Real-world evidence of finerenone in Type 2 diabetes: addressing confounding and follow-up duration

Abstract

Full text

We thank Dong and colleagues for their insightful interest in our article regarding the association between non-steroidal mineralocorticoid receptor antagonists (MRAs) and cardiovascular events in Type 2 diabetes,^1^ and for their letter outlining key methodological considerations.^2^ We appreciate the opportunity to clarify the context of our findings and discuss the future of real-world evidence (RWE) in this domain.

First, regarding the concerns about unmeasured confounders—specifically medication adherence and baseline left ventricular ejection fraction—we agree that these are inherent limitations of retrospective electronic health record analyses. Unlike randomized controlled trials that enforce strict adherence protocols to determine ‘efficacy’, our study aimed to evaluate ‘effectiveness’ in a real-world setting where adherence varies.^3^ To mitigate selection bias, we employed robust propensity score matching including distinct pharmacological proxies [e.g. use of renin–angiotensin system blockade, beta-blockers, and sodium-glucose co-transporter 2 (SGLT2) inhibitor], cardiovascular comorbidities, and laboratory values, which serve as surrogate markers for heart failure severity. While residual confounding is possible, the consistency of our results across sensitivity analyses suggests a genuine treatment effect.

Second, the authors noted that our follow-up of 24 months may be insufficient to capture long-term cardiovascular and renal outcomes. We respectfully submit that this duration is clinically relevant. The pivotal finerenone in reducing kidney failure and disease progression in diabetic kidney disease (FIDELIO-DKD) trial^4^ demonstrated significant benefits with a median follow-up of 2.6 years, which is comparable to our observation window. Our findings essentially provide an early real-world validation of the trends observed in FIDELIO-DKD, proving that the divergence in event rates begins relatively early. As finerenone is a newly approved drug, real-world follow-up is currently limited by its short time on the market. Future longitudinal studies will be vital to assess ‘legacy effects’ and long-term renal preservation as clinical experience matures.

Third, regarding the attenuated effect sizes observed in women and patients with preserved renal function, we agree this heterogeneity warrants further investigation. This observation may reflect reduced statistical power within smaller subgroups or a ‘ceiling effect’ where lower-risk populations demonstrate smaller absolute risk reductions. These findings align with the clinical imperative to prioritize high-risk phenotypes who derive the greatest absolute benefit from MRA therapy.^4,5^

Finally, concerning safety, we concur that monitoring discontinuation is vital. However, we highlight that our study explicitly analysed safety endpoints, demonstrating a significantly lower risk of hyperkalaemia with finerenone compared with spironolactone. This safety signal is a critical differentiator for non-steroidal MRAs and supports their use in routine practice.

In conclusion, we are grateful for the constructive dialogue initiated by Dong et al.^2^ Despite the limitations of retrospective data, we believe our study adds crucial RWE that complements pivotal clinical trials, supporting the early and safe initiation of finerenone in eligible patients with Type 2 diabetes.