Proteomic and clinical determinants of long-term outcome in patients with dilated cardiomyopathy

Background

Dilated cardiomyopathy (DCM) is the most common cause of heart transplantation. However, the prediction of prognosis is difficult and disease-specific biomarkers are lacking.

Purpose

To identify clinical determinants and proteins associated with long-term outcomes in DCM.

Methods

Patients ≥18 years with DCM from the Swedish Registry for Dilated Cardiomyopathy (1997–2006) were included. DCM was diagnosed according to contemporary criteria; left ventricular ejection fraction (LVEF) <50% and exclusion of ischemic, valvular, arrhythmic, inflammatory, endocrine, and hypertensive causes. Randomly selected healthy individuals with normal cardiac function were included as a control group (1988-1990). Baseline clinical data and plasma samples were collected and stored in -80°C. Follow-up for a composite endpoint of heart transplantation or death was obtained through national registries until March 2024. Plasma proteins were analyzed using targeted cardiovascular proteomics panels comprising 184 proteins, with protein levels reported as NPX values. Differential expression was assessed using age- and sex adjusted linear models with false discovery rate correction. Cox proportional hazards models were used for the composite outcome and adjusted for age, sex, LVEF and NYHA-class, whereas models assessing protein associations were adjusted for MAGGIC score variables (except creatinine and smoking).

Results

In all, 248 patients with a mean age of 52.4 ± 10.9 years (12.5% women) and 84 controls with a mean age of 49.1 ± 16.9 years (47.5% women) were included. Mean LVEF in DCM patients was 29.1 ± 13.6 (table 1). The endpoint rates at ten and twenty years were 33.5% and 66.7% respectively, with a total of 178 events (71.8%) occurring during follow-up (in mean 16.1 ±years).

Proteomic data were available for analysis in 202 DCM patients and in 64 controls. DCM patients showed marked protein alterations with PARP-1, ANGPT1, and IGFBP-2 most up-regulated and EGFR, COL1A1, and CHIT-1 most down-regulated (p<0.0001) (figure 1). Older age, longer symptom duration, lower LVEF and higher NYHA class were associated with higher event rates although the 25-year follow-up violated proportional hazards assumptions due to rising age-related mortality.

In Cox model adjusted for MAGGIC score a one unit increase in TRAIL-R, GAL-9 and PGF were associated with a more than twofold increase in HR for heart transplantation or death.

Conclusion

Long-term follow-up of DCM patients revealed a distinct proteomic signature reflecting immune dysregulation, cell adhesion abnormalities, and inflammatory activation. TRAIL-R, Gal-9, and PGF independently predicted adverse outcomes, defining a molecular profile associated with progression to end-stage DCM.Table 1) Baseline characteristicsFor image description, please refer to the figure legend and surrounding text.Figure 1.Volcano PlotFor image description, please refer to the figure legend and surrounding text.