The n e w e ng l a n d j o u r na l of m e dic i n e

Original Article

Empagliflozin after Acute

Myocardial Infarction
J. Butler, W.S. Jones, J.A. Udell, S.D. Anker, M.C. Petrie, J. Harrington,
M. Mattheus, I. Zwiener, O. Amir, M.C. Bahit, J. Bauersachs, A. Bayes‑Genis,
Y. Chen, V.K. Chopra, G. Figtree, J. Ge, S.G. Goodman, N. Gotcheva, S. Goto,
T. Gasior, W. Jamal, J.L. Januzzi, M.H. Jeong, Y. Lopatin, R.D. Lopes, B. Merkely,
P.B. Parikh, A. Parkhomenko, P. Ponikowski, X. Rossello, M. Schou, D. Simic,
P.G. Steg, J. Szachniewicz, P. van der Meer, D. Vinereanu, S. Zieroth,
M. Brueckmann, M. Sumin, D.L. Bhatt, and A.F. Hernandez​​

A BS T R AC T

BACKGROUND
Empagliflozin improves cardiovascular outcomes in patients with heart failure, The authors’ full names, academic de‑
patients with type 2 diabetes who are at high cardiovascular risk, and patients with grees, and affiliations are listed in the Ap‑
pendix. Dr. Butler can be contacted at
chronic kidney disease. The safety and efficacy of empagliflozin in patients who ­javed​.­butler@​­bswhealth​.­org or at Baylor
have had acute myocardial infarction are unknown. Scott and White Research Institute, 3434
Live Oak St., Dallas, TX 75204.
METHODS This article was published on April 6, 2024,
In this event-driven, double-blind, randomized, placebo-controlled trial, we assigned, at NEJM.org.
in a 1:1 ratio, patients who had been hospitalized for acute myocardial infarction DOI: 10.1056/NEJMoa2314051
and were at risk for heart failure to receive empagliflozin at a dose of 10 mg daily Copyright © 2024 Massachusetts Medical Society.
or placebo in addition to standard care within 14 days after admission. The primary
end point was a composite of hospitalization for heart failure or death from any cause
as assessed in a time-to-first-event analysis.
RESULTS
A total of 3260 patients were assigned to receive empagliflozin and 3262 to receive
placebo. During a median follow-up of 17.9 months, a first hospitalization for
heart failure or death from any cause occurred in 267 patients (8.2%) in the em-
pagliflozin group and in 298 patients (9.1%) in the placebo group, with incidence
rates of 5.9 and 6.6 events, respectively, per 100 patient-years (hazard ratio, 0.90;
95% confidence interval [CI], 0.76 to 1.06; P = 0.21). With respect to the individual
components of the primary end point, a first hospitalization for heart failure oc-
curred in 118 patients (3.6%) in the empagliflozin group and in 153 patients
(4.7%) in the placebo group (hazard ratio, 0.77; 95% CI, 0.60 to 0.98), and death
from any cause occurred in 169 (5.2%) and 178 (5.5%), respectively (hazard ratio,
0.96; 95% CI, 0.78 to 1.19). Adverse events were consistent with the known safety
profile of empagliflozin and were similar in the two trial groups.
CONCLUSIONS
Among patients at increased risk for heart failure after acute myocardial infarc-
tion, treatment with empagliflozin did not lead to a significantly lower risk of a first
hospitalization for heart failure or death from any cause than placebo. (Funded
by Boehringer Ingelheim and Eli Lilly; EMPACT-MI ClinicalTrials.gov number,
NCT04509674.)

n engl j med  nejm.org 1

The New England Journal of Medicine
Downloaded from nejm.org by DHIRAJ KUMAR on April 6, 2024. For personal use only.
No other uses without permission. Copyright © 2024 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

A
fter acute myocardial infarction, oversight on the development of the statistical
patients are at increased risk for heart analysis plan (available with the protocol), patient
failure and death, particularly if they pre­ recruitment and follow-up, and data analysis. An
sent with congestion or a decreased left ventricular independent data monitoring committee reviewed
ejection fraction.1-3 Treatment with sodium–glu- the safety data. Statistical analyses were performed
cose cotransporter 2 (SGLT2) inhibitors improves by employees of Boehringer Ingelheim with over-
cardiovascular outcomes in high-risk patients with sight by the executive committee, and key analy-
type 2 diabetes, those with chronic kidney dis- ses were verified by an independent statistician.
ease, and those with heart failure with a reduced The first author prepared the first draft of the
or preserved left ventricular ejection fraction.4 In submitted manuscript, which was reviewed and
the EMMY trial (Impact of Empagliflozin on Car- edited by all the authors. The first and last authors
diac Function and Biomarkers of Heart Failure in vouch for the accuracy and completeness of the
Patients with Acute Myocardial Infarction), pa- data and for the fidelity of the trial to the protocol
tients who received empagliflozin after an acute and the statistical analysis plan.
myocardial infarction had a reduced natriuretic
peptide concentration, an increased left ventric- Patients
ular ejection fraction, and a decreased cardiac Patients were men and women 18 years of age or
volume; however, this trial was not designed to older who had been hospitalized with an acute
assess clinical outcomes.5 The DAPA-MI trial myocardial infarction within 14 days before ran-
(Dapagliflozin Effects on Cardiometabolic Out- domization and had either evidence of a newly
comes in Patients with an Acute Heart Attack) was developed left ventricular ejection fraction of less
limited by the small number of clinical events dur- than 45% or signs or symptoms of congestion
ing the trial and therefore was unable to assess that resulted in treatment during the index hos-
the effects of SGLT2 inhibitor therapy after myo- pitalization (or both). Patients needed to have at
cardial infarction on rates of death or hospitaliza- least one additional enrichment factor (a clinical
tions for heart failure.6 factor that was known to be associated with hos-
Here, we report the results of the EMPACT-MI pitalization for heart failure or death from any
trial (Study to Evaluate the Effect of Empagliflozin cause), including an age of 65 years or older; a
on Hospitalization for Heart Failure and Mortality newly developed ejection fraction of less than
in Patients with Acute Myocardial Infarction), in 35%; a history of myocardial infarction, atrial fi-
which empagliflozin was compared with placebo brillation, or type 2 diabetes; an estimated glo-
with respect to the risk of hospitalization for merular filtration rate (GFR) of less than 60 ml
heart failure or death among patients with acute per minute per 1.73 m2 of body-surface area; an
myocardial infarction and either a new reduction elevated natriuretic peptide or uric acid level; an
in left ventricular ejection fraction or signs or elevated pulmonary artery or right ventricular sys-
symptoms of congestion (or both). tolic pressure; three-vessel coronary artery disease;
peripheral artery disease; or no revascularization
for the index myocardial infarction. Patients with
Me thods
a previous diagnosis of heart failure, as well as
Trial Oversight those who were taking or planning to take SGLT2
The EMPACT-MI trial was an international, event- inhibitors, were excluded. A full list of eligibility
driven, double-blind, randomized, placebo-con- criteria is provided in the Supplementary Appendix
trolled trial. The trial design has been described (available at NEJM.org) and was published pre-
previously.7 The trial was approved by the ethics viously.7
committee at each trial site, and all the patients
provided written informed consent. The trial spon- Trial Design
sors were Boehringer Ingelheim and Eli Lilly. The Patients who met the eligibility criteria were ran-
trial protocol (available with the full text of this domly assigned in a 1:1 ratio to receive empa-
article at NEJM.org) was developed and amended gliflozin at a dose of 10 mg daily or placebo in
by the executive and steering committees, which addition to standard care. Randomization was
included employees of Boehringer Ingelheim (who performed with the use of interactive response
represented the sponsors) and provided scientific technology and was stratified according to type

2 n engl j med  nejm.org

The New England Journal of Medicine

Downloaded from nejm.org by DHIRAJ KUMAR on April 6, 2024. For personal use only.
No other uses without permission. Copyright © 2024 Massachusetts Medical Society. All rights reserved.
 Empagliflozin after Acute Myocardial Infarction

2 diabetes status and geographic region (North received training in reviewing end-point events.
America, Latin America, Europe, or Asia). The Investigator-reported end-point events were veri-
EMPACT-MI trial had a streamlined design, with fied according to the algorithm described in the
the collection of essential data only, including Supplementary Appendix.
information about specific safety events, and
mainly remote follow-up of patients (by means of Statistical Analysis
a Web-based application or a telephone call) with In this event-driven trial, we estimated that 532
only a few face-to-face visits; the trial assessed patients with a primary end-point event would
investigator-reported end-point events rather than provide the trial with 85% power to detect a 23%
centrally adjudicated end-point events. lower risk of an event in the empagliflozin group
After randomization, patients had a remote than in the placebo group, with a two-sided type
visit at 2 weeks, a face-to-face visit at 6 months, I error of 0.05. The original protocol planned for
and remote visits every 6 months thereafter until the enrollment of 3312 patients, with an option
the end of the trial, when a final telephone call to increase enrollment to 5000 patients if the ac-
was performed. During these visits, data on pre- crual of events was slower than expected. The
specified end points, safety events, and adher- sample size was further increased to 6500. These
ence to the trial regimen were collected. Data on decisions were made on the basis of blinded trial
all concomitant medications were collected for data, with no change to the target number of
6 months after randomization; thereafter, med- events or revisions to effect-size projections or
ication data were collected only on open-label power calculations. No interim efficacy analyses
initiation of treatment with SGLT2 inhibitors or were performed.
combined treatment with SGLT1 and SGLT2 in- The analyses of the primary composite end
hibitors. Because of the established safety profile point and its components were performed accord-
of empagliflozin,8-10 we used focused safety re- ing to the intention-to-treat principle and included
porting, in which the investigators reported only all the patients who underwent randomization.
serious adverse events, adverse events that led to The differences between the empagliflozin and
discontinuation of the trial regimen for at least placebo groups in the risk of a primary end-point
7 consecutive days, and adverse events of special event were assessed with the use of a Cox pro-
interest. All the patients who underwent ran- portional-hazards model that included the base-
domization were followed for the duration of the line covariates of age, geographic region, estimat-
trial, regardless of whether they received empa- ed GFR (<45, 45 to <60, 60 to <90, or ≥90 ml per
gliflozin or placebo. minute per 1.73 m2 according to the Chronic
Kidney Disease Epidemiology Collaboration for-
Trial End Points mula), left ventricular ejection fraction (<35% or
The primary end point was a composite of hos- ≥35%), type 2 diabetes status, persistent or per-
pitalization for heart failure or death from any manent atrial fibrillation, previous myocardial
cause as assessed in a time-to-first-event analy- infarction, peripheral artery disease, and smoking
sis. The key secondary end points in the pre- status. Data for patients who did not have a pri-
specified hierarchical testing strategy were the mary end-point event were censored on the last
total number of hospitalizations for heart failure day they were known to have been free of the
or death from any cause, the total number of event, which may have been the last time point
nonelective cardiovascular hospitalizations or before the patient was lost to follow-up (under the
death from any cause, the total number of non- assumption of noninformative censoring).
elective hospitalizations for any cause or death A prespecified hierarchical testing procedure
from any cause, and the total number of hospi- was used, beginning with the primary end point
talizations for myocardial infarction or death from and then proceeding to the set of key secondary
any cause. Additional prespecified end points are end points. A Hochberg step-up procedure was
described in the Supplementary Appendix. In lieu used to assess the first and second key secondary
of central adjudication, end-point events were re- end points at the same level of hierarchy, with the
viewed and categorized according to prespecified next two key secondary end points subsequently
definitions by investigators at the trial sites who tested in a hierarchical manner.
were unaware of trial-group assignment and had All key secondary end points were analyzed

n engl j med  nejm.org 3

The New England Journal of Medicine
Downloaded from nejm.org by DHIRAJ KUMAR on April 6, 2024. For personal use only.
No other uses without permission. Copyright © 2024 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

with the use of a negative binomial regression The trial regimen was stopped prematurely
model that included the same covariates as the for reasons other than death in 684 patients
primary model and the logarithm of time as an (21.2%) in the empagliflozin group and in 716
adjustment for observation time. The observation patients (22.2%) in the placebo group. A total of
time started on the day of randomization and 436 patients (6.7%) started treatment with an
ended on the last day when information about end- open-label SGLT2 inhibitor during the trial, in-
point events was collected for an individual pa- cluding 201 (6.2%) in the empagliflozin group
tient, which may have been the last time point and 235 (7.2%) in the placebo group. A total of
before the patient was lost to follow-up. Post hoc 6328 patients (97.0%) were followed until the end
sensitivity analyses that accounted for the com- of the trial for the occurrence of a primary end-
peting risks of death from any cause and death point event, and 6467 patients (99.2%) had data
from cardiovascular causes were performed with on vital status available at the end of the trial
the use of Fine and Gray models for time to a first (Fig. S1). The median duration of follow-up was
hospitalization for heart failure, time to death 17.9 months, and the median duration of expo-
from cardiovascular causes, and time to a first sure to empagliflozin or placebo was similar in
hospitalization for heart failure or death from the two trial groups (Table S4). Adherence to the
cardiovascular causes. Safety analyses included trial regimen is shown in Table S5.
all the patients who received at least one dose of
empagliflozin or placebo. The confidence intervals End Points
for the secondary and exploratory end points were A primary end-point event — a first hospitaliza-
not adjusted for multiplicity and should be inter- tion for heart failure or death from any cause
preted as exploratory. — occurred in 267 of 3260 patients (8.2%) in the
empagliflozin group and in 298 of 3262 patients
(9.1%) in the placebo group, with incidence rates
R e sult s
of 5.9 and 6.6 events, respectively, per 100 pa-
Patients tient-years (hazard ratio, 0.90; 95% confidence
From December 2020 through March 2023, a interval [CI], 0.76 to 1.06; P = 0.21). With respect
total of 6610 patients at 451 sites in 22 countries to the individual components of the primary end
were screened, of whom 6522 were randomly as- point, a first hospitalization for heart failure oc-
signed to receive empagliflozin at a dose of 10 mg curred in 118 patients (3.6%) in the empagliflozin
daily (3260 patients) or placebo (3262 patients). group and in 153 patients (4.7%) in the placebo
The median time from admission to randomiza- group (hazard ratio, 0.77; 95% CI, 0.60 to 0.98),
tion was 5 days (interquartile range, 3 to 8). The and death from any cause occurred in 169 (5.2%)
characteristics of the patients at baseline were and 178 (5.5%), respectively (hazard ratio, 0.96;
similar in the two trial groups (Table 1).11 A total 95% CI, 0.78 to 1.19) (Table 2 and Fig. 1). Re-
of 78.4% of the patients had a left ventricular ejec- sults for the primary end point, a first hospital-
tion fraction of less than 45%, and 57.0% had ization for heart failure, and death from any
signs or symptoms of congestion that resulted in cause were consistent across subgroups (Fig. 2
treatment during the index hospitalization. Among and Figs. S2 and S3). Results for the primary end
the patients with signs or symptoms of conges- point were consistent across sensitivity analyses,
tion, 20.6% had a left ventricular ejection fraction which included additional categories of hospital-
of at least 45%. The most common enrichment ization for heart failure (Fig. S4). Causes of death
factors included an age of 65 years or older (in are shown in Table S6.
50.0% of the patients), type 2 diabetes (in 31.9%), Results for key secondary end points are shown
and three-vessel coronary artery disease (in 31.0%); in Table 2. The rate ratio (empagliflozin vs. pla-
70.5% of the patients had more than one enrich- cebo) was 0.87 (95% CI, 0.68 to 1.10) for the total
ment factor (Table S1 in the Supplementary Ap- number of hospitalizations for heart failure or
pendix). Nearly 75% of the patients who under- death from any cause, 0.92 (95% CI, 0.78 to 1.07)
went randomization presented with ST-segment for the total number of nonelective cardiovas-
elevation myocardial infarction (STEMI), and cular hospitalizations or death from any cause,
revascularization for the index myocardial in- 0.87 (95% CI, 0.77 to 1.0) for the total number
farction was performed in 89.3%. of nonelective hospitalizations for any cause or

4 n engl j med  nejm.org

The New England Journal of Medicine

Downloaded from nejm.org by DHIRAJ KUMAR on April 6, 2024. For personal use only.
No other uses without permission. Copyright © 2024 Massachusetts Medical Society. All rights reserved.
 Empagliflozin after Acute Myocardial Infarction

Table 1. Characteristics of the Patients at Baseline.*

Empagliflozin Group Placebo Group

Characteristic (N = 3260) (N = 3262)
Age — yr 63.6±11.0 63.7±10.8
Female sex — no. (%) 812 (24.9) 813 (24.9)
Race or ethnic group — no. (%)†
White 2730 (83.7) 2721 (83.4)
Black 44 (1.3) 48 (1.5)
Asian 421 (12.9) 413 (12.7)
Other or missing 65 (2.0) 80 (2.5)
Geographic region — no. (%)
North America 431 (13.2) 433 (13.3)
Latin America 290 (8.9) 288 (8.8)
Europe 2153 (66.0) 2154 (66.0)
Asia 386 (11.8) 387 (11.9)
Index myocardial infarction type — no. (%)‡
STEMI 2444 (75.0) 2401 (73.6)
NSTEMI 814 (25.0) 861 (26.4)
Revascularization for the index myocardial infarction — no. (%) 2911 (89.3) 2911 (89.2)
Thrombolytic therapy for the index myocardial infarction — no. (%) 345 (10.6) 355 (10.9)
Lowest left ventricular ejection fraction — no. (%)§
<25% 126 (3.9) 126 (3.9)
≥25 to <35% 721 (22.1) 699 (21.4)
≥35 to <45% 1724 (52.9) 1716 (52.6)
≥45 to <55% 438 (13.4) 468 (14.3)
≥55% 227 (7.0) 225 (6.9)
Signs or symptoms of congestion that resulted in treatment — no. (%)
Overall 1852 (56.8) 1863 (57.1)
Lowest left ventricular ejection fraction of <45%¶ 1172 (36.0) 1151 (35.3)
Lowest left ventricular ejection fraction of ≥45%¶ 657 (20.2) 684 (21.0)
Cardiovascular disease history and risk factors — no. (%)
Previous myocardial infarction 388 (11.9) 459 (14.1)
Atrial fibrillation 358 (11.0) 361 (11.1)
Diabetes mellitus type 2 1035 (31.7) 1046 (32.1)
Hypertension 2262 (69.4) 2276 (69.8)
Peripheral artery disease 172 (5.3) 180 (5.5)

* Plus–minus values are means ±SD. Percentages may not sum to 100 because of rounding. The trial groups are com‑
posed of patients hospitalized for acute myocardial infarction and at risk for heart failure who were randomly assigned
within 14 days after admission to receive empagliflozin at a dose of 10 mg daily or placebo in addition to standard care.
NSTEMI denotes non–ST-segment elevation myocardial infarction, and STEMI ST-segment elevation myocardial infarc‑
tion.
† Race and ethnic group were reported by the patients; “other” includes patients who reported mixed race. Data were
missing for 56 patients in the empagliflozin group and 73 patients in the placebo group.
‡ Data on index myocardial infarction type were missing for 2 patients in the empagliflozin group.
§ Data on lowest left ventricular ejection fraction were missing for 24 patients in the empagliflozin group and 28 patients
in the placebo group.
¶ Data on lowest left ventricular ejection fraction (<45% or ≥45%) among patients with signs or symptoms of congestion
that resulted in treatment were missing for 23 patients in the empagliflozin group.

n engl j med  nejm.org 5

The New England Journal of Medicine
Downloaded from nejm.org by DHIRAJ KUMAR on April 6, 2024. For personal use only.
No other uses without permission. Copyright © 2024 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 2. Primary, Secondary, and Other End Points.

Empagliflozin Group Placebo Group Effect

End Point (N = 3260) (N = 3262) (95% CI)*

Value Incidence Value Incidence

no. of events per no. of events per

100 patient-yr† 100 patient-yr†
Primary composite end point
A first hospitalization for heart failure or 267 (8.2) 5.9 298 (9.1) 6.6 0.90 (0.76–1.06)‡
death from any cause — no. (%)
A first hospitalization for heart failure 118 (3.6) 2.6 153 (4.7) 3.4 0.77 (0.60–0.98)
Death from any cause 169 (5.2) 3.6 178 (5.5) 3.8 0.96 (0.78–1.19)
Key secondary end points§
Total no. of hospitalizations for heart failure 317 7.1¶ 385 8.3¶ 0.87 (0.68–1.10)‖
or death from any cause
Total no. of nonelective cardiovascular hos‑ 666 15.5¶ 730 16.9¶ 0.92 (0.78–1.07)‖
pitalizations or death from any cause
Total no. of nonelective hospitalizations for 998 23.0¶ 1138 26.3¶ 0.87 (0.77–1.0)‖
any cause or death from any cause
Total no. of hospitalizations for myocardial 276 6.7¶ 274 6.3¶ 1.06 (0.83–1.35)‖
infarction or death from any cause
Other secondary and prespecified explor-
atory end points
Death from cardiovascular causes — no. (%) 132 (4.0) 2.8 131 (4.0) 2.8 1.03 (0.81–1.31)
A first hospitalization for heart failure or 231 (7.1) 5.1 259 (7.9) 5.7 0.90 (0.75–1.07)
death from cardiovascular causes
— no. (%)
Total no. of hospitalizations for heart failure 148 2.4¶ 207 3.6¶ 0.67 (0.51–0.89)‖

* The effects are presented as hazard ratios estimated with the use of a Cox proportional-hazards model unless indicated otherwise. The con‑
fidence intervals for secondary and exploratory outcomes were not adjusted for multiplicity and should be interpreted as exploratory.
† Values are the number of patients with an end-point event per 100 patient-years (as calculated by dividing the number of patients with at
least one event by the time at risk and multiplying the quotient by 100), unless indicated otherwise.
‡ P = 0.21 for the comparison of the empagliflozin group with the placebo group.
§ The key secondary end points were assessed as part of a hierarchical confirmatory testing procedure.
¶ Shown is the adjusted number of events per 100 patient-years, as calculated with the use of negative binomial regression analysis.
‖ The effect is presented as the rate ratio estimated with the use of negative binomial regression analysis.

death from any cause, and 1.06 (95% CI, 0.83 group, with a rate of 2.4 and 3.6 events, respec-
to 1.35) for the total number of hospitalizations tively, per 100 patient-years (rate ratio, 0.67; 95%
for myocardial infarction or death from any CI, 0.51 to 0.89). Results of sensitivity analyses
cause. that accounted for the competing risks of death
With respect to exploratory end points, death from any cause or death from noncardiovascular
from cardiovascular causes occurred in 132 pa- causes were consistent with those from Cox regres-
tients (4.0%) in the empagliflozin group and in sion models (data not shown).
131 patients (4.0%) in the placebo group (hazard
ratio, 1.03; 95% CI, 0.81 to 1.31). The time to Safety
death from cardiovascular causes and the time A similar percentage of patients in the two trial
to a first hospitalization for heart failure or death groups had a serious adverse event (23.7% in the
from cardiovascular causes were similar in the empagliflozin group and 24.7% in the placebo
two trial groups (Figs. S5 and S6). The total num- group) or an adverse event that resulted in perma-
ber of hospitalizations for heart failure was 148 in nent discontinuation of the trial regimen (3.8% in
the empagliflozin group and 207 in the placebo each group) (Table 3). Contrast-induced acute kid-

6 n engl j med  nejm.org

The New England Journal of Medicine

Downloaded from nejm.org by DHIRAJ KUMAR on April 6, 2024. For personal use only.
No other uses without permission. Copyright © 2024 Massachusetts Medical Society. All rights reserved.
 Empagliflozin after Acute Myocardial Infarction

Figure 1. Kaplan–Meier Estimates and Cumulative A First Hospitalization for Heart Failure or Death from Any Cause
Incidence Function for the Composite Primary End 12
100 Hazard ratio, 0.90 (95% CI, 0.76–1.06)
Point and Its Components.
P=0.21
Shown are the time to a first hospitalization for heart 90 10
Placebo
failure or death from any cause (composite primary 80

Percentage with Event

8
end point; Panel A), the time to a first hospitalization 70 Empagliflozin
for heart failure (Panel B), and the time to death from 6
60
any cause (Panel C). Patients hospitalized for acute
50 4
myocardial infarction and at risk for heart failure were
randomly assigned within 14 days after admission to 40
2
receive empagliflozin at a dose of 10 mg daily or placebo 30
in addition to standard care. The insets show the same 20 0
0 90 180 270 360 450 540 630 720 810
data on an expanded y axis. 10
0
0 90 180 270 360 450 540 630 720 810
ney injury occurred in 8 patients (0.2%) in the Trial Day
empagliflozin group and in 9 patients (0.3%) in
No. at Risk
the placebo group. Placebo 3262 3092 3044 2832 2486 2071 1556 1040 551 137
Empagliflozin 3260 3111 3060 2881 2532 2107 1566 1048 531 134

Discussion B First Hospitalization for Heart Failure

6
In the EMPACT-MI trial, empagliflozin treat- 100 Hazard ratio, 0.77 (95% CI, 0.60–0.98)

Estimated Cumulative Incidence (%)

ment did not lead to a significantly lower risk of 90 5
Placebo
a composite primary end-point event — a first 80 4
hospitalization for heart failure or death from 70
3 Empagliflozin
any cause — than placebo among patients pre- 60
senting with an acute myocardial infarction and 50 2
an increased risk of heart failure. The rates of 40
1
prespecified key secondary end-point events did 30
0
not differ substantially in the two trial groups. 20
0 90 180 270 360 450 540 630 720 810
Recently, the DAPA-MI trial, which excluded 10
patients with diabetes, did not show a lower risk 0
0 90 180 270 360 450 540 630 720 810
of death from cardiovascular causes or hospital-
Trial Day
ization for heart failure with dapagliflozin thera-
py than with placebo after acute myocardial in- No. at Risk
Placebo 3262 3092 3044 2832 2486 2071 1556 1040 551 137
farction.6 The prespecified composite primary Empagliflozin 3260 3111 3060 2881 2532 2107 1566 1048 531 134
end point in the DAPA-MI trial was changed to
a seven-level win ratio. The actual numbers of C Death from Any Cause
heart-failure events or deaths were too few to 100
7
Hazard ratio, 0.96 (95% CI, 0.78–1.19)
allow for any meaningful conclusion.12 The data 90
6
Placebo
from the EMPACT-MI trial help fill the gap in 80 5
Percentage with Event

knowledge about the effect of SGLT2 inhibitors 70 4

Empagliflozin
in patients after acute myocardial infarction. 60 3
Certain factors may have contributed to the 50
2
lack of an effect of empagliflozin on the primary 40
composite end point in the EMPACT-MI trial. 30
1

Deaths from any cause composed 52% of the 20 0

0 90 180 270 360 450 540 630 720 810
primary end-point events and occurred in a simi- 10
lar percentage of patients in the two trial groups. 0
By design, we enrolled patients soon after acute 0 90 180 270 360 450 540 630 720 810

myocardial infarction, a time when several Trial Day

mechanisms that may not be amenable to mod- No. at Risk
Placebo 3262 3186 3159 2975 2632 2207 1660 1111 593 148
ification with SGLT2 inhibition, which include Empagliflozin 3260 3177 3148 2995 2639 2218 1658 1119 572 153
cardiac causes (e.g., stent thrombosis, recurrent

n engl j med  nejm.org 7

The New England Journal of Medicine
Downloaded from nejm.org by DHIRAJ KUMAR on April 6, 2024. For personal use only.
No other uses without permission. Copyright © 2024 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Subgroup Empagliflozin Placebo Hazard Ratio (95% CI)

no. of patients with event/total no. (%)
Overall 267/3260 (8.2) 298/3262 (9.1) 0.90 (0.76–1.06)
Age
<65 yr 92/1639 (5.6) 102/1623 (6.3) 0.88 (0.66–1.16)
≥65 yr 175/1621 (10.8) 196/1639 (12.0) 0.92 (0.75–1.13)
Sex
Male 176/2448 (7.2) 219/2449 (8.9) 0.81 (0.66–0.99)
Female 91/812 (11.2) 79/813 (9.7) 1.14 (0.85–1.55)
Geographic region
North America 34/431 (7.9) 42/433 (9.7) 0.91 (0.58–1.43)
Latin America 38/290 (13.1) 52/288 (18.1) 0.74 (0.49–1.13)
Europe 179/2153 (8.3) 177/2154 (8.2) 0.99 (0.80–1.22)
Asia 16/386 (4.1) 27/387 (7.0) 0.58 (0.31–1.08)
Ethnic group
Not Hispanic or Latinx 223/2866 (7.8) 237/2859 (8.3) 0.94 (0.78–1.13)
Hispanic or Latinx 40/338 (11.8) 55/331 (16.6) 0.73 (0.48–1.10)
Race
White 236/2730 (8.6) 249/2721 (9.2) 0.94 (0.79–1.13)
Black 7/44 (15.9) 11/48 (22.9) 0.86 (0.33–2.23)
Asian 20/421 (4.8) 30/413 (7.3) 0.64 (0.36–1.13)
Other, including mixed race 0/9 2/7 (28.6) NC
Median time from index myocardial infarction diagnosis
to randomization
≤5 days 140/1870 (7.5) 163/1915 (8.5) 0.86 (0.68–1.08)
>5 days 127/1388 (9.1) 135/1347 (10.0) 0.95 (0.74–1.21)
Index myocardial infarction type
STEMI 194/2444 (7.9) 195/2401 (8.1) 0.96 (0.79–1.18)
NSTEMI 73/814 (9.0) 103/861 (12.0) 0.77 (0.57–1.04)
Type 2 diabetes at baseline
No 159/2225 (7.1) 175/2216 (7.9) 0.85 (0.69–1.06)
Yes 108/1035 (10.4) 123/1046 (11.8) 0.97 (0.75–1.25)
History of myocardial infarction
No 225/2872 (7.8) 239/2803 (8.5) 0.91 (0.76–1.10)
Yes 42/388 (10.8) 59/459 (12.9) 0.83 (0.56–1.24)
Baseline estimated GFR
≥60 ml/min/1.73 m2 156/2540 (6.1) 171/2524 (6.8) 0.89 (0.72–1.11)
<60 ml/min/1.73 m2 111/720 (15.4) 127/738 (17.2) 0.92 (0.71–1.18)
Baseline systolic blood pressure
<110 mm Hg 73/719 (10.2) 82/723 (11.3) 0.91 (0.66–1.24)
110 to <130 mm Hg 115/1605 (7.2) 136/1570 (8.7) 0.84 (0.65–1.08)
≥130 mm Hg 79/935 (8.4) 80/969 (8.3) 1.00 (0.73–1.37)
Baseline LVEF
<35% 101/705 (14.3) 109/688 (15.8) 0.93 (0.71–1.22)
≥35% 163/2531 (6.4) 188/2546 (7.4) 0.87 (0.71–1.07)
Lowest LVEF during index hospitalization
<45% 222/2571 (8.6) 243/2541 (9.6) 0.92 (0.76–1.10)
≥45% 42/665 (6.3) 54/693 (7.8) 0.78 (0.52–1.17)
Signs or symptoms of congestion that resulted in treatment
during index hospitalization
No 79/1408 (5.6) 89/1399 (6.4) 0.84 (0.62–1.14)
Yes 188/1852 (10.2) 209/1863 (11.2) 0.93 (0.77–1.14)
Lowest LVEF and signs or symptoms of congestion that resulted
in treatment during index hospitalization
Lowest LVEF of <45% and congestion 143/1172 (12.2) 154/1151 (13.4) 0.97 (0.77–1.22)
Lowest LVEF of <45% and no congestion 79/1399 (5.6) 89/1390 (6.4) 0.85 (0.63–1.15)
Lowest LVEF of ≥45% 42/665 (6.3) 54/693 (7.8) 0.78 (0.52–1.17)
Baseline use of loop or high-ceiling diuretics
No 118/2126 (5.6) 149/2184 (6.8) 0.84 (0.66–1.06)
Yes 149/1134 (13.1) 149/1078 (13.8) 0.93 (0.74–1.17)
Baseline use of beta-blocker
No 64/745 (8.6) 74/728 (10.2) 0.85 (0.61–1.18)
Yes 203/2515 (8.1) 224/2534 (8.8) 0.92 (0.76–1.11)
Baseline use of mineralocorticoid-receptor antagonist
No 149/1992 (7.5) 165/1957 (8.4) 0.87 (0.69–1.08)
Yes 118/1268 (9.3) 133/1305 (10.2) 0.95 (0.74–1.22)
Baseline use of ACE inhibitor, ARB, or ARNI
No 86/865 (9.9) 106/931 (11.4) 0.88 (0.66–1.17)
Yes 181/2395 (7.6) 192/2331 (8.2) 0.92 (0.75–1.13)
0.25 1.00 4.00

Empagliflozin Better Placebo Better

8 n engl j med  nejm.org

The New England Journal of Medicine

Downloaded from nejm.org by DHIRAJ KUMAR on April 6, 2024. For personal use only.
No other uses without permission. Copyright © 2024 Massachusetts Medical Society. All rights reserved.
 Empagliflozin after Acute Myocardial Infarction

Figure 2 (facing page). Composite Primary End Point, gliflozin or placebo in the DAPA-HF trial (Dapa-
According to Prespecified Subgroups. gliflozin and Prevention of Adverse Outcomes in
Shown is the risk of a first hospitalization for heart Heart Failure), 549 were hospitalized for heart
failure or death from any cause in the trial groups. Eth‑ failure, 33 had a heart-failure event that resulted
nic group and race were reported by the patient. The in the receipt of intravenous diuretic therapy in
glomerular filtration rate (GFR) was estimated with the
the outpatient setting, and 604 had a worsening
use of the Chronic Kidney Disease Epidemiology Col‑
laboration formula. The size of the boxes indicates the heart-failure event that resulted in the initiation
size of the subgroup. The confidence intervals were or intensification of oral diuretic therapy in the
not adjusted for multiplicity and should be interpreted outpatient setting.21 Whether the inclusion of a
as exploratory. ACE denotes angiotensin-converting broader measure of the burden of heart failure
enzyme, ARB angiotensin-receptor blocker, ARNI an‑
in the primary end point would have affected the
giotensin receptor–neprilysin inhibitor, LVEF left ven‑
tricular ejection fraction, NC not calculated, NSTEMI results of our trial is unclear.
non–ST-segment elevation myocardial infarction, and In our trial, some of the patients with a lower
STEMI ST-segment elevation myocardial infarction. left ventricular ejection fraction or congestion at
the time of randomization may have had a stunned
myocardium that was reversible; further improve-
myocardial infarction, mechanical complications, ment after revascularization is unlikely in this
and scar-related ventricular arrhythmias) and non- lower-risk population.3,22,23 This might be the
cardiac causes within the first 30 days, contribute case particularly in patients with STEMI, who
to mortality.13 composed nearly 75% of the patients in the
As in our trial, the sample size in the EMPACT-MI trial, in which approximately 90%
PARADISE-MI trial (Prospective ARNI versus of the patients underwent early revascularization.
ACE Inhibitor Trial to Determine Superiority in Previous trials involving patients with estab-
Reducing Heart Failure Events after Myocardial lished heart failure or with type 2 diabetes and
Infarction) was increased because of low rates of atherosclerotic cardiovascular disease have shown
primary end-point events — 6.7 and 7.4 events reductions of 29 to 35% in the relative risk of
per 100 patient-years in the valsartan–sacubitril hospitalization for heart failure among patients
and ramipril groups, respectively.14,15 These rates treated with SGLT2 inhibitors as compared with
and the rates in our trial are lower than those patients who received placebo.8-10,24,25 The find-
observed in previous trials and observational ings of our exploratory analyses of a first hospi-
studies.16,17 The reasons for this may be related to talization for heart failure and the total number
multiple factors, including the widespread use of of hospitalizations for heart failure in the empa-
medical therapies, timely access to revasculariza- gliflozin group as compared with the placebo
tion after myocardial infarction, and the coronavi- group appear to be consistent with the results of
rus disease 2019 (Covid-19) pandemic, as well as these previous trials, and further study of the
regional wars in the case of the EMPACT-MI trial.18 effects of SLGT2 inhibitors on heart-failure out-
The number and percentage of heart-failure comes in high-risk patients after myocardial in-
events that contributed to the primary end point farction may be warranted.
may have been affected by several factors. Our We did not observe evidence of increased
trial was conducted during the Covid-19 pan- rates of serious adverse events, adverse events that
demic, when the number of hospitalizations for resulted in permanent discontinuation of the trial
heart failure decreased substantially.19 Patients regimen, or adverse events of special interest.
with less severe symptoms may not have sought The data from EMPACT-MI trial further build on
care or may have been treated in the outpatient the safety profile of SGLT2 inhibitors in patients
setting. In addition, two of the regions where across the spectrum of cardiovascular risks and
our trial was conducted were affected by war.20 provide evidence for the safety of these agents in
Heart-failure events other than hospitalization hospitalized patients.8,26,27
were not included in the primary end point. In Our trial has limitations. The end-point events
some other trials, outpatient heart-failure events were not centrally adjudicated but were assessed
have contributed meaningfully to the total bur- by site investigators according to prespecified defi-
den of heart-failure events. For example, of the nitions. Outpatient heart-failure events were not
4744 patients randomly assigned to receive dapa- analyzed as clinical end points. Despite our at-

n engl j med  nejm.org 9

The New England Journal of Medicine
Downloaded from nejm.org by DHIRAJ KUMAR on April 6, 2024. For personal use only.
No other uses without permission. Copyright © 2024 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 3. Adverse Events in the Safety Population.*

Empagliflozin Group Placebo Group

Event (N = 3234) (N = 3229)

Value Incidence Value Incidence

no. of events per no. of events per

100 patient-yr 100 patient-yr
Any adverse event — no. (%) 891 (27.6) 25.37 883 (27.3) 25.38
Serious adverse event — no. (%) 765 (23.7) 21.43 798 (24.7) 22.69
Adverse event that led to permanent discontinuation of 122 (3.8) 2.93 122 (3.8) 2.96
empagliflozin or placebo — no. (%)
Adverse events of special interest
Ketoacidosis
Overall — no. (%) 2 (0.1) 0.05 1 (<0.1) 0.02
Type 2 diabetes present at baseline — no./total no. (%) 2/1024 (0.2) 0.16 1/1032 (0.1) 0.08
Type 2 diabetes absent at baseline — no./total no. (%) 0/2210 0 0/2197 0
Adverse event that led to lower-limb amputation — no. (%) 9 (0.3) 0.19 5 (0.2) 0.11
Hepatic injury — no. (%)† 8 (0.2) 0.19 2 (0.1) 0.05
Contrast-induced kidney injury — no. (%) 8 (0.2) 0.19 9 (0.3) 0.22
Other relevant adverse events
Acute renal failure — no. (%)‡ 43 (1.3) 1.04 59 (1.8) 1.44
Acute kidney injury — no. (%)§ 27 (0.8) 0.65 43 (1.3) 1.05
Volume depletion — no. (%)¶ 35 (1.1) 0.84 40 (1.2) 0.98
Hypoglycemia‡
Overall — no. (%) 4 (0.1) 0.10 5 (0.2) 0.12
Type 2 diabetes present at baseline — no./total no. (%) 4/1024 (0.4) 0.31 5/1032 (0.5) 0.39
Type 2 diabetes absent at baseline — no./total no. (%) 0/2210 0 0/2197 0

* Patients who received at least one dose of empagliflozin or placebo were included in the safety population. Shown are adverse events ana‑
lyzed up to 7 days after the discontinuation of the trial regimen, except for lower-limb amputations, which were analyzed up to the end of
the trial. Adverse events that were to be reported in the trial included serious adverse events, adverse events that led to discontinuation of
the trial regimen for at least 7 days, and adverse events of special interest, defined as ketoacidosis, adverse events leading to lower-limb
amputation, hepatic injury, and contrast-induced kidney injury.
† Hepatic injury was defined as an aspartate aminotransferase (AST) level or an alanine aminotransferase level (ALT) (or both) of at least 3
times the upper limit of the normal range, combined with a total bilirubin level of at least 2 times the upper limit of the normal range, as
measured in the same blood sample or in blood samples obtained within 30 days of each other; or an ALT level or AST level (or both) of at
least 10 times the upper limit of the normal range. Hepatic injury as defined by these criteria and reported by the investigator occurred in
three patients in the empagliflozin group and in no patients in the placebo group.
‡ Events were identified with the use of a standardized Medical Dictionary for Regulatory Activities (MedDRA), version 26.1, query.
§ “Acute kidney injury” is a MedDRA, version 26.1, preferred term.
¶ Events were identified with the use of a Boehringer Ingelheim–customized MedDRA, version 26.1, query.

tempts to improve the representation of women, death from any cause — in patients with acute
older adults, and historically underrepresented ra- myocardial infarction who were at increased risk
cial and ethnic minorities in this trial, their repre- for heart failure.
sentation remained suboptimal. Some patients in Supported by Boehringer Ingelheim and Eli Lilly.
these groups are at increased risk for heart fail- Disclosure forms provided by the authors are available with
ure after myocardial infarction, and further work the full text of this article at NEJM.org.
A data sharing statement provided by the authors is available
is needed to improve their representation (Table with the full text of this article at NEJM.org.
S3). Only focused safety data were collected.
1,28
We thank the patients for participating in the trial and
In the current trial, empagliflozin did not re- the investigators and staff at the trial sites, the members of
EMPACT-MI trial committees, and the trial personnel from
duce the risk of the composite primary end-point the Duke Clinical Research Institute, Fortrea, and Boehringer
event — a first hospitalization for heart failure or Ingelheim.

10 n engl j med  nejm.org

The New England Journal of Medicine

Downloaded from nejm.org by DHIRAJ KUMAR on April 6, 2024. For personal use only.
No other uses without permission. Copyright © 2024 Massachusetts Medical Society. All rights reserved.
 Empagliflozin after Acute Myocardial Infarction

Appendix
The authors’ full names and academic degrees are as follows: Javed Butler, M.D., M.P.H., W. Schuyler Jones, M.D., Jacob A. Udell, M.D.,
M.P.H., Stefan D. Anker, M.D., Ph.D., Mark C. Petrie, M.D., Josephine Harrington, M.D., Michaela Mattheus, Dipl.Biomath., Isabella
Zwiener, Ph.D., Offer Amir, M.D., M. Cecilia Bahit, M.D., Johann Bauersachs, M.D., Antoni Bayes‑Genis, M.D., Yundai Chen, M.D.,
Vijay K. Chopra, M.D., Gemma Figtree, M.D., Junbo Ge, M.D., Shaun G. Goodman, M.D., Nina Gotcheva, M.D., Shinya Goto, M.D.,
Tomasz Gasior, M.D., Waheed Jamal, M.D., James L. Januzzi, M.D., Myung Ho Jeong, M.D., Yuri Lopatin, M.D., Renato D. Lopes,
M.D., Ph.D., Béla Merkely, M.D., Puja B. Parikh, M.D., M.P.H., Alexander Parkhomenko, M.D., Piotr Ponikowski, M.D., Xavier Ros-
sello, M.D., Morten Schou, M.D., Dragan Simic, M.D., P. Gabriel Steg, M.D., Joanna Szachniewicz, M.D., Peter van der Meer, M.D.,
Dragos Vinereanu, M.D., Shelley Zieroth, M.D., Martina Brueckmann, M.D., Mikhail Sumin, M.D., Deepak L. Bhatt, M.D., M.P.H., and
Adrian F. Hernandez, M.D.
The authors’ affiliations are as follows: Baylor Scott and White Research Institute, Dallas (J. Butler); the Department of Medicine,
University of Mississippi, Jackson (J. Butler); the Division of Cardiology, Department of Medicine, Duke University Medical Center
(W.S.J., J.H., A.F.H.), and the Duke Clinical Research Institute (R.D.L.) — both in Durham, NC; Women’s College Hospital (J.A.U.),
the Peter Munk Cardiac Centre, University Health Network, University of Toronto (J.A.U., S.G.G.), and the Division of Cardiology,
Department of Medicine, St. Michael’s Hospital, Unity Health Toronto (S.G.G.), Toronto, the Canadian VIGOUR Centre, University of
Alberta, Edmonton (S.G.G.), and the Section of Cardiology, Max Rady College of Medicine, University of Manitoba, Winnipeg (S.Z.)
— all in Canada; the Department of Cardiology, German Heart Center Charité, Berlin Institute of Health Center for Regenerative
Therapies, German Center for Cardiovascular Research Partner Site Berlin, Charité Universitätsmedizin, Berlin (S.D.A.), Boehringer
Ingelheim Pharma (M.M., I.Z.) and Boehringer Ingelheim International (T.G., W.J., M.B., M. Sumin), Ingelheim, the Department of
Cardiology and Angiology, Hannover Medical School, Hannover (J. Bauersachs), and the First Department of Medicine, Faculty of
Medicine Mannheim, University of Heidelberg, Mannheim (M.B.) — all in Germany; the School of Cardiovascular and Medical Sci-
ences, British Heart Foundation Glasgow Cardiovascular Research Centre, University of Glasgow, Glasgow, United Kingdom (M.C.P.);
the Heart Institute, Hadassah Medical Center, Hebrew University of Jerusalem, Jerusalem (O.A.); Instituto de Neurología (INECO)
Neurociencias Oroño, Fundación INECO, Rosario, Argentina (M.C.B.); the Heart Institute, Hospital Universitari Germans Trias i Pujol,
and the Department of Medicine, Universitat Autònomoa de Barcelona, Barcelona (A.B.-G.), and Son Espases University Hospital,
Health Research Institute of the Balearic Islands, University of the Balearic Islands, Palma de Mallorca (X.R.) — all in Spain; the Depart-
ment of Cardiology, First Medical Center of Chinese People’s Liberation Army General Hospital, Beijing (Y.C.), and the Department of
Cardiology, Zhongshan Hospital, Fudan University, and the Shanghai Institute of Cardiovascular Diseases, National Clinical Research
Center for Interventional Medicine, Shanghai (J.G.) — both in China; the Department of Cardiology, Medanta, Gurgaon, India (V.K.C.);
the Faculty of Medicine and Health, University of Sydney, Sydney (G.F.); the Department of Cardiology, National Cardiology Hospital,
Sofia, Bulgaria (N.G.); Tokai University School of Medicine, Isehara, Japan (S.G.); the Division of Cardiology, Harvard Medical School
and Massachusetts General Hospital, Boston (J.L.J.); Chonnam National University Hospital and Medical School, Gwangju, South Korea
(M.H.J.); Volgograd State Medical University, Volgograd, Russia (Y.L.); Heart and Vascular Center, Semmelweis University, Budapest,
Hungary (B.M.); the Division of Cardiovascular Medicine, Department of Medicine, State University of New York at Stony Brook, Stony
Brook (P.B.P.), and Mount Sinai Fuster Heart Hospital, Icahn School of Medicine at Mount Sinai, New York (D.L.B.) — both in New
York; the Ukrainian Institute of Cardiology M.D. Strazhesko, National Academy of Medical Sciences, Kyiv, Ukraine (A.P.); Collegium
Medicum–Faculty of Medicine, WSB University, Dąbrowa Górnicza (T.G.), and the Institute for Heart Diseases Medical University (P.P.)
and Jan Mikulicz-Radecki University Clinical Hospital (J.S.), Wrocław — all in Poland; the Department of Cardiology, Herlev and Gen-
tofte University Hospital, Copenhagen (M. Schou); the Department of Cardiovascular Diseases, University Clinical Center Belgrade,
Belgrade, Serbia (D.S.); Université Paris–Cité, French Alliance for Cardiovascular Trials, INSERM Unité 1148, Assistance Publique–
Hôpitaux de Paris, Hôpital Bichat, Paris (P.G.S.); the Department of Cardiology, University of Groningen, University Medical Center
Groningen, Groningen, the Netherlands (P.M.); and University and Emergency Hospital of Bucharest, Bucharest, Romania (D.V.).

References
1. Harrington J, Jones WS, Udell JA, et dial infarction: the EMMY trial. Eur Heart 11. Harrington J, Udell JA, Jones WS, et
al. Acute decompensated heart failure in J 2022;​43:​4421-32. al. Baseline characteristics of patients en-
the setting of acute coronary syndrome. 6. James S, Erlinge D, Storey RF, et al. rolled in the EMPACT-MI trial. Eur J Heart
JACC Heart Fail 2022;​10:​404-14. Dapagliflozin in myocardial infarction Fail 2023;​25:​1708-15.
2. Steg PG, Dabbous OH, Feldman LJ, et without diabetes or heart failure. NEJM Evid 12. James S, Erlinge D, Storey RF, et al.
al. Determinants and prognostic impact 2024;​3(2). Rationale and design of the DAPA-MI tri-
of heart failure complicating acute coro- 7. Harrington J, Udell JA, Jones WS, et al. al: dapagliflozin in patients without dia-
nary syndromes: observations from the Empagliflozin in patients post myocardial betes mellitus with acute myocardial in-
Global Registry of Acute Coronary Events infarction rationale and design of the farction. Am Heart J 2023;​266:​188-97.
(GRACE). Circulation 2004;​109:​494-9. EMPACT-MI trial. Am Heart J 2022;​253:​ 13. Zaman S, Kovoor P. Sudden cardiac
3. Frantz S, Hundertmark MJ, Schulz- 86-98. death early after myocardial infarction:
Menger J, Bengel FM, Bauersachs J. Left 8. Packer M, Anker SD, Butler J, et al. pathogenesis, risk stratification, and pri-
ventricular remodelling post-myocardial Cardiovascular and renal outcomes with mary prevention. Circulation 2014;​ 129:​
infarction: pathophysiology, imaging, empagliflozin in heart failure. N Engl J 2426-35.
and novel therapies. Eur Heart J 2022;​43:​ Med 2020;​383:​1413-24. 14. Pfeffer MA, Claggett B, Lewis EF, et
2549-61. 9. Zinman B, Wanner C, Lachin JM, et al. Angiotensin receptor-neprilysin inhi-
4. Usman MS, Siddiqi TJ, Anker SD, et al. Empagliflozin, cardiovascular out- bition in acute myocardial infarction.
al. Effect of SGLT2 inhibitors on cardio- comes, and mortality in type 2 diabetes. N Engl J Med 2021;​385:​1845-55.
vascular outcomes across various patient N Engl J Med 2015;​373:​2117-28. 15. Pfeffer MA. PARADISE MI presenta-
populations. J Am Coll Cardiol 2023;​81:​ 10. Anker SD, Butler J, Filippatos G, et al. tion slides. In:​Proceedings and Abstracts
2377-87. Empagliflozin in heart failure with a pre- of the American College of Cardiology
5. von Lewinski D, Kolesnik E, Tripolt served ejection fraction. N Engl J Med Annual Scientific Sessions, May 15–17,
NJ, et al. Empagliflozin in acute myocar- 2021;​385:​1451-61. 2021. Virtual:​American College of Cardi-

n engl j med  nejm.org 11

The New England Journal of Medicine
Downloaded from nejm.org by DHIRAJ KUMAR on April 6, 2024. For personal use only.
No other uses without permission. Copyright © 2024 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

ology, 2021 (https://www​.­acc​.­org/​­education​ Catastrophic disruptions in clinical trials. ent patient populations. Eur Heart J 2021;​
-­and​­​-­meetings/​­image​-­and​-­slide​-­gallery/​­​ Circulation 2022;​146:​369-71. 42:​4887-90.
­media​-­detail?id=​­2729468B6E934​­65DB15­ 21. Docherty KF, Jhund PS, Anand I, et al. 25. Butler J, Packer M, Greene SJ, et al.
A1AAE6E3A44AE). Effect of dapagliflozin on outpatient Heart failure end points in cardiovascular
16. Dickstein K, Bebchuk J, Wittes J. The worsening of patients with heart failure outcome trials of sodium glucose co-
high-risk myocardial infarction database and reduced ejection fraction: a prespeci- transporter 2 inhibitors in patients with
initiative. Prog Cardiovasc Dis 2012;​ 54:​ fied analysis of DAPA-HF. Circulation 2020;​ type 2 diabetes mellitus: a critical evalua-
362-6. 142:​1623-32. tion of clinical and regulatory issues. Cir-
17. Desta L, Jernberg T, Spaak J, Hofman- 22. Sjöblom J, Muhrbeck J, Witt N, Alam culation 2019;​140:​2108-18.
Bang C, Persson H. Risk and predictors of M, Frykman-Kull V. Evolution of left ven- 26. Wanner C, Inzucchi SE, Lachin JM, et
readmission for heart failure following a tricular ejection fraction after acute myo- al. Empagliflozin and progression of kid-
myocardial infarction between 2004 and cardial infarction: implications for implant- ney disease in type 2 diabetes. N Engl J
2013: A Swedish nationwide observation- able cardioverter-defibrillator eligibility. Med 2016;​375:​323-34.
al study. Int J Cardiol 2017;​248:​221-6. Circulation 2014;​130:​743-8. 27. The EMPA-KIDNEY Collaborative
18. Harrington J, Butler J. Heart failure 23. Chew DS, Heikki H, Schmidt G, et al. Group. Empagliflozin in patients with
after myocardial infarction: glass emptier Change in left ventricular ejection frac- chronic kidney disease. N Engl J Med
than full. Eur J Heart Fail 2023;​25:​1225-7. tion following first myocardial infarc- 2023;​388:​117-27.
19. Hall ME, Vaduganathan M, Khan MS, tion and outcome. JACC Clin Electrophysiol 28. Lowry MTH, Doudesis D, Wereski R,
et al. Reductions in heart failure hospital- 2018;​4:​672-82. et al. Influence of age on the diagnosis of
izations during the COVID-19 pandemic. 24. Butler J, Anker SD, Filippatos G, et al. myocardial infarction. Circulation 2022;​
J Card Fail 2020;​26:​462-3. Sodium glucose co-transporter inhibitors 146:​1135-48.
20. Harrington J, Felker GM, Mentz RJ. and heart failure outcomes across differ- Copyright © 2024 Massachusetts Medical Society.

12 n engl j med  nejm.org

The New England Journal of Medicine

Downloaded from nejm.org by DHIRAJ KUMAR on April 6, 2024. For personal use only.
No other uses without permission. Copyright © 2024 Massachusetts Medical Society. All rights reserved.