U-TXM: a novel biomarker for early detection of diabetic kidney disease

Figures

• 

  Figure 1. Correlations between U-TXM and renal function markers.(a) Scatter plot of log(U-TXM) vs. log(UACR); (b) scatter plot of log(U-TXM) vs. log(BUN); (c) scatter plot of log(U-TXM) vs. log(serum creatinine, Cr); (d) scatter plot of log(U-TXM) vs. log(microalbumin, mAlb); (e) violin plot of UACR across low/middle/high U-TXM groups; (f) violin plot of BUN across low/middle/high U-TXM groups; (g) violin plot of Cr across low/middle/high U-TXM groups; (h) violin plot of mAlb across low/middle/high U-TXM groups.Panels (a–d) display Spearman correlation coefficients (r) with P values; panels (e–h) show distributions stratified by U-TXM. Variables were log-transformed where applicable. Differences in the plots are indicated by asterisks: ns denotes P>0.05, * denotes P<0.05, ** denotes P<0.01, *** denotes P<0.001, **** denotes P<0.0001.

• 

  Figure 2. Forest plot of univariate and multivariate logistic regression analysis. This chart shows the effect of different clinical variables on the target outcome in univariate (left) and multifactorial (right) Logistic regression analyses. The dot represents the OR and the horizontal line represents the 95% CI. Chart on the left (single factor analysis): Including Sex, Age, DD, SBP, DBP, BMI, FPG, HbA1c, logBUN, logCr, UA, CysC, FPG, DBil, Alb, logLDL, logTC, logTG, logPLT, and LOU-TXM. Among them, the variables with a P-value <0.05 were statistically significant. Figure on the right (multivariate analysis): After adjusting for other variables, SBP, logBUN, logCr, and logU-TXM were still significantly associated with the target outcome (P<0.05), while the effects of BMI, FPG, and HbA1c were no longer significant after adjustment. Alb, albumin; BMI, body mass index; BUN, blood urea nitrogen; CI, confidence interval; Cr, creatinine; CysC, cystatin C; DBP, diastolic blood pressure; DBil, direct bilirubin; DD, disease duration; FPG, fasting plasma glucose; HbA1c, glycated hemoglobin; LDL, low-density lipoprotein cholesterol; logBUN, log-transformed blood urea nitrogen; logCr,…

• 

  Figure 3. ROC curve analysis. This figure presents the ROC curves evaluating the predictive performance of different variables for the target outcome. The AUC was 0.623 in the training set (a) and 0.625 in the validation set (e). SimultaneousEquation 1: it(P) = -0.424 + 0.768 × Cr + 0.685 × U-TXM. This model achieved an AUC of 0.712 in the training set (b) and 0.706 in the validation set (f). Simultaneous Equation 2: it(P) = -4.438 + 0.903 × Cr - 0.251 × BUN + 0.563 × U-TXM + 0.031 × SBP. The AUC of this model was 0.736 in the training set (c) and 0.731 in the validation set (g). Simultaneous Equation 3: it(P) = -5.733 + 1.051 × Cr - 0.420 × BUN + 0.576 × U-TXM + 0.031 × SBP + 0.054 × BMI - 0.137 × HbA1c + 0.101 × FPG. This model achieved an AUC of 0.769 in the training set (d) and 0.803 in the validation set (h). The x-axis represents 1 - Specificity, while the y-axis represents Sensitivity. Different colored curves indicate different models or variables, with a larger AUC signifying stronger predictive ability. The ROC curve is used to assess the classification performance of different biomarkers or models. AUC values closer to 1 indicate higher predictive accuracy. AUC, area under the…

• 

  Figure 4. Association between urinary U-TXM levels and UACR groups, and diagnostic utility of U-TXM for DKD progression. Left panel: Violin plot showing the distribution of urinary U-TXM levels across UACR groups: A1 (<30 mg/g), A2 (30–300 mg/g), and A3 (>300 mg/g). U-TXM levels increase progressively from A1 to A3, with statistically significant differences observed between A1 and A2 (P<0.0001), as well as between A1 and A3 (P<0.0001). Right panel: ROC curve for U-TXM in discriminating patients with elevated albuminuria (A2/A3, UACR ≥30 mg/g) from those with normal albumin levels (A1). The AUC is 0.629, indicating statistical significance. These findings highlight the potential of U-TXM as a supplementary biomarker for assessing DKD progression. “****” indicates p < 0.0001, as per our statistical analysis.

• 

  Figure 5. Schematic illustration of the pathways linking urinary 11-dehydrothromboxane B2 (U-TXM) to diabetic kidney disease (DKD). Platelet activation increases thromboxane A2 (TXA2) generation, which promotes renal inflammation and contributes to endothelial/podocyte dysfunction and fibrotic signaling, thereby facilitating DKD progression (6, 7, 12–14). TXA2 is rapidly hydrolyzed to TXB2 and further metabolized to the stable urinary metabolite U-TXM, which reflects in-vivo TXA2 biosynthesis and platelet activation (15). Prior studies associate elevated U-TXM with heightened inflammatory/vascular risk and adverse prognosis, supporting its biomarker potential (21, 23, 24). In diabetes, platelet hyperreactivity and treatment factors may modulate TXA2/U-TXM levels (25, 26, 28).Solid arrows indicate confirmed mechanisms, while dashed arrows represent hypothetical links.