Association of an Exploratory Iron Index Score With Observed NT-proBNP Trajectory in SGLT2-Treated Heart Failure and Clinical Outcomes in a Broader Cohort.

OBJECTIVE: Iron-related risk in heart failure is multidimensional, yet routine assessment still relies primarily on single laboratory-based definitions. This study evaluated whether an exploratory composite Iron Index Score was associated with the observed routine-care NT-proBNP trajectory after SGLT2 inhibitor therapy and with adverse clinical outcomes in a broader score-complete cohort. METHODS: This retrospective real-world cohort study used a nested analytic design. The primary biomarker analysis included a final SGLT2-treated cohort with paired baseline and follow-up NT-proBNP measurements and complete score components (n=238). The supportive clinical analysis included a broader score-complete cohort regardless of SGLT2 use (n=472). The cohort predominantly comprised patients with heart failure with reduced ejection fraction. The Iron Index Score assigned 1 point each for transferrin saturation <20%, hemoglobin <12 g/dL, ferritin <100 ng/mL, albumin <3.5 g/dL, and red cell distribution width >14.5%. The primary outcome was the change in log-transformed NT-proBNP from baseline to the lowest recorded on-treatment value. Secondary analyses examined NT-proBNP response, heart failure hospitalization, and exploratory all-cause mortality. RESULTS: In the final SGLT2-treated biomarker cohort, the adjusted ordinal score showed a directionally less favorable delta log NT-proBNP estimate but did not reach conventional statistical significance (beta 0.15, 95% CI -0.00 to 0.30; P=0.054). The high-risk group showed a stronger exploratory adverse signal compared with the low-risk group (beta 1.00, 95% CI 0.40 to 1.59; P<0.001). NT-proBNP responder status was not independently associated with score burden. In the broader score-complete cohort, higher score burden was associated with heart failure hospitalization in the unadjusted analysis, but this association was attenuated after adjustment (hazard ratio 1.04, 95% CI 0.85 to 1.27; P=0.690). All-cause mortality analyses were interpreted as exploratory because only 21 deaths occurred. CONCLUSION: A higher exploratory Iron Index Score showed a hypothesis-generating signal toward a less favorable observed NT-proBNP trajectory in a predominantly HFrEF SGLT2-treated cohort. Broader cohort analyses showed supportive but less stable clinical outcome signals.