Plasma Proteomic Profiling Identifies Prognostic Biomarkers in Mantle Cell Lymphoma

Introduction
Mantle cell lymphoma (MCL), a rare and incurable B-cell malignancy, is characterized by extensive biological and clinical heterogeneity with disease courses ranging from indolent to highly aggressive. Patients who progress within 12 months (POD12⁺) of first-line intensive chemoimmunotherapy and/or Bruton tyrosine kinase inhibitors (BTKi) represents a subgroup with particularly aggressive MCL. Plasma proteins have emerged as promising biomarkers with potential utility in the early detection of relapse in various cancers. We propose that this minimally invasive approach can be used to enhance risk stratification in MCL, thereby reducing the risks of over- or under-treatment of various MCL subgroups, while offering valuable biological insights.

Methods
We conducted high-throughput proteomic profiling in plasma samples from 90 newly diagnosed Swedish MCL patients by simultaneous evaluation of 1460 unique proteins using the Olink® Explore 1536 platform. From this cohort, 75 patients met the inclusion criteria based on the quality control of the Olink® platform and documented clinical parameters. The primary outcome was highly aggressive disease, defined as POD12⁺ following first-line intensive chemoimmunotherapy and/or BTKi. Differentially expressed proteins (DEPs) between indolent and aggressive cases were first identified using the Olink® Statistical Analysis App. Enrichment analysis of DEPs was then performed using Enrichr. Finally, statistical models, such as area under the curve (AUC) and Cox proportional hazard, were used to evaluate the levels of plasma proteins at diagnosis and their association with POD12.

Results
Differential expression analysis identified 252 DEPs (p<0.05), with LRRN1 and IL-15 being among the most significant. These two plasma proteins showed superior predictive power for POD12 compared to the MCL international predictive index (MIPI). The individual hazard ratios (HR) were 18.1 (3.6-328) for LRRN1 and 17.4 (3.5-316) for IL-15, compared to an HR of 8.0 (1.6-147) for high risk versus low risk MIPI. Importantly, both proteins remained significant when sex and MIPI or biological MIPI (MIPIb) were included as covariates (p < 0.05), suggesting their potential as independent prognostic factors. The combination of LRRN1 and IL-15 generated an AUC of 0.92 (0.84-1.0), compared to 0.74 (0.59-0.89) for MIPI, and classified 87% of patients correctly as either POD12^- or POD12⁺.

Additionally, enrichment analysis of all DEPs revealed a significant enrichment of proteins involved in “receptor tyrosine kinase activity” (n = 10) in POD12⁺ patients. Receptor tyrosine kinases, known to play key roles in biological processes such as cell proliferation, migration, and survival, are well-established therapeutic targets in various cancers. Among the 10 enriched DEPs, fibroblast growth factor receptor 2 (FGFR2), a receptor tyrosine kinase not previously described in the context of MCL, added the greatest benefit to the prognostic model. When combined, the three proteins LRRN1, IL-15, and FGFR2 correctly classified 91% of patients as POD12^- or POD12⁺.

Conclusions
Using plasma proteomic profiling, we identified potential biomarkers of clinical and biological relevance in MCL. Plasma levels of LRRN1, IL-15, and FGFR2 at diagnosis outperformed MIPI in predicting disease aggressiveness, highlighting the potential of this minimally invasive approach to improve risk stratification and guide therapeutic decisions in MCL.