Predictors of worsening kidney function and their relation to SGLT2 inhibitor dapagliflozin administration in type 2 diabetes mellitus patients with chronic heart failure

Abstract Background Sodium-glucose cotransporter 2 inhibitors (SGLT2i) had a favorable impact on the kidney function in heart failure (HF) patients, while there is no clear evidence of what factors predict this effect. The aim of the study was to identify plausible predictors for kidney function outcome among HF patients and investigate their association with SGLT2i. Methods We prospectively enrolled 480 patients with established type 2 diabetes mellitus (T2DM) and concomitant chronic HF I-IV New York Heart Association functional classes and followed them for 52 weeks. In the study, we determined the kidney outcome as a composite of a sustained declined in estimated glomerular filtration rate by 40% from baseline, or newly end-stage of kidney disease, or kidney replacement therapy. The relevant medical information, measurement of the biomarkers (N-terminal natriuretic pro-peptide, irisin, apelin, adropin, C-reactive protein, tumor necrosis factor-alpha) were collected at baseline and at the end of the study. Results The composite kidney outcome was detected in 88 (18.3%) patients of entire population. Amongst these individuals, 51 patients demonstrated declined in eGFR by 40% from baseline at the end of the follow-up period, 35 individuals had newly diagnosed ESKD and consequently two patients required kidney replacement therapy (hemodialysis). All patients received guideline-recommended optimal therapy, which was adjusted to phenotype / severity of HF, CV risk and comorbidity profiles and fasting glycaemia. The changes of the biomarkers levels during the dapagliflosin administration are reported Figure 1. We used Received Operation Curve analysis to identify the well-balanced cut-off points for the biomarkers (Figure 2). The multivariate logistic regression revealed that use of SGLT2i (OP = 0.92; p = 0.048), baseline serum levels of irisin less 4.50 ng/mL (OR = 1.51; p = 0.001), adropin less 2.10 ng/mL (OR = 1.15; p = 0.001) along with increase in the levels of these biomarkers less 15% (OR = 1.60; p = 0.001) and less 6% (OR = 1.21; p = 0.001) respectively retained an independent predictor for composite kidney end-point. We noticed that irisin serum levels less 4.50 ng/mL at the baseline and the increase in irisin less 15% added more valuable predictive information than the reference variable. Yet, the combination of irisin less 4.50 ng/mL at the baseline and the increase in irisin serum levels less 15% (AUC = 0.91; 95% CI = 0.87–0.95) improved discriminative value of each biomarker alone. Conclusions we suggest that low levels of irisin and its inadequate increase during administration of SGLT2i are promising predictors for unfavorable kidney outcome among T2DM patients with concomitant HF. The dynamic changes of the biomarkers ROC curve analysis