Multi-Omics: Plasma Clues to Kidney Cancer Relapse on PD-1
Between 30 and 50 percent of patients with high-risk clear cell renal cell carcinoma (ccRCC) relapse within five years of nephrectomy, and adjuvant PD-1 blockade helps only some of them. No blood test yet says which. Wang and colleagues went after that gap with plasma multi-omics built into a phase II trial of the PD-1 inhibitor penpulimab, pairing MS-based proteomics with untargeted metabolomics to find circulating signals that separate early relapsers from durable responders.
Key Takeaways
Adjuvant penpulimab was associated with longer disease-free survival than surveillance in very-high-risk ccRCC, though the trial was not randomized.
Plasma multi-omics revealed weaker B-cell immunity and fewer gut microbe-derived metabolites in kidney cancer patients who relapsed early on PD-1 therapy.
The candidate blood markers come from 23 treated patients and stay hypothesis-generating until an independent cohort confirms them.
Plasma Multi-Omics Inside a Phase II Adjuvant Trial
The trial enrolled 247 patients at Chinese centers from 2022 to 2025. Because each patient chose an arm together with their physician, the authors used 1:1 propensity score matching (PSM), a statistical pairing on baseline features, to build two balanced groups of 87. Treated patients gave plasma before their first and second doses. Proteins were quantified by data-independent acquisition (DIA) mass spectrometry after bead enrichment, and metabolites by LC-MS on reversed-phase and HILIC columns.
Key Findings
Fewer relapses on penpulimab: 7 of 87 treated patients relapsed or died versus 19 of 87 under surveillance (HR 0.37; 95% CI 0.16 to 0.89), with 2-year disease-free survival (DFS) of 88.7% versus 75.4%.
Relapsers started with an immune-poor proteome: 53 baseline proteins separated early relapsers from non-progressors, with complement and coagulation pathways up and B-cell receptor signaling down.
Microbial metabolites fell as lipid signals rose: Among 101 differential baseline metabolites, relapsers showed depleted gut microbiota-derived compounds and higher lipid-remodeling species such as DG (18:2/22:6).
A 14-feature panel came out of cross-omics analysis: LASSO-penalized Cox regression kept four proteins (MS4A1, NDUFAF7, RPS13, PSAPL1) and ten metabolites linked to DFS.
Figure 1. Trial design and patient flow. Panel A lists key inclusion criteria and the two postoperative strategies, penpulimab every three weeks or routine surveillance. Panel B traces 247 enrolled patients through propensity score matching into two arms of 87, plus six discontinuations for grade 3 or 4 adverse events. Adapted from Wang et al. (2026), Journal for ImmunoTherapy of Cancer.
Reading a Non-Randomized Efficacy Signal
Confounding is the first worry when patients pick their own arm. Matching balanced every baseline variable (standardized mean differences below 0.2), and penpulimab stayed independently tied to longer DFS after multivariable adjustment (HR 0.32; p=0.011). Overall survival is immature at a median 20 months of follow-up, so the authors' comparison with KEYNOTE-564 deserves a loose grip. Grade 3 to 4 treatment-related events reached 6.9%, with no treatment-related deaths.
Immune Loss and Microbial Metabolites in Relapsers
Only 23 treated patients met the strict progression definitions (8 relapsed within 12 months; 15 stayed disease-free beyond it). Their plasma told the same story before and after the first dose: weaker antigen presentation and B-cell signals, plus lower diet-microbe co-metabolites such as 3-hydroxyhippuric acid. That fits Routy and colleagues (2018), who tied gut microbiome composition to PD-1 response in kidney and lung cancer, and it mirrors our earlier write-up on plasma markers in lung cancer immunotherapy.
Keeping Both Layers on the Same Patients
Look closely at the sample counts: three post-treatment specimens were dropped from the metabolomics set for hemolysis but kept in the proteomics set, so the two layers weren't describing quite the same patients. Drift like that creeps in whenever each layer gets its own aliquot and its own QC call. In our integrated multi-omics workflow at Dalton, proteins and metabolites come from one plasma aliquot in a single Omni-MS injection, so a sample passes or fails for both layers at once. With only 23 patients feeding multi-omics biomarker discovery, a complete matrix beats extra depth in either layer.
Frequently Asked Questions
What is multi-omics used for in cancer immunotherapy research?
Multi-omics combines two or more molecular layers, such as proteins and metabolites, measured in the same patients. In immunotherapy studies it helps separate responders from non-responders and points to the biology behind the gap.
Can plasma multi-omics predict relapse after adjuvant PD-1 therapy in kidney cancer?
Not yet. In this trial, four proteins and ten metabolites tracked disease-free survival in 23 penpulimab-treated patients, but the analysis was exploratory and lacked an untreated comparison group.
How much plasma does a proteomics and metabolomics study need?
Very little. This trial drew 5 mL of EDTA blood per time point, and most LC-MS proteomics and metabolomics assays use only tens of microliters of plasma. Fast processing and a single freeze-thaw cycle matter more than volume.
Conclusion
Penpulimab after nephrectomy looks like a credible alternative PD-1 option for very-high-risk ccRCC, at least on DFS. The plasma panel isn't a predictive test yet; it rests on 23 patients from one arm. Sponsors planning adjuvant trials should run the omics in both arms, because that's the only way to tell a marker of drug benefit from a general prognostic one.
Related Reading
See how we run these analyses in one lab: Dalton's multi-omics CRO services.
Citation
Wang, Y., Wang, D., Zhao, H., Zhou, Y., Liang, Q., Huang, Q., Peng, C., Xu, J., Guo, G., Yang, G., Chen, L., Xun, Y., Du, J., Ma, X., Zhang, X., & Gu, L. (2026). Adjuvant penpulimab in very-high risk clear cell renal cell carcinoma: A prospective, non-randomized, controlled phase II trial with integrated plasma multi-omics analyses. Journal for ImmunoTherapy of Cancer, 14(7), e015686. https://doi.org/10.1136/jitc-2026-015686
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.