Highlights from the MMRF’s Meeting on AI and Data in Multiple Myeloma

For more than a decade, the Multiple Myeloma Research Foundation® (MMRF®) has built and maintained the largest and most comprehensive multiple myeloma dataset in the world. Importantly, the organization has shared this data with the field and invested in new studies and data types to keep it relevant.

Throughout the year, the MMRF hosts meetings and workshops to discuss the potential clinical impact of these data. The world’s leading researchers, clinicians, companies, and data scientists come together to discuss challenges, examine how emerging technologies can generate new insights, and find ways to put the MMRF’s data to work on questions that matter most to patients.

One of those meetings is the MMRF’s highly anticipated Immunity Workshop, which convenes the research community to discuss the role of the immune system in myeloma. The MMRF recently hosted an Immunity Workshop focused on one of the biggest questions in the field right now: How do we harness new AI tools and data analytics, along with everything we’re learning about the immune system, to improve patient outcomes?

Over a full day of presentations and panels, MMRF leaders, researchers and clinicians, and industry representatives from companies like Johnson & Johnson, Tempus AI, and One Biosciences shared how they’re using MMRF data and cutting-edge tools to improve our understanding of myeloma and the immune system, identify new biomarkers, refine risk stratification, and, ultimately, cure every patient.

Below, read the biggest takeaways from the day.

1. MMRF data and collaboration made the day’s biggest ideas possible

The clearest throughline of the entire workshop was that no single institution, company, AI agent, or dataset can solve myeloma alone. Speakers repeatedly emphasized the need for the kinds of large-scale shared data sets and field-wide collaborative efforts that are hallmarks of the MMRF’s model.

Throughout the workshop, presenters referenced the MMRF’s groundbreaking CoMMpass℠ Study and Immune Atlas as foundational datasets that the entire field is building on.

“We are using Immune Atlas data as a foundation to build on to understand the biology of myeloma and identify new biomarkers,” shared Manoj Bhasin, PhD, MS of Emory University School of Medicine, a co-principal investigator on the MMRF’s Immune Atlas project.

The MMRF team shared the evolution of its data initiatives culminating in the recently launched MMRF Virtual Lab®, along with updates from the Horizon Clinical Trials Program and Translational Research Umbrella Program that are designed to generate new data for the field. Looking ahead, MMRF leaders shared plans to introduce new data challenges that would invite researchers to leverage Virtual Lab and MMRF data to answer some of the questions raised throughout the day.

As MMRF leadership noted throughout the day, progress doesn’t happen in isolation. By bringing together partners who are all working on a shared goal, building on MMRF’s foundational data and tools, and integrating data sets from across the field, we’ll accelerate a cure for every patient faster.

2. The importance of the immune system in myeloma cannot be overstated’

Another recurring theme was that long-term remission not only depends on killing cancer cells, but on whether a patient’s immune system can keep the disease under control over time.

Researchers presented evidence that immune dysfunction—particularly among an important type of white blood cell called T cells—may encourage relapse and that the immune environment surrounding myeloma cells may play an important role in disease progression.

Sessions such as these highlighted the need for further research on immune health and ways to strengthen a patient’s defenses. In other sessions, researchers called for new biomarkers to capture the strength of a patient’s immune system and track how they respond to treatment in the clinic and in clinical trials.

3. Cutting-edge technologies are mapping the immune microenvironment

Researchers have long studied myeloma cells in isolation, but increasingly, they’re looking at the surrounding immune cells, stromal cells, and bone marrow, which are collectively known as the tumor or immune microenvironment.

Using advanced single-cell sequencing and spatial mapping technologies, researchers are now able to see, cell by cell, how myeloma and immune cells interact and influence each other, revealing subtle changes and patterns that were once invisible. Speakers were candid that no single technology has all the answers. What’s needed instead is a combination of tools and high-quality data to generate more insights.

4. AI is identifying new biomarkers

Several speakers described using AI models trained on large datasets to predict which patients are likely to respond to a given therapy and estimate a patient’s “immune age” as a measure of immune fitness.

Workshop attendees expressed measured optimism about AI, with several noting that buzz-worthy AI models haven’t been validated widely enough to impact care. This is especially important for a disease as heterogeneous and complex as myeloma. Attendees nevertheless shared that AI is a powerful tool that will work best alongside human oversight and rigorous science.

5. “AI co-scientists” are helping researchers generate hypotheses and drug targets

Another talk from a researcher at Mass General explored “agentic AI,” or AI systems that can independently search scientific literature, spot gaps in current knowledge, generate new hypotheses, and even design experiments to test them. These tools are already helping automate literature reviews and surface research questions that might otherwise take months to surface, freeing researchers’ time for other tasks.

Relying too heavily on AI to guide scientific direction, however, can narrow the range of ideas researchers pursue. These tools work best with a scientist firmly at the helm, deliberately pushing for novel ideas.

Beyond prediction, some teams are putting AI to work on drug discovery. A researcher at Emory University described a group of AI agents that can propose entirely new combinations of treatment targets. It has already identified a promising new pairing alongside BCMA (a well-established myeloma target) that is now headed to laboratory testing to see if it holds up.

6. AI is turning messy medical records and scientific literature into usable knowledge

A team from Mount Sinai shared two AI systems tackling a critical problem: Most of what we know about patients is buried in messy, unstructured clinical notes, and most of what we know from published research takes months or years to reach the databases doctors use.

One system, built on large-language models, reads through electronic health records and structures the information into research-ready data in just a few hours. A companion system continuously scans scientific literature. It found that today’s most widely used precision-oncology databases are missing recently discovered mutations that have been linked to resistance against newer myeloma therapies. Identifying such knowledge gaps could directly inform a patient’s treatment.

7. AI-guided pathology tools are speeding up clinical trial enrollment

A researcher from Johnson & Johnson described an AI tool the company developed to support one of its clinical trials for a bladder cancer therapy. The tool scans routine pathology images to help predict whether a patient likely carries a mutation required for trial eligibility before formal genetic testing comes back. It is not meant to replace such testing but to help patients and their doctors make a faster decision about whether to pursue the trial.

The researcher noted that building the model was not the biggest challenge. It was integrating the tool into real clinical workflows and earning physician trust in its results.

The approach was developed for bladder cancer, but several attendees raised the possibility of adapting it for myeloma.

8. AI’s real-world performance still often trails the hype

Throughout the meeting, several speakers offered sobering, real-world counterpoints to the AI hype in cancer research. One widely used pathology-image model, for example, had only modest impact in the clinic—a reminder that AI tools trained on narrow data don’t always hold up in the messier real world. Another AI system built to flag disease progression and match patients to clinical trials performed about the same as oncologists doing this on their own.

These weren’t presented as failures so much as honest examples of the fact that while AI is a genuine accelerant in research, it isn’t magic. Just as with any new myeloma treatment or test, the field needs to rigorously test AI tools and find out where they add value.

9. The new definition of cure is top of mind

Just a few months ago, the International Myeloma Society hosted a historic meeting to agree on a new definition of a cure for myeloma. At the MMRF’s Immunity Workshop, a panel of leading myeloma physicians discussed the new definition and what the field is focusing on going forward.

Panelists described treating several patients who had had deep, treatment-free remissions for a decade or more and may be functionally cured. While it’s clear that the immune system plays an important role in keeping cancer under control for such patients, panelists highlighted that we still don’t fully understand the interplay between the immune system and myeloma. Data, such as the MMRF’s Immune Atlas, are key to unlocking the immune system’s complexity and understanding how patients achieve functional cures.

Data will also help physicians better stratify patients who have a small number of abnormal cells to predict which patients will progress to myeloma and need treatment and which will be able to safely live in an MGUS-like state indefinitely and without treatment.

Now that the field has defined what a cure looks like for myeloma, researchers aim to gain a greater understanding of why some patients relapse after long remissions, how patients’ immune systems can be strengthened, and how more patients can achieve cures.