PANGEA-SMM: A Novel Prediction Model for Smoldering Multiple Myeloma Progression Using Dynamic Biomarkers and Large International Cohorts

A new study published today in Nature Medicine details a more precise method for predicting when smoldering multiple myeloma (SMM) – a precursor condition to the blood cancer multiple myeloma – will progress to active disease. The research, led by scientists at Dana-Farber/Harvard Cancer Center and involving an international collaboration of researchers, focuses on tracking changes in key biomarkers over time, rather than relying on a single snapshot assessment. This refined approach, dubbed PANGEA-SMM, could help clinicians better identify patients who would benefit from early intervention and avoid unnecessary treatment for those at lower risk.

Understanding Smoldering Multiple Myeloma and the Necessitate for Better Risk Assessment

Smoldering multiple myeloma is a condition where abnormal plasma cells are present in the bone marrow, but aren’t yet causing the symptoms that define active multiple myeloma. It’s a heterogeneous disease, meaning its progression varies significantly from person to person. Currently, risk stratification – determining how likely someone is to develop active myeloma – relies on models like the 20/2/20 criteria, which assesses levels of M-protein, involvement of the bone marrow, and light chain ratios. However, these models are static, meaning they don’t account for how biomarkers change over time. This can lead to overtreatment in some patients and delayed treatment in others.

The PANGEA project, a long-term cohort study, was initiated to address this gap. Researchers assembled data from a large cohort of 2,344 patients with SMM from seven international centers, collecting longitudinal clinical and biological data between March 2021 and October 2024. The core hypothesis was that incorporating dynamic biomarker changes would improve prediction accuracy.

Key Biomarkers and Dynamic Risk Factors

The study identified four evolving biomarkers that were significantly associated with a shorter time to progression: an increase of 0.2 g/dL or more in M-protein levels, a 20% or greater increase in the involved/uninvolved serum free light chain ratio, a greater than 25% increase in creatinine levels, and a decrease of 1.5 g/dL or more in hemoglobin. These aren’t just the current levels of these markers that matter, but the *trend* – whether they are increasing or decreasing, and at what rate.

Researchers tested various definitions for these “dynamic” changes, ultimately settling on criteria that best improved the model’s predictive power. For example, they considered whether a biomarker increased by a certain percentage compared to previous values, or by a specific absolute amount. The optimal definitions varied for each biomarker, highlighting the complexity of the disease.

PANGEA-SMM: Outperforming Existing Models

The PANGEA-SMM model, incorporating these dynamic biomarkers, demonstrated superior performance compared to established models like the 20/2/20 and IMWG models. It achieved a C-statistic of 0.79 in predicting progression, even when biomarker history was incomplete (C-statistic of 0.78) or a recent bone marrow biopsy wasn’t available (C-statistic of 0.78). The C-statistic is a measure of how well a model can discriminate between patients who will and will not progress to active myeloma; a higher number indicates better performance.

The study involved a robust validation process, utilizing five independent cohorts of patients from six international centers, including institutions in Greece, the UK, Germany, Spain, and Italy. Data collection adhered to strict ethical guidelines, with approval from the Dana-Farber/Harvard Cancer Center institutional review board (no. 21-127) and, in some cases, requiring informed consent from patients. Lakshman et al. (2018) previously established risk stratification criteria for SMM, and the PANGEA-SMM model’s performance was compared to these established methods.

What This Means for Patients and Clinical Practice

The development of PANGEA-SMM represents a significant step forward in personalized risk assessment for SMM. By considering the trajectory of key biomarkers, clinicians can gain a more nuanced understanding of each patient’s individual risk profile. This could lead to more informed decisions about when to initiate treatment, potentially avoiding unnecessary interventions for patients at low risk and ensuring timely treatment for those at higher risk.

The researchers have made PANGEA-SMM an open-access tool, along with validation tools for comparison with existing models. They’ve as well developed a clinical calculator to facilitate its use in everyday practice. This accessibility is crucial for widespread adoption and potential benefit to patients worldwide. The study data were managed using Research Electronic Data Capture (REDCap), a secure, web-based platform designed for research data management. Harris et al. (2009) describe the capabilities of REDCap in detail.

Limitations and Future Directions

While promising, the PANGEA-SMM model isn’t without limitations. The study was retrospective, meaning it analyzed data collected in the past. Prospective studies, where patients are followed forward in time, are needed to confirm these findings. The model doesn’t include all potential biomarkers or genetic factors that might influence progression. Further research is needed to explore the role of cytogenetic markers, such as FISH findings, in refining risk prediction.

The researchers acknowledge that the model’s performance may vary depending on the population studied and the frequency of biomarker measurements. They conducted additional analyses using a simulated dataset with less frequent observations, demonstrating that the model remained robust even with less frequent monitoring. The PANGEA project, as highlighted in The Lancet Haematology, is a continuing effort to improve risk stratification in precursor conditions to multiple myeloma.

Next Steps: The research team is continuing to refine the PANGEA-SMM model and explore its potential applications in clinical trials. They are also working to develop new biomarkers and incorporate additional data sources, such as genetic information, to further improve its accuracy. Ongoing validation studies in diverse populations will be crucial to ensure its generalizability and effectiveness.