboratory values including presence of measurable disease. The interface includes a series of structured questions regarding disease characteristics and previous treatments, and a placeholder for uploaded anonymized snapshots of recent laboratory reports. In the backend, a 'disease map' is created and eligibility is evaluated against currently recruiting MM clinical trials listed on ClinicalTrials.gov at a given location, generating a personalized list of potential studies. Each trial is categorized as “full-match” or “partial-match” (if some information is missing for full matching), both considered as “match”, or “doesn’t-fit” (exclusion criteria were detected). The output includes the rational of the categorization. The application is a secure web platform deployed on a cloud-based server, developed entirely using Claude Code (Anthropic). Uploaded laboratory reports (PDF/ image files) are automatically read using AI, which extracts the values into a structured template. Each clinical trial has a 37-item eligibility template, based on ClinicalTrials.gov and designed by the MM experts. A rule-based matching engine compares the pts ‘disease map’ with each trial's publicly available protocol to create the final matching.

The development phase required iterative validation using a range of theoretical clinical scenarios representative of pts with MM. Following optimization of the matching algorithm, the system was evaluated with a prospective validation study (approved by the institutional ethics committee), which enrolled MM patients at a treatment decision time-point. MMimy’s categorization output for the participating pts were subsequently compared with an independent and blinded “classical” trial-matching process performed by the MM-experienced study coordinators, who used the same input data. Gold standard for eligibility was based on assessment by two MM physicians. The predefined endpoints were sensitivity, positive predictive value (PPV) and accuracy.

Results

Thirty-two recruiting MM trials were identified in Israel at the time of validation. Ten pts enrolled into the prospective validation study: 4 NDMM and 6 RRMM; 5 entered their data independently, and 5 required assistance from a Health Care Professional. All pts received at least one “match” trial suggestion (range: 1-14). Comparing MMimy’s “match” suggestions to the gold standard showed 94.8% sensitivity, 100% PPV and 99% accuracy. Comparing “classical” trial matching to the gold standard showed 96.3% sensitivity, 86% PPV and 97% accuracy. Validation of an expansion cohort is underway and will be presented at the meeting.

Conclusion

Prospective validation of MMimy, an AI-powered web application for mapping relevant clinical trials to MM pts, demonstrated high sensitivity, PPV, and accuracy that were comparable to those of the clinical research team. It provides pts and their caregivers with an accessible and user-friendly interface for independently identifying potentially relevant clinical trials. Nevertheless, important challenges remain, including technological barriers among users with limited digital literacy, and the incomplete eligibility criteria publicly available in clinicaltrials.gov. MMimy can support pts discussing relevant options with their treating physician. Evaluation of complete medical records vs. full trial protocols is still essential to confirm eligibility.

Ori Grossberger, MD: No relevant conflicts of interest to declare.