The total blood n-glycome landscape of hematological malignancies

Abstract

Abstract Introduction Hematological malignancies (HM) are heterogeneous with complex pathophysiology. Glycans, the third major informational biopolymer, surpass nucleic acids and proteins in structural complexity and information density. As a prevalent post-translational modification, glycosylation serves as a functional readout of genetic variants and drives HM development. However, HM glycomics research is at an early stage compared to other omics fields. Therefore, we conducted the largest multi-center clinical study to date in the field of HM glycomics, mapping the total blood N-glycome (TBNG) landscape across HM subtypes and treatment phases. Methods Patients were enrolled from a multi-center observational clinical study registered with the Chinese Clinical Trial Registry (ChiCTR) (Registration number: ChiCTR2400089864). Participants were recruited from 8 medical centers. TBNG features in HM were investigated using retrospective samples and prospective longitudinal samples. The study was approved by the Medical Ethics Committee of the Affiliated Drum Tower Hospital of Nanjing University Medical School (IRB No. 2024-654-01) and conducted in accordance with the Declaration of Helsinki. Serum N-glycan profiling was performed using capillary electrophoresis. Serum N-glycan detection followed previously described methods (Su R et al. Hepatology 2025). Nine specific serum N-glycan peaks were identified per sample. Results TBNG analysis was performed on 979 HM patients from 8 centers (April 2016 - April 2024), including 348 MM (35.5%), 370 lymphoma (37.8%), and 261 acute leukemia (AL) (26.7%) cases. The retrospective cohort included 923 patients (94.3%) (samples only at diagnosis or a single post-treatment follow-up point) while the prospective longitudinal cohort included 56 patients (5.7%) (samples at diagnosis and complete remission (CR) follow-up). Analysis covered 1,035 samples: 641 at diagnosis (ND), 59 at PR/VGPR, 257 at CR, and 78 at relapse/refractory (R/R). Healthy controls (n=150) were included. Firstly, we identified distinct TBNG profiles differentiated HM types. All HM patients (MM, lymphoma, AL) showed elevated triantennary glycan Peak9 (NA3Fb) vs. controls (p < 0.0001). TBNG also correlated with subtypes and risk stratification. In MM, IgG type showed increases in two monogalactosyl N-glycans (Peak3 and Peak4, NG1A2F; p < 0.0001), IgA type showed increase in biantennary glycan Peak7 (NA2FB, p < 0.0001), and light chain (LC) type showed increase in Peak9 (p < 0.0001). In DLBCL, elevated Peak9 correlated with aggressiveness and tumor burden: higher in aggressive vs. indolent lymphomas (p=0.022), Group B vs. A (p=0.0002), and stage III/IV vs. I/II (p=0.018). These findings support TBNG's potential as a biomarker for HM diagnosis, subtyping, and risk stratification. More importantly, we observed that the above-mentioned characteristic glycan changes vary with different disease remission statuses. Retrospective analysis confirmed significant differences in TBNG between ND and CR states. For IgG MM, Peak3 and Peak4 significantly decreased at CR compared to ND (p < 0.0001). For IgA MM, Peak7 decreased at CR (p < 0.0001). For LC MM, Peak9 decreased at CR (p = 0.0002). DLBCL and AML patients showed decreases in Peak9 at CR (p = 0.0006). TBNG profile at CR tends to be closer to that of healthy individuals, while the R/R group exhibited TBNG characteristics more similar to the ND group. To validate the findings from the retrospective study, we conducted an additional analysis focusing on the longitudinal changes in N-glycan patterns before and after treatment in 56 patients. The prospective cohort results confirmed retrospective TBNG trends between ND and CR states, supporting TBNG's potential as a biomarker for treatment response and MRD monitoring. Conclusion Using the largest HM blood cohort analyzed glycomically to date, this study delineates the comprehensive TBNG landscape in real-world patients. The findings suggest that different disease entities, subtypes, and stages of HM exhibit characteristic TBNG alterations. Furthermore, these TBNG signatures dynamically change in response to treatment and corresponding disease remission status. This TBNG landscape establishes its utility as a novel biomarker for HM diagnosis, subtyping, risk stratification, treatment response, and MRD monitoring.