Received: 15 October 2018 Revised: 12 February 2019 Accepted: 1 March 2019

DOI: 10.1002/ccd.28194

ORIGINAL STUDIES

Long-term impact of diabetes in patients with ST-segment

elevation myocardial infarction: Insights from the
EXAMINATION randomized trial
Pilar Jimenez-Quevedo MD, PhD1 | Salvatore Brugaletta MD, PhD2 |
Angel Cequier MD, PhD3 | Andrés Iñiguez MD, PhD4 | Antonio Serra MD, PhD5 |
Vicente Mainar MD, PhD6 | Gianluca Campo MD, PhD7 | Maurizio Tespili MD, PhD8 |
Luis Nombela-Franco MD, PhD1 | Maria del Trigo MD1 | Nieves Gonzalo MD, PhD1 |
Javier Escaned MD, PhD1 | Pablo Salinas MD, PhD1 | Ivan Nuñez-Gil MD, PhD1 |
Cristina Fernandez-Perez MD, PhD1 | Antonio Fernández-Ortiz MD, PhD1 |
Carlos Macaya MD, PhD1 | Patrick W. Serruys MD, PhD9 | Manel Sabate Tenas MD, PhD2

1
Interventional Cardiology Department, Clinico
San Carlos University Hospital, IdISSC, Madrid, Abstract
Spain Background: Long-term outcomes of diabetic patients suffering from ST-segment elevation
2
Interventional Cardiology Department, myocardial infarction (STEMI) and treated with second-generation drug-eluting stent have been
University Hospital Clínic, Institut
scarcely evaluated. The aim of this posthoc subanalysis of the EXAMINATION trial was to com-
d'Investigacions Biomèdiques August Pi i
Sunyer (IDIBAPS), Barcelona, Spain pare 5-year outcomes according to the presence of diabetes mellitus.
3
Interventional Cardiology Department, Methods: From a total of 1,497 patients included in the trial, 258 were diabetics (n = 137, received
University Hospital of Bellvitge, Barcelona, everolimus-eluting stent (EES) and n = 121 bare-metal stent (BMS); whereas 1,239 were nondia-
Spain betics (n = 613 treated with EES and n = 626 with BMS). Patient-oriented combined endpoint
4
Interventional Cardiology Department, (POCE) defined as all-cause death, any MI or any revascularization, and other clinical parameters
Hospital do Meixoeiro, Vigo, Spain
5
were collected up to 5-years. All results were adjusted for various potential confounders.
Interventional Cardiology Department,
Results: At 5-years, patients with diabetes showed similar rates of POCE between diabetics
University Hospital of Sant Pau, Barcelona,
Spain treated with EES and those treated with BMS (32.8% vs. 32.2%; p = 0.88). However, rates of TLR
6
Interventional Cardiology Department, were significantly lower in the EES group (4.4% vs. 9.9%; HR 0.52 (0.29–0.94); P = 0.03). In non-
Hospital General of Alicante, Alicante, Spain diabetics, the use of EES was associated with a significant improvement in all-clinical parameters
7
Interventional Cardiology Department, except for MI rate: POCE: [10.0% vs. 12.6%; HR 0.78(0.62–0.98); P = 0.038], all cause death:
University Hospital Ferrara, Ferrara, Italy
[7.0% vs. 12.1%; HR 0.62(0.42–0.90); P = 0.014], and [TLR: 4.2 vs. 6.7; HR 0.60 (0.37–0.98);
8
Interventional Cardiology Department,
P = 0.04]. Overall, diabetics showed higher rate of POCE at 5-years (32.6% vs. 21.5% in nondia-
University Hospital Bolognini Seriate,
Bergamo, Italy betics HR1.45[1.03–2.04];p = 0.03) driven by increased rates of MI and the need for revasculariza-
9
Interventional Cardiology Department, tion that occurred in coronary segments remote from target lesions [2.7% vs. 1.1%; HR: 2.27
International Centre of Circulatory Health, (1.12–5.23); P = 0.02 and 14% vs. 6.2%; HR: 2.11 (1.38–3.22); P = 0.001, respectively].
Imperial College London, London, United
Conclusions: Diabetics had worse clinical outcomes than nondiabetics after STEMI mainly due
Kingdom
to atherosclerosis progression. At 5-years, the treatment with EES did not reduce the rate of
Correspondence
Pilar Jimenez-Quevedo, MD, PhD, POCE in diabetics but reduced the need for revascularization compared with BMS.
Interventional Cardiology Department,
Hospital Clinico San Carlos, IdISSC, c/Martín KEYWORDS
Lagos s/n, 28040 Madrid, Spain.
Email: patropjq@gmail.com atherosclerosis progression, diabetes mellitus, drug eluting stent, myocardial infarction,

Funding information percutaneous coronary interventions

Spanish Heart Foundation

Catheter Cardiovasc Interv. 2019;1–9. wileyonlinelibrary.com/journal/ccd © 2019 Wiley Periodicals, Inc. 1

2 JIMENEZ-QUEVEDO ET AL.

1 | I N T RO D UC T I O N or contrast material, chronic antivitamin K agent treatment and STEMI

secondary to stent thrombosis. The local ethics committee from each
The metabolic disorders that occur in patients with diabetes lead to center approved the study protocol and all patients signed the informed
an increase in vascular inflammation, the development of platelet, and consent. The primary end point of the study was the POCE a combina-
1 tion of all cause of death, any myocardial infarction (MI) or any revascu-
endothelial dysfunction and a prothrombotic state. All of these mech-
anisms may be involved in the increased risk of diabetics for develop- larization at 1 year follow-up.17 Secondary endpoints included DOCE
ing coronary artery disease and presenting worse clinical outcomes (cardiac death, target vessel myocardial infarction, or target lesion revas-
after acute coronary syndromes and following any mode of revascu- cularization) and other clinical parameters. Diabetic patients were
2,3
larization. Several studies have assessed the impact of diabetes in defined as those patients who were treated with insulin or hypoglyce-
patients with ST elevation myocardial infarction (STEMI) and primary mic agents at the time of the PCI. For the purpose of this subanalysis,
percutaneous coronary intervention (PCI). Overall, the presence of we compared the 1- and 5-year outcomes of diabetic patients treated
diabetes appeared to increase the risk of restenosis and conferred a with EES or BMS and compared to those of nondiabetics.
4–8
poorer clinical outcome than patients without diabetes.
Previous randomized trials performed in diabetic patients have 2.2 | Procedure
demonstrated that the use of first generation drug-eluting stent (DES)
At the index procedure unfractionated heparin or bivalirudin was used
reduced both angiographic and clinical restenosis when compared
for procedural anticoagulation. Glycoprotein IIb-IIIa inhibitors were
with bare metal stent (BMS).9,10 However, data comparing second
administrated at the discretion of the investigator. Manual thrombectomy
generation DES (everolimus-eluting stent, EES) with first generation
before stent implantation was recommended. Dual antiplatelet treatment
DES in diabetics showed the superiority of EES to paclitaxel-eluting
with aspirin and clopidogrel was prescribed for at least 1 year in both
stent in terms of in-stent late loss11 and target-vessel failure.12 Only
groups. In the case of staged procedures, the recommendation was to
one small randomized study showed similar late loss and clinical out-
implant the same type of stent as per randomization and perform the
come between EES and sirolimus eluting stent at 4-years.13 Thus, a
revascularization within the first month following discharge.
metanalisis performed in patients with diabetes and coronary artery
disease undergoing stenting, EES was the most efficacious and safe
DES.14 2.3 | Statistical analyses
The EXAMINATION (clinical Evaluation of the Xience-V stent in Continuous variables were expressed as mean ± standard deviation and
Acute Myocardial Infarction) study was a multicenter prospective ran- categorical variables as percentages. All comparisons between groups
domized all-comer controlled trial that compared EES versus BMS in were performed by Student's t-test for quantitative variables and Fish-
patients with STEMI.15 At 1 year, the patient-oriented combined end- er's exact test for categorical variables. The analyses of the clinical out-
point (POCE), was similar between groups, however, target lesion comes at 1- and 5-year was stratified with backward Cox regression
revascularization and stent thrombosis rates were lower in patients models including those variables with a P < 0.1 or those clinically rele-
treated with EES. Furthermore, at 5 years there was a significant vant. The assumption of the proportional hazard was verified. To ana-
reduction in the POCE in the EES group, mainly driven by a reduction lyze clinical outcome in the diabetic subgroup, we have used the
in all-cause mortality, as well as, in the device-oriented combined end- propensity score adjustment. In the nondiabetic subgroup the variables
point (DOCE), owing to a reduction in target lesion revascularization finally included in the model that compared EES versus BMS were: type
(TLR).16 To the best of our knowledge, there is not any study that has of stent, sex, hypercholesterolemia, previous MI, clinical status at admis-
compared the long-term performance of current generation stents in sion. The variables finally included in the model that compared diabetics
diabetic patients with STEMI. Therefore, this substudy is aimed to versus nondiabetics were: age, BMI, smoking status, hypertension,
assess the long-term clinical impact of diabetes in patients with STEMI hypercholesterolemia, family history, previous stroke, cardiogenic shock,
treated with EES or BMS. and multivessel disease. All data were presented by hazard ratio and
95% confidence interval. All statistical analyses were performed using
STATA (version 9.0) software, and all reported P-values were two-sided.
2 | MATERIAL AND METHODS We assumed significance at the 5% level (P < 0.05).

2.1 | Patients and study design

3 | RE SU LT S
The EXAMINATION trial was a multicenter prospective controlled
trial that compared clinical performance of EES versus BMS in patients
with STEMI. The study design has been reported elsewhere.17 The
3.1 | Baseline and procedural characteristics
current study involves a post-hoc analysis (Trial Registration. Baseline and procedural characteristics according to the type of stent
NCT00828087). In brief, any patient with a STEMI within the first and the presence of diabetes are presented in Table 1. In the diabetic
48 hr after symptom onset that required emergent percutaneous group, all variables were well matched except for the presence of
revascularization in a vessel size between 2.25 and 4 mm was preinfarction angina, which occurred more commonly in the BMS
included. Main exclusion criteria were age younger than 18, known group. In the nondiabetic group, hypercholesterolemia was more fre-
intolerance to aspirin, clopidogrel, heparin, stainless steel, everolimus, quently observed in the BMS group. Out of the 1,498 patients
JIMENEZ-QUEVEDO ET AL. 3

TABLE 1 Baseline and procedural characteristics of diabetic and non-diabetic patients according to the type of stent implanted

Diabetic patients (n = 258) Non diabetic patients (n = 1,235)

EES group BMS group EES group BMS group
CLINICAL VARIABLES (n = 137) (n = 121) P value (n = 611) (n = 624) P value
Age-years 65.3 ± 11. 5 66.3 ± 11.7 0.47 59.7 ± 12.1 60.6 ± 12-5 0.18
Male sex 112 (81.8) 100 (82.6) 0.85 521 (85) 510 (81. 5) 0.97
BMI 27.8 ± 4.1 28.3 ± 3.9 0.37 27.1 ± 3.6 27.2 ± 3.8 0.50
Previous smoker 78 (56.9) 77 (63.6) 0.27 466 (76.0) 461 (73.6) 0.33
Hypertension 86 (62.8) 89 (73.6) 0.06 261 (42.6) 289 (46.2) 0.20
Hypercholesterolemia 75 (54.7) 60 (49.6) 0.40 279 (45. 5) 385 (61. 5) 0.01
Prior MI 9 (6.6) 8 (6.6) 0.98 24 (3.9) 39 (6.2) 0.06
Prior PCI 8 (5.8) 6 (5.0) 0.75 21 (3. 4) 26 (4.2) 0.50
Prior CABG 1 (0.7) 0 (0) >0.99 2 (0.3) 7 (1.1) 0.17
Prior stroke 2 (1. 5) 7 (5.8) 0.08 10 (1.6) 12 (1.9) 0.70
Pre-MI angina 36 (26.3) 48 (39.7) 0.02 169 (27.6) 180 (28.8) 0.63
Primary PCI 108 (78.8) 105 (86.8) 0.09 521 (85) 533 (85.3) 0.88
Clinical status on admission, n(%) 0.07 0.83
Killip I 111 (81) 109 (90.1) 557 (91.2) 559 (89.7)
Killip II 19 (13.9) 10 (8.3) 40 (6. 5) 46 (7. 4)
Killip III 2 (1.5) 2 (1.7) 8 (1.3) 11 (1.8)
Killip IV 5 (3.6) 0 (0) 38 (6.2) 42 (6.7)
Multivessel disease, n(%) 18 (13.1) 19 (15.7) 0.55 82 (13.4) 69 (11.0) 0.20
LAD as infarcted related artery 63 (46) 47 (38.8) 0. 24 259 (42.3) 258 (41.2) 0.71
EES group BMS group EEEs group BMS group
PROCEDURAL CHARACTERISTICS (n = 137) (n = 121) P value (n = 611) (n = 624) P value
TIMI flow before PCI 0–1 86 (62.8) 70 (58.3) 0.46 407 (66.7) 430 (69.2) 0.34
Unfractioned heparin 108 (100) 94 (100) 490 (100) 496 (100)
Aspirin before PCI 129 (100) 114 (100) 584 (100) 584 (100)
Clopidogrel before PCI 127 (100) 117 (100) 564 (100) 580 (100)
Glycoprotein IIb-IIIa inhibitors 61 (44.5) 54 (44.6) 0.98 339 (55.3) 331 (52.9) 0.39
Manual thrombectomy 86 (62.8) 69 (57) 0.34 409 (66.7) 412 (65.8) 0.73
Direct stenting 72 (53.7) 56 (46.7) 0. 26 378 (62.6) 378 (62) 0.82
Postdilatation 17 (12.4) 21 (17.4) 0. 26 101 (16.5) 82 (13.1) 0.09
Overlapping stent 36 (26.3) 41 (33.8) 0.47 161 (26. 4) 165 (26.4) 0.07
Number stent, mean ± SD 1.41 ± 0.72 1.45 ± 0.69 0.66 1.39 ± 0.64 1.36 ± 0.62 0.51
Total stent length median (IQR) 23.0 (18.0–35.0) 23.0 (18.0–35.0) 0.80 23.0 (18–35) 23.0 (18.0–32. 5) 0.43
TIMI flow after PCI 2-3 132 (96.4) 120 (99.2) 0.21 596 (97.5) 606 (97.3) 0.76

BMI, body mass index; MI, myocardial infarction; PCI, percutaneous coronary intervention; CABG, coronary artery bypass graft; LAD, left anterior
descending,

randomized in the trial, one patient was not finally included in this were performed between 1 and 5 years. After 1 year the rate of repeat
study due to de lack of information regarding the presence of diabe- revascularization were similar between groups as well as other clinical
tes. This patient had not any event during the follow-up. Complete parameters. (Supporting Information Table ).
5-years follow-up were obtained in 97% of both groups. Mean follow-
up duration was 1,699 ± 360 days.
3.3 | Clinical outcomes at 1 and at 5 years in
nondiabetic patients according to the type of stent
3.2 | Clinical outcomes at 1 and at 5 years in diabetic implanted
patients according to the type of stent implanted In nondiabetic patients, at 1 year, there was a trend towards lower
At 1 year, rates of POCE, DOCE, as well as of other clinical parameters rate of POCE in the EES arm whereas rates of DOCE, death, and MI
such as death, MI and stent thrombosis were similar between diabetic were comparable between groups (Table 2). However, rates of repeat
patients treated EES versus BMS (Table 2). In contrast, diabetic patients revascularization (TLR), target vessel revascularization (TVR) and
treated with EES showed a significantly reduced rate of repeat revascu- definitive and probable thrombosis were significantly lower in the EES
larization compared with BMS. The beneficial effect in terms of need for arm at 1 year. Likewise, the rate of bleeding was lower in the EES
revascularization persisted up to 5 years (Figure 1). Landmark analyses group mainly due to a significant reduction in minor bleeding.
 4

TABLE 2 Clinical outcomes of diabetic and nondiabetic patients treated with Everolimus eluting stent compare with bare metal stent at 1 and 5 years (adjusted comparison)

1 year outcomes of diabetic patients 1 year outcomes of nondiabetic patients

EES Group BMS group Adjusted HR EEEs group BMS group Adjusted HR
1-year outcomes (n = 137) (n = 121) (95% IC) P value (n = 611) (n = 624) (95% IC) P value
Patient oriented combined endpoint, n (%) 26 (19.0) 26 (21.5) 0.83 (0.49–1.40) 0.49 61 (10.0) 79 (12.6) 0.78 (0.70–1.00) 0.05
Device oriented combined endpoint, n (%) 13 (9. 5) 12 (9.9) 0.98 (0.49–1.96) 0.95 29 (4.7) 49 (7.8) 0.61 (0.38–0.97) 0.40
All-cause death n (%) 10 (7.3) 4 (3.3) 1.54 (0.45–5.12) 0.48 16 (2.6) 22 (3. 5) 0.88 (0.45–1.70) 0.71
Cardiac death, n (%) 9 (6.6) 3 (2.5) 1.71 (0.43–6.80) 0.44 15 (2. 4) 18 (2.9) 1.01 (0.50–2.04) 0.95
Myocardial infarction (WHO ext definition), n (%) 2 (1.5) 5 (4.1) 0.33 (0.06–1.81) 0.20 8 (1.3) 10 (1.6) 0.78 (0.30–1.99) 0.60
Revascularization, n (%) 16 (11.7) 22 (18.2) 0.59 (0.40–0.95) 0.04 44 (7.2) 57 (9.1) 0.76 (0.67–0.99) 0.04
Target lesion 4 (2.9) 9 (7. 4) 0.41 (0.21–0.74) 0.03 12 (2.0) 28 (4.5) 0.40 (0.20–0.80) 0.01
Target vessel 6 (4.4) 11 (9.1) 0.53 (0.25–0.95) 0.04 22 (3.6) 40 (6.4) 0.52 (0.31–0.89) 0.017
Stent thrombosis (ARC) 6 (4.4) 3 (2.5) 2.39 (0.59–9.74) 0.22 10 (1.6) 19 (3.0) 0.55 (0.25–1.20) 0.13
Definite/probable ST, n (%) 3 (2.2) 2 (1.7) 1.76 (0.29–10.78) 0.54 4 (0.7) 17 (2.7) 0.23 (0.08–0.71) 0.01
Definite 2 (1. 5) 1 (0.8) 2.20 (0.19–25.99) 0.52 2 (0.3) 13 (2.1) 0.15 (0.03–0.68) 0.014
Probable 1 (0.7) 1 (0.8) 1.23 (0.07–20.33) 0.88 2 (0.3) 4 (0.6) 0.51 (0.09–2.80) 0.43
Possible 3 (2.2) 1 (0.8) 4.49 (0.35–34.17) 0.28 6 (1.0) 2 (0.3) 3.5 (0.70–17.59) 0.12
Bleeding, n (%) 9 (6.6) 6 (5.0) 1.29 (0.44–3.72) 0.63 20 (3.3) 33 (5.3) 0.60 (0.37–0.96) 0.03
Major 4 (2.9) 2 (1.7) 1.72 (0.30–9.83) 0.53 5 (0.8) 9 (1.4) 0.58 (0.19–1.75) 0.33
Minor 6 (4.4) 4 (3.3) 1.32 (0.36–4.82) 0.67 15 (2.4) 26 (4.2) 0.57 (0.26–0.97) 0.04
5-year outcomes EES group BMS group 95% IC P value EES group BMS group 95% IC P value
Patient oriented endpoint, n (%) 45 (32.8) 39 (32.2) 0.97 (0.62–1.50) 0.88 61 (10.0) 79 (12.6) 0.78 (0.62–0.98) 0.038
Device oriented endpoint, n (%) 28 (20.4) 21 (17.3) 1.11 (0.62–1.97) 0.72 29 (4.7) 49 (7.8) 0.60 (0.38–0.96) 0.035
All-cause death n (%) 22 (16.1) 12 (9.9) 1.47 (0.71–3.02) 0.29 43 (7.0) 76 (12.1) 0.62 (0.42–0.90) 0.014
Cardiac death, n (%) 18 (13.1) 7 (5.8) 1.90 (0.78–4.68) 0.16 29 (4.7) 48 (7.7) 0.67 (0.48–0.99) 0.049
Myocardial infarction (WHO ext definition), n (%) 9 (6.6) 7 (5.8) 1.14 (0.41–3.12) 0.80 26 (4.2) 20 (3.2) 1.29 (0.72–2.33) 0.38
Revascularization, n (%) 23 (16.8) 29 (24.0) 0.71 (0.54–0.92) 0.012 70 (11.4) 87 (13.9) 0.79 (0.68–1.00) 0.05
Target lesion 6 (4.4) 12 (9.9) 0.52 (0. 29–0.94) 0.03 26 (4.2) 42 (6.7) 0.60 (0.37–0.98) 0.04
Target vessel 10 (7.3) 16 (13.2) 0.63 (0.41–0.98) 0.04 39 (6. 4) 60 (9.6) 0.63 (0.42–0.94) 0.024
Stent thrombosis 11 (8.0) 6 (5.0) 1.99 (0.72–5.47) 0.14 28 (4.6) 49 (7.8) 0.60 (0.37–0.96) 0.035
Definite/probable stent thrombosis, n (%) 4 (2.9) 2 (1.7) 2.13 (0.38–11.93) 0.36 11 (1.8) 21 (3. 4) 0.52 (0.34–1.00) 0.05
Definite 3 (2.2) 1 (0.8) 2.9 (0.30–28.92) 0.36 9 (1. 5) 17 (2.7) 0.52 (0.23–1.19) 0.12
Probable 1 (0.7) 1 (0.8) 1.23 (0.75–20.33) 0.85 3 (0. 5) 4 (0.6) 0.75 (0.17–3.40) 0.71
Possible 7 (5.1) 4 (3.3) 1.85 (0.53–6.42) 0.33 17 (2.8) 28 (4.5) 0.69 (0.37–1.27) 0.23
Bleeding, n (%) 9 (6.6) 8 (6.6) 1.03 (0.39–2.73) 0.85 41 (6.7) 44 (7.0) 0.95 (0.62–1.46) 0.82
Major 4 (2.9) 2 (1.7) 1.73 (0.30–9.83) 0.53 8 (1.3) 11 (1.8) 0.75 (0.30–1.89) 0.55
Minor 6 (4. 4) 6 (5.0) 0.96 (0.30–3.04) 0.95 34 (5.5) 35 (5.6) 0.98 (0.61–1.58) 0.95

Abbreviations: ARC, American research consortium; Myocardial infarction (WHO extended definition), myocardial infarction (World health organization extended definition).
JIMENEZ-QUEVEDO ET AL.
JIMENEZ-QUEVEDO ET AL. 5

FIGURE 1 Kaplan–Meier event curves comparing everolimus-eluting stents (EES) and bare metal stent (BMS) in diabetic patients for target lesion
revascularization (propensity score adjustment)

At 5-years a beneficial effect of using EES in nondiabetic patients was lower than in nondiabetic patients (Tables 3 and 4 and Figure 3).
was observed in most clinical parameters, including POCE (Figure 2), Conversely, nondiabetics were more often previous smokers, had
DOCE, all cause and cardiac death, TLR, TVR, and overall stent throm- more frequently family history of cardiovascular disease and TIMI
bosis. In addition, there was a trend toward lower definitive and prob- 0 or I prior to revascularization. At 5-year follow-up, incidence of
able stent thrombosis in the EES group. POCE was significantly higher in diabetic patients (32.6% vs. 21. 5%;
HR 1.45(1.03–2.04); P = 0.03) driven by an increase in the rates of
MI and revascularization in coronary segments remote from target
3.4 | Five-year clinical outcomes of diabetic
lesions (2.7% vs. 1.1%; HR: 2.27 (1.12–5.23); P = 0.02 and 14%
compared with nondiabetic patients with STEMI vs. 6.2%; HR: 2.11 (1.38–3.22); P = 0.001, respectively). Of note,
Overall, Diabetic patients were older, with higher BMI and more fre- rates of TLR and TVR were similar in diabetics and nondiabetics at
quently hypercholesterolemic. In addition, the use of IIb-IIIa inhibitors 5 years.

FIGURE 2 Kaplan–Meier event curves comparing everolimus-eluting stents (EES) and bare metal stent (BMS) in non-diabetic patients for the
composite patient-oriented combined endpoint (POCE) of all death, any myocardial infarction, and any revascularization (adjusted comparison)
6 JIMENEZ-QUEVEDO ET AL.

TABLE 3 Baseline characteristics of diabetic patients compared with nondiabetic patients

BASELINE CHARACTERISTICS Diabetic group (n = 258) Nondiabetic group (n = 1,239) P value

Age-year 65.8 ± 11.6 60.2 ± 12.3 <0.001
Male sex 212 (82.2) 1,031 (83.2) 0.71
BMI 28.18 ± 4.1 27.2 ± 3.7 0.001
Previous smoker 155 (60.1) 927 (74.8) <0.001
Hypertension 175 (67.8) 550 (44. 4) <0.001
Hypercholesterolemia 135 (52.3) 520 (42.0) 0.002
Family history 30 (11.6) 223 (18.0) 0.013
Prior MI 17 (6.6) 62 (5.1) 0.36
Prior PCI 14 (5. 4) 47 (3.8) 0.22
Prior CABG 1 (0. 4) 9 (0.7) 0.54
Prior stroke 9 (3. 5) 22 (1.8) 0.09
PreMI angina 84 (32.6) 349 (28.2) 0.17
Primary PCI 213 (82.6) 1,054 (85.1) 0.29
Clinical status on admission, n (%) 0.07
Killip I 220 (85.3) 1,116 (90.4)
Killip II 29 (11.2) 86 (7.0)
Killip III 4 (1.6) 19 (1.5)
Killip IV 5 (1.9) 13 (1.1)
Multivessel disease, n (%) 37 (14.3) 151 (12.2) 0.35
LAD as infarcted-related artery 110 (42.6) 517 (41.7) 0.83
PROCEDURAL CHARACTERISTICS DM group (n = 258) No DM group (n = 1,239) P value
TIMI flow before PCI 0–1 156 (60.7) 837 (68.0) 0.02
Unfractioned heparin 202 (78.3) 986 (79.6) 0.67
Aspirin before PCI 137 (100) 613 (100)
Clopidogrel before PCI 137 (100) 613 (100)
Glycoprotein IIb-IIIa inhibitors 115 (44.6) 670 (54.1) 0.006
Manual thrombectomy 155 (60.1) 821 (66.3) 0.06
Direct stenting 128 (50. 4) 756 (62.3) 0.001
Postdilatation 38 (14.7) 183 (14.8) >0.99
Overlapping stent 77 (29.8) 326 (26.3) 0.46
Number stent (mean [SD]) 1.43 ± 0.70 1.38 ± 0.63 0.25
Total stent length (mean [SD]) 28.05 ± 15.10 27.39 ± 13.77 0.49
TIMI flow after PCI 2-3 252 (97.7) 1,202 (97.4) >0.99

BMI, Body mass index; MI, myocardial infarction; PCI, percutaneous coronary intervention; CABG, Coronary artery bypass grafting; LAD, left anterior
descending.

4 | DISCUSSION First generation DES demonstrated to reduce the need for repeat
revascularization compared to BMS in patients with STEMI.18 The use
Main findings of the present study can be summarized as follows: first, of second generation DES has extended the benefit beyond target
in the diabetic group POCE at 1 year (primary end-point of the study) vessel revascularization. Indeed, a pooled analysis of the EXAMINA-
and POCE at 5-years were not different in patients treated with EES TION and COMFORTABLE-AMI trials demonstrated also a reduction
compared with BMS. Conversely, diabetic patients with STEMI in target vessel myocardial infarction rate by the use of newer genera-
treated with EES had lower rate of repeat revascularization at target tion DES.19 In the diabetic STEMI population, however, data compar-
lesion up to 5-year follow-up. Second, in nondiabetics POCE at 1-year ing outcomes of different stent types is scarce. A subgroup analysis of
was similar between patients treated with EES compared with BMS the HORIZONS AMI trial (“the Harmonizing outcome with revasculari-
but at 5 years was significantly lower in patients treated with EES. In zation and stents in acute myocardial infarction”) revealed that dia-
addition, nondiabetic patients treated with EES presented significant betic patients treated with paclitaxel-eluting stent had a reduction in
reductions in most clinical events both at 1 and at 5 years. Third, dia- TLR compared to BMS at 1 year without an increase in the rates of
betic patients as compared to nondiabetics, showed worse clinical death, reinfarction, stroke, or thrombosis. A pooled analysis of individ-
outcomes due to progression of atherosclerosis as expressed by an ual patient data from seven randomized trials including a total of
increase rates in non-target vessel MI and non-TVR. 389 diabetic patients showed that first generation DES reduced the
JIMENEZ-QUEVEDO ET AL. 7

TABLE 4 Five years follow-up of diabetic versus nondiabetic patients (adjusted comparison)

Five-year follow-up clinical events Diabetics (n = 258) Nondiabetics (n = 1,239) Adjusted HR [95% IC] P value
Patient oriented endpoint, n (%) 84 (32.6) 267 (21.5) 1.45 (1.03–2.04) 0.03
Device oriented endpoint, n (%) 49 (19.0) 162 (13.1) 1.16 (0.7–1.88) 0.53
All-cause death, n (%) 34 (13.2) 119 (9.6) 1.04 (0.69–1.58) 0.82
Cardiac death, n (%) 25 (9.7) 77 (6.2) 1.13 (0.69–1.86) 0.61
Myocardial infarction (WHO extended definition), n (%) 16 (6.2) 46 (3.7) 1.58 (1.00–2.92) 0.048
Non TV-MI 7 (2.7%) 14 (1.1) 2.27 (1.12–5.23) 0.02
Revascularization, n (%) 52 (20.2) 157 (12.7) 1.58 (1.13–2.20) 0.007
Target lesion 18 (7.0) 68 (5.5) 1.28 (0.74–2.20) 0.36
Target vessel 26 (10.1) 99 (8.0) 1.26 (0.80–1.98) 0.30
Non-TV revascularization 36 (14%) 77 (6.2%) 2.11 (1.38–3.22) 0.001
Stent thrombosis 17 (6.6) 77 (6.2) 0.79 (0.44–1.39) 0.42
Definite/probable stent thrombosis, n (%) 6 (2.3) 32 (2.6) 0.84 (0.34–2.10) 0.72
Definitive 4 (1.6) 26 (2.1) 0.78 (0.26–2.32) 0.66
Probable 2 (0.8) 7 (0.6) 0.99 (0.19–5.20) 0.99
Possible 11 (4.3) 45 (3.6) 0.82 (0.39–1.72) 0.61
Bleeding, n (%) 17 (6.6) 85 (6.9) 0.78 (0.45–1.36) 0.39
Major 6 (2.3) 19 (1. 5) 1.03 (0.37–2.86) 0.94
Minor 12 (4.7) 69 (5.6) 0.74 (0.39–1.41) 0.37

Myocardial infarction (WHO ext definition), Myocardial infarction World Health Organization extended definition; Non TV-MI, Myocardial infarction no
target vessel related; Non-TV revascularization, no target vessel revascularization.

FIGURE 3 Kaplan–Meier event curves comparing diabetic patients (DM) and nondiabetic patients (no-DM) for the composite patient-oriented
combined endpoint (POCE) of all death, any myocardial infarction, and any revascularization (adjusted comparison)

need for repeat revascularization up to 24 months follow-up.20 We overall stent thrombosis was also reduced in the EES group. This dif-
herein have reported that this benefit may be extended up to 5 years ferential effect of the stent between diabetics and nondiabetics may
by the use of second generation EES with no concerns on late catch- be explained by several factors. First, the number of known diabetics
up phenomenon.21 at the time of the STEMI presentation is limited (and usually lower than
In the analyses restricted to the nondiabetic population, results in stable coronary artery disease). This may contribute to the lack of
were consistent with those of the entire EXAMINATION trial showing power to provide significant differences between groups. Second,
a greater clinical effect by the use of EES at 5 years. Indeed, a signifi- from the pathophysiological point of view, coronary artery disease in
cant reduction in POCE, DOCE, all-cause death and revascularization diabetic patients is more aggressive. The exaggerated amount of
(TLR, TVR) rates were observed in the EES group. In addition, rate of neointima hyperplasia after stent implantation that characterizes
8 JIMENEZ-QUEVEDO ET AL.

diabetics is only the “tip of the iceberg.”22 Diabetics have a profound high risk group of patients. Progression of atherosclerosis analysis was
23
metabolic disorder that causes extensive coronary atherosclerosis only limited to clinical events that occurred at nontarget vessels.
with coronary plaques, that are eventually prone to rupture or disease Actual atherosclerosis progression not leading to revascularization or
progression. Therefore, long-term events cannot be fully preventable MI could not be assessed by invasive or noninvasive techniques.
only by improved stent design.
Studies assessing the impact of diabetes in patients with STEMI
showed an increase in mortality, MI, stent thrombosis and 5 | CONC LU SION
4–8 6
revascularization in this subgroup of patients. In a meta-analysis
In conclusion, this study reported for the first time that percutaneous
that compared the outcomes between diabetics and nondiabetics with
coronary intervention with second generation DES in diabetic patients
STEMI, the data from 11 randomized trials comparing DES with BMS
with STEMI was not associated with a reduction of POCE or DOCE,
were analyzed. In this study, 972 diabetic patients and 5,326 nondia-
however, a significant reduction of the rate of repeat revascularization
betics were included, with 63% treated with DES in both groups. In
at target lesion at 1- and 5-year were observed. In the non-diabetic
keeping with other studies death, re-MI, stent thrombosis and TLR,
population the use of EES was associated with a reduction of POCE,
occurred more frequently in diabetic patients. However, when
DOCE, and stent thrombosis at 5-years follow-up. In addition, the
patients treated with DES were analyzed no difference in TVR rate
overall diabetic population included in this study experienced worse
was observed in diabetics compared to nondiabetics. In this study, dia-
clinical outcomes compared to nondiabetic patients due to events that
betics had higher rate of nontarget vessel MI and repeat revasculariza-
occurred in coronary segments remote from target lesions. These
tion, but TLR and TVR rates were similar among diabetics and
findings emphasize the important role of atherosclerosis progression
nondiabetics suggesting that EES is able to inhibit neointimal prolifer-
in the long-term clinical outcome in this subpopulation. Dedicated
ation but indeed cannot prevent MI or revascularization in other terri-
studies in diabetic patients are warranted to further improve out-
tories remote from target lesion.
comes in this challenging group of patients.
Progression of atherosclerosis is an important cause of repeat
revascularization after percutaneous revascularization especially
beyond 1 year when restenosis is almost overcome.24,25 In this regard,
ACKNOWLEDGMENT
at 5-year follow-up of the SIRIUS trial26 almost a third of the MI
reported was MI's located in the nontarget vessel (28%,) in the SES The sponsor of the study had no role in the study design. Funding

group, and 17%, in the BMS group. In addition, the cumulative revas- Spanish Heart Foundation. The promoter funded an independent data

cularization incidence was 32.3% in the SES group but only 12.5% management and analysis centre (Cardialysis, Rotterdam, Netherlands)

were due to TLR. In the same way, but in a dedicated study performed for database management.

in diabetics, half of all revascularizations were due to disease progres-

sion and one third of all MI were at nontarget sites at 5 years.27 Our
DISC LOSURES
data is in keeping with the increased risk of atherosclerosis progres-
sion in diabetics observed in other studies28 In patients after STEMI, Pilar Jimenez-Quevedo: no relationship with industry. Manel Sabate:
diabetic patients experienced more events not related to the treated consultant for Abbott vascular Dr. Brugaletta has received speaker's
vessel and this played an important role in the poorer outcome fees from Abbott vascular. All other authors have reported that they
observed in this population. have no relationships relevant to the contents of this paper to disclose.

ORCID
4.1 | Limitations
Pilar Jimenez-Quevedo https://orcid.org/0000-0001-7562-8677
This study represents a post hoc analysis of a randomized trial and the
Andrés Iñiguez https://orcid.org/0000-0002-7758-6252
results should be interpreted with caution. In addition, the number of
Luis Nombela-Franco https://orcid.org/0000-0003-3438-8907
diabetics is relatively limited and therefore the study is underpowered
Pablo Salinas https://orcid.org/0000-0003-4040-4029
to demonstrate significant differences in clinical events. Potential for
Ivan Nuñez-Gil https://orcid.org/0000-0002-1779-3102
type 1 or type 2 errors due to the very small numbers of patients,
Patrick W. Serruys https://orcid.org/0000-0002-9636-1104
especially when comparing infrequent outcomes. Previous studies
have shown that up to one-third of patients who underwent PCI are
misclassified as nondiabetic patient and have a higher risk compared RE FE RE NC ES
to nondiabetic pts.29,30 In this study only known diabetic patient 1. Creager MA, Lüscher TF, Cosentino F, Beckman JA. Diabetes and vas-
treated with oral agents or insulin were defined as diabetics and this cular disease: Pathophysiology, clinical consequences, and medical
therapy: part I. Circulation. 2003;108:1527-1532.
may constitute a limitation that has to be addressed in future studies.
2. Malmberg K, Yusuf S, Gerstein HC, et al. Impact of diabetes on long-term
Data regarding glycaemic control or requirement for insulin in the dia- prognosis in patients with unstable angina and non-Q-wave myocardial
betic population are not available and, therefore, we cannot define infarction: Results of the OASIS (organization to assess strategies for
ischemic syndromes) registry. Circulation. 2000;102:1014-1019.
the influence of these factors on the 5-year clinical outcomes. This
3. Barsness GW, Peterson ED, Ohman EM, et al. Relationship between
study in STEMI patients was performed in the “clopidogrel-era” and diabetes mellitus and long-term survival after coronary bypass and
therefore this may influence in the clinical outcomes specially in this angioplasty. Circulation. 1997;96:2551-2556.
JIMENEZ-QUEVEDO ET AL. 9

4. De Luca G, Sauro R, Varricchio A, et al. Impact of diabetes on long- elevation myocardial infarction: A pooled analysis of the EXAMINA-
term outcome in STEMI patients undergoing primary angioplasty with TION (clinical evaluation of the Xience-V stent in acute myocardial
glycoprotein IIb-IIIa inhibitors and BMS or DES. J Thromb Thromboly- INfArcTION) and COMFORTABLE-AMI (comparison of Biolimus
sis. 2010;30:133-141. eluted from an erodible stent coating with bare metal stents in acute
5. Brener SJ, Mehran R, Dressler O, Cristea E, Stone GW. Diabetes ST-elevation myocardial infarction) trials. JACC Cardiovasc Interv.
mellitus, myocardial reperfusion, and outcome in patients with acute 2014;7:55-63.
ST-elevation myocardial infarction treated with primary angioplasty 20. Iijima R, Byrne RA, Dibra A, et al. Drug-eluting stents versus bare-
(from HORIZONS AMI). Am J Cardiol. 2012;109:1111-1116. metal stents in diabetic patients with ST-segment elevation acute
6. De Luca G, Dirksen MT, Spaulding C, et al. Impact of diabetes on long- myocardial infarction: A pooled analysis of individual patient data from
term outcome after primary angioplasty: Insights from the DESERT seven randomized trials. Rev Esp Cardiol. 2009;62:354-364.
cooperation. Diabetes Care. 2013;36:1020-1025. 21. Nakagawa Y, Kimura T, Morimoto TNomura M, et al. Incidence and
7. Sanidas EA, Brener SJ, Maehara A, et al. Outcomes in diabetic patients risk factors of late target lesion revascularization after sirolimus-
undergoing primary percutaneous coronary intervention for acute eluting stent implantation (3-year follow-up of the j-cypher registry).
anterior myocardial infarction: Results from the INFUSE-AMI study. Am J Cardiol. 2010;106:329-336.
Catheter Cardiovasc Interv. 2014;83:704-710. 22. Kornowski R, Mintz GS, Kent KM, et al. Increased restenosis in diabe-
8. Jensen LO, Maeng M, Thayssen P, et al. Influence of diabetes mellitus tes mellitus after coronary interventions is due to exaggerated intimal
on clinical outcomes following primary percutaneous coronary inter- hyperplasia. A serial intravascular ultrasound study. Circulation. 1997;
vention in patients with ST-segment elevation myocardial infarction. 95:1366-1369.
Am J Cardiol. 2012;109:629-635. 23. Nicholls SJ, Tuzcu EM, Kalidindi S, et al. Nissen SE effect of diabetes
9. Sabaté M, Jiménez-Quevedo P, Angiolillo DJ, et al. Randomized com- on progression of coronary atherosclerosis and arterial remodeling: A
parison of sirolimus-eluting stent versus standard stent for percutane- pooled analysis of 5 intravascular ultrasound trials. J Am Coll Cardiol.
ous coronary revascularization in diabetic patients: The diabetes and 2008;52:255-262.
sirolimus-eluting stent (DIABETES) trial. Circulation. 2005;112:2175- 24. Zellweger MJ, Kaiser C, Jeger R. Coronary artery disease progression
2183. late after successful stent implantation. J Am Coll Cardiol. 2012;59:
10. Baumgart D, Klauss V, Baer F, et al. SCORPIUS study; one-year results 793-799.
of the SCORPIUS study: A German multicenter investigation on the 25. Cutlip DE, Chhabra AG, Baim DS. Beyond restenosis: Five-year clinical
effectiveness of sirolimus-eluting stents in diabetic patients. J Am Coll outcomes from second-generation coronary stent trials. Circulation.
Cardiol. 2007;50:1627-1634. 2004;110:1226-1230.
11. Grube E, Chevalier B, Guagliumi G, et al. The SPIRIT V diabetic study: 26. Chacko R, Mulhearn M, Novack V. Impact of target lesion and nontar-
A randomized clinical evaluation of the XIENCE V everolimus-eluting get lesion cardiac events on 5-year clinical outcomes after sirolimus-
stent vs the TAXUS Liberté paclitaxel-eluting stent in diabetic patients eluting or bare-metal stenting. J Am Coll Cardiol Intv. 2009;2:
with de novo coronary artery lesions. Am Heart J. 2012;163:867-875. 498-503.
12. Kaul U, Bangalore S, Seth A, et al. TUXEDO–India investigators 27. Jiménez-Quevedo P, Hernando L, Gómez-Hospital JA, et al. Sirolimus-
paclitaxel-eluting versus Everolimus-eluting coronary stents in diabe- eluting stent versus bare metal stent in diabetic patients: The final
tes. N Engl J Med. 2015;373:1709-1719. five-year follow-up of the DIABETES trial. EuroIntervention. 2013;9:
13. Maeng M, Baranauskas A, Christiansen EH, et al. A 10-month angio- 328-335.
graphic and 4-year clinical outcome of everolimus-eluting versus 28. Rozenman Y, Sapoznikov D, Gotsman MS. Restenosis and progression
sirolimus-eluting coronary stents in patients with diabetes mellitus of coronary disease after balloon angioplasty in patients with diabetes
(the DiabeDES IV randomized angiography trial). Catheter Cardiovasc mellitus. Clin Cardiol. 2000;23:890-894.
Interv. 2015;86:1161-1167. 29. Kok MM, von Birgelen C, Sattar N, et al. Prediabetes and its impact on
14. Bavishi C, Baber U, Panwar S, et al. Efficacy and safety of everolimus clinical outcome after coronary intervention in a broad patient popula-
and zotarolimus-eluting stents versus first-generation drug-eluting tion. EuroIntervention. 2018;14:e1049-e1056.
stents in patients with diabetes: A meta-analysis of randomized trials. 30. von Birgelen C, Kok MM, Sattar N, et al. "Silent" diabetes and clinical
Int J Cardiol. 2017;230:310-318. outcome after treatment with contemporary drug-eluting stents: The
15. Sabate M, Cequier A, Iñiguez A, et al. Everolimus-eluting stent versus BIO-RESORT silent diabetes study. JACC Cardiovasc Interv. 2018;11:
bare-metal stent in ST-segment elevation myocardial infarction 448-459.
(EXAMINATION): 1 year results of a randomised controlled trial. Lan-
cet. 2012;380:1482-1490. SUPPOR TI NG I NFORMATION
16. Sabaté M, Brugaletta S, Cequier A, et al. Clinical outcomes in patients
with ST-segment elevation myocardial infarction treated with Additional supporting information may be found online in the
everolimus-eluting stents versus bare-metal stents (EXAMINATION): Supporting Information section at the end of this article.
5-year results of a randomised trial. Lancet. 2016;387:357-366.
17. Sabaté M, Cequier A, Iñiguez A, et al. Rationale and design of the
EXAMINATION trial: A randomised comparison between everolimus-
How to cite this article: Jimenez-Quevedo P, Brugaletta S,
eluting stents and cobalt-chromium bare-metal stents in ST-elevation
myocardial infarction. EuroIntervention. 2011;7:977-984. Cequier A, et al. Long-term impact of diabetes in patients with
18. Kastrati A, Dibra A, Spaulding C, et al. Meta-analysis of randomized ST-segment elevation myocardial infarction: Insights from the
trials on drug-eluting stents vs. bare-metal stents in patients with EXAMINATION randomized trial. Catheter Cardiovasc Interv.
acute myocardial infarction. Eur Heart J. 2007;28:2706-2713.
19. Sabaté M, Räber L, Heg D, et al. Comparison of newer-generation
2019;1–9. https://doi.org/10.1002/ccd.28194
drug-eluting with bare-metal stents in patients with acute ST-segment