JACC VOL. 85, NO.

15, 2025

ª 2025 THE AUTHORS. PUBLISHED BY ELSEVIER ON BEHALF OF THE AMERICAN

COLLEGE OF CARDIOLOGY FOUNDATION. THIS IS AN OPEN ACCESS ARTICLE UNDER

THE CC BY-NC-ND LICENSE ([Link]

Early Ezetimibe Initiation After

Myocardial Infarction Protects Against
Later Cardiovascular Outcomes in the
SWEDEHEART Registry
Margret Leosdottir, MD, PHD,a,b Jessica Schubert, MD,c Julia Brandts, MD,d,e Stefan Gustafsson, PHD,f
Thomas Cars, PHD,f Johan Sundström, MD, PHD,c Tomas Jernberg, MD, PHD,g Kausik K. Ray, MBCHB, MD, FMEDSCI,d,h
Emil Hagström, MD, PHDc,i

ABSTRACT

BACKGROUND Combination lipid-lowering therapy (LLT) after myocardial infarction (MI) achieves lower low-density
lipoprotein cholesterol (LDL-C) levels and better cardiovascular outcomes vs statin monotherapy. As a result, global
guidelines recommend lower LDL-C but, paradoxically, advise treatment through a stepwise approach. Yet the need for
combination therapy is inevitable as <20% of patients achieve goals with statins alone. Whether combining ezetimibe
with a statin early vs late after MI results in better outcomes is unknown.

OBJECTIVES In this study, the authors sought to assess the impact of delayed treatment escalation on outcomes by
comparing early vs late oral combination LLT (statins plus ezetimibe) in patients with MI.

METHODS LLT-naïve patients (SWEDEHEART registry) hospitalized for MI (2015-2022) and discharged on statins were
included. Using clone-censor-weight and Cox proportional hazards models, we compared differences in risks of MACE
(death, MI, stroke), components of MACE, and cardiovascular death between patients with ezetimibe added to
statins #12 weeks after discharge as reference (early combination therapy), from 13 weeks to 16 months (late combi-
nation therapy), or not at all.

RESULTS Of 35,826 patients (median age 65.1 years, 26.0% women), 6,040 (16.9%) received ezetimibe early, 6,495
(18.1%) ezetimibe late, and 23,291 (65.0%) received no ezetimibe. High-intensity statin use was $98% in all groups. Over a
median 3.96 years (Q1-Q3: 2.15-5.81 years), 2,570 patients had MACE (440 cardiovascular deaths). One-year MACE inci-
dences were 1.79 (early), 2.58 (late), and 4.03 (none) per 100 patient-years. Compared with early combination therapy,
weighted risk differences in MACE for late combination therapy at 1, 2, and 3 years were 0.6% (95% CI: 0.1%-1.1%;
P < 0.01), 1.1% (95% CI: 0.3%-2.0%; P < 0.01), and 0.7% (95% CI:
0.2% to 1.3%; P ¼ 0.18), and 3-year HR was 1.14
(95% CI: 0.95-1.41). For those receiving no ezetimibe, risk differences were 0.7% (95% CI: 0.2%-1.3%), 1.6% (95% CI:
0.8%-2.5%), and 1.9% (95% CI: 0.8%-3.1%; P for all <0.01; 3-year HR: 1.29 [95% CI: 1.12-1.55]). Similar differences in risk
of cardiovascular death at 3 years were observed (HRs vs early: late: 1.64 [95% CI: 1.15-2.63]; none: 1.83 [95% CI:
1.35-2.69]).

CONCLUSIONS MI care pathways should implement early combination therapy with statins and ezetimibe as
standard care, because delaying use of combination LLT or using high-intensity statin monotherapy is associated
with avoidable harm. (JACC. 2025;85:1550–1564) © 2025 The Authors. Published by Elsevier on behalf of the
American College of Cardiology Foundation. This is an open access article under the CC BY-NC-ND license
([Link]
Listen to this manuscript’s
audio summary by
Editor Emeritus
Dr Valentin Fuster on
From the aDepartment of Cardiology, Skåne University Hospital, Malmö, Sweden; bDepartment of Clinical Sciences Malmö, Lund
[Link]/journal/jacc.
University, Malmö, Sweden; cDepartment of Medical Sciences, Uppsala University, Uppsala, Sweden; dDepartment of Primary
Care and Public Health, Imperial Centre for Cardiovascular Disease Prevention, School of Public Health, Imperial College London,
London, United Kingdom; eDepartment of Internal Medicine, University Hospital RWTH Aachen, Aachen, Germany; fSence
Research, Uppsala, Sweden; gDepartment of Clinical Sciences, Danderyd Hospital, Karolinska Institute, Stockholm, Sweden;
h
Imperial Clinical Trials Unit, Imperial College London, London, United Kingdom; and the iUppsala Clinical Research Center,
Uppsala, Sweden.

ISSN 0735-1097 [Link]

JACC VOL. 85, NO. 15, 2025 Leosdottir et al 1551
APRIL 22, 2025:1550–1564 Timing of Oral LLT Intensification After MI

O bservational data from registries show that ezetimibe, with the remainder predicted to ABBREVIATIONS

at least two-thirds of patients with a achieve LDL-C goals if PCSK9i were added. If AND ACRONYMS

myocardial infarction (MI) are statin naïve the LDL-C goals currently recommended are
ACS = acute coronary
at the time of presentation.1,2 Thus, the first time to be stringently implemented, the use of syndrome
lipid-lowering therapy (LLT) is initiated for many is combination therapies is unavoidable. A
LDL-C = low-density
after a MI. Typically, this LLT consists of a high- stepwise approach inevitably means delays in lipoprotein cholesterol
intensity statin before hospital discharge, based on attaining recommended LDL-C goals. LLT = lipid-lowering therapy
clinical trials demonstrating that early treatment Whether this delay in treatment escalation is MACE = major adverse
with high-intensity statins is better than moderate- inferior to a proactive approach of early cardiovascular events

intensity statins after a MI, 3 with cardiovascular ben- combination LLT is unknown. Given the po- MI = myocardial infarction
efits appearing early and sustained if treatment is tential ethical dilemma of using a formal PCSK9i = proprotein
continued.4 Guidelines adopted this evidence, recom- randomized trial, we used the SWEDEHEART convertase subtilisin/kexin
type 9 inhibitor
mending early use of high-intensity statins and registry to provide insights into this question
providing a quality metric of treatment effectiveness by using a trial emulation clone-censor-weight
by recommending low-density lipoprotein cholesterol framework overcoming challenges with confounding
(LDL-C) goals of <1.8 mmol/L (<70 mg/dL).5 Subse- and immortal time bias. We hypothesized that the
quent trials have moved from statin monotherapy to delayed addition of ezetimibe to statins after a MI was
combination therapies, showing first that the addition associated with worse cardiovascular outcomes
of ezetimibe to statins before hospital discharge not compared with an approach using early combination
only lowered LDL-C levels beyond what was achiev- therapy with statins and ezetimibe.
able with monotherapy, but more importantly also
resulted in better cardiovascular outcomes.6 Trials of METHODS
proprotein convertase subtilisin/kexin type-9 inhibi-
tors (PCSK9is) showed that among patients with a The aim of this study was to compare the incidences
recent MI if LDL-C was $1.8 mmol/L ($70 mg/dL) of major adverse cardiovascular events (MACE), the
despite high-intensity statins, the addition of PCSK9i components of MACE, and cardiovascular death
reduced LDL-C and cardiovascular outcomes. 7,8
The between patients with MI discharged on statin ther-
evidence that lower LDL-C levels in the post-MI apy and receiving ezetimibe add-on early, patients
setting resulted in better outcomes were adopted by receiving ezetimibe add-on late, and those not
both European and U.S. guidelines. 9-11 However, the receiving ezetimibe at all.
recommendations on how to implement this con- DATA SOURCES. In this observational study, pro-
sisted of a stepwise approach, starting with the initia- spectively collected data from four Swedish national
tion of high-intensity statins as close to the MI as registries were used. The SWEDEHEART registry col-
possible, with escalation of treatment taking place lects data on all patients hospitalized for an MI at car-
after subsequent LDL-C measurements, with the addi- diac care units in Sweden and includes data on patient
tion of ezetimibe first and then PCSK9i for those characteristics, ongoing and discharge medications,
failing to achieve LDL-C goals. and acute coronary care. 13 When audited against the
Swedish National Inpatient Registry, SWEDEHEART
SEE PAGE 1565
captures >90% of all patients <80 years of age with MI
However, data from the SWEDEHEART (Swedish in the country. 14 Furthermore, data quality are moni-
Web-System for Enhancement and Development of tored yearly in SWEDEHEART and shows complete-
Evidence-Based Care in Heart Disease Evaluated Ac- ness and accuracy of >95%.14 Data on cardiac
cording to Recommended Therapies) registry showed rehabilitation interventions and outcomes, including
that approximately 75% to 80% of MI patients failed medication and lipid measures, are also registered in
to achieve LDL-C goals with statin monotherapy. 12 SWEDEHEART during the first year after a MI, with
The same study used modeling of expected LDL-C data available in w80% of all MI cases.15 The Pre-
reductions from treatment escalation showing that scribed Drug Registry contains data on all prescribed
goal attainment would double with the addition of and filled drug prescriptions in Sweden. The National

The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’
institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information,
visit the Author Center.

Manuscript received November 30, 2024; revised manuscript received January 30, 2025, accepted February 7, 2025.
1552 Leosdottir et al JACC VOL. 85, NO. 15, 2025

Timing of Oral LLT Intensification After MI APRIL 22, 2025:1550–1564

Patient Registry includes diagnoses from all non- Treatment with PCSK9i was rare and included in a
primary care outpatient visits and hospital discharges, sensitivity analysis. Use of other LLTs during the
and the Cause of Death Registry registers causes of study period was negligible and not included in the
death for all deaths in the country. Data were linked analysis.
with the use of the unique personal identity number OUTCOMES. The primary outcome was major adverse
assigned to all Swedish residents. cardiovascular events (MACE), consisting of all-cause
STUDY POPULATION. All patients aged 18 to <80 death, nonfatal MI (ICD-10 I21-I22), and nonfatal
years registered with an MI diagnosis (International stroke (ICD-10 I60-I64). Secondary outcomes were
Classification of Disease-10th Revision [ICD-10] codes cardiovascular death (ICD-10 I00-I99), the individual
I21-I22) from January 1, 2015, to September 23, 2022, components of MACE, and LDL-C target achievement.
were included. Earlier data were not included, The study index date was the date of discharge from
because ezetimibe use was sparse in Sweden before MI hospitalization. Complete follow-up was until a
2015.14 Patients on statin, ezetimibe, or PCSK9i before nonfatal event or death occurred, or at the end of data
admission and patients not discharged on statin capture (December 31, 2021, for nonfatal events and
(defined as prescription not filled within 7 days of cause-specific death and September 30, 2022, for all-
discharge) were excluded. Also, those with a baseline cause death), whichever occurred first.
LDL-C of <1.4 mmol/L (<55 mg/dL), a truncated Dutch To minimize the effect of temporal trends in MI
Lipid Clinic Network score of $6 points (using LDL-C treatment, including use of combination LLT, and to
level and age at index event only), or a diagnosis of ensure proportionality across outcomes, all outcomes
familial hypercholesterolemia (ICD-10 F78.0A) were are reported at 1, 2, and 3 years.
excluded because they were deemed likely to violate STATISTICAL ANALYSIS. Descriptive baseline data
the assumption that all patients should be potential are presented as median (Q1-Q3) or n (%).
candidates for being allocated to any treatment group When assessing treatment effects based on obser-
during the study. The age limit of <80 years was also vational data, there are several major challenges.
applied for this reason, ie, to minimize treatment- First, the absence of randomization to treatment
allocation bias on the account of frailty, comorbid- allocation introduces confounding bias due to unbal-
ities, and high age. The study complies with the anced patient characteristics, baseline risks, and
Declaration of Helsinki and was approved by the clinical choices based on those characteristics. Sec-
Regional Ethics Committee in Stockholm (approval ond, immortal time bias is introduced when timing of
no. 2012/6031/2) and the Swedish Ethical Review treatment initiation and start of study follow-up do
Authority (approval nos. 2019-04277 and 2021/00019). not coincide. A strategy to overcome these challenges
EXPOSURES. Three exposure groups were defined, is the clone-censor-weight method, a methodology for
all with statin treatment at discharge: 1) early com- emulating randomized clinical trials in observational
bination LLT with ezetimibe dispensed #12 weeks studies.16,17 A visual representation of this approach,
after discharge (reference group); 2) late combination as used in the present study, is provided in
LLT with ezetimibe dispensed from 13 weeks Supplemental Figure 1. The target trial emulation
to #16 months after discharge; and 3) no ezetimibe framework entails using observational data to mimic a
dispensed during the first 16 months after discharge. hypothetical randomized clinical trial addressing a
Too few patients had in-hospital initiation of combi- research question/hypothesis and includes a trial
nation LLT, further restricted by limited follow-up protocol with inclusion and exclusion criteria, treat-
time, to assess the effect of in-hospital initiation of ments, study outcomes, and analyses used.
combination LLT separately. Instead, the 12-week and For each index MI discharge, 3 clones (copies) were
16-month time points were used because standard created, corresponding to each treatment strategy.
cardiac rehabilitation programs in Sweden include 2 We evaluated adherence of the clones to their
time-fixed follow-up visits at 2 months (6-10 weeks) assigned treatment strategies at weekly intervals. If a
and 1 year (11-14 months) after discharge, aligning clone deviated from its assigned strategy, ie, changed
with information being registered in SWEDEHEART. 15 LLT, it was censored at that point. For example, if a
During these visits, LDL-C is sampled with an clone assigned to the “no ezetimibe initiation within
accepted time window of 2 weeks before or after the 12 weeks” strategy began treatment at 6 weeks, it was
2-month visit (4-12 weeks) and 2 months before or censored at that time (informative censoring). To
after the 1-year visit (9-16 months). Titration of LLT is adjust for potential selection bias due to informative
usually performed within this time frame. censoring, we applied weight models based on the
JACC VOL. 85, NO. 15, 2025 Leosdottir et al 1553
APRIL 22, 2025:1550–1564 Timing of Oral LLT Intensification After MI

probability of informative censoring. In cases where Swedish residents, there are no missing data in these
informative censoring occurred at a single time point, registries. In SWEDEHEART, a small proportion of
a logistic regression model was used; when it could variables had missing information, which was highest
occur at multiple time points, a pooled logistic for left ventricular ejection fraction (7.5%) and LDL-C
regression model was used. We calculated the cu- (6.5%) (Table 1). In the analyses, missing data were
mulative probability of remaining uncensored and handled with the use of single imputation via chained
derived inverse probability of censoring weights. To equations, under the assumption that data were
limit the influence of extreme weights, we truncated missing at random. All analyses were performed with
weights at the 99.5th percentile. Covariate balance the use of R software, version 3.6.0 (R Foundation).18
was assessed with the use of standardized mean dif-
ferences. Covariates and their definitions are listed in RESULTS
Supplemental Table 1.
To estimate a treatment effect using the clone- PATIENT CHARACTERISTICS. In total, 35,826 patients
censor-weight framework, risk differences were were included, of whom 6,040 (16.9%) received eze-
generated. Kaplan-Meier survival probability esti- timibe combination therapy early, 6,495 (18.1%)
mates were calculated for all outcomes. These ana- received ezetimibe combination therapy late, and
lyses were stratified by treatment exposure, and 23,291 (65.0%) did not receive ezetimibe during the
weighted risk differences were evaluated at 1, 2, and 3 first 16 months after discharge (Supplemental
years of follow-up. As a complementary analysis, HRs Figures 2 and 3). The median age at the index MI
and 95% CIs were estimated with the use of weighted was 65.1 years (Q1-Q3: 57.0-72.1 years), and 26.0%
Cox proportional hazards regression. CIs were calcu- were women (Table 1). Compared with patients
lated by means of nonparametric bootstrapping with receiving ezetimibe combination therapy early or
the percentile method, and P values were calculated late, patients not receiving add-on therapy with eze-
through permutation testing, in both cases with 1,000 timibe during the first 16 months were older, had
iterations. Patients receiving add-on therapy with more comorbidities, and had a lower baseline LDL-C.
ezetimibe early were used as the reference group in Prescription of high-intensity statins at discharge was
all models. high in all groups ($98%). Patient characteristics
In descriptive baseline analyses, the group with no stratified by sex are presented in Supplemental
ezetimibe dispensed during the first 16 months after Tables 3 and 4. The proportion of patients attending
discharge includes not only those not receiving eze- at least 1 of the 2 cardiac rehabilitation visits during
timibe during the first 16 months, but also patients the first year after MI was higher among patients
who had an event or died before having the chance of receiving ezetimibe early (93.1%) or late (95.6%) than
being allocated to the ezetimibe treatment groups. in those not receiving add-on therapy with ezetimibe
Those patients contributed to event rates in the early (86.6%) (Supplemental Table 5). At 1 year after
and late treatment groups in the clone-censor-weight discharge, 85.0% had filled a prescription for statins
model, but not in descriptive data analyses, possibly (any intensity) during their most recent dispensing
causing exaggerated between-group differences in period, accounting for a 25% grace period. Differences
baseline data. between groups were minimal (Supplemental
Sensitivity analyses were conducted by assessing Table 6). Fewer than 1.0% had filled a prescription
between-group risk differences and HRs with the use for PCSK9i, with some variation among groups,
of a negative control outcome. This was performed to ranging from 2.0% in the early ezetimibe group to
identify potential bias in group allocation, using an 1.3% in the late ezetimibe group and 0.3% in the no
outcome not causally influenced by the treatment ezetimibe group. The use of other LLTs was minimal
under investigation. To detect unmeasured con- (Table 1, Supplemental Table 6).
founding and treatment bias, we used skeletal frac-
tures (ICD-10 codes listed in Supplemental Table 2) as TEMPORAL TRENDS IN THE USE OF STATINS AND
a “negative control outcome.” A negative control EZETIMIBE. The proportion of patients receiving
outcome is an event that shares the same potential combination therapy up to 16 months after the index
sources of bias with the study outcome but cannot event increased from approximately 14% in 2015 to
plausibly be related to the treatment of interest. almost 60% in 2021 (Supplemental Figure 4). The
Furthermore, a sensitivity analysis censoring for proportion of patients discharged from hospital after
PCSK9i use during follow-up was performed. the index MI on statin plus ezetimibe was negligible
Because participation in the National Patient and before 2018. Before 2019, it was more common for
Prescribed Drug Registries is mandatory for all patients to receive add-on therapy with ezetimibe
1554 Leosdottir et al JACC VOL. 85, NO. 15, 2025

Timing of Oral LLT Intensification After MI APRIL 22, 2025:1550–1564

T A B L E 1 Patient Characteristics

Treatment Group

Overall Early Ezetimibe Late Ezetimibe No Ezetimibe

(N ¼ 35,826) Missing (n ¼ 6,040) (n ¼ 6,495) (n ¼ 23,291)

Admission characteristics
Age, y 65.1 (57.0-72.1) 0 63.6 (56.0-70.6) 63.4 (55.9-70.5) 66.1 (57.7-72.8)
Female 9,302 (26.0) 0 1,395 (23.1) 1,688 (26.0) 6219 (26.7)
Body mass index, kg/m2 26.9 (24.5-29.9) 1,471 27.0 (24.6-30.0) 27.1 (24.7-30.1) 26.8 (24.4-29.8)
Current smoker 9,683 (27.7) 895 1,647 (27.9) 1,39 (28.9) 6,197 (27.3)
Hypertension 12,519 (35.1) 179 1,835 (30.6) 2,025 (31.3) 8,659 (37.4)
Diabetes mellitus 3,543 (9.9) 0 451 (7.5) 476 (7.3) 2,616 (11.2)
Charlson comorbidity index 1.0 (1.0-2.0) 0 1.0 (1.0-2.0) 1.0 (1.0-2.0) 1.0 (1.0-2.0)
Medical history
Previous MI 946 (2.6) 0 125 (2.1) 152 (2.3) 669 (2.9)
Heart failure 549 (1.5) 0 58 (1.0) 64 (1.0) 427 (1.8)
Laboratory and physiological variables
LDL-C, mmol/L 3.3 (2.7-4.0) 2,327 3.8 (3.2-4.4) 3.6 (3.0-4.2) 3.1 (2.6-3.7)
eGFR, mL/min/1.73 m2 75.8 (66.5-85.1) 0 77.4 (68.2-86.4) 77.2 (68.1-86.2) 75.1 (65.7-84.5)
Fasting glucose, mmol/L 6.7 (5.8-8.0) 5,239 6.6 (5.8-7.8) 6.6 (5.8-7.9) 6.7 (5.8-8.1)
Systolic blood pressure, mm Hg 150 (133-170) 235 150 (132-170) 150 (133-170) 150 (133-170)
In-hospital characteristics and management
STEMI 16,609 (46.4) 1 3,038 (50.3) 3,114 (47.9) 10,457 (44.9)
Reduced LVEFa 3,742 (11.3) 2,698 610 (10.7) 678 (11.2) 2,454 (11.5)
Coronary angiography 35,030 (97.8) 0 5,944 (98.4) 6,408 (98.7) 22,678 (97.4)
Multivessel diseaseb 14,067 (39.3) 0 2,659 (44.0) 2,671 (41.1) 8,737 (37.5)
Revascularized during hospitalization 30,036 (83.8) 0 5,403 (89.5) 5,673 (87.3) 18,960 (81.4)
Discharge medication
High-intensity statin 35,285 (98.5) 0 6,015 (99.6) 6,437 (99.1) 22,833 (98.0)
Ezetimibe 1,058 (3.0) 0 1,058 (17.5) N/A N/A
PCSK9i #5 0 #5 #5 #5
Other LLTc 8 (0.0) 0 #5 #5 #5
ACE inhibitor/ARB 29,747 (83.0) 0 4,995 (82.7) 5,462 (84.1) 19,290 (82.8)
Platelet inhibitord 35,561 (99.3) 0 6,019 (99.7) 6,468 (99.6) 23,074 (99.1)

Values are median (Q1-Q3), n, or n (%). aLeft ventricular ejection fraction <40%. bTwo or more main coronary arteries (left anterior descending artery, circumflex artery, right coronary artery,
or left main) with significant ($50%) stenoses. cFibrates, resins, or nicotinic acid. dAcetylsalicylic acid or P2Y12 inhibitor.
ACE ¼ angiotensin-converting enzyme; ARB ¼ angiotensin II receptor blocker; eGFR ¼ estimated glomerular filtration rate; LDL-C ¼ low-density lipoprotein cholesterol; LLT ¼ lipid-
lowering therapy; LVEF ¼ left ventricular ejection fraction; MI ¼ myocardial infarction; PCSK9i ¼ proprotein convertase subtilisin/kexin type 9 inhibitor; STEMI ¼ ST-segment elevation
myocardial infarction.

late (ie, $13 weeks after the index event), whereas The proportions of patients with at least 1 mea-
toward the end of the study, initiation of combination surement of LDL-C during follow-up who achieved
therapy early was more common. the LDL-C target of <1.4 mmol/L (<55 mg/dL), ac-
PROPORTION OF PATIENTS AT LDL-C GOAL. Data cording to the clone-censor-weight framework, are
were available on LDL-C levels for 33,499 patients shown in Figure 2. Despite having the highest base-
(93.5%) at baseline and for 20,351 patients (56.8%) at line LDL-C levels, approximately 55% of patients
both cardiac rehabilitation follow-up visits (6-10 weeks receiving ezetimibe combination therapy early had
and 1 year). A larger proportion of patients with early reached the target at 1 year after discharge. Of pa-
ezetimibe combination therapy had achieved the tients receiving ezetimibe late or not having received
LDL-C target at 1 year compared with patients ezetimibe up to 1 year, fewer were at goal. Over 3
receiving ezetimibe therapy late and those not having years, although the proportion at goal increased in all
received ezetimibe up to the 1-year visit (time frame for 3 groups, it remained better with combination ther-
LDL-C sampling, 9-16 months) (Figure 1). Median and apy and was highest in the early combination therapy
LDL-C changes and proportions of patients at the LDL- group (Figure 2).
C target at the 6-10–week and 1-year follow-up visits for EARLY VS LATE COMBINATION LLT AND OUTCOMES.

each treatment group are presented in During a median follow-up of 3.96 years (Q1-Q3:
Supplemental Table 7. 2.15-5.81 years), 2,570 patients had a MACE.
JACC VOL. 85, NO. 15, 2025 Leosdottir et al 1555
APRIL 22, 2025:1550–1564 Timing of Oral LLT Intensification After MI

Unadjusted incidence rates at 1 year were 1.79

F I G U R E 1 Proportion of Patients at LDL-C Target (LDL-C <1.4 mmol/L,
(early ezetimibe combination therapy), 2.58 (late <55 mg/dL and ‡50% Reduction) at 1-Year Follow-Up
ezetimibe combination therapy), and 4.03 (no ezeti-
mibe) per 100 person-years. Covariate balance was 100%
assessed after 16 months and was within the
90%
preferred standardized mean difference of 0.1 for
all covariates considered (Supplemental Figure 5). 80%
Compared with patients initiated on ezetimibe early,
70% 61.6%
the risk of MACE for patients receiving ezetimibe 71.2% 74.7%
add-on therapy late was higher at 1 and 2 years 60%

after MI but was not statistically significant at 3 years 50%

after MI (Figure 3). For those not receiving ezetimibe
40%
at all, the risk was higher at 1, 2, and 3 years after
MI (Figure 3). The adjusted HR for MACE over 3 years 30%
for patients receiving ezetimibe late was 1.14
20% 38.4%
(95% CI: 0.95-1.41) and for patients not 28.8% 25.3%
receiving ezetimibe was 1.29 (95% CI: 1.12-1.55) 10%
(Supplemental Figure 6). 0%
The cumulative incidence of the 4 prespecified Early Ezetimibe Late Ezetimibe No Ezetimibe
secondary outcomes followed a similar pattern, with
Not at Target
patients not receiving ezetimibe at all having the
At Target (LDL-C <1.4 mmol/L, <55 mg/dL and ≥50% Reduction)
highest risk for adverse outcomes compared with
patients receiving add-on treatment with ezetimibe
Only patients attending cardiac rehabilitation with lipid levels measured are
early (Figure 4). For cardiovascular death, the HRs at 3
included. LDL-C ¼ low-density lipoprotein cholesterol.
years for patients receiving ezetimibe late or not at all
were 1.64 (95% CI: 1.15-2.63) and 1.83 (95% CI: 1.35-
2.69), respectively (Supplemental Figure 6).
Using the current cohort (35,826 patients and 2,570
events) and the 3-year risk differences in MACE, statistically significant difference in the cumulative
moving from statin monotherapy to an early combi- incidence of skeletal fractures for those receiving
nation strategy would require a number needed to ezetimibe late or not at all (Supplemental Figure 7).
treat (NNT) of 53 (95% CI: 32-125) to prevent 1 MACE. In the sensitivity analysis censoring for PCSK9i use
If we moved from a delayed to early combination during follow-up, no major differences in any of the
strategy the NNT to prevent one MACE would be 143 outcomes were observed (Supplemental Table 8).
(95% CI: 44-167). Applying these NNTs to the current DISCUSSION
study population means that instead of the observed
16.9% receiving early combination LLT, 18.1% In this study of 35,826 statin-treated patients with a
receiving late combination LLT, and 65.0% receiving recent MI, there were 3 key findings regarding timing
no ezetimibe, if all patients received early combina- of initiation of combination LLT by the addition of
tion LLT then this approach could prevent an addi- ezetimibe to high-intensity statins. First, as expected,
tional 477 events. Table 2 provides hypothetical the use of ezetimibe combination therapy early or late
scenarios of potential MACE averted through imple- resulted in a greater proportion of patients achieving
mentation of early combination therapies in pop- an LDL-C of <1.4 mmol/L (<55 mg/dL) compared
ulations of varying size. with not receiving add-on therapy with ezetimibe at
SENSITIVITY ANALYSES. To evaluate potential bias all. Second, early combination therapy by adding
or confounding related to frailty, where patients with ezetimibe to background high-intensity statin was
more severe health conditions might not have been associated with greater absolute and relative benefits
offered combination therapy, we performed a sensi- in terms of MACE compared with a strategy of
tivity analysis using fractures as a negative control delayed initiation of combination therapy (Central
outcome. Compared with patients receiving early Illustration). Third, approximately two-thirds of pa-
combination with ezetimibe, the analysis showed no tients had not received add-on therapy with
1556 Leosdottir et al JACC VOL. 85, NO. 15, 2025

Timing of Oral LLT Intensification After MI APRIL 22, 2025:1550–1564

F I G U R E 2 Proportion of Patients Attaining the LDL-C Target of <1.4 mmol/L (<55 mg/dL) During Follow-Up

60%
Proportion at LDL-C Target

40%

20%

0%

0 1 2 3

Years Since Discharge

Risk differences (95% CI) (95% CI) (95% CI)

Reference Reference Reference
−12.6% (−15.1% to −10.8%) −9.4% (−12.1% to −7.2%) −7.3% (−10.3% to −4.5%)
−15.4% (−17.5% to −13.5%) −18.1% (−20.3% to −16.0%) −17.3% (−19.7% to −14.7%)

Early Ezetimibe Late Ezetimibe No Ezetimibe

Patients with at least 1 measure of LDL-C are included (varying time points and frequency of measurements). LDL-C ¼ low-density lipoprotein
cholesterol.

ezetimibe by 16 months after MI, and those patients ($70 mg/dL) despite high-intensity or maximally
experienced the highest risk of MACE and cardio- tolerated statins, combination therapy using a PCSK9i
vascular death. resulted in lower achieved LDL-C levels and a lower
The addition of nonstatin LLT at the time of an MI, risk of cardiovascular events.7,8 Collectively, those
with ezetimibe or PCSK9i, results in lower achieved trials provide the evidence base for most global
LDL-C levels than possible through statin mono- guidelines recommending lower LDL-C goals
therapy and improves plaque morphology.19-21 This (<1.4 mmol/L, <55 mg/dL) in such patients. 10
provided a mechanistic rationale for the reduction in Although the “destination” of achieving lower
atherosclerotic events observed in clinical outcome LDL-C levels for post-MI patients seems to follow the
trials of combination LLTs. 6,22 For example, in the trial evidence, the path through which the destina-
IMPROVE-IT (Examining Outcomes in Subjects tion is reached does not. For example, guidelines
With Acute Coronary Syndrome: Vytorin [Ezetimibe/ such as the 2019 European Society of Cardiology
Simvastatin] vs Simvastatin; NCT00202878) trial, (ESC)/European Atherosclerosis Society dyslipidemia
adding ezetimibe to simvastatin in patients with guidelines paradoxically recommend the addition of
acute coronary syndromes (ACS) improved cardio- ezetimibe if LDL-C goals are not achieved with the
vascular prognosis.6 In that trial, ezetimibe (in addi- maximum-tolerated dose of statin alone.10 This was
tion to statins) was initiated within 10 days of the reiterated in the 2020 non–ST-segment elevation MI
acute event, and 66% of the patients were statin guidelines. 23 A review of the literature suggests that
naïve at study baseline. Thus, an early combination- the time frame of 4 to 6 weeks was not based on ev-
LLT regimen was initiated in two-thirds of the pa- idence of prognostic superiority within that period,
tients, providing evidence for the benefit of early but rather for practical reasons, because the maximal
combination oral LLT after ACS. Subsequently, trials effect of statins on LDL-C lowering occurs 2 to
assessing the added value of PCSK9i after MI or ACS 4 weeks after initiating therapy. However, in real-life
showed that among patients with LDL-C $1.8 mmol/L scenarios, it is unlikely that lipid measurements are
JACC VOL. 85, NO. 15, 2025 Leosdottir et al 1557
APRIL 22, 2025:1550–1564 Timing of Oral LLT Intensification After MI

F I G U R E 3 Cumulative Incidence of the Primary Outcome MACE for Patients Initiated on Combination Therapy With Statin and Ezetimibe Early, Late,
or No Ezetimibe

7.5%
Proportion With MACE

5.0%

2.5%

0.0%

0 1 2 3

Years Since Discharge

No. at risk
35,826 23,098 18,785 16,002
35,826 27,131 4,852 3,381
35,826 4,831 3,301 2,301

Risk differences (95% CI) P Value (95% CI) P Value (95% CI) P Value
0.7% (0.2%-1.3%) <0.01 1.6% (0.8%-2.5%) <0.01 1.9% (0.8%-3.1%) <0.01
0.6% (0.1%-1.1%) <0.01 1.1% (0.3%-2.0%) <0.01 0.7% (−0.6% to 2.3%) 0.18
Reference Reference Reference

No Ezetimibe Late Ezetimibe Early Ezetimibe

MACE ¼ major adverse cardiovascular events (all-cause death, nonfatal myocardial infarction, and nonfatal stroke).

undertaken at 4 to 6 weeks, meaning that up-titration simultaneous LLT combination therapy with high-
follows clinical routines such as outpatient visits intensity statin and ezetimibe was introduced. 30 The
during the year after an MI, so it can take much recommendation was based on IMPROVE-IT as well
longer. 14,24,25 as the assumption that many statin-naïve patients
This stepwise approach of LLT escalation has been who have LDL-C levels that make it unlikely that they
questioned in a position paper from the ESC Societies would reach the LDL-C goal with statin monotherapy
of Acute Cardiovascular Care and Preventive Cardi- would benefit long term from initiation of statin and
ology.26 That position paper recommends the use of a ezetimibe combination therapy during ACS hospital-
combination regimen of a high-intensity statin in ization to decrease the time taken to achieve LDL-C
conjunction with ezetimibe before discharge after goals.30
ACS. This was based on the observed treatment Further support for the pragmatic use of early
inertia among physicians in up-titration of LLT during combination therapy before discharge was provided in
the post-ACS follow-up period in several surveys and a recent publication from a Polish ACS registry, where
consequent low rates of LDL-C target achievement, mortality was studied with the use of propensity score
thus leaving patients at high residual risk. 26,27 This matching of patients initiated on combination LLT
approach has been recommended by others as (ezetimibe plus statins) during hospitalization vs
well.28,29 In the 2023 ESC guidelines on ACS, a new statin monotherapy (atorvastatin or rosuvastatin).31
Class II Level of Evidence: B recommendation on Early use of combination therapy was associated
1558 Leosdottir et al JACC VOL. 85, NO. 15, 2025

Timing of Oral LLT Intensification After MI APRIL 22, 2025:1550–1564

F I G U R E 4 Cumulative Incidence of the 4 Prespecified Secondary Outcomes

1.5%
Proportion With CV Death

1.0%

0.5%

0.0%

0 1 2 3
Years Since Discharge

No. at risk
34,213 21,940 17,789 14,614
34,213 25,414 3,869 2,794
34,213 3,728 2,630 1,769

Risk differences (95% CI) P Value (95% CI) P Value (95% CI) P Value
0.2% (−0.1% to 0.4%) 0.24 0.6% (0.2%-0.8%) <0.01 0.8% (0.5%-1.1%) <0.01
0.1% (−0.1% to 0.4%) 0.40 0.4% (0.0%-0.6%) 0.05 0.7% (0.2%-1.3%) 0.01
Reference Reference Reference

No Ezetimibe Late Ezetimibe Early Ezetimibe

4.0%
Proportion With All-Cause Death

3.0%

2.0%

1.0%

0.0%

0 1 2 3

Years Since Discharge

No. at risk
35,826 23,543 19,467 16,768
35,826 27,698 4,774 3,566
35,826 4,897 3,383 2,379

Risk differences (95% CI) P Value (95% CI) P Value (95% CI) P Value
0.5% (0.1%-0.8%) <0.01 1.1% (0.3%-1.6%) <0.01 1.6% (0.7%-2.2%) <0.01
0.4% (0.0%-0.6%) <0.01 0.6% (0.1%-1.2%) 0.03 0.7% (−0.3% to 1.6%) 0.12
Reference Reference Reference

No Ezetimibe Late Ezetimibe Early Ezetimibe

(A) Cardiovascular (CV) death, (B) all-cause death, (C) nonfatal myocardial infarction (MI), and (D) nonfatal stroke.

Continued on the next page

JACC VOL. 85, NO. 15, 2025 Leosdottir et al 1559
APRIL 22, 2025:1550–1564 Timing of Oral LLT Intensification After MI

F I G U R E 4 Continued

3.0%
Proportion With Nonfatal MI

2.0%

1.0%

0.0%

0 1 2 3

Years Since Discharge

No. at risk
35,826 23,226 18,957 16,204
35,826 27,286 4,624 3,426
35,826 4,850 3,316 2,312

Risk differences (95% CI) P Value (95% CI) P Value (95% CI) P Value
0.2% (−0.5% to 0.6%) 0.37 0.5% (0.2%-1.0%) 0.03 0.5% (−0.3% to 1.3%) 0.07
0.2% (−0.5% to 0.6%) 0.36 0.4% (−0.4% to 0.9%) 0.10 0.3% (−0.8% to 1.1%) 0.37
Reference Reference Reference

No Ezetimibe Late Ezetimibe Early Ezetimibe

D
Proportion With Nonfatal Stroke

1.0%

0.5%

0.0%

0 1 2 3

Years Since Discharge

No. at risk
35,826 23,411 19,285 16,555
35,826 27,539 4,731 3,519
35,826 4,878 3,368 2,368

Risk differences (95% CI) P Value (95% CI) P Value (95% CI) P Value
0.2% (−0.1% to 0.4%) 0.24 0.6% (0.2%-0.8%) <0.01 0.8% (0.5%-1.1%) <0.01
0.1% (−0.1% to 0.4%) 0.40 0.4% (0.0%-0.6%) 0.05 0.7% (0.2%-1.3%) 0.01
Reference Reference Reference

No Ezetimibe Late Ezetimibe Early Ezetimibe

1560 Leosdottir et al JACC VOL. 85, NO. 15, 2025

Timing of Oral LLT Intensification After MI APRIL 22, 2025:1550–1564

T A B L E 2 Potential Health Gains Estimated as Predicted Number of MACE Avoided Through Increasing the Proportion of Patients Receiving
Early Combination Therapy vs Other Scenarios in Populations of Different Sizes

Study
Populationa Hypothetical Population (Predicted)b

No. at risk 35,826a 10,000 50,000 100,000 250,000

MACE events occurring over 3 y 2,570 717 3,587 7,174 17,934
MACE avoided over 3 y by increasing
the proportion of patients receiving early
ezetimibe as combination therapy vs other scenarios
30% early, 15% late, 55% no ezetimibe 68 19 94 189 472
50% early, 15% late, 35% no ezetimibe 203 57 283 566 1415
70% early, 15% late, 15% no ezetimibe 338 94 472 943 2,358
100% receive early ezetimibe 477 133 666 1,331 3,328

a
Represents the observed study population. bThis scenario reflects the status quo, ie, MACE events predicted to occur if the population at risk increased in size but the current
proportion of use of different LLTs remained the same, namely, 16.9% receiving early combination LLT, 18.1% receiving late combination LLT, and 65.0% receiving no
ezetimibe.
LLT ¼ lipid-lowering therapy; MACE ¼ major adverse cardiovascular events.

with reduced all-cause death, with an OR of 0.526 clone-censor-weight method. 16 Here, for the first time
(95% CI: 0.378-0.733) and an absolute risk difference of using this method on secondary prevention treat-
nearly 5% after 3 years of follow-up. The Polish regis- ments after MI, we assessed whether the timing of
try has several differences from our study in terms of LLT intensification was prognostically important and,
design (the Polish registry study included patients on by default, whether current stepwise approaches
previous LLT, and subsequent up-titration or changes risked harm in patients with a recent MI. Our findings
in statin intensity were not accounted for), but their indicate potential harms of the current approach to
results lend support to the present findings that early guideline implementation by way of a delayed step-
combination therapy of ezetimibe with statins after wise approach. Moreover, the present findings add
ACS is beneficial. More recently, efforts in Italy (AT support for how to implement our recent observa-
TARGET IT [Evaluation of Adherence, Persistence and tions assessing the timing of both absolute and rela-
Efficacy of Treatment With PCSK9 Inhibitors in Italy; tive changes in non-HDL-C, as well as the time to
NCT05430828]) and Germany (Jena auf Ziel-JaZ) achieve target levels on cardiovascular outcomes.34,35
showed that redesigning care pathways to implement In the latter study, the best outcomes, including
early combination therapies resulted in more patients reduction in death, were observed when non-HDL-C
achieving LDL-C goals and that those achieving the targets were both achieved early (within 2 months)
lowest levels had the lowest risk of adverse and sustained over 1 year compared with all other
outcomes.32,33 scenarios, with statistically significant benefits
The potential impact on health outcomes of the observed within 1 year that increased over time,
timing of treatment escalation through use of com- supporting lower cumulative atherogenic lipid
bination LLTs after an MI has not been evaluated in exposure. 34
clinical practice. An ideal scenario would be a ran- Ezetimibe is generically available in most countries
domized controlled trial, but to test this hypothesis and reduces LDL-C by 20% to 25%, but it is underu-
may present ethical challenges where the design tilized. The reasons could reflect perceptions that
would consciously require offering more effective ezetimibe is a “weak” LDL-C–reducing treatment,
combination LLTs and thus achieve early LDL-C reinforced by trial evidence of a single-digit RR
lowering in one group vs delaying it in another. reduction.6 Such views do not take into account the
Although observational studies can be helpful to fill fact that RR reduction in cardiovascular events with
in the data gaps, the lack of randomization introduces LLTs is proportional to the absolute reduction in
confounding because of differences in patient char- LDL-C. In general populations, patients typically have
acteristics, risks, and treatment choices based on a higher LDL-C, and thus derive a greater absolute
these characteristics. Furthermore, when the timing lowering in LDL-C and would be predicted to derive
of a treatment initiation, in this case ezetimibe, is greater RR reductions than observed in IMPROVE-IT.
used to define treatment groups, immortal time bias There is also little evidence of safety concerns when
is introduced. A strategy to overcome these chal- very low levels of LDL-C are achieved with combina-
lenges is to use a target trial emulation based on the tion therapies. 36-38
JACC VOL. 85, NO. 15, 2025 Leosdottir et al 1561
APRIL 22, 2025:1550–1564 Timing of Oral LLT Intensification After MI

C ENTR AL I LL U STRA T I O N Early Ezetimibe Initiation After Myocardial Infarction Protects Against Later
Cardiovascular Outcomes

Leosdottir M, et al. JACC. 2025;85(15):1550–1564.

LDL-C ¼ low-density lipoprotein cholesterol; LLT ¼ lipid-lowering therapy; MACE ¼ major adverse cardiovascular events; MI ¼ myocardial infarction; Ref ¼ reference.

Implementing complex multistep care pathways Moreover, for those who fail to achieve LDL-C goals
inevitably may result in variations in care. In the on 2 oral drugs, our proposed approach considerably
latest SWEDEHEART annual report, a disturbingly reduces the time taken to assess eligibility for PCSK9i
wide variation in use of combination LLT at site level treatment, which would then be consistent with the
was observed, ranging from 10% to 65% at the 6- to timing of PCSK9i treatment in the ODYSSEY Out-
10-week follow-up and from 30% to 85% at the 1-year comes (Evaluation of Cardiovascular Outcomes After
follow-up. 14 When all evidence is considered, we an Acute Coronary Syndrome During Treatment With
think that implementation of cholesterol manage- Alirocumab; NCT01663402) trial, ie, w2 months after
ment after an MI, reducing variability in care, and ACS.8 Where cost limits use of injectable therapies,
improving prognosis are best served by moving away bempedoic acid could be considered as a third oral
from a stepwise strategy, which only delays the agent without necessarily increasing pill burden with
inevitable ie the need for combination therapy for the use of fixed-dose combinations with ezeti-
most patients to achieve guideline-recommended mibe.39,40 Therefore, there seems to be little cost,
LDL-C goals. Public health is best served by a move biological, or safety argument against a high-intensity
toward early combination therapy, which is prag- statin plus ezetimibe combination approach as the
matically achievable with a combination of high- standard of care for all MI patients. The present study
intensity statins and ezetimibe before hospital suggests that the price of delay is high and one that
discharge, as suggested in the ESC 2022 position we can no longer afford to pay. For patients already
paper.26 on statins admitted with an MI, ongoing trials such as
1562 Leosdottir et al JACC VOL. 85, NO. 15, 2025

Timing of Oral LLT Intensification After MI APRIL 22, 2025:1550–1564

EVOLVE-MI (EVOLocumab Very Early After Myocar- same sample size limitation. Because the prescription
dial Infarction; NCT05284747) with PCSK9i added of ezetimibe (both early and late) was low in the first
during hospitalization may provide insights into early half of the study period, the follow-up period had to
use of PCSK9i during hospital admission in the near be limited to 3 years in all groups. Whether the
future, but for now their use may need to be consid- observed risk differences extend beyond 3 years is
ered on a case-by-case basis. therefore unknown. Statin treatment at discharge was
STUDY STRENGTHS AND LIMITATIONS. In studies defined as a binary variable. Although statin intensity
where treatment effects are based on observational or at baseline and time-varying statin intensity were
registry data, there are known substantial risks of adjusted for in the cloning models, residual con-
confounding by indication. The target trial emulation founding because of differences in type and dose of
approach can mitigate some confounding by stating statin prescribed and filled as well as actual adher-
inclusion and exclusion criteria and time-zero clearly. ence over time, not captured by these variables,
Further bias minimization can be sought by using the cannot be excluded. However, w98% of patients were
novel clone-censor-weight framework, which is discharged on a high-intensity statin, reducing the
particularly useful when early vs late treatment impact of adding discharge statin treatment as a bi-
strategies are compared. Such comparisons have been nary variable. Although patients being discharged
impossible until now, because of unsurmountable from a university hospital or not was adjusted for in
problems with immortal time bias (patients treated the models, there might have been other site-level
with a late treatment strategy need to survive until differences in early prescription of ezetimibe not
that “late” time point and are thus inherently accounted for. Because this was an observational
different from early-treatment patients). The clone- study, confounding due to nonrandom treatment
censor-weight method allows for such comparisons allocation needs to be acknowledged. The clone-
to be unbiased. This is a unique strength of the pre- censor-weight method, however, eliminates a large
sent study. The external validity of our findings is part of that bias. Lack of between-group differences
large, because this study is based on a large unse- in the negative control outcome analysis further
lected consecutive cohort of clinical-practice MI pa- strengthens this assumption. Finally, as with all an-
tients with nationwide coverage, including >90% of alyses on observational data, the risk of residual
all MI cases in patients aged <80 years in the country confounding needs to be acknowledged.
during the time period, reducing the selection bias
CONCLUSIONS
present in randomized clinical trials.41 By excluding
patients likely to violate the assumption of random
The need for combination therapy is inevitable for
group allocation, the risk of selection bias inherent to
most patients after an MI. A delayed approach to LLT
observational studies was reduced, further strength-
escalation is associated with avoidable harms. Care
ened by the adequate covariate balance obtained
pathways can be streamlined with tangible health
(including baseline LDL-C and time-varying statin
benefits if the standard of care for LLT after an MI
intensity) and no significant differences observed in
consists of an early combination of high intensity
negative control noncardiac outcome analyses.
statins with ezetimibe.
Finally, the possibility of linking data between reg-
DATA SHARING STATEMENT SWEDEHEART does not
istries, with minimal missing information and virtu-
allow individual data sharing to a third party. Access
ally no loss-to-follow-up, improves the reliability and
to aggregated data might be granted after review by
completeness of the data and enhances the overall
the SWEDEHEART steering committee.
generalizability of our findings.
There are, however, limitations. As previously ACKNOWLEDGMENTS Dr Leosdottir thanks Lund
stated, our real-life patient cohort, albeit large, had University, Dr Ray thanks the Imperial NIHR
too few patients discharged on combination therapy Biomedical Research Centre, and Dr Hagström thanks
with sufficient follow-up time to assess the effect on Uppsala University.
outcomes of the in-hospital initiation of combination
LLT separately. Because up-titration of LLT usually FUNDING SUPPORT AND AUTHOR DISCLOSURES
takes place around the 6- to 10-week follow-up visit
at Swedish cardiac rehabilitation centers, limiting the This work was funded by Skåne and Uppsala Regions, Sweden, and
the Swedish Heart and Lung Foundation (20190390). Dr Leosdottir
time frame of early combination treatment to the
has received institutional grants and honoraria from Amarin, Amgen,
guideline-recommended 4 to 6 weeks instead of AstraZeneca, Bonnier Health Care, Novo Nordisk, and Sanofi, outside
12 weeks as used here would have resulted in the the submitted work. Dr Schubert has received institutional grants and
JACC VOL. 85, NO. 15, 2025 Leosdottir et al 1563
APRIL 22, 2025:1550–1564 Timing of Oral LLT Intensification After MI

honoraria from Amgen, Pfizer, and Sanofi, outside the submitted Esperion, Cargene Therapeutics, Resverlogix, Novartis, Silence
work. Dr Brandts has received institutional grants and honoraria from Therapeutics, NewAmsterdam Pharma, Scribe Therapeutics, CRISPR
Amgen, Boehringer Ingelheim, Daiichi-Sankyo, Novartis, Novo Nor- Therapeutics, Vaxxinity, Amarin, CSL Behring, Bayer, Cleerly Health,
disk, Lily, and Sanofi, outside the submitted work. Dr Gustafsson is Emendobio and Stock Options PEMI31, New Amsterdam Pharma, and
employed by Sence Research, an independent company in epidemi- SCRIBE, outside the submitted work. Dr Hagström has received
ology and biostatistics at which the statistical analyses within this institutional grants and honoraria from Amarin, Amgen, AstraZeneca,
research project were conducted. Dr Cars is employed by and has co- Bayer, Novartis, Novo Nordisk, Pfizer, and Sanofi, outside the sub-
ownership in Sence Research. Dr Sundström has co-ownership and mitted work.
stock in Sence Research. Dr Jernberg has received grants from MSD;
and has received a consulting fee from Amgen to his institution. Dr ADDRESS FOR CORRESPONDENCE: Dr Margret Leos-
Ray has received institutional grants from Amarin, Amgen, Daiichi-
dottir, Department of Cardiology, Skåne University Hos-
Sankyo, Sanofi, and Ultragenix; and has received consultancy fees
from Amgen, Sanofi, Regeneron, Pfizer, Viatris, Abbott, AstraZeneca, pital, Jan Waldenströms gata 15, plan 3, 205 02 Malmö,
Lilly, Kowa Pharmaceutics, Novo Nordisk, Boehringer Ingelheim, Sweden. E-mail: [Link]@[Link].

REFERENCES

1. Ersboll AK, Kristensen MS, Nybo M, et al. Trends 10. Mach F, Baigent C, Catapano AL, et al. 2019 21. Raber L, Ueki Y, Otsuka T, et al. Effect of
in low-density lipoprotein cholesterol goal ESC/EAS guidelines for the management of dysli- alirocumab added to high-intensity statin therapy
achievement and changes in lipid-lowering ther- pidaemias: lipid modification to reduce cardiovas- on coronary atherosclerosis in patients with acute
apy after incident atherosclerotic cardiovascular cular risk. Eur Heart J. 2020;41:111–188. myocardial infarction: the PACMAN-AMI random-
disease: Danish cohort study. PLoS One. 2023;18: ized clinical trial. JAMA. 2022;327:1771–1781.
11. Visseren FLJ, Mach F, Smulders YM, et al. 2021
e0286376.
ESC guidelines on cardiovascular disease prevention 22. Schwartz GG, Fayyad R, Szarek M, DeMicco D,
2. Leosdottir M, Hagstrom E, Hadziosmanovic N, in clinical practice. Eur Heart J. 2021;42:3227–3337. Olsson AG. Early, intensive statin treatment
et al. Temporal trends in cardiovascular risk fac- reduces “hard” cardiovascular outcomes after
12. Allahyari A, JernbergT Hagstrom E,
tors, lifestyle and secondary preventive medica- acute coronary syndrome. Eur J Prev Cardiol.
Leosdottir M, Lundman P, Ueda P. Application of the
tion for patients with myocardial infarction 2017;24:1294–1296.
2019 ESC/EAS dyslipidaemia guidelines to nation-
attending cardiac rehabilitation in Sweden 2006-
wide data of patients with a recent myocardial 23. Collet JP, Thiele H, Barbato E, et al. 2020 ESC
2019: a registry-based cohort study. BMJ Open.
infarction: a simulation study. Eur Heart J. 2020;41: guidelines for the management of acute coronary
2023;13:e069770.
3900–3909. syndromes in patients presenting without persis-
3. Ray KK, Cannon CP. Intensive statin therapy in tent ST-segment elevation. Eur Heart J. 2021;42:
13. Jernberg T, Attebring MF, Hambraeus K, et al.
acute coronary syndromes: clinical benefits and 1289–1367.
The Swedish Web-System for Enhancement and
vascular biology. Curr Opin Lipidol. 2004;15:637–
Development of Evidence-Based Care in Heart 24. Kotseva K, de Backer G, de Bacquer D, et al.
643.
Disease Evaluated According to Recommended Lifestyle and impact on cardiovascular risk factor
4. Ray KK, Cannon CP, McCabe CH, et al. Early and Therapies (SWEDEHEART). Heart. 2010;96:1617– control in coronary patients across 27 countries:
late benefits of high-dose atorvastatin in patients 1621. results from the European Society of Cardiology
with acute coronary syndromes: results from the ESC-EORPEUROASPIRE V registry. Eur J Prev
14. Vasko P, Alfredsson J, Bäck M, et al. SWEDE-
PROVE IT-TIMI 22 trial. J Am Coll Cardiol. Cardiol. 2019;26:824–835.
HEART annual report. 2023. [Link]
2005;46:1405–1410.
se/swedeheart/dokument-sh/arsrapporter-sh 25. Ray KK, Aguiar C, Arca M, et al. Use of
5. Grundy SM, Cleeman JI, Merz CN, et al. Impli- combination therapy is associated with improved
15. Back M, Leosdottir M, Hagstrom E, et al. The
cations of recent clinical trials for the National LDL-cholesterol management: 1-year follow-up
SWEDEHEART secondary prevention and cardiac
Cholesterol Education Program Adult Treatment results from the European observational SAN-
rehabilitation registry (SWEDEHEART CR registry).
Panel III guidelines. Circulation. 2004;110:227– TORINI study. Eur J Prev Cardiol. 2024;31:1792–
Eur Heart J Qual Care Clin Outcomes. 2021;7:431–
239. 1803.
437.
6. Cannon CP, Blazing MA, Giugliano RP, et al. 26. Krychtiuk KA, Ahrens I, Drexel H, et al. Acute
16. Hernan MA, Robins JM. Using big data to
Ezetimibe added to statin therapy after acute LDL-C reduction post ACS: strike early and strike
emulate a target trial when a randomized trial is
coronary syndromes. N Engl J Med. 2015;372: strong: from evidence to clinical practice. A clinical
not available. Am J Epidemiol. 2016;183:758–764.
2387–2397. consensus statement of the Association for Acute
17. Suissa S. Immortal time bias in pharmaco-
7. Gencer B, Mach F, Murphy SA, et al. Efficacy of CardioVascular Care (ACVC), in collaboration with
epidemiology. Am J Epidemiol. 2008;167:492–
evolocumab on cardiovascular outcomes in pa- the European Association of Preventive Cardiology
499.
tients with recent myocardial infarction: a pre- (EAPC) and the European Society of Cardiology
specified secondary analysis from the FOURIER 18. R Core Team. The R project for statistical Working Group on Cardiovascular Pharmaco-
trial. JAMA Cardiol. 2020;5:952–957. computing. [Link] therapy. Eur Heart J Acute Cardiovasc Care.
2022;11:939–949.
8. Schwartz GG, Steg PG, Szarek M, et al. Alir- 19. Hougaard M, Hansen HS, Thayssen P, et al.
ocumab and cardiovascular outcomes after acute Influence of ezetimibe on plaque morphology in 27. Ramsaran E, Preusse P, Sundaresan D, et al.
coronary syndrome. N Engl J Med. 2018;379: patients with ST elevation myocardial infarction Adherence to blood cholesterol treatment guide-
2097–2107. assessed by optical coherence tomography: an lines among physicians managing patients with
OCTIVUS sub-study. Cardiovasc Revasc Med. atherosclerotic cardiovascular disease. Am J Car-
9. Grundy SM, Stone NJ, Bailey AL, et al. 2018
2020;21:1417–1424. diol. 2019;124:169–175.
AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/
APhA/ASPC/NLA/PCNA guideline on the manage- 20. Nicholls SJ, Kataoka Y, Nissen SE, et al. Effect 28. Sabouret P, Puymirat E, Kownator S, et al.
ment of blood cholesterol: a report of the Amer- of evolocumab on coronary plaque phenotype and Lipid-lowering treatment up to one year after
ican College of Cardiology/American Heart burden in statin-treated patients following acute coronary syndrome: guidance from a French
Association Task Force on Clinical Practice Guide- myocardial infarction. JACC Cardiovasc Imaging. expert panel for the implementation of guidelines
lines. J Am Coll Cardiol. 2019;73:e285–e350. 2022;15:1308–1321. in practice. Panminerva Med. 2023;65:244–249.
1564 Leosdottir et al JACC VOL. 85, NO. 15, 2025

Timing of Oral LLT Intensification After MI APRIL 22, 2025:1550–1564

29. Ya’Qoub L, Mansoor H, Elgendy IY. Upfront 34. Schubert J, Leosdottir M, Lindahl B, et al. 39. Ballantyne CM, Laufs U, Ray KK, et al. Bem-
combination of statin and ezetimibe for patients Intensive early and sustained lowering of nonhigh- pedoic acid plus ezetimibe fixed-dose combination
with acute coronary syndrome: time for a new density lipoprotein cholesterol after myocardial in patients with hypercholesterolemia and high
approach? J Am Heart Assoc. 2023;12:e031615. infarction and prognosis: the SWEDEHEART reg- CVD risk treated with maximally tolerated statin
istry. Eur Heart J. 2024;45:4204–4215. therapy. Eur J Prev Cardiol. 2020;27:593–603.
30. Byrne RA, Rossello X, Coughlan JJ, et al. 2023
ESC guidelines for the managementof acute coro- 35. Schubert J, Lindahl B, Melhus H, et al. Low- 40. Nissen SE, Lincoff AM, Brennan D, et al.
nary syndromes. Eur Heart J. 2023;44:3720–3826. density lipoprotein cholesterol reduction and Bempedoic acid and cardiovascular outcomes in
statin intensity in myocardial infarction patients tatin-intolerant patients. N Engl J Med. 2023;388:
31. Lewek J, Niedziela J, Desperak P, et al. Inten-
and major adverse outcomes: a Swedish nation- 1353–1364.
sive statin therapy versus upfront combination
wide cohort study. Eur Heart J. 2021;42:243–
therapy of statin and ezetimibe in patients with 41. Ohm J, JernbergT Johansson D, et al. Associ-
252.
acute coronary syndrome: a propensity score ation of clinical trial participation after myocardial
matching analysis based on the PL-ACS data. J Am 36. Zimerman A, O’Donoghue ML, Ran X, et al. infarction with socioeconomic status, clinical
Heart Assoc. 2023;12:e030414. Long-term cognitive safety of achieving very low characteristics, and outcomes. Eur Heart J Open.
LDL cholesterol with evolocumab. NEJM Evid. 2021;1:oeab020.
32. Gargiulo P, Basile C, Galasso G, et al. Strike
2025;4:evidoa2400112.
early–strike strong lipid-lowering strategy with
proprotein convertase subtilisin/kexin type 9 in- 37. Giugliano RP, Wiviott SD, Blazing MA, et al.
hibitors in acute coronary syndrome patients: real- Long-term safety and efficacy of achieving very KEY WORDS combination therapy,
world evidence from the AT-TARGET-IT registry. low levels of low-density lipoprotein cholesterol: ezetimibe, lipid-lowering therapy,
Eur J Prev Cardiol. 2024;31:1806–1816. a prespecified analysis of the IMPROVE-IT trial. low-density lipoprotein cholesterol,
JAMA Cardiol. 2017;2:547–555. myocardial infarction, statin
33. Makhmudova U, Samadifar B, Maloku A, et al.
Intensive lipid-lowering therapy for early 38. Gaba P, O’Donoghue ML, Park JG, et al. As-
achievement of guideline-recommended LDL- sociation between achieved low-density lipopro-
cholesterol levels in patients with ST-elevation tein cholesterol levels and long-term A PPE NDI X For supplemental tables, figures,
myocardial infarction (“Jena auf Ziel”). Clin Res cardiovascular and safety outcomes: an analysis of and references, please see the online version of
Cardiol. 2023;112:1212–1219. FOURIER-OLE. Circulation. 2023;147:1192–1203. this paper.