Counting days: Cardiovascular and kidney event postponement with glucagon‐like peptide 1 receptor agonists in individuals with type 2 diabetes

Glucagon-like peptide 1 receptor agonists (GLP-1RA) with ≥24-h pharmacodynamic activity delay the development and progression of cardiovascular and kidney disease in individuals with type 2 diabetes, and the benefits are consistent across subcutaneous and oral formulations, as recently demonstrated in a systematic review and meta-analysis of 10 large, randomized controlled trial.1 Specifically, GLP-1RA reduced the incidence rate of major adverse cardiovascular events (MACE), defined as first occurrence of cardiovascular death, nonfatal myocardial infarction or nonfatal stroke, by 14% (hazard ratio 0.86 [95% CI 0.81, 0.90]) with a number needed to treat (NNT) of 63 based on a weighted average median trial follow-up of 3.2 years. As with any other medication, even when prescribed according to guidelines, the person with diabetes receiving a GLP-1RA should be adequately informed about the benefits and risks of the therapy. However, although relative risks and NNT are statistically meaningful, they remain challenging for many people to interpret in a personal context. An alternative, potentially more intuitive measure that has been suggested is the postponement of outcomes.2, 3 The concept aligns with the underlying physiology: GLP-1RA therapy does not confer absolute protection against cardiovascular and kidney events, but slows disease progression, thereby delaying the onset of critical endpoints.4 However, outcome postponement estimates have not been widely adopted, perhaps because the underlying calculations are complex or time-consuming. To investigate the postponement of cardiovascular and kidney events in individuals with type 2 diabetes treated with GLP-1RA, we extracted the following data from the 10 trials that were included in the recently published systematic review and meta-analysis (LEADER, SUSTAIN-6, EXSCEL, Harmony Outcomes, REWIND, PIONEER 6, AMPLITUDE-O, FREEDOM-CVO, FLOW, SOUL): (1) study duration, (2) mortality rate in the placebo group, and (3) hazard ratio for MACE and its individual components, a composite kidney outcome (kidney failure [kidney replacement therapy or persistent estimated glomerular filtration rate <15 mL/min/1.73 m2], sustained ≥50% eGFR decline or nearest equivalent, or kidney-related death), kidney failure, and worsening of kidney function. Based on these parameters, we estimated the postponement of first occurrence of MACE, cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, the composite kidney outcome, kidney failure or worsening of kidney function achieved by GLP-1RA treatment over a weighted average median follow-up of 3.2 years. The outcome postponement model has previously been described in detail.3 95% CIs were calculated by Taylor-series error propagation (delta method). Meta-analysis was performed using a random-effects model (DerSimonian–Laird). Calculations were performed and the figure was generated using R (version 4.4. R Foundation for Statistical Computing, Vienna, Austria). MACE data were extracted from all 10 trials; however, PIONEER 6 and FREEDOM-CVO did not provide data on kidney failure (a component of the composite kidney endpoint), and Harmony Outcomes and FREEDOM-CVO did not provide data on worsening kidney function. In Figure 1 the estimated 3.2-year standardized postponement of MACE and its three components for each of the 10 trials are depicted. The weighted mean postponement was 10.4 [95% CI 6.1–14.6] days for MACE, 4.6 [95% CI 2.0–7.2] days for cardiovascular death, 4.0 [95% CI 1.3–6.7] days for nonfatal myocardial infarction, and 2.4 [95% CI 1.2–3.7] days for nonfatal stroke. Figure 2 depicts the estimated 3.2-year standardized postponement of the composite kidney outcome, kidney failure and worsening of kidney function. The weighted mean postponement was 2.3 [95% CI 0.4–4.2] days, 0.5 [95% CI −0.1 to 1.1] days, and 3.1 [95% CI 1.3–5.0] days, respectively. These outcome postponement estimates complement and extend previously published findings on the effects of GLP-1RA in individuals with type 2 diabetes who have, or are at high risk for, cardiovascular and kidney disease. Because the postponement of events with only a few days after several years of treatment will likely be perceived by many as surprisingly small, incorporating this type of analysis is particularly important in the benefit–risk assessment of GLP-1RA therapy and can provide a more nuanced understanding of treatment effects. Several limitations should be acknowledged when interpreting the outcome postponement estimates presented here. First, the outcome postponement model assumes stable event rates in all treatment arms throughout trials. Although we found no indication of changing dynamics by visual inspection, this assumption may not fully capture real-world variability. Second, the included studies differed in participants' baseline risk profile and other characteristics, which influence the magnitude of the estimated postponements. Finally, estimated postponements should be considered an underestimate of the full benefit of treatments, as these most likely extend beyond trial durations. However, predicting these effects is prone to substantial uncertainties, limiting the prognostic value. There are multiple proven benefits of GLP-1RA in individuals with type 2 diabetes, including a reduction of cardiovascular risk. However, as we have demonstrated here, the absolute postponement of MACE and kidney outcomes in individuals with type 2 diabetes at high risk of such events was modest. Although outcome postponement estimates should not be interpreted in isolation, they may strengthen conventional benefit–risk assessments of GLP-1RA therapy by providing a complementary measure of treatment effect. S.S., T.S.P, and K.K. were involved in the conception, design, and conduct of the study and the analysis and interpretation of the results. S.S. wrote the first draft of the manuscript, and all authors edited, reviewed, and approved the final version of the manuscript. S.S. and K.K. are the guarantors of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The authors thank Palle Mark Christensen, MD, PhD, and Jesper Hallas, MD, DMSc, for insightful discussions on outcome postponements. None. S.S. reports receiving consulting and speaker honoraria from Novo Nordisk A/S and was employed by Novo Nordisk A/S from 2022 to 2023. K.K. and T.S.P. declare no potential conflicts of interest relevant to this article. The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/dom.70589. The data that support the findings of this study are available from the corresponding author upon reasonable request.