Cardiorenal therapeutics in a polluted world: interpreting treatment benefit in an environmental context

Fine particulate matter (PM_2.5) is a recognized environmental risk factor contributing to several million premature deaths each year. Despite this well-established burden, clinical trials of cardiorenal therapies have historically been designed and interpreted as if the environmental context of participants were uniform and clinically irrelevant. However, air pollution has rarely entered trial design, stratification, or interpretation despite the resounding evidence that it can contribute to disease pathogenesis and may modify treatment response. Two related questions therefore arise for preventive cardiology: whether established therapies remain effective in polluted environments, and whether therapies can attenuate environmental assault through mechanism-specific pathways.

In this issue of the American Journal of Preventive Cardiology, Al-Kindi and colleagues [1] address this gap by examining cardiorenal therapy across the spectrum of ambient particulate exposure. This post hoc analysis was based on the FInerenone in chronic kiDney diseasE and type 2 diabetes: Combined FIDELIO-DKD and FIGARO-DKD Trial programme analYsis (FIDELITY), pooling the two pivotal phase 3 finerenone trials in patients with chronic kidney disease and type 2 diabetes, with blinded endpoint adjudication and a median follow-up of approximately three years. Among 12,990 participants from 48 countries, the median PM_2.5 concentration was 15.5 µg/m³, well above the revised U.S. National Ambient Air Quality Standard (NAAQS) for annual PM_2.5 of 9.0 µg/m³. Higher PM_2.5 exposure was independently associated with composite cardiovascular, kidney, and combined endpoints. More importantly, finerenone retained its treatment benefit across exposure strata, reducing these outcomes by 14%, 24%, and 17%, respectively, with no statistical interaction across PM_2.5 quartiles. These benefits on UACR and eGFR decline were maintained whether PM_2.5 was below or above the NAAQS threshold of 9.0 µg/m³. This indicated that the clinical benefits of mineralocorticoid receptor (MR) antagonism are preserved in settings of higher ambient particulate exposure. Meanwhile, when PM_2.5 was modeled as a continuous variable, the effect estimates were similar in the finerenone and placebo arms, suggesting that finerenone does not appear to neutralize the cardiorenal effects induced by PM_2.5.

A notable strength of the study is its grounding in mechanistic plausibility. PM_2.5 can promote inflammation and fibrosis, pathways that partially overlap with those targeted by non-steroidal MR antagonism [2,3]. This overlap provides a reasonable basis for testing whether finerenone modifies pollution-related cardiorenal risk, yet it does not by itself establish a drug–environment interaction. As a recent review noted, pharmacological studies of air pollution may help distinguish pollutant-specific pathway interruption from general cardiovascular protection [4]. Whether such interactions emerge may therefore depend on where along the exposure-to-injury pathway a therapy acts. For example, in a randomized trial of traffic-related air pollution, inhaled bronchodilators blunted acute electrocardiographic disturbances in patients with chronic obstructive pulmonary disease, whereas beta-blockers and other cardiovascular medications offered no comparable protection in patients with ischemic heart disease [5]. This contrast suggests that therapies acting closer to the route of entry or early target tissue may be more likely to attenuate exposure-specific assault than therapies acting on shared downstream.

A consistent relative effect across exposure strata does not imply a uniform absolute benefit. For kidney events specifically, absolute benefit appeared more favorable in higher-exposure subgroups. The number needed to treat to prevent one event over 3.5 years fell from 123 in the lower-exposure subgroup to 39 in the higher-exposure subgroup. While this gradient should be interpreted cautiously given its post hoc nature, it illustrates that relative effect estimates alone may understate absolute benefit in highly exposed populations, a distinction with direct relevance to how clinical trial findings are translated to settings with substantially different baseline risk.

This study also represents a methodological precedent for environmental pharmacoepidemiology. Most prior investigations of drug–environment interaction have relied on observational comparisons of existing medication use, in which pollution-attributable risk is contrasted across strata defined by patients’ background prescriptions. Such designs are inherently susceptible to confounding by indication, as patients receiving beta-blockers, statins, or MR antagonists are typically sicker than those who are not. Their apparent vulnerability to environmental stressors may therefore reflect both medication use and the underlying disease severity that prompted the prescription, and disentangling these influences in observational data is rarely feasible. Al-Kindi and colleagues capitalize on the FIDELITY design, in which finerenone was randomly allocated independently of baseline cardiorenal risk, breaking this confounding for the treatment comparison. Thus, the estimates of effect modification by finerenone are less susceptible to indication-driven bias.

Randomization, however, governs the assignment of finerenone, not the assignment of pollution exposure, which remains limitations on the usual concerns of observational exposure assessment. First, PM_2.5 quartiles in FIDELITY partly function as proxies for geography, with the highest-exposure stratum dominated by participants from Asia and Eastern Europe and the lowest-exposure stratum anchored in North America and Western Europe. Within this design, it is difficult to fully separate ambient particulates from co-pollutants such as NO₂, ozone, and black carbon, or from regional differences in diet, healthcare access, ancestry, and cumulative social exposures. Second, the baseline profile of the highest-exposure quartile warrants careful interpretation. Participants in this group were younger and leaner, with lower hs-CRP and a lower prevalence of prior cardiovascular disease, yet had more advanced albuminuria. This pattern may reflect regional differences in disease ascertainment and divergent cardiorenal phenotypes rather than a monotonic dose-response gradient. The modest but statistically significant heterogeneity in all-cause mortality across PM_2.5 quartiles, with a non-monotonic pattern favoring mid-exposure strata, further supports cautious interpretation. Third, PM_2.5 exposure was assessed at the trial-site level using a satellite-blended model rather than a direct individual measurement, an approach that may have introduced exposure misclassification. Taken together, these constraints temper causal inference and point to priorities for future drug–environment interaction studies in cardiorenal medicine.

This post hoc FIDELITY analysis brings the environmental context of cardiorenal therapeutics into clearer view. Their findings raise a broader question: how should preventive cardiology incorporate environmental exposure into therapeutic reasoning? At present, the field has only begun to do so. Existing cardiovascular guidance on PM_2.5 has given limited attention to the therapeutic implications of environmental exposure, aside from brief discussions of dietary supplements such as omega-3 fatty acids [6]. A recent American Heart Association scientific statement on nonoptimal ambient temperature and cardiovascular health identified drug–environment interactions as a priority research gap, including effects of medications on thermoregulation and the stability of pharmacologic agents under heat stress [7]. Comparable gaps remain for air pollution and other environmental stressors. Addressing these gaps does not require redesigning the cardiovascular trial as an environmental cohort. A more practical approach is collaborative: environmental epidemiologists working with trial teams to harmonize site-level exposure data, incorporate environmental context at the design and stratification stage, and prespecify drug–environment interaction analyses where scientifically warranted. The data infrastructure to support such analyses is largely in place. What has been lacking is explicit recognition that participants in preventive cardiology trials are continuously exposed to ambient pollutants and climate-related stressors, and that treatment effects are measured against an environmental backdrop that has too often been overlooked. In a world where polluted air and climate-related stressors increasingly define the conditions of chronic disease, therapeutic strategies and environmental prevention can no longer be kept in separate conversations.

CRediT authorship contribution statement

Xin Meng: Conceptualization, Investigation, Writing – original draft. Kai Chen: Conceptualization, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.