Characteristics of patients with chronic kidney disease and Type 2 diabetes initiating finerenone in the USA: a multi-database, cross-sectional study

Our study, analyzing real-world data from the USA, shows that finerenone has been used across a broad range of clinical and demographic characteristics during the initial months of availability after its marketing-authorization approval. The results point toward the use of finerenone as a complementary mode of action to other renal and CV protective drugs recommended for patients with CKD and T2D such as ACEi, ARB and SGLT2i.

Of note, our study aimed to describe patients who started finerenone, stratifying by relevant characteristics such as concurrent SGLT2i use and UACR categories, where available. To incorporate elements of the US FDA label [12], we restricted the Entry Cohort Event (initiation of finerenone) using the inclusion criteria specified in the methods (1. aged 18 years or older, 2. having CKD, 3. having T2D, and 4. not having T1D). The more elements we included to restrict our cohort definition, such as eGFR between 25 and 90 ml/min/1.73 m², the closer we would get to the definition used in the clinical trials and to the approved indication but at the expense of the sensitivity of a cohort that contained the patients using finerenone in real-world practice in the US. In a comparative setting, one would aim to narrow the definition to the specific label indication for larger internal validity, but the descriptive nature of our study warranted a focus on external validity to enhance generalizability.

Clinical practice shortly after finerenone availability seems to be in line with KDIGO recommendations, which position finerenone on top of RAS inhibition, with and without SGLT2i, as a drug to reduce CV and kidney risk in patients with T2D and CKD with albuminuria (UACR ≥30 mg/g) [1]. Furthermore, our study showed that around 40–50% of patients initiating finerenone received an SGLT2i during the 90 days before or the 30 days after initiating finerenone. In fact, a sensitivity analysis indicated that 10% of patients received an SGLT2i during the 30 days after finerenone initiation, suggesting a potential co-initiation of therapies.

Early intervention and intensive treatment is essential to reduce both CV and renal complications in patients with CKD and T2D [18]. Different clinical trials have shown the benefits of ACEi, ARB and SGLT2i to decrease CV outcomes and CKD progression among patients with CKD and T2D [5,19,20]. However, inflammation and fibrosis promoted by MR overactivation are generally not targeted by these therapies, leading to a persistent residual risk [21,22]. Consequently, it is necessary to perform a comprehensive approach that acts on the broader pathogenesis of CV and kidney disease. For example, SGLT2i act primarily on the hemodynamic and metabolic drivers of CKD progression, whereas finerenone appears to have a wider role on inflammation and fibrosis [23–25]. Despite the CONFIDENCE trial providing definite information about the effects of combining SGLT2i and finerenone in this population [26], it is important to ascertain how finerenone is being used in clinical practice, particularly with the concomitant use of both drugs.

Information about the use of finerenone in routine clinical practice is warranted. Consequently, our results are of great interest. In our study, mean age was 68–72, years, median eGFR 44 ml/min/1.73 m², and median UACR 110–339 mg/g. Previous studies performed in European populations of patients with CKD and T2D have shown a similar clinical profile to that observed in our study [27,28]. In the FIDELITY pooled analysis, which combined data from the FIGARO and FIDELIO trials, mean age was 64.8 years, 30.2% of patients were female, baseline eGFR was 57.6 ml/min/1.73 m², and baseline UACR 515 mg/g [10]. Therefore, compared with FIDELITY, our study's cohort of finerenone users seemed older, with worse renal function, but lower UACR on average.

Regarding baseline treatments, around 70.5% of patients were taking ACEi or ARB, 42–54% SGLT2i and 35–42% GLP-1RA. Compared with the overall population, patients taking concomitant SGLT2i had lower eGFR and UACR. This proportion of ACEi and ARB use is similar to that observed in prior observational studies of patients with CKD and T2D [28–30]. In FIDELITY, almost all patients were receiving RAS inhibitors, 7% SGLT2i and GLP-1RA, and the CV and renal benefits of finerenone were independent of treatment with SGLT2i and GLP-1RA [10,31]. Our data indicate that the combined use of finerenone and SGLT2i was significant in clinical practice independent of background therapies and clinical characteristics, potentially suggesting a healthcare practice acceptance of the complementary modes of action of finerenone and SGLT2i. Of note, the CONFIDENCE trial is a phase 2 study that is currently investigating the safety, tolerability and efficacy of the combination of empagliflozin with finerenone in 807 adults with T2D, stage 2–3 CKD, and UACR ≥300–<5000 mg/g [26].

Our study has some limitations which need to be considered when interpreting the results. This was a cross-sectional study using secondary databases, thus only anonymized data that had been recorded in the database could be collected. These data may not include all relevant information about a patient's health status or medical history, particularly if the patient has received care from multiple healthcare providers or institutions. Optum Claims contains records for drug dispensations which supposedly accurately reflect actual drug exposures, but we could not determine if patients were actually taking the drugs as indicated. Of note, the only element in our study design that contained data post-index was when we defined concomitant use of SGLT2i as an inclusion criterion for cohort number 2. We did so because concomitant use in the context of this cohort definition means evidence of use of SGLT2i around the initiation of finerenone; not necessarily only before but also shortly after. In addition, there could be a misspecification of the date in the database of origin thus a patient may have been indicated to start SGLT2i treatment on a given date but may not be recorded in the database until days or weeks later. Consequently, we included a 30-day post-index allowance to identify a record for SGLT2i exposure closely related to finerenone initiation. On the other hand, Optum EHR contains information on drug prescriptions which may or may not reflect actual dispensation and exposure to the drug. Moreover, drug prescriptions for chronic treatments in the US may be revised with a low frequency, since prescribers often rely on the patient getting refills periodically, thus a prescription may be a proxy for actual drug exposure even months later. When it comes to condition diagnoses, claims databases reflect the administrative management of the reimbursement schema of the patient, whereas EHR database reflect actual diagnoses made in a healthcare setting where patients are managed. In addition, Optum Claims has limited data availability when it comes to laboratory measurements, thus variables using eGFR or UACR constructs may not be complete. On the other hand, Optum EHR contains a more systematic collection of laboratory values. This notably affected our stratification by UACR, which is already impacted by low adherence to UACR monitoring recommendations in the US [32]; from the large proportion of patients without an identifiable record for UACR at baseline, we could not disentangle whether it is because of it not being monitored or because of it not being recorded in the database. This, along with the high intrapersonal variability of UACR, may cause a certain degree of misclassification of patients according to baseline UACR levels. Our study uses data from July 2021, when finerenone was approved in the US, until the most recent data available (September 2022 for Optum EHR and March 2022 for Optum Claims). This has two immediate consequences on our study. First, there is limited uptake appearing in the databases, therefore limiting the sample size of our study. Second, our study reflects a population of early adopters of finerenone. Regarding the former, despite the limited sample size compared with other studies using secondary real-world data, our results show consistency between databases, and we provided a comprehensive overview of the precision of continuous variables. In fact, we did not attempt to test comparative hypotheses, thus not affecting type II error probability due to low sample size. For the latter, we anticipate that the patient profile may change as more evidence and experience of real-world finerenone use becomes available. This is partially compensated by using the OMOP CDM and the OHDSI tools, as we were able to replicate the results of our analysis in two complementary databases, one administrative claims and one EHR, without affecting study design and assumptions, and, therefore, obtained a more comprehensive picture of the characteristics of patients. Furthermore, the analysis can easily be reproduced in other suitable data sources converted into OMOP CDM. To further enhance the transparency of our research conduct, our cohort definitions, concept sets, and Json scripts are available as Supplementary Materials to this publication. Retrospective claims database studies are observational in nature, so they cannot establish causality between exposure and outcome. Lack of generalizability may influence the results, and alternative explanations for the findings cannot be ruled out. Our cohorts were generated using the most recently available data and it provides understanding on the characteristics of patients as finerenone becomes available in other countries. Although this information may be especially relevant to a specific patient population and not fully generalizable outside the US, it provides contemporary insights on the attitude of prescribers toward the use of finerenone in patients for whom this therapy could be beneficial.

This contemporary study showed that the use of finerenone in clinical practice in the US is mostly aligned with international clinical guidance in terms of patient profile. Current management of CKD and T2D patients potentially reflects the use of finerenone independently from background therapies and clinical characteristics, suggesting implementation of therapeutic strategies based on different modes of action to address underlying pathologies.

D Vizcaya, C Kovesdy, A Reyes, E Pessina, P Pujol, G James, NG Oberprieler, contributed extensively to the work presented in this paper. All authors have contributed significantly to the conception, design, or acquisition of data, or analysis and interpretation of data. All authors have participated in drafting, reviewing, and/or revising the manuscript and have approved its submission.

The authors thank O Piqué-Duró, from University Pompeu Fabra, for assistance in reviewing the initial draft and quality control.

The study is funded by Bayer AG. D Vizcaya, E Pessina, P Pujol, G James, A Reyes and NG Oberprieler are full-time employees of Bayer and may own stock. OP-D was master's intern student at Bayer Pharmaceuticals, Sant Joan Despí, the time of this study. CP Kovesdy has received consulting fees from Abbott, Akebia, Astra Zeneca, Bayer, Boehringer Ingelheim, Cara Therapeutics, CSL Vifor, GSK and ProKidney. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Writing and editorial assistance was provided by Content Ed (Madrid, Spain) with funding from Bayer.

This study was conducted in accordance with Good Pharmacoepidemiology Practices. This analysis was planned as part of a larger study evaluating the real-world effectiveness of finerenone (NCT05703880; https://clinicaltrials.gov/ct2/show/NCT05703880).

This work is licensed under the Attribution-NonCommercial-NoDerivatives 4.0 Unported License. To view a copy of this license, visit https://creativecommons.org/licenses/by-nc-nd/4.0/