Multi-Omics: Plasma Shifts Track T-DXd in Breast Cancer
No blood test yet tells oncologists which patients with advanced breast cancer will progress early on trastuzumab deruxtecan (T-DXd), an antibody-drug conjugate (ADC) that delivers a topoisomerase I payload to HER2-expressing cells. A September 2026 study in Frontiers in Oncology went after that gap with paired plasma multi-omics. The authors profiled the proteome and metabolome of 16 patients before treatment and after two cycles, then screened baseline features against progression-free survival (PFS). Sixteen patients is a small cohort, and the paper says so.
Key Takeaways
Paired plasma multi-omics, sampling each patient before and during therapy, can expose systemic drug effects that a single baseline draw would miss.
Plasma acute-phase and coagulation proteins fell after two cycles of T-DXd in advanced breast cancer, consistent with lower tumor burden and quieter inflammation.
Baseline plasma SLC16A1 tracked with shorter progression-free survival on T-DXd, but only 16 patients were profiled and no candidate was externally validated.
Paired Plasma Multi-Omics Before and After Two Cycles
Of 114 patients screened, 16 had matched, QC-passing plasma from baseline and after cycle two; 13 were HER2-low, and ten had received more than three prior lines. Proteins were quantified by 4D data-independent acquisition (DIA) LC-MS/MS, which fragments all precursors in sequential mass windows instead of only the most intense ions, returning 4,444 proteins. Untargeted metabolomics on a Q Exactive HF-X added 3,596 features. Thresholds were loose: with nominal P below 0.05 and no multiple-testing correction, some protein hits will be noise.
Key Findings
A broad plasma proteome shift: 508 proteins changed after two cycles (136 up, 372 down). C-reactive protein and serum amyloid A fell sharply, while complement C4 and apolipoprotein A-I rose.
Much of the decline involved liver-made proteins: over 60% of changed proteins were secreted, mostly complement, coagulation, and apolipoprotein species. Fibrinogen chains and C3 fell together, and albumin was the top network hub.
Metabolite changes were narrower: 70 of 3,596 features qualified. Isoleucine rose about 1.74-fold and tryptophan fell, while BCAT1 and BCKDHB increased in step with branched-chain amino acid shifts.
Six proteins and four metabolites flagged for PFS: baseline SLC16A1 carried a hazard ratio of 3.16 per standard deviation (95% CI 1.15 to 8.72) and was higher in HER2-low patients.
Figure 1. Study workflow. Twenty-nine T-DXd-treated patients with advanced breast cancer formed the clinical cohort, and 16 supplied paired pre- and post-treatment plasma. Proteomics (4D-DIA LC-MS/MS) and untargeted LC-MS metabolomics ran in parallel, followed by differential screening, enrichment analysis, and cross-omics integration. The final arms covered PFS-driven selection of baseline biomarkers and correlation testing among prognostic proteins and metabolites. Adapted from Sang et al. (2026), Frontiers in Oncology.
What the Protein Changes Probably Reflect
The proteome response reads mostly as a host signal. Falling acute-phase proteins fit reduced tumor burden, but a responder on chemotherapy might look the same, so specificity to T-DXd is unproven. Geyer and colleagues (2019) showed that coagulation proteins can shift from partial clotting during plasma handling alone, which makes the fibrinogen drop worth a quality check. A troponin T rise in a few patients merits follow-up too, given interest in cardiotoxicity biomarkers during breast cancer therapy.
Cross-Omics Analysis: Useful Leads, Soft Annotations
Correlations between the two layers formed structured modules rather than one tidy axis, and purine metabolism dropped at both the enzyme (ADA) and metabolite (hypoxanthine) level. Metabolite identity is the weak point. A few PFS-linked features, dibenzylamine among them, look like spectral-library hits that need authentic standards before anyone builds a targeted assay around them.
Why One-Sample Integration Matters Here
A paired design works only if the treatment effect outruns technical drift between timepoints, and that margin shrinks when proteins and metabolites are extracted separately and run on two instruments. Under those conditions, a BCAT1 and isoleucine correlation could reflect biology or simply run order. In our lab at Dalton, the Omni-MS workflow reads both layers from one plasma injection, which is how we'd run paired on-treatment multi-omics studies when volume per timepoint is tight. That logic carries into our multi-omics biomarker discovery work: fewer handling steps, fewer reasons to doubt a fold change.
Frequently Asked Questions
What is plasma multi-omics used for in drug development?
It measures proteins, metabolites, or other molecular layers from the same blood samples. Drug developers use it to track pharmacodynamic response and to find candidate biomarkers of benefit or toxicity without a biopsy.
Can a blood test predict who will respond to trastuzumab deruxtecan?
Not yet. This study found baseline protein and metabolite candidates linked to progression-free survival, but with 16 patients and eight progression events, they need validation in larger independent cohorts.
How much plasma does a multi-omics study need per sample?
Less than many teams expect. Here, metabolomics extraction used 100 microliters, and MS-based proteomics of neat plasma often starts from a few microliters. Banking two or more aliquots per timepoint matters more than total volume.
Conclusion
It's now reasonable to expect that two cycles of T-DXd leave a measurable plasma footprint, carried mostly by liver-derived proteins. Whether any baseline marker predicts benefit isn't established. For teams planning ADC trials, the practical step is banking paired pre- and on-treatment plasma now so candidates like these can be tested prospectively.
Related Reading
See how we run these analyses in one lab: Dalton's multi-omics CRO services.
Citation
Sang, D., Yang, W., Liu, J., Li, Y., Zhao, S., Wang, X., Zhang, Y., Wang, B., Li, M., & Yuan, P. (2026). Plasma multi-omics reveals molecular remodeling and prognostic biomarkers in advanced breast cancer treated with trastuzumab deruxtecan. Frontiers in Oncology, 16, 1973097. https://doi.org/10.3389/fonc.2026.1973097
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.