Early CRAB-like Biomarker Signatures Reveal a Preclinical Susceptibility Continuum for Multiple Myeloma

2025-12-16

AI Summary

Multiple myeloma (MM) lacks clinically applicable pre-diagnostic tools for long-term risk prediction. Method: Leveraging 10-year longitudinal biochemical data from 379,000 UK Biobank participants, we systematically characterized the dynamic evolution of CRAB-like biomarkers—hemoglobin, erythrocyte parameters, total protein, and albumin/globulin ratio—prior to MM diagnosis. We employed multivariable time-to-event modeling and time-dependent ROC analysis to develop a clinically feasible early-risk signature. Contribution/Results: We identified nonlinear, dose–response alterations in anemia and protein metabolism dysregulation beginning up to 10 years before MM diagnosis—earlier than previously reported. The resulting biomarker panel significantly improved 5- and 10-year MM risk prediction, elevating the C-index from 0.66 to 0.76. Crucially, erythrocyte morphological and protein metabolic abnormalities enabled high-risk stratification a decade prior to clinical diagnosis, substantially extending the conventional early-detection window.

Technology Category

• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Machine Learning: Graph-based Machine Learning

Application Category

• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Web Mining and Content Analysis: Models for Web evolution

Abstract

Multiple myeloma (MM) evolves over decades, yet robust tools for identifying individuals at risk long before clinical onset remain limited. Using data from 378,930 UK Biobank participants, we systematically characterized the longitudinal dynamics and predictive value of routinely measured "CRAB-like" biomarkers, including hematologic indices, protein metabolism markers, renal function, and serum calcium. Across multivariable models, biomarkers reflecting anemia and protein imbalance (including hemoglobin, red blood cell indices, total protein, albumin, and the albumin/globulin ratio) showed strong and consistent associations with future MM, independent of demographic, lifestyle, clinical, and genetic risk factors. These markers displayed pronounced non-linear dose-response relationships and contributed substantially to 5- and 10-year MM risk discrimination, with the C-index improving from 0.66 to 0.76. Longitudinal analyses revealed progressive shifts in red cell morphology and protein metabolism profiles up to a decade before diagnosis, supporting the existence of a preclinical susceptibility continuum detectable in the general population. Our findings suggest that subtle yet quantifiable deviations in common laboratory tests reflect early microenvironmental changes that precede malignant plasma cell expansion, offering opportunities for risk stratification and targeted surveillance.

Problem

• Identifies early biomarkers for multiple myeloma risk prediction
• Analyzes longitudinal dynamics of CRAB-like biomarkers before diagnosis
• Improves risk stratification using common laboratory test deviations

Innovation

• Longitudinal analysis of CRAB-like biomarkers for early detection
• Non-linear dose-response relationships improve risk discrimination
• Common lab tests reveal preclinical susceptibility continuum