Multi-Omics: Plasma Score Flags Cardiovascular Risk in PAD
Statins have blunted what a standard lipid panel can tell a vascular surgeon. With low-density lipoprotein cholesterol (LDL-C) under control, patients with peripheral artery disease (PAD) look alike on paper, yet some still go on to a heart attack, stroke, or cardiovascular death after revascularization. Writing in BMC Cardiovascular Disorders, Cao and colleagues applied plasma multi-omics, pairing untargeted metabolomics with data-independent acquisition (DIA) proteomics, to 165 such patients and built a five-feature score tied to major adverse cardiovascular events (MACE).
Key Takeaways
In statin-treated patients with peripheral artery disease, LDL-C and a hyperlipidemia label miss part of the lipid-related risk left after revascularization.
Plasma multi-omics pairing mass spectrometry proteomics with untargeted metabolomics can sort patients with peripheral artery disease into lipid subtypes with different event rates.
A one-protein, four-metabolite plasma score linked to cardiovascular events in peripheral artery disease comes from one cohort with 23 events and needs external validation.
How the Plasma Multi-Omics Design Was Built
Fasting EDTA plasma went down two LC-MS paths. Untargeted profiling on a Q Exactive HF-X yielded about 3,000 metabolites, of which 645 lipid and lipid-like species were kept, while nanoparticle-enriched (PTM-Max) samples ran in DIA mode on an Orbitrap Astral. The statistics were phenotype-anchored: weighted gene co-expression network analysis (WGCNA) isolated a lipid module that tracked hyperlipidemia but not statin use, and clustering then defined patient subtypes before any cross-omics analysis began.
Key Findings
A statin-independent lipid module: The 70-metabolite MEbrown module tracked hyperlipidemia (partial r = -0.173, P = 0.031) but not statin use, with enrichment in sphingolipid, bile acid, and steroid hormone pathways.
Two subtypes, different outcomes: Subtype C1 (n = 64) had more smokers than C2 (n = 101) and worse MACE-free survival (log-rank P = 0.001).
Proteins moved with the lipids: After statin adjustment, 28 plasma proteins differed between subtypes, enriched for coagulation and redox functions.
Five features carried the prognostic signal: QDPR plus four putatively annotated metabolites gave a cross-validated 15-month AUC of 0.871 for MACE and an adjusted hazard ratio of 1.25 (P = 0.032).
Figure 1. The MEbrown lipid module associated with hyperlipidemia in PAD. Panel A plots module membership against metabolite significance for hyperlipidemia (r = 0.53). Panel B shows enriched KEGG pathways, led by sphingolipid metabolism. Panel C is a donut chart of lipid classes (inner ring) and subclass counts (outer ring), dominated by fatty acyls and prenol lipids. Adapted from Cao et al. (2026), BMC Cardiovascular Disorders.
Why LDL-C Alone Misses Residual Risk
Residual risk on statins is old news. In the FOURIER trial, Sabatine and colleagues (2017) pushed median LDL-C to 30 mg/dL with evolocumab and still recorded primary endpoint events in nearly one in ten treated patients. Overall lipid profiles in this PAD cohort barely separated patients with and without hyperlipidemia. The brown module did not behave that way; its tie to hyperlipidemia was strongest when LDL-C already sat below 1.8 mmol/L (P = 0.008) and vanished above that cutoff (interaction P = 0.023).
QDPR and Acylcarnitines Point to Vascular Stress
QDPR is the most interpretable of the five features. It recycles tetrahydrobiopterin, which keeps endothelial nitric oxide synthase coupled, so circulating QDPR could plausibly report oxidative strain in the vessel wall. Succinylcarnitine and glutarylcarnitine hint at disturbed mitochondrial fuel use. None of the four metabolites was matched to an authentic standard (roughly MSI level 2), so the biology is a working hypothesis until targeted assays settle identity.
One Plasma Aliquot, Two Omics Layers
The two layers in this paper came off different instruments and separate preparations, so every protein-metabolite edge in the correlation network rests on two rounds of batch correction across plates. In our own work at Dalton, we pull both layers from one plasma aliquot with the Omni-MS workflow in a single injection, which drops that alignment step and spares biobank volume. That matters most at replication. Most multi-omics biomarker discovery efforts don't stall on the first cohort, they stall on the second, and single-sample plasma multi-omics profiling removes one common reason why.
Frequently Asked Questions
What is multi-omics used for in cardiovascular research?
Multi-omics measures two or more molecular layers, such as proteins and metabolites, in the same patients. Cardiovascular teams use it to define patient subtypes, build biomarker panels, and catch pathway changes that routine lipid tests miss.
Can plasma multi-omics predict cardiovascular events in peripheral artery disease?
Early data suggest it can add information. In 165 revascularized PAD patients, a one-protein, four-metabolite score reached a cross-validated 15-month AUC of 0.871 for MACE. With only 23 events and no external cohort, it is not ready for clinical decisions.
How much plasma does a proteomics and metabolomics biomarker study need?
Untargeted LC-MS metabolomics typically uses 50 to 100 microliters of plasma, and neat-plasma proteomics needs only a few microliters, though enrichment steps can require more. Consistent fasting collection, fast processing, and storage at -80 degrees Celsius matter as much as volume.
Conclusion
It's now believable that statin-treated PAD patients carry a lipid-linked molecular risk state that routine lipid testing doesn't register. Whether QDPR and these four putative metabolites are the right readout is unsettled, given a cross-validated AUC interval stretching from 0.573 to 1.000. The practical next step is a targeted assay with authentic standards, tested in an independent multicenter cohort.
Related Reading
See how we run these analyses in one lab: Dalton's multi-omics CRO services.
Citation
Cao, S., Zhang, B., Li, P., Luo, J., Tan, S., Lu, C., Yu, F., Yang, C., Chen, Z., Li, Y., & Diao, Y. (2026). Residual lipid-related proteomic and metabolomic signatures associated with major adverse cardiovascular events after revascularization in peripheral artery disease. BMC Cardiovascular Disorders, 26(1), 833. https://doi.org/10.1186/s12872-026-06621-y
Note
This blog post summarizes findings from the above-cited research. Figures are adapted from the original publication. For full details, please refer to the source article.
By Seungjun Yeo, CEO at Dalton Bioanalytics. Specializing in multi-omics mass spectrometry for drug discovery and biomarker research.