CoMMpass Dataset Illuminates Immune Cells’ Role in Multiple Myeloma

As the race for better treatment options for multiple myeloma goes on, focus has turned to the dysfunction present in patients’ immune cells as a view into the disease biology that goes beyond tumor characteristics and high-risk cytogenetic features. This effort has been represented by the publication of a single-cell atlas looking at over 1 million cells in the bone marrow microenvironment of hundreds of patients with multiple myeloma before and during treatment.¹

Finding correlations between particular immune signatures and outcomes in these patients can reveal better ways to classify each patient’s individual disease. Distinct clusters in immune cell populations showed the potential to predict progression-free survival and overall survival in this dataset. Additionally, treating multiple myeloma employs a combination of agents with various mechanisms of action. Understanding how the immune microenvironment might be impacted by each type of treatment could allow physicians to select the right regimen for each patient.

Crucially, the dataset from the Immune Atlas is being shared on the Multiple Myeloma Research Foundation (MMRF) virtual platform as part of its CoMMpass (Clinical Outcomes in Multiple Myeloma to Personal Assessment of Genetic Profiles) registry study (NCT01454297) dataset which has provided a basis for deeper analyses into the disease for over a decade.² The data collected by the MMRF in partnership with 5 leading academic centers will give research groups the ability to generate hypotheses and identify new immune-based approaches on a cellular level.

In an interview with Targeted Oncology, George Mulligan, PhD, chief scientific officer of the Multiple Myeloma Research Foundation, discussed the key takeaways from the Cell Atlas project and what will come next now that this dataset has been made available.

Targeted Oncology: What were the goals of creating the Immune Atlas based on the CoMMpass dataset?

George Mulligan, PhD: Our goals are to try and better understand the immune system across multiple myeloma. It has generally been complex to get a biological insight in the disease because it’s so heterogeneous, so we set out to analyze a large number of samples. We’ve been biobanking specimens and running these types of studies for more than 10 years, and we had in our archives a number of patient samples from our CoMMpass study, which was started in 2011 and helped define and clarify the biological basis of multiple myeloma and the heterogeneity and the many different tumor subtypes that exist.

We felt we could use that resource to better understand the immune system, because we had stored samples that were remaining from those patients that captured the bone marrow milieu, not just the tumor cells. We therefore built up a partnership of sites…and worked out some key technical details with them, as well as some strategic considerations….

The first batch of samples was 337 individual patients, and more than 100 of those have sequential samples, not just baseline samples from diagnosis, but samples from when they had a response to their treatment or had a relapse from a response and the progressive disease sample was captured.

We have clinical follow-up on these patients, not just how they were treated at diagnosis, [but] how they were treated at subsequent relapse. We followed all the patients for at least 8 years and therefore understand their series of outcomes and the heterogeneity of how they were treated and how the outcomes resulted from that. Now we have this single-cell dataset capturing that whole spectrum of patients from the time of diagnosis.

Why is the immune microenvironment so important in myeloma research?

Studying the immune system in this setting has a couple of purposes. As cancer develops in any setting…we increasingly understand that for that to happen, there has to be a fundamental breakdown in the immune system that allows this clone of cells to emerge and take over a portion of the body. In most healthy people, our immune system is able to suppress this formation, even if it gets off to a start. We felt that we could understand how this contributes to the disease emergence.

We wanted to try to also understand whether there’s a link with outcome. This is particularly pressing in myeloma now because there are a number of approved immune-based therapies…improving outcomes significantly for patients. That’s by no means a cure; at this stage, the disease remains very difficult to treat, and most patients relapse.

One would think that...the immune system is critical when they’re treated with immune agents…[but] it’s not the cornerstone of this study. Patients in this study were generally treated with non-immune therapies which targeted the plasma cell and are very effective. …The immune system is important even with these treatments that don’t use immune agents. It seems like that function contributes to outcomes with all therapies, not just when we’re activating the immune system. That will shed an increasing light on the immune system in emergence of the disease in response to standard therapies. This will be critical in response to newly developed agents which either activate T cells or bring T cells in contact with the tumor or other immune-mediated agents that are emerging. We shed light on these questions and more….

What were the key takeaways from the publication in Nature Cancer?

The Atlas paper first provides this foundational dataset which characterizes 337 patients at diagnosis. It deeply looks at the single cell population, so all the immune components of the bone marrow microenvironment, more than 1.3 million cells. We can see even rare cell types, and we can analyze their outcomes as relates to biological categories that these patients are in or that their tumors are in, and how that affects the immune compartment, and think about how the immune compartment and its variety that we observed might relate to clinical outcome.

We did see that there was information that was prognostic and related to how the patients do long-term with therapy. We also were able to show that this prognostic information is distinct from what we know about the tumor, which clearly contributes to prognosis. One can combine these 2 and have more information than one would separately.

There are a variety of observations…about specific components of the immune system, interferon signaling, what appear to be dysfunctional T cells, which again relate to the clinical outcome of these patients when treated with standard of care….

Finally, we will make this dataset immediately available for public research as a key component of the longstanding CoMMpass data set. People will be able to go to our portal, called Virtual Lab…and they will be able to access the…tumor characteristics and demographics of the patients, which people have analyzed for a long time, and now they will…see these more than 300 patients worth of immune data….

We have a single-cell Shiny app which will let you separate and look at those individual cell types and expression of different genes. The data can be downloaded in various forms, and [they] can also be analyzed on the platform for those who might not have analytical capabilities locally. We will build upon that alongside other researchers globally to uncover more insights about how the immune system relates to key parameters like outcomes in different racial subgroups [and] changes in the biology of the immune system over time as samples get followed.

Those are studies that are in progress. One of them around longitudinal sampling was also just published as well,³ so we expect this to emerge as another cornerstone to how we understand myeloma, and we hope that there are a whole series of findings that emerge from this from multiple creative groups across the globe.

How do you see this dataset being used in the future?

We have started to share the data with groups that are experts in myeloma biology and groups that are experts in immune biology. We have new requests starting to come in as well around ideas we hadn’t yet considered. Those will be interesting analyses to see how they emerge. There are additional ways we think CoMMpass will serve utility in this way. From that bone marrow milieu, where we’ve now done this single-cell RNA sequencing approach, we’re growing a dataset around a set of cellular protein markers that will also emerge and become part of the CoMMpass dataset which people will be able to download and analyze.

We also have additional samples from this same study which is so rich and already has outcome data. If people have proposals to generate new data, as long as we can bring the data back into the dataset, we’d be willing to consider providing those samples as well. That describes some future emerging research as well. [We are] sharing this data set across the academic world so that partners can analyze it in any way they want. That’s basic biology, that’s clinical subtypes, and that’s also understanding the prognostic impact, because we understand their clinical outcomes.

How can the dataset inform the use of both standard and novel therapies in myeloma?

When we think about how it would inform therapies, there are a couple important angles. One is the therapies that these patients were given as at diagnosis at the time the study started, and that includes standard triplet regimens like proteasome inhibitor, IMID, plus dexamethasone, which have very good outcomes. A significant fraction of this patient population was also treated with transplant, which allows you to analyze the effect plus and minus transplant.

It’s important that this was very much a real-world representation of American [patients with] myeloma. You have patients of varying ages, which you might not see in a clinical trial, including patients up to their late 80s. There is roughly 20% of the population that is African American, which also rarely happens in studies. This was by intent, and we made sure that those patients were included because we need to understand that biology of any group that has a threefold increase in incidence. There’s something for us to learn there as basic scientists, and so that’s an important part of the study.

Each of those aspects we expect to be teased apart. We started to look at some of this in this study, either pooling the treatments or focusing on that very common triplet…. [or] patients who specifically received a different triplet, or the few that got a quadruplet therapy or a doublet. The fact that there are patients [for whom] a triplet plus a [stem cell] transplant gives remarkable results, more than 7 years of progression-free survival, is an outstanding question that we and others will dig into, because those patients can benefit from very active therapy, and may not want to go into experimental therapy when they’re going to do so well with therapies that are so well characterized.

You raise the point of newer therapies, which include active antibodies that target CD38 and then CAR [chimeric antigen receptor] T [cells] and bispecifics, which aim to supercharge the immune system to eliminate or at least get rid of most of the tumor and keep it in check. This is where we think the dataset will be very important, because we can now understand the landscape of those immune cells at diagnosis. We have to understand how to best use CAR T and bispecifics in relapsed disease.

I think ultimately, they will move into the earliest lines of therapy. This gives us a chance to think, maybe there are some patients we should investigate because they’re going to do spectacularly on these drugs, and maybe we should keep an eye out for patients that may not do so well regardless of any immune therapy they get, because they may need to get a combination therapy, or maybe they’re just not suited, and they should find a different track with other active myeloma agents. Not only are the clinical centers that we work with engaged in these kind of strategies, but increasingly, the biopharma industry understands that these are tracks of strategic drug development that can they can leverage to best effect to get their agents used either earlier or in the right segment population first, before they branch into the broadest patient population that they can make a case for.

It’s a very exciting time to do this in myeloma, yet I remind people that still these agents do not appear to be curative, and so it’s also a bit of a desperate time in myeloma, where we want new agents, we want new therapies, and we want new strategies, and we want them to be scientifically driven, so that we can test ideas and move on from ideas that don’t work out and stick with the strongest ideas. There’s been great success with many therapies approved in the absence of a cure. These therapies can’t just be combined sequentially and randomly. Mathematically, that’s just a poor drug development strategy. These are the kind of datasets that we can provide and others can lean into. There are other interesting data sets globally that provide a rationale for strategic drug development across myeloma so that we can get to a cure broadly.

REFERENCES

1. Pilcher WC, Yao L, Gonzalez-Kozlova E, et al. A single-cell atlas characterizes dysregulation of the bone marrow immune microenvironment associated with outcomes in multiple myeloma. Nat Cancer. 2026;7(1). doi:10.1038/s43018-025-01072-4

2. Groundbreaking New MMRF-Led Studies Shed Light on Immune System’s Role in Multiple Myeloma. News release. Multiple Myeloma Research Foundation. January 9 2026. Accessed January 19, 2026. https://tinyurl.com/3hmen8un