Summary: In a manufacturer-sponsored exploratory analysis nested within the Phase 2b MOSAIC randomised trial in diabetic kidney disease, selonsertib (an ASK1/MAP3K5 inhibitor) produced a distinct plasma proteome signature on the SomaScan platform that implicated fibrosis, inflammation and oxidative-stress pathways. The signature was most pronounced in a subset of patients with poor baseline kidney function. This is a biomarker and mechanism-of-action study; it reports no new clinical efficacy estimates for kidney outcomes.
PICO Summary
| Element | Detail |
|---|---|
| Population | Adults with diabetic kidney disease enrolled in the Phase 2b MOSAIC randomised controlled trial. This report is a secondary plasma-proteome analysis of trial participants; the article does not restate the trial enrolment total, follow-up or country. |
| Intervention | Selonsertib, an oral inhibitor of apoptosis signal-regulating kinase 1 (ASK1 / MAP3K5), an upstream mediator of oxidative-stress signalling. Specific arm size and dose not reported in this proteomic sub-study. |
| Comparison | Placebo arm of the same MOSAIC trial. Arm size not reported in this sub-study. |
| Outcome | A plasma proteome signature of selonsertib activity was identified, implicating signalling pathways governing fibrosis, inflammation and oxidative-stress response. The proteomic effect was most pronounced in a subset of patients with poor baseline kidney function who responded well to treatment. No new effect estimates, 95% confidence intervals, p-values, or ARR/NNT for clinical kidney endpoints are reported in this exploratory analysis; the MOSAIC efficacy on kidney function is cited only as prior background. |
Expert Commentary
This paper should be read as a mechanistic and biomarker study, not as evidence that selonsertib works. A plasma proteome signature was derived on the SomaScan platform from MOSAIC trial samples, and it was found to map onto fibrosis, inflammation and oxidative-stress pathways, with the largest changes seen in patients whose baseline kidney function was poor. That is biologically coherent and consistent with the drug’s proposed ASK1 target, but the analysis is exploratory and hypothesis-generating. No new clinical effect estimates, confidence intervals, or p-values for kidney outcomes are presented here, so the underlying efficacy claim still rests on the earlier MOSAIC and prior selonsertib trials rather than on this report. The principal limitation is the subgroup framing: a signature that is strongest in responders with poor baseline function is exactly the pattern that can arise from post-hoc selection, and it needs prospective confirmation before it is treated as a predictive marker. Industry sponsorship is material, since every listed author is employed by the manufacturer of selonsertib, which warrants cautious interpretation of a favourable mechanistic narrative. Can I use this with my patients? Not yet. There is nothing here to change practice for a person with diabetic kidney disease today. The constructive read is that future ASK1-inhibitor programmes should pre-specify and validate this proteomic signature in an enriched, poor-baseline-function population before efficacy is claimed.
References
Petrovic V, Whiteman A, Peach M, Kim S, Malkov VA, Budas G, et al. Plasma proteome signatures of ASK1 inhibition by selonsertib associate with efficacy in the MOSAIC randomized trial for diabetic kidney disease. BMC Nephrol. 2025;26(1):244. doi:10.1186/s12882-025-04166-4
