Early SGLT2 Inhibitor Use in Takotsubo Syndrome

Key Takeaways

• Early SGLT2 inhibitor use was associated with lower all-cause mortality in the propensity-matched analysis.
• No significant associations were reported for heart failure hospitalization, cardiogenic shock, cardiac arrest, or major adverse cardiovascular events.
• The mortality signal remained directionally consistent in sensitivity analyses excluding patients with COVID-19 and, separately, diabetes mellitus.

Early SGLT2 inhibitor initiation within 14 days of incident Takotsubo syndrome was associated with lower all-cause mortality in a multicenter real-world cohort, with a hazard ratio of 0.71. Investigators evaluated adults after a first recorded event and examined whether early postdiagnosis exposure was associated with subsequent outcomes. The main signal was all-cause mortality, while other measured cardiovascular endpoints were reported separately. The comparison reflected observed treatment timing across routine care settings rather than randomized assignment. It compared early use with no early use after incident Takotsubo syndrome.

The study used the TriNetX US Collaborative Network for a multicenter real-world observational cohort, which the authors described as an observational landmark analysis. Adults with incident Takotsubo syndrome were identified from 2015 through 2025 across US health care organizations in the network. Early SGLT2 inhibitor use was defined as treatment initiation within 14 days of Takotsubo syndrome, and patients without early use served as comparators. Among 54,701 identified patients, early users were matched 1:1 with patients without early use, yielding 1,803 matched pairs for analysis. Time-to-event follow-up began on day 14, establishing the landmark for subsequent clinical outcomes in both matched groups.

All-cause mortality was the primary outcome in the matched cohort and the main endpoint for the time-to-event analysis. Investigators reported both event rates and a hazard estimate for this endpoint in the matched cohort. Event rates were 8.1% with early SGLT2 inhibitor use and 13.6% without early use after the landmark period. The hazard ratio was 0.71, with a 95% confidence interval of 0.58 to 0.87 and a p value of 0.001. In the matched cohort, the observed difference was limited to mortality rather than a broader composite outcome signal.

Secondary outcomes included heart failure hospitalization, cardiogenic shock, cardiac arrest, and major adverse cardiovascular events in the matched comparison. Associations for those endpoints were not significant, separating them from the mortality finding. The mortality signal remained directionally consistent in sensitivity analyses excluding patients with COVID-19 from the cohort. A separate sensitivity analysis excluding patients with diabetes mellitus showed the same directional pattern for mortality. Because the design was observational, the findings reflect association rather than causation within the reported real-world cohort.