Risk Assessment of Kidney Disease Progression and Efficacy of SGLT2 Inhibition in Patients With Type 2 Diabetes Free

Chronic kidney disease (CKD) affects approximately one-third of patients with type 2 diabetes (T2D), and its prevalence has remained high worldwide (1,2). The rate of decline in kidney function, as measured by the estimated glomerular filtration rate (eGFR), and the risk of progression to advanced kidney disease are highly variable in patients with T2D (3,4). Early identification of patients with T2D that are at higher risk of kidney disease progression can help inform patient-provider communication, timing of referral to nephrology, and selection of optimal medical management. This is particularly helpful at a time of availability of several pharmacotherapeutic strategies that can modify the course of CKD, such as renin-angiotensin system inhibitors (5,6), sodium–glucose cotransporter 2 (SGLT2) inhibitors (7–10), selective mineralocorticoid receptor antagonists (11), and glucagon-like peptide 1 receptor agonists (12).

Although other kidney risk scores have been developed, they were developed using a general population without a specific attention to T2D, focused on predicting an eGFR <60 mL/min/1.73 m², which is applicable only to patients with preserved renal function, or were generated to predict kidney failure, a late and less common outcome in patients without advanced CKD (13–16). Given limitations in these singular outcomes, the U.S. Food and Drug Administration and the European Medicines Agency have embraced a composite kidney disease progression outcome of sustained ≥40–50% decline in eGFR, development of end-stage kidney disease (ESKD), or kidney death (17). This was recently adopted by the CKD Prognosis Consortium in their development and validation of a robust prediction model for the composite of ≥40% eGFR decline or kidney failure in patients with and without T2D (18).

Thus, we leveraged data from multiple cardiovascular clinical trials enrolling patients with T2D to develop and validate a clinical risk model for predicting the risk of kidney disease progression as defined by the composite outcome advocated by regulatory authorities.

This study consisted of patients with T2D from a pooled cohort of four multinational, randomized, placebo-controlled, cardiovascular clinical trials: Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus (SAVOR)–Thrombolysis in Myocardial Infarction (TIMI) 53 (19), Dapagliflozin Effect on Cardiovascular Events (DECLARE)-TIMI 58 (20), Further Cardiovascular Outcomes Research With PCSK9 Inhibition in Subjects With Elevated Risk (FOURIER) (TIMI 59) (21), and Cardiovascular and Metabolic Effects of Lorcaserin in Overweight and Obese Patients (CAMELLIA)-TIMI 61 (22). SAVOR-TIMI 53 was a randomized, placebo-controlled outcomes trial that evaluated the dipeptidyl peptidase 4 inhibitor saxagliptin in 16,492 patients with T2D with multiple risk factors for or established atherosclerotic cardiovascular disease (ASCVD) followed for a median of 2.1 years. DECLARE-TIMI 58 was a randomized, placebo-controlled, outcomes trial that evaluated the SGLT2 inhibitor dapagliflozin in 17,160 patients with T2D with multiple risk factors for or established ASCVD followed for a median of 4.2 years. Only the placebo arm (n = 8,578) was used for the purposes of clinical risk score development given the known effect of dapagliflozin on kidney outcomes. FOURIER (TIMI 59) was a randomized, placebo-controlled, outcomes trial of the proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitor evolocumab in 27,564 patients with stable ASCVD on statin therapy followed for a median of 2.2 years, of whom 9,318 were available for this analysis and had T2D. Lastly, CAMELLIA-TIMI 61 was a randomized, placebo-controlled, outcomes trial of the selective serotonin 2C receptor agonist lorcaserin in 12,000 patients with overweight or obesity with multiple risk factors for or established ASCVD followed for a median of 3.3 years, of whom 6,816 had T2D.

The derivation cohort consisted of 70% of patients (n = 28,842) selected at random from the pooled cohort. The validation cohort consisted of the remaining 30% of patients (n = 12,362) from the pooled cohort. The ethics committees from participating study centers approved the protocols for each of the trials. Written informed consent was obtained from all patients prior to enrollment.

The primary outcome for this analysis was kidney disease progression, defined as the composite of sustained ≥40% decline in eGFR, ESKD (defined as dialysis for ≥90 days, kidney transplantation, or sustained eGFR <15 mL/min/1.73 m²), or kidney death. eGFR was estimated with the Chronic Kidney Disease Epidemiology Collaboration equation using serum creatinine (23). This outcome was selected given current clinical trial and drug development standards following discussions between the National Kidney Foundation and U.S. Food and Drug Administration, in which a confirmed and sustained ≥40% decline in eGFR was deemed to be an acceptable surrogate end point for the subsequent development of kidney failure (17). ESKD and kidney death were prospectively collected and centrally adjudicated by the TIMI Clinical Events Committee using standard definitions in SAVOR-TIMI 53, DECLARE-TIMI 58, and CAMELLIA-TIMI 61. In FOURIER (TIMI 59), ESKD was identified by search of the adverse event reporting safety clinical trial database using the Medical Dictionary of Regulatory Affairs (MedDRA) preferred terms (“end stage renal disease” and “diabetic end stage renal disease”). In DECLARE-TIMI 58, we also assessed chronic eGFR slope (6 months to 48 months) using previously published methodology (24).

Based on clinical relevance and availability in the clinical trial record of participants treated in the cohort, 23 baseline candidate risk variables were considered for inclusion in the risk model. The candidates included age, sex, coronary artery disease, ASCVD (defined as prior myocardial infarction, prior ischemic stroke, and peripheral artery disease), percutaneous coronary intervention, coronary artery bypass grafting, history of heart failure, atrial fibrillation, hypertension, systolic blood pressure (SBP), diastolic blood pressure (DBP), heart rate, current smoker status, duration of T2D, baseline insulin use, eGFR, urine albumin-to-creatinine ratio (UACR), BMI, waist circumference, glycated hemoglobin (HbA_1c), HDL-cholesterol (-C), LDL-C, and hemoglobin. Race was not included as a candidate variable to align with the broader initiative of removing race from clinical algorithms.

Kidney disease progression was evaluated as a time-to-event end point using multivariable Cox regression. Schoenfeld residuals were used to test the proportional hazards assumption, which was met for all evaluated variables. Clinical predictors of kidney disease progression were selected using backward selection, with a threshold of P < 0.001 so that only highly significant variables were included in the final model. Prior to the selection procedure, potential nonlinearities in the association between candidate continuous predictors and composite kidney end points were assessed with spline curves. If the linearity assumption was not met, categorization based on graphically determined thresholds was considered. Accordingly, age, DBP, heart rate, HbA_1c, UACR, waist circumference, HDL-C, LDL-C, and hemoglobin were modeled continuously, whereas SBP, duration of T2D, eGFR, and BMI were modeled categorically.

The discriminatory performance of the final risk model was assessed in both the derivation and validation sets using the Harrell c-index. In addition, the cumulative incidence of kidney disease progression was assessed according to categories of predicted risk. Extrapolating from a 10-year risk framework, with <2.5%, 2.5 to <10%, and ≥10% 10-year risk of kidney disease progression corresponding to low-, intermediate-, and high-risk, respectively, 2-year risk bins were defined: <0.5% (low), 0.5 to <2% (intermediate), and ≥2% (high). Calibration was assessed graphically in the validation cohort by calculating the slope with 95% CI of the comparison between the predicted 2-year risk (x-axis) and observed Kaplan-Meier cumulative incidence at 2 years (y-axis).

The performance of the final risk model was assessed separately in patients with and without eGFR <60 mL/min/1.73 m² as well as in patients with and without albuminuria (UACR <30 mg/g). In addition, discrimination for the individual components of the composite end point was assessed. We also assed the discriminatory performance of previous risk models including the kidney failure risk equation (14), the CKD Prognosis Consortium risk equation for incident CKD (13), and the CKD Prognosis Consortium risk model for decline in kidney function (18). We assessed the Harrell c-index for each of the models in the validation cohort using the published coefficients in each of the original manuscripts as well as coefficients resulting from refitting the models of the existing scores in our derivation cohort.

We assessed for heterogeneity in the relative and absolute treatment effect of dapagliflozin versus placebo on kidney end points in DECLARE-TIMI 58 according to baseline predicted risk. To test for heterogeneity in the hazard of the event, the interaction between treatment and the predicted risk was included in the Cox regression model. To compare absolute differences in the treatment effect of dapagliflozin versus placebo on absolute risk reduction (ARR) in kidney disease progression, the 4-year Kaplan-Meier cumulative incidence in patients randomly assigned to dapagliflozin was subtracted from that of patients randomly assigned to placebo across each risk category. To assess the trend of ARR in kidney disease progression with dapagliflozin by baseline risk category, we used an inverse-variance weighted least squares model, regressing ARR on risk category. Hence, the P-interaction was calculated as an ARR trend over the risk groups. In addition, we also assessed for heterogeneity of treatment effect in the herein described manner according to baseline predicted risk using the CKD Prognosis Consortium risk model for decline in kidney function (18). Finally, to assess the treatment effect of dapagliflozin on chronic eGFR slope, we assessed differences in least squares means eGFR between dapagliflozin and placebo from a mixed model as well as differences in chronic eGFR slope (generated from linear regression using measured values at months 6, 12, 24, 36, and 48) according to baseline predicted risk (24).

All statistical analyses were performed in R 4.2.2 software. All P values are two-sided, unless otherwise specified.

We encourage parties interested in collaboration and data sharing to contact the corresponding author directly for further discussions.

The baseline characteristics of the study population are summarized in Table 1. In the derivation cohort (n = 28,842), the median age was 64 years (interquartile range [IQR] 59–70), 34% were women, and 21% were non-White. The median duration of diabetes was 9.2 years (IQR 4.1–15), and median HbA_1c was 7.5% (IQR 6.7–8.5; 58 mmol/mol [IQR 50-69]); 35.6% of patients were receiving insulin therapy at baseline. The median eGFR at baseline was 76 mL/min/1.73 m² (IQR 62–91), with 77.6% of patients having preserved kidney function (eGFR ≥60 mL/min/1.73 m²). The median UACR was 13.3 mg/g (IQR 5.1–49), with 67% of patients having normal or mildly increased albuminuria (UACR <30 mg/g). During a median follow-up of 2.4 years, kidney disease progression occurred in 481 patients (Supplementary Table 1).

The baseline characteristics of the validation cohort (n = 12,362) were very similar (Table 1). The median duration of diabetes was 9.0 years, with 77.4% of patients having preserved kidney function and 66% of patients having normal or mildly increased albuminuria. Kidney disease progression occurred in 214 patients during a median follow-up of 2.4 years (Supplementary Table 1).

From the 23 candidate variables, 8 independent predictors of kidney disease progression were selected for inclusion in the final clinical risk model, which is presented in Table 2. The model had very good discrimination in both the derivation and validation cohorts, with Harrell c-indices of 0.798 (95% CI, 0.774–0.821) and 0.798 (95% CI, 0.765–0.831), respectively. Discrimination for the individual components of the kidney disease progression composite was also very good, with Harrell c-indices of 0.799 for sustained ≥40% eGFR decline, 0.894 for ESKD or kidney death, 0.891 for ESKD, and 0.867 for kidney death in the validation cohort (Supplementary Table 2). Finally, the model predicted kidney disease progression in the validation cohort both in patients with and without baseline eGFR <60 mL/min/1.73 m² (c-indices of 0.845 and 0.763, respectively) (Supplementary Table 3) as well as patients with and without baseline albuminuria (c-indices of 0.729 and 0.803, respectively) (Supplementary Table 3).

As shown in Fig. 1A, the model identified strong gradients of risk for kidney disease progression in both the derivation and validation cohorts. Moreover, the 2-year cumulative incidence fell within the appropriate predefined 2-year risk categories, indicating appropriate calibration of the risk model as well. Further supporting the risk model’s calibration, observed Kaplan-Meier cumulative incidence of kidney disease progression closely corresponded to predicted risk (across risk deciles), resulting in a calibration slope of 0.98 (95% CI, 0.93–1.04) in the validation cohort (Fig. 1B).

In terms of comparative discriminatory performance, our final risk model in general compared favorably to the kidney failure risk equation (14), the CKD Prognosis Consortium risk equation for incident CKD (13), and the CKD Prognosis Consortium risk model for decline in kidney function (18) when using both the published formulas and refit models (Supplementary Tables 4 and 5). Supplementary Table 6 demonstrates the kidney failure risk equation (14) is well calibrated in our cohort of clinical trials, while the CKD Prognosis Consortium risk model for decline in kidney function (18) tends to slightly overestimate risk and the CKD Prognosis Consortium risk equation for incident CKD (13) strongly underestimates risk.

Figure 2 demonstrates the relative and absolute reductions in the risk of kidney disease progression with dapagliflozin according to predicted 4-year risk of <1%, 1 to <4%, and ≥4% (corresponding to 2-year risk categories of <0.5%, 0.5 to <2%, and ≥2%). The relative rate reductions with dapagliflozin were generally consistent across risk categories (15%, 49%, and 47%, respectively; P-interaction = 0.70), but the ARRs were greater across increasing risk categories (0.2%, 1.0%, and 3.5%, respectively; P-interaction = 0.01). The balanced numbers of patients in the dapagliflozin and placebo arms in each risk category are reflective of the 1:1 randomization in DECLARE-TIMI 58. Similarly, Supplementary Fig. 1 demonstrates the same pattern of consistent relative rate reduction and greater ARRs across increasing previously suggested 3-year kidney risk categories adapted to the ∼4-year follow-up of DECLARE-TIMI 58 patients (4.2 year median) (25). Supplementary Fig. 2 demonstrates that using the CKD Consortium risk model for decline in kidney function, the relative rate reductions with dapagliflozin were also consistent across risk categories (53%, 42%, and 48%, respectively; P-interaction = 0.58), but the ARRs were greater across increasing risk categories (0.5%, 0.8%, and 3.7%, respectively; P-interaction < 0.001).

Supplementary Fig. 3 presents mean eGFR decline over the span of 4 years in the placebo group from DECLARE-TIMI 58 according to predicted risk of kidney disease progression. The mean chronic eGFR slope was steeper in higher-risk compared with lower-risk groups (−3.97 vs. −2.48 vs. −2.13 mL/min/1.73 m²/year in the high-, intermediate-, and low-risk groups, respectively; P < 0.001 for each pairwise comparison). There was a slower decline in eGFR in patients treated with dapagliflozin versus placebo within every stratum of predicted baseline risk (Supplementary Fig. 4). The improvement in eGFR slope with dapagliflozin was +0.65, +0.43, and +0.33 mL/min/1.73 m²/year in the high-, intermediate-, and low-risk groups, respectively.

We derived and validated a clinical risk model that predicts kidney disease progression, defined as sustained ≥40% decline in eGFR, ESKD, or kidney death, in a large cohort of patients with T2D with a broad spectrum of kidney function across four contemporary clinical trials. The TIMI Risk Score for Kidney Disease Progression in Type 2 Diabetes (https://timi.org/timi-calculators/) had very good discrimination and calibration in the validation cohort and performed well in important subgroups, including those with and without eGFR <60 mL/min/1.73 m² as well as with and without albuminuria. The CKD Prognosis Consortium risk model for decline in kidney function also discriminates kidney disease progression in our clinical trial cohort population quite well. Uniquely, we also showed how risk prediction with these tools enabled identifying the magnitude of absolute clinical benefit from SGLT2 inhibition with dapagliflozin in patients with T2D.

The risk model leverages eight variables that are routinely used in the care of patients with T2D. Naturally, indicators related to kidney disease were relevant variables, including baseline eGFR and UACR, both of which are established strong indicators for kidney disease, and hemoglobin, likely reflecting the severity of baseline CKD. Baseline glycemia measured by HbA_1c and duration of diabetes were both independently associated with progression of kidney disease, likely representing the negative effects of longstanding diabetes on renal function. Notably, SBP ≥160 mmHg stood out as having one of the strongest associations with kidney disease progression. Lastly, the presence of vascular disease and heart failure were also independent predictors, reflecting the interdependence of these organ systems.

Currently, comprehensive care of patients with T2D at risk for CKD involves adequate control of glycemia, blood pressure, lipids, smoking cessation, promoting healthy lifestyle choices, appropriate referrals to kidney specialists, and initiation of medications that have consistently been shown to slow kidney disease progression such as SGLT2 inhibitors. Delivery of risk-based and personalized care requires both accurate risk prediction models and relevant absolute risk thresholds for clinical decision making. This is particularly important in an era where there is a sizeable and growing set of pharmacotherapeutic options to combat kidney disease, especially in patients with T2D, in whom there is high interpersonal variability in kidney disease risk (26).

Other kidney risk assessment tools have been published (13,14,18). However, there are a few key differences worth highlighting. First, our risk model was developed in a population of patients with T2D with less advanced kidney disease compared with other model derivation cohorts (27). Historically, eGFR and albuminuria have been regarded as the most important indicators of kidney disease progression. However, in a population without advanced kidney disease, as in the case of our own cohort where <10% of patients had UACR ≥300 mg/g or eGFR <45 mL/min/1.73 m², the discrimination of eGFR and albuminuria is reduced. More established risk models, such as the CKD Prognosis Consortium risk equation for incident CKD developed by Nelson et al. (13), although with potential for use in the general population, were developed to predict new-onset CKD (eGFR <60 mL/min/1.73 m²) in patients with preserved renal function, a strategy that has limited applicability for risk stratification and therapy selection for patients with CKD or eGFR that approximates the eGFR threshold of 60 mL/min/1.73 m². This was demonstrated by the score’s tendency to underestimate risk in our clinical trial population. Other risk models that have been more widely validated, such as the kidney failure risk equation (14), have focused on predicting kidney failure in patients with established CKD, which is a late and less common short- and intermediate-term outcome in patients without advanced CKD. Despite performing well in predicting kidney end points in our own cohort, these risk scores were not superior to the TIMI Risk Score for Kidney Disease Progression in Type 2 Diabetes. This is not surprising, considering that the kidney failure risk equation (14) and CKD Prognosis Consortium model for new-onset CKD (eGFR <60 mL/min/1.73 m²) (13) were each developed to predict different end points (i.e., kidney failure and new-onset CKD, respectively) than our own model. However, the comparison was made in context of their wide recognition and the shift of focus to include changes in eGFR in defining kidney disease progression. Moreover, we found that the CKD Prognosis Consortium model for new-onset CKD also did well predicting ESKD.

Lastly, our focus on predicting broad and earlier outcomes is in line with the recent shift toward use of a sustained ≥40% eGFR decline as an accepted major surrogate end point of progression to kidney failure (17,20,28,29), allowing earlier recognition of risk and greater opportunity to slow the progression of kidney disease. The CKD Prognosis Consortium has recently developed and validated a prediction model for the composite of ≥40% eGFR decline or kidney failure (defined as kidney replacement therapy) in patients with and without T2D and stratified by presence of CKD (18). We add to this by requiring events of eGFR decline to be confirmed, which is a relevant distinction when using the risk score for tailoring drug therapies such as SGLT2 inhibitors and renin-angiotensin system inhibitors, which are known to cause transient declines in eGFR due to intraglomerular hemodynamic changes. Nonetheless, we further add to the literature by showing that the CKD Prognosis Consortium risk model (18) not only validates well as a risk stratification tool for a composite kidney end point that includes eGFR reduction of ≥40% that is sustained but also discriminates treatment heterogeneity with dapagliflozin. Overall, these data support the ability of risk prediction models for kidney disease progression to discriminate risk and the magnitude of treatment benefit.

As contemplated in existing guidelines and expert consensus pathways, particularly in T2D, there is ample room for tailoring kidney disease-modifying therapy, but these efforts are often guided by abnormalities of single parameters such as eGFR or albuminuria (30,31). Consequently, there may be a proportion of patients with higher risk but without overt abnormalities in these two indicators who may be identified with a multivariable risk stratification tool. Using the TIMI Risk Score for Kidney Disease Progression in Type 2 Diabetes, we demonstrate that there is a clear gradient of absolute risk reduction in progression of kidney disease with SGLT2 inhibition as a function of being able to detect those that have higher baseline risk. In addition, despite experiencing the greatest decline in eGFR over time in DECLARE, patients at the higher predicted risk category also experienced the greatest numerical attenuation of decline in eGFR slope with dapagliflozin. This is particularly helpful in addressing potential prescriber hesitancy in patients with more advanced CKD given clinicians’ concern for acute reduction in eGFR with initiation of dapagliflozin, which is now known to be related to acute glomerular hemodynamic changes rather than a true decline in kidney function (32).

There are limitations to this study. First, there are other known established biomarkers of kidney disease progression that were not measured, such as phosphate levels, cystatin C, and fibroblast growth factor 23. However, the objective of this study was to generate a clinical stratification tool that would be more widely implemented and thus based on readily accessible clinical markers used in routine management of patients with T2D.

Second, our risk model was developed in clinical trial patients, which may impact generalizability to nontrial cohorts. Nevertheless, the patients in these trials were chosen for their overall cardiovascular risk rather than their CKD risk, and, in fact, in comparison with other existing clinical trials, there is ample representation of patients with normal eGFR and normal or mildly increased albuminuria. Therefore, in cardiovascular trials in which albuminuria is traditionally used as an enrichment factor for cardiovascular outcomes, future use of the TIMI Risk Score for Kidney Disease Progression in Type 2 Diabetes would allow simultaneous enrichment for kidney outcomes in trials not traditionally designed to do so. An additional strength of developing this score in a clinical trial cohort is the protocol-driven measurement and confirmation of eGFR changes, which add to the fidelity of the end point of eGFR change.

Lastly, because the median follow-up time across the four trial cohorts was only 2.4 years and the decline in eGFR required confirmation with a subsequent scheduled blood draw, the primary time horizon for risk prediction was limited to 2 years. Nevertheless, the large number of events supported the development of a robust model, and the model performed very well when applied to 4-year risk prediction in DECLARE-TIMI 58.

Management of patients with T2D aimed at modifying kidney complications will benefit from accurate risk prediction and can align risk and therapeutic decision making while personalizing the care of these patients. Such an approach is highly appealing with the availability of established and a growing number of disease-modifying options to reduce the risk of kidney disease progression. Clinical risk models that help identify individuals at high risk of kidney disease progression might be used to tailor effective intervention to those who could benefit most.

Acknowledgments. The authors thank Giorgio E.M. Melloni (TIMI Study Group) for his valuable work in developing the interactive web application for this risk score on the TIMI Calculators page (https://timi.org/timi-calculators/).

Funding. F.A.M., S.D.W., B.M.S., D.A.M., and M.S.S. were supported for the present analysis by the American Heart Association Cardiometabolic Health & Type 2 Diabetes Mellitus Strategically Focused Research Network (20SFRN35120087). F.A.M. was supported by a T32 postdoctoral training grant from the National Institutes of Health, National Heart, Lung, and Blood Institute (T32HL007604).

Duality of Interest. SAVOR-TIMI 53 and DECLARE-TIMI 58 were supported by institutional research grants to Brigham and Women’s Hospital from AstraZeneca. FOURIER was supported by an institutional research grant to Brigham and Women’s Hospital from Amgen. CAMELLIA-TIMI 61 was supported by an institutional research grant to Brigham and Women’s Hospital from Eisai. F.A.M. reports consulting fees from Janssen. D.D.B. has received research grant support to his institution from AstraZeneca and Pfizer, consulting fees from AstraZeneca, Mobility Bio, Inc., and Youngene Therapeutics, honoraria from the Medical Education Speakers Network (MESN) and USV Private Limited, and participates on clinical end point committees for studies sponsored by Kowa Pharmaceuticals. J.P.D. reports fees from AstraZeneca, Ionic, Bayer, Boehringer Ingelheim, Eli Lilly and Company, Silence Therapeutics, Novo Nordisk, Sanofi, Cincor, CSL Behring, Reata, Genentech, and GlaxoSmithKline for clinical trial design, conduct, analysis, or data safety monitoring board participation. O.M. reports advisory board for Novo Nordisk, Eli Lilly and Company, Sanofi, Merck Sharp & Dohme, Boehringer Ingelheim, Bayer, and AstraZeneca, research grant support through Hadassah Hebrew University Hospital from Novo Nordisk and AstraZeneca, and speakers bureau for AstraZeneca, Novo Nordisk, Eli Lilly and Company, Sanofi, Merck Sharp & Dohme, and Boehringer Ingelheim. B.M.S. reports institutional research grants to Brigham and Women’s Hospital from Better Therapeutics, Merck, Novo Nordisk, and Pfizer, and consulting fees from Allergan, Boehringer Ingelheim, Better Therapeutics, Elsevier PracticeUpdate Cardiology, Esperion, Hanmi, Lexicon, Novo Nordisk, and equity in health [at] Scale and Doximity. S.D.W. reports grants from Amgen, AstraZeneca, Daiichi Sankyo, Eisai, Janssen, Merck, and Pfizer, and consulting fees from AstraZeneca, Boston Clinical Research Institute, Icon Clinical, and Novo Nordisk. The spouse of S.D.W., Dr. Caroline Fox, is an employee of Merck. D.L.B. discloses the following relationships—advisory board: Angiowave, Bayer, Boehringer Ingelheim, Cardax, CellProthera, Cereno Scientific, Elsevier PracticeUpdate Cardiology, High Enroll, Janssen, Level Ex, McKinsey, Medscape Cardiology, Merck, MyoKardia, NirvaMed, Novo Nordisk, PhaseBio, PLx Pharma, Regado Biosciences, and Stasys; board of directors: Angiowave (stock options), Boston VA Research Institute, Bristol-Myers Squibb (stock), DRS.LINQ (stock options), High Enroll (stock), Society of Cardiovascular Patient Care, TobeSoft; chair: inaugural chair, American Heart Association Quality Oversight Committee; consultant: Broadview Ventures, Hims; data monitoring committees: Acesion Pharma, Assistance Publique-Hôpitaux de Paris, Baim Institute for Clinical Research (formerly Harvard Clinical Research Institute, for the PORTICO trial, funded by St. Jude Medical, now Abbott), Boston Scientific (Chair, Pulmonary Embolism Thrombolysis [PEITHO] trial), Cleveland Clinic (including for the CENTERA THV System in Intermediate Risk Patients Who Have Symptomatic, Severe, Calcific, Aortic Stenosis [ExCEED] trial, funded by Edwards Livesciences), Contego Medical (Chair, Protection Against Emboli During Carotid Artery Stenting Using the Neuroguard IEP System [PERFORMANCE 2]), Duke Clinical Research Institute, Mayo Clinic, Mount Sinai School of Medicine (for the Edoxaban Compared to Standard Care After Heart Valve Replacement Using a Catheter in Patients With Atrial Fibrillation [ENVISAGE] trial, funded by Daiichi Sankyo, for the Randomized Comparison of Abluminus DES+ Sirolimus-Eluting Stents Versus Everolimus-Eluting Stents in Coronary Artery Disease Patients With Diabetes Mellitus Global [ABILITY-DM] trial, funded by Concept Medical), Novartis, Population Health Research Institute, and Rutgers University (for the National Institutes of Health-funded Myocardial Ischemia and Transfusion [MINT] Trial); honoraria: American College of Cardiology (Senior Associate Editor, Clinical Trials and News, ACC.org, Chair, ACC Accreditation Oversight Committee), Arnold and Porter law firm (work related to Sanofi/Bristol-Myers Squibb clopidogrel litigation), Baim Institute for Clinical Research (formerly Harvard Clinical Research Institute, Evaluation of Dual Therapy With Dabigatran vs. Triple Therapy With Warfarin in Patients With AF That Undergo a PCI With Stenting [RE-DUAL PCI] clinical trial steering committee funded by Boehringer Ingelheim, ApoA-I Event Reducing in Ischemic Syndromes II [AEGIS-II] executive committee funded by CSL Behring), Belvoir Publications (Editor in Chief, Harvard Heart Letter), Canadian Medical and Surgical Knowledge Translation Research Group (clinical trial steering committees), Cowen and Company, Duke Clinical Research Institute (clinical trial steering committees, including for the A Trial Comparing Cardiovascular Safety of Degarelix Versus Leuprolide in Patients With Advanced Prostate Cancer and Cardiovascular Disease [PRONOUNCE] trial, funded by Ferring Pharmaceuticals), HMP Global (Editor in Chief, Journal of Invasive Cardiology), Journal of the American College of Cardiology (Guest Editor, Associate Editor), K2P (Co-Chair, interdisciplinary curriculum), Level Ex, Medtelligence/ReachMD (continuing medical education [CME] steering committees), MJH Life Sciences, Oakstone CME (Course Director, Comprehensive Review of Interventional Cardiology), Piper Sandler, Population Health Research Institute (for the COMPASS operations committee, publications committee, steering committee, and USA national co-leader, funded by Bayer), Slack Publications (Chief Medical Editor, Cardiology Today’s Intervention), Society of Cardiovascular Patient Care (Secretary/Treasurer), WebMD (CME steering committees), and Wiley (steering committee); other: Clinical Cardiology (Deputy Editor), National Cardiovascular Data Registry (NCDR)-ACTION Registry Steering Committee (Chair), Veterans Affairs Clinical Assessment Reporting and Tracking (VA CART) Research and Publications Committee (Chair); patent: sotagliflozin (named on a patent for sotagliflozin assigned to Brigham and Women’s Hospital who assigned to Lexicon, neither I nor Brigham and Women’s Hospital receive any income from this patent); research funding: Abbott, Acesion Pharma, Afimmune, Aker Biomarine, Amarin, Amgen, AstraZeneca, Bayer, Beren, Boehringer Ingelheim, Boston Scientific, Bristol-Myers Squibb, Cardax, CellProthera, Cereno Scientific, Chiesi, CinCor, Cleerly, CSL Behring, Eisai, Ethicon, Faraday Pharmaceuticals, Ferring Pharmaceuticals, Forest Laboratories, Fractyl, Garmin, HLS Therapeutics, Idorsia, Ironwood, Ischemix, Janssen, Javelin, Lexicon, Eli Lilly and Company, Medtronic, Merck, Moderna, MyoKardia, NirvaMed, Novartis, Novo Nordisk, Owkin, Pfizer, PhaseBio, PLx Pharma, Recardio, Regeneron, Reid Hoffman Foundation, Roche, Sanofi, Stasys, Synaptic, The Medicines Company, Youngene, and 89Bio; royalties: Elsevier (Editor, Braunwald’s Heart Disease); site co-investigator: Abbott, Biotronik, Boston Scientific, CSI, Endotronix, St. Jude Medical (now Abbott), Philips, SpectraWAVE, Svelte, and Vascular Solutions; trustee: American College of Cardiology; and unfunded research: FlowCo and Takeda. I.R. discloses the following relationships—advisory board: AstraZeneca, Eli Lilly and Company, Merck Sharp & Dohme, Novo Nordisk, Inc., and Sanofi; consultant: AstraZeneca, Insuline Medical, Medial EarlySign Ltd, CamerEyes Ltd, Exscopia, Orgenesis Ltd, BOL, Glucome Ltd, DarioHealth, Diabot, and Concenter BioPharma; speakers bureau: AstraZeneca, Eli Lilly and Company, Merck Sharp & Dohme, Novo Nordisk, Inc., and Sanofi; and stock/shareholder: Glucome Ltd, Orgenesis Ltd, DarioHealth, CamerEyes Ltd, Diabot, and BOL. E.B. has received research grants through his institution from Astra Zeneca, Daiichi Sankyo, Merck, and Novartis, and consultancy fees from Amgen, Bristol-Myers Squibb, Boehringer Ingelheim/Lilly, Cardurion, and Verve. D.A.M. discloses the following relationships—consultant: Abbott Laboratories, InCardia, Inflammatix, Merck & Co, Novartis, and Roche Diagnostics. M.S.S. reports research grant support through Brigham and Women’s Hospital from Abbott, Amgen, Anthos Therapeutics, AstraZeneca, Boehringer Ingelheim, Daiichi-Sankyo, Eisai, Intarcia, Ionis, Merck, Novartis, and Pfizer, and consulting for Althera, Amgen, Anthos Therapeutics, AstraZeneca, Beren Therapeutics, Boehringer Ingelheim, Bristol-Myers Squibb, Dr. Reddy's Laboratories, Fibrogen, Intarcia, Merck, Moderna, Novo Nordisk, Precision BioSciences, and Silence Therapeutics. F.A.M., D.D.B., A.B., B.M.S., S.D.W., E.A.B., R.P.G., J.-G.P., E.B., D.A.M., and M.S.S. are members of the TIMI Study Group, which has received institutional research grant support through Brigham and Women’s Hospital from Abbott, Amgen, Anthos Therapeutics, ARCA Biopharma, Inc., AstraZeneca, Bayer HealthCare Pharmaceuticals, Inc., Daiichi-Sankyo, Eisai, Intarcia, Ionis Pharmaceuticals, Inc., Janssen Research and Development, LLC, MedImmune, Merck, Novartis, Pfizer, Quark Pharmaceuticals, Regeneron Pharmaceuticals, Inc., Roche, Siemens Healthcare Diagnostics, Inc., Softcell Medical Limited, The Medicines Company, and Zora Biosciences. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. F.A.M., B.M.S., S.D.W., D.A.M., and M.S.S. conceptualized and designed the study. F.A.M. and M.S.S. provided input on data analysis. A.B. analyzed the data. F.A.M. and M.S.S. drafted the manuscript. F.A.M., D.D.B, A.B., J.P.D., O.M., B.M.S., S.D.W., D.L.B., I.R., M.W.F., E.B., D.A.M., and M.S.S. were involved in data interpretation and revised the manuscript. All authors approved the final submitted version and agreed to be accountable for the report. All authors confirm that they had full access to all the data in the study and vouch for the integrity of the study. M.S.S. is the guarantor of this work and, as such, had full access to the data in the study and takes responsibility of the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented in abstract form at the American College of Cardiology (ACC) 72nd Annual Scientific Sessions/World Congress of Cardiology (WCC) 2023, New Orleans, LA, 4–6 March 2023.