Circulating Soluble Thrombomodulin Is Elevated in Early-Stage CKD and Is Associated with Renal Dysfunction: A Transcriptomic and Retrospective Cohort Study

Figures

• 

  Figure 1. THBD expression is increased in renal injury and is predominantly localized to renal endothelial cells. (A) Thbd mRNA expression in kidneys from sham-operated mice and mice 7 days after unilateral ureteral obstruction (UUO) in the GSE217650 dataset. Data are shown as mean ± SD. The two groups were compared using Welch’s t test (two-tailed). ** p < 0.01. (B) Pearson correlation between Thbd expression and a renal fibrosis-related gene score in GSE217650. (C) Thbd mRNA expression in kidney samples from C57BL/6 mice subjected to mild (20 min) or severe (35 min) ischemia–reperfusion injury (IRI) 7 days in the GSE281539 dataset, and was used to investigate transcriptional alterations associated with the AKI-to-CKD transition. The two groups were compared using Welch’s t test (two-tailed). * p < 0.05. (D) Uniform manifold approximation and projection (UMAP) visualization of major renal cell populations in the human kidney single-cell RNA-sequencing dataset GSE183276. (E) Distribution of THBD-detected and THBD-undetected cells in healthy, acute kidney injury (AKI), and chronic kidney disease (CKD) samples from GSE183276. (F) UMAP visualization of major renal cell populations in the…

• 

  Figure 2. Circulating soluble thrombomodulin is elevated in patients with early-stage CKD. (A–D) Comparison of circulating soluble thrombomodulin (sTM) concentration (A), serum creatinine (CREA) level (B), estimated glomerular filtration rate (eGFR) (C), and urinary albumin-to-creatinine ratio (UACR) (D) between participants without CKD and patients with stage 1 CKD. Individual data points are shown together with the median (IQR). Two-group comparisons were performed using the Mann–Whitney U test (see Section 2.5 for details). * p < 0.05; *** p < 0.001; ns, not significant. CKD stages were assigned based on longitudinal clinical diagnosis (see Section 2.3).

• 

  Figure 3. Associations between circulating soluble thrombomodulin and renal function parameters in the study cohort. (A–D) Pearson correlation analyses between circulating soluble thrombomodulin (sTM) and serum creatinine (CREA) (A), serum urea (B), estimated glomerular filtration rate (eGFR) (C), and urinary albumin-to-creatinine ratio (UACR) (D). Given the exploratory nature of the correlation analyses, p values were adjusted using the Benjamini–Hochberg false discovery rate (FDR) procedure across 20 tested associations; all panels shown survived FDR correction (q < 0.05). Correlation coefficients (r) and corresponding q values are shown in each panel.

• 

  Figure 4. Circulating soluble thrombomodulin increases progressively with CKD severity and discriminates CKD. (A–D) Comparison of circulating soluble thrombomodulin (sTM) concentration (A), serum creatinine (CREA) (B), serum urea (C), and estimated glomerular filtration rate (eGFR) (D) among the non-CKD group and patients with CKD stages 1–2, 3, 4, and 5. Individual data points are shown together with the mean ± SD. Group comparisons in (A–D) were analyzed by Welch’s ANOVA followed by Dunnett’s T3 post-hoc tests (see Section 2.5); asterisks denote Dunnett’s T3-adjusted p values: * p < 0.05, *** p < 0.001, and **** p < 0.0001; (E–G) Receiver operating characteristic (ROC) curves for sTM, eGFR, and UACR in discriminating CKD, including CKD1-5 from non-CKD (E), CKD1-2 from non-CKD (F), and CKD4-5 from non-CKD and CKD1-2 combined (G). Area under the curve (AUC), confidence interval (CI), optimal cutoffs, sensitivities, specificities, and DeLong test comparisons are reported in Table A1.

• 

  Figure 5. Associations between circulating soluble thrombomodulin and thromboelastography, coagulation, and fibrinolytic parameters in the study cohort. (A–D) Pearson correlation analyses between circulating soluble thrombomodulin (sTM) and thromboelastography (TEG) parameters, including reaction time (R value) (A), clot formation time (K value) (B), α angle (C), and maximum amplitude (MA) (D). (E–I) Pearson correlation analyses between circulating sTM and conventional coagulation and fibrinolytic parameters, including prothrombin time (PT) (E), activated partial thromboplastin time (APTT) (F), fibrinogen (FIB) (G), D-dimer (H), and fibrin(ogen) degradation products (FDP) (I). p values were adjusted using the Benjamini–Hochberg false discovery rate (FDR) procedure across all 20 correlation tests shown in Figure 3, Figure 5 and Figure 6; R did not survive FDR correction (q = 0.559) and is shown for transparency. Correlation coefficients (r) and corresponding q values are shown in each panel.

• 

  Figure 6. Associations between circulating soluble thrombomodulin and cardiac biomarkers in the study cohort. (A–D) Pearson correlation analyses between circulating soluble thrombomodulin (sTM) and myoglobin (Mb) (A), creatine kinase (CK) (B), creatine kinase-MB (CK-MB) (C), and high-sensitivity cardiac troponin I (hs-TnI) (D). (E–H) Comparison of serum Mb (E), CK (F), CK-MB (G), and hs-TnI (H) levels among the non-CKD group and patients with CKD stages 1–2, 3, 4, and 5. p values were adjusted using the Benjamini–Hochberg false discovery rate (FDR) procedure across all 20 correlation tests shown in Figure 3, Figure 5 and Figure 6. Correlation coefficients (r) and corresponding q values are shown in the correlation plots. Individual data points are presented together with the mean ± SD in the group comparisons. Group comparisons in (E–H) were analyzed by Welch’s ANOVA followed by Dunnett’s T3 post-hoc tests (see Section 2.5); asterisks denote Dunnett’s T3-adjusted p values: * p < 0.05, ** p < 0.01, *** p < 0.001.