OUP user menu

Prognostic impact of a chronic total occlusion in a non-infarct-related artery in patients with ST-segment elevation myocardial infarction: 3-year results from the HORIZONS-AMI trial

Bimmer E. Claessen, George D. Dangas, Giora Weisz, Bernhard Witzenbichler, Giulio Guagliumi, Martin Möckel, Sorin J. Brener, Ke Xu, José P.S. Henriques, Roxana Mehran, Gregg W. Stone
DOI: http://dx.doi.org/10.1093/eurheartj/ehr471 768-775 First published online: 12 January 2012

Abstract

Aims We sought to investigate the impact of multivessel disease (MVD) with and without a chronic total occlusion (CTO) in a non-infarct-related artery (IRA) on mortality in patients with ST-segment elevation myocardial infarction (STEMI) undergoing primary percutaneous coronary intervention (PCI).

Methods and results In the HORIZONS-AMI trial, of 3283 patients undergoing primary PCI, 1524 patients (46.4%) had single-vessel disease (SVD), 1477 (45.0%) had MVD without a CTO, and 283 (8.6%) had MVD with a CTO in a non-IRA. Compared with SVD patients and MVD patients without a CTO, patients with a non-IRA CTO were significantly less likely to achieve post-procedural TIMI 3 flow (P = 0.0003), more often had absent myocardial blush (P = 0.0002), and less frequently achieved complete ST-segment resolution (P = 0.0001). By multivariable analysis, MVD with CTO in a non-IRA was an independent predictor of both 0- to 30-day mortality [hazard ratio (HR) 2.88, 95% confidence interval (CI) 1.41–5.88, P = 0.004] and 30-day to 3-year mortality (HR 1.98, 95% CI 1.19–3.29, P= 0.009), while MVD without a CTO was a significant predictor for 0- to 30-day mortality (HR 2.20, 95% CI 1.00–3.06, P = 0.049) but not late mortality.

Conclusion In patients with STEMI undergoing primary PCI in the HORIZONS-AMI trial, MVD with or without a CTO in a non-IRA was an independent predictor of early mortality. The presence of a CTO in a non-IRA was also an independent predictor of increased late mortality to 3 years.

  • ST-segment elevation myocardial infarction
  • Chronic total occlusion
  • Multivessel disease

See page 695 for the editorial comment on this article (doi:10.1093/eurheartj/ehr412)

Introduction

Acute ST-segment elevation myocardial infarction (STEMI) typically arises from sudden thrombotic occlusion of a coronary artery.1 Prompt restoration of epicardial blood flow reduces infarct size and mortality.2 Mechanical reperfusion by primary percutaneous coronary intervention (PCI) with stent implantation is currently the preferred treatment for STEMI patients.3 Approximately 40–60% of the STEMI patients have multivessel disease (MVD) and ∼10% of the patients have a chronic total occlusion (CTO) in a non-infarct-related artery (IRA).2,4 Multivessel disease is regarded as a risk factor associated with worse outcome after STEMI.5 However, some studies have suggested that MVD may only be of prognostic importance if a CTO in a non-IRA is present.4,6,7

We therefore investigated the impact of MVD with and without a CTO in a non-IRA on markers of reperfusion and 3-year mortality in patients presenting with STEMI and treated with primary PCI in the large-scale randomized HORIZONS-AMI (Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction) trial.

Methods

The design and primary results of the HORIZONS-AMI trial have been described previously.810 Briefly, 3602 STEMI patients were enrolled internationally at 123 medical centres and randomized in a 1:1 ratio before coronary angiography to unfractionated heparin with a glycoprotein IIb/IIIa inhibitor (GPI, either abciximab or eptifibatide) vs. bivalirudin with provisional use of a GPI for refractory thrombotic or ischaemic complications in the cardiac catheterization laboratory. After coronary angiography, 3006 patients with lesions eligible for stenting were randomized again, in a 3:1 ratio, to either a paclitaxel-eluting stent (PES, TAXUS EXPRESS) or an otherwise identical uncoated bare-metal stent (BMS, EXPRESS), respectively.

Patient eligibility criteria have also been published previously.810 In summary, consecutive patients ≥18 years of age with symptom onset within 12h of duration and ST-segment elevation of ≥1 mm in two or more contiguous leads, new left bundle branch block, or true posterior MI were considered for enrolment. Principal exclusion criteria included contraindications to any of the study medications; prior administration of fibrinolytic therapy, bivalirudin, GPI, low-molecular-weight heparin, or fondaparinux for the present admission (prior unfractionated heparin was allowed); current use of coumadin; history of bleeding diathesis, conditions predisposing to haemorrhagic risk, or refusal to receive blood transfusions; stroke or transient ischaemic attack within 6 months or any permanent neurological deficit; recent or known platelet count <100 000 cells/mm3 or Hgb <10 g/dL; planned elective surgical procedure that would necessitate interruption of thienopyridines during the first 6 months after enrolment; coronary stent implantation within 30 days; and non-cardiac co-morbid conditions with life expectancy <1 year or that might result in protocol non-compliance.

The study was approved by the institutional review board or Ethics Committee at each participating centre, and all patients signed informed consent. HORIZONS-AMI was registered at clinicaltrials.gov (identifier: NCT00433966). Angiograms were analysed at an independent core laboratory (Cardiovascular Research Foundation, New York, NY, USA) with the use of validated methods by technicians who were unaware of the treatment assignments and clinical outcomes.

Definitions and endpoints

The angiographic core laboratory assessed each baseline and post-procedure angiogram for standard vessel and lesion-related measures, including Thrombolysis in Myocardial Infarction (TIMI) epicardial flow and myocardial blush grades (MBGs), as described previously.11 Electrocardiograms (ECG) obtained pre-procedure and at 60min post-procedure were analysed as pairs by the ECG core laboratory by independent readers who were blinded to the clinical and angiographic data. ST-segment deviation was evaluated using standardized techniques.12 ST-segment elevation resolution (STR) was categorized as complete (>70%), partial (30–70%), or none (<30%). All trial patients in whom both baseline and 60 min post-PCI ECG were available and interpretable were included in this analysis. Patients with left bundle branch block at baseline, paced ventricular rhythm, and patients without ST-segment deviation in two or more contiguous leads were excluded from electrocardiographic analysis.

Multivessel disease was defined as the presence of at least one lesion with diameter stenosis ≥50% in at least two major epicardial coronary arteries, as determined by the angiographic core laboratory, and further subclassified as double- or triple-vessel disease. A CTO in a non-IRA was defined as the presence of TIMI 0 or 1 flow grade in a non-treated vessel not related to the acute infarct episode. Clinical endpoints for the current analysis included all-cause mortality, major adverse clinical events (MACE, the composite of death, reinfarction, target vessel revascularization for ischaemia, or stroke) and stent thrombosis (definite or probable according to the Academic Research Consortium definition) at 30-day and 3-year follow-up.13 An independent clinical events committee blinded to treatment assignment adjudicated all primary endpoint and stent thrombosis events using original source documents. Anaemia was defined using WHO criteria as baseline haematocrit less than 39% for men and less than 36% for women.14

Statistical methods

Outcomes were examined in groups of patients with single-vessel disease (SVD), MVD without a CTO, and MVD with a CTO in a non-IRA. Categorical variables are presented as percentages and were compared with the χ2 test or Fisher's exact test. The χ2 test was used to calculate trend P-values. Continuous variables are presented as medians with inter-quartile ranges and were compared using the Mann–Whitney U-test. The primary event analyses were performed with the use of time-to-event data (with data censored at the time of a patient's withdrawal from the study or at the time of last follow-up). Cumulative event rates were calculated according to the Kaplan–Meier method and compared using the log-rank test. Landmark survival analysis was performed to examine the impact of MVD with and without a CTO in a non-IRA on early (0–30 days) and late (30 days–3 years) clinical outcomes. Multivariable Cox's proportional hazards regression was performed to identify predictors of overall 3-year mortality, early (0–30 days) and late (30 days–3 years) mortality. The multivariable model was built by stepwise variable selection with entry and exit criteria set at the P = 0.01 level. The following patient level candidate predictors were evaluated: age, gender, hypertension, hyperlipidaemia, current smoking, diabetes mellitus, creatinine clearance, congestive heart failure (CHF), prior MI, prior PCI, prior coronary artery bypass graft surgery, white blood cell count, time from symptom onset to first balloon inflation, Killip class >1, anterior infarction, stent implantation, pre-procedural TIMI flow grade 0/1, post-procedural TIMI flow grade <3, SVD vs. MVD without a non-IRA CTO, and SVD vs. MVD with a non-IRA-CTO. As post-randomization variables were included in this model, the randomized therapies were not. However, exploratory interaction tests were performed between the severity of coronary artery disease (SVD, MVD without a non-IRA CTO, or MVD with a non-IRA-CTO) and the two randomized treatment assignments (bivalirudin vs. heparin + a GPI, and BMS vs. PES) with regard to mortality at 3 years. All statistical analyses were performed by SAS V9.2, SAS Institute Inc., Cary, NC, USA. A value of P < 0.05 was considered statistically significant, and all P-values are two-sided.

Results

Patients and procedures

Of 3602 enrolled patients, primary PCI was performed in 3340 (92.7%), 3283 of whom (98.3%) had complete core laboratory angiographic analysis, comprising the current study population. Single-vessel disease was present in 1524 patients (46.4%), while MVD without or with a CTO in the non-IRA was present in 1477 patients (45.0%) and 283 patients (8.6%), respectively. In the latter group, the CTO was in the left anterior descending, left circumflex, and right coronary artery in 93 (32.9%), 96 (33.9%), and 148 (52.3%) of the patients, respectively (54 patients had two non-IRA CTOs). Baseline, angiographic, and procedural characteristics according to the complexity of coronary artery disease are shown in Table 1. Less favourable baseline characteristics (advanced age, female gender, hypertension, hyperlipidaemia, diabetes mellitus, prior MI, prior PCI, prior CABG, had a lower left ventricular ejection fraction (LVEF), history of CHF, and renal insufficiency) were more common with greater severity of coronary artery disease. However, patients with SVD were more often current smokers compared with MVD patients with or without a CTO. Patients with a CTO in a non-IRA (compared with those with SVD and MVD without a CTO) had longer symptom onset to balloon times, higher Killip class, and less often underwent stent implantation. Patients with MVD without a CTO less often had the left anterior descending artery treated during the primary PCI, more often had pre-procedural TIMI flow grade 0/1, and had more and longer stents implanted compared with patients with SVD and MVD with a CTO. Among patients with MVD, patients with a CTO more often had triple-vessel disease compared with those without a CTO.

View this table:
Table 1

Baseline clinical and angiographic characteristics

SVD (n= 1524, 46.4%)MVD, no CTO (n= 1477, 45.0%)MVD with CTO (n= 283, 8.6%)P-value
Baseline characteristics
 Age (median, IQR)58.3 (50.4, 67.9)61.8 (54.0, 71.0)62.2 (55.5, 71.8)<0.0001
 Female25.9% (394/1524)21.5% (318/1477)14.8% (42/283)<0.0001
 Body mass index (kg/m2)26.9 (24.5, 30.1)27.2 (24.4, 30.1)27.6 (24.8, 30.4)0.50
 Anaemia10.3% (145/1413)10.6% (151/1419)10.9% (29/265)0.92
 WBC count, baseline (giga/L)11.1 (9.0, 13.5)10.4 (8.6, 13.4)10.6 (8.9, 13.6)0.13
 Hypertension46.5% (708/1522)56.4% (833/1477)66.4% (188/283)<0.0001
 Hyperlipidaemia41.0% (624/1522)42.5% (628/1477)54.4% (154/283)0.0001
 Current smoking50.5% (764/1514)43.4% (640/1474)45.2% (126/279)0.0005
 Diabetes mellitus13.3% (203/1522)17.2% (254/1477)27.6% (78/283)<0.0001
  Insulin-dependent3.4% (52/1522)5.1% (75/1477)7.4% (21/283)0.004
 Prior MI7.1% (108/1522)10.6% (157/1477)29.3% (83/283)<.0001
 Prior PCI8.4% (128/1522)10.9% (161/1477)21.6% (61/282)<.0001
 Prior CABG0.6% (9/1522)1.0% (15/1477)21.9% (62/283)<.0001
 History of CHF1.4% (22/1522)3.0% (45/1477)6.4% (18/283)<0.0001
 Renal insufficiency14.1% (200/1416)18.2% (250/1377)20.1% (52/259)<0.004
 Killip class 2–47.3% (11/1521)8.2% (121/1477)16.6% (47/283)<0.001
 Randomization to bivalirudin49.0% (746/1524)51.3% (758/1477)50.9% (144/283)0.42
 Symptom onset to hospital arrival [h, median (IQR)]2.08 (1.25, 3.75)2.17 (1.25, 3.92)2.42 (1.17, 4.33)0.39
Angiographic and procedural characteristics
 Number of diseased vessels
  One100% (1524/1524)0% (0/1477)0% (0/283)
  Two0% (0/1524)58.9% (870/1477)47.3% (134/283)<0.0001
  Three0% (0/1524)41.1% (607/1477)52.7% (149/283)<0.0001
 TIMI flow grade 0–1 pre-PCI67.8% (1038/1532)63.2% (1053/1667)66.2% (192/290)0.02
 Randomization to paclitaxel-eluting stent75.2% (1051/1398)74.7% (990/1325)76.4% (178/233)0.85
 Use of glycoprotein IIb/IIIa inhibitors51.0% (778/1524)48.7% (719/1477)49.1% (139/283)0.42
 LVEF [%, median (IQR)]50 (43–59)50 (45–58)45 (35–54)<0.0001
 Index PCI vessel = LAD46.6% (717/1537)34.9% (582/1668)42.4% (123/290)<0.001
 Symptom onset to first balloon inflation [h, median (IQR)]3.57 (2.58, 5.42)3.78 (2.72, 5.60)4.04 (2.83, 6.63)0.002
Index procedure
  Stent implanted94.2% (1435/1524)93.4% (1380/1477)89.8% (254/283)0.02
  Stent number, per patient1.38 ± 0.681.66 ± 0.921.51 ± 0.78<0.0001
  Stent total length (mm)24 (16, 32)28 (20, 44)24 (18, 36)<0.0001
 Peak creatine phosphokinase (units, median, IQR)1568 (703, 2892)1492 (692, 2751)1700 (659, 3362)0.34
  • WBC, white blood cell; MI, myocardial infarction; PCI, percutaneous coronary intervention; CABG, coronary artery bypass graft; IQR, inter-quartile range; LAD, left anterior descending artery; LVEF, left ventricular ejection fraction; TIMI, Thrombolysis In Myocardial Infarction.

Patients with a CTO in a non-IRA less often achieved post-procedural TIMI 3 flow in the IRA (SVD 91.7%, MVD without a CTO 92.6%, and MVD with a CTO 85.5%, P for trend = 0.0003), were significantly more likely to have absent post-procedural myocardial blush in the IRA (SVD 7.9%, MVD without 7.0%, and MVD with a CTO 12.2%, P for trend = 0.0002), and less often achieved complete STR at 1 h after the index procedure (CTO 48.9%, MVD without a CTO 52.9%, and MVD with a CTO 39.7%, P for trend = 0.001). Moreover, a trend was present towards greater infarct size in patients with vs. without a CTO in a non-IRA, as expressed by higher peak creatine phosphokinase levels (median 1700 vs. 1534 U, P = 0.27; Table 1).

Clinical outcomes

At 30 days, from 30 days to 3 years and at 3 years overall, the rates of MACE and the individual endpoints of mortality, cardiac mortality, and non-target vessel revascularization increased significantly with greater coronary artery disease complexity (Table 2). During the overall 3-year follow-up duration and during the 30-day to 3-year follow-up interval, the rates of reinfarction and ischaemia-driven target vessel revascularization were also higher in patients with more complex coronary artery disease.

View this table:
Table 2

Clinical outcomes

SVD (n= 1524, 46.4%)MVD, no CTO (n= 1477, 45.0%)MVD with CTO (n= 283, 8.6%)P-value
30-day clinical outcomes
 Major adverse cardiovascular events3.9% (60)6.0% (88)8.5% (24)0.002
 Death1.4% (21)2.7% (40)5.7% (16)<0.0001
  Cardiac1.2% (19)2.6% (38)5.7% (16)<0.0001
  Non-cardiac0.1% (2)0.1% (2)0.0% (0)0.83
 Reinfarction1.4% (21)2.3% (34)1.8% (5)0.16
 Stroke0.5% (7)0.5% (8)0.7% (2)0.84
 Major bleeding6.6% (101)7.9% (116)9.0% (25)0.24
 Ischaemia-driven target vessel revascularization2.0% (31)2.9% (42)2.2% (6)0.32
 Non-target vessel revascularization1.1% (17)2.4% (35)2.9% (8)0.01
 Definite/probable stent thrombosis2.0% (29)2.8% (39)2.3% (6)0.38
30-day to 3-year clinical outcomes
 Major adverse cardiovascular events14.3% (209)19.4% (269)27.7% (72)<0.0001
 Death3.2% (46)4.1% (56)10.2% (26)<0.0001
  Cardiac1.2% (17)1.4% (19)4.0% (10)0.002
  Non-cardiac2.0% (29)2.7% (37)6.4% (16)0.0003
 Reinfarction4.0% (57)6.4% (86)10.1% (25)<0.0001
 Stroke0.8% (12)1.3% (18)1.7% (4)0.34
 Major bleeding1.5% (22)2.2% (30)3.0% (7)0.23
 Ischaemia-driven target vessel revascularization9.9% (143)12.5% (171)17.6% (44)0.001
 Non-target vessel revascularization3.1% (45)15.1% (208)12.2% (30)<0.001
 Definite/probable stent thrombosis0.2% (3)0.3% (4)0.0% (0)0.33
Overall 3-year clinical outcomes
 Major adverse cardiovascular events17.4% (259)24.1% (346)34.3% (95)<0.0001
 Death4.5% (67)6.7% (96)15.3% (42)<0.0001
  Cardiac2.4% (36)3.9% (57)9.5% (26)<0.0001
  Non-cardiac2.1% (31)2.9% (39)6.4% (16)0.0008
 Reinfarction5.3% (77)8.5% (118)11.8% (30)<0.0001
 Stroke1.3% (19)1.8% (25)2.4% (6)0.35
 Major bleeding7.7% (116)9.6% (139)11.6% (31)0.06
 Ischaemia-driven target vessel revascularization10.8% (157)13.2% (182)17.6% (45)0.005
 Non-target vessel revascularization4.2% (61)16.9% (236)14.6% (37)<0.0001
 Definite/probable stent thrombosis4.3% (61)5.8% (78)6.1% (15)0.16

Figure 1 shows time-to-event curves for overall 3-year mortality and with a landmark set at 30 days. Between 0 and 30 days, mortality was significantly higher in patients with MVD and a CTO in a non-IRA (5.7%) than in those with MVD without a CTO (2.7%, P < 0.01) or SVD (1.4%, P < 0.01). Between 30 days and 3 years, the mortality curves continued to diverge, especially for patients with a non-IRA CTO. Long-term mortality in 30-day survivors was significantly higher in patients with MVD and a CTO in a non-IRA (10.2%) compared with MVD without a CTO (4.1%, P < 0.01) and SVD (3.2%, P < 0.01). As shown in Figure 2, a CTO portended relatively higher 3-year mortality in both patients with double- and triple-vessel disease to a similar degree (P for interaction = 0.14). However, 3-year mortality was comparable in patients with triple-vessel disease and double-vessel disease in whom a CTO was not present (7.0 vs. 6.5%, respectively, P = 0.72), whereas 3-year mortality was significantly higher in patients with triple-vessel compared with double-vessel disease when a CTO in a non-IRA was present (10.7 vs. 19.4%, P = 0.047).

Figure 1

Time-to-event curves for overall 3-year mortality (A) and mortality between 0–30 days and 30 days–3 years (B) in patients with single-vessel disease, MVD without a CTO, and MVD with a CTO in a non-infarct-related artery. CTO, chronic total occlusion; IRA, infarct-related artery; MVD, multivessel disease, SVD, single-vessel disease. P-values shown in figure are three-way. Pair-wise P-values for overall 3-year mortality are: CTO vs. SVD, P < 0.0001; MVD without a CTO vs. SVD, P = 0.01; CTO vs. MVD without a CTO, P < 0.0001. Pair-wise P-values for 30-day mortality are: CTO vs. SVD, P < 0.0001; MVD without a CTO vs. SVD, P< 0.0001; CTO vs. MVD without a CTO, P = 0.02. Pair-wise P-values for mortality between 30 days–3 years are: CTO vs. SVD, P < 0.0001; CTO MVD without a CTO vs. SVD, P = 0.20; CTO vs. MVD without a CTO, P < 0.0001.

Figure 2

Three-year mortality in patients with double-vessel disease (left pair of bars) and triple-vessel disease (right pair of bars) according to the presence or absence of a CTO. MVD, multivessel disease, CTO, chronic total occlusion; HR, hazard ratio; 95% CI, 95% confidence interval.

Table 3 shows the independent multivariable predictors of overall 3-year mortality, mortality between 0 and 30 days, and mortality between 30 days to 3 years. Multivessel disease with a CTO in a non-IRA was an independent predictor of overall 3-year [hazard ratio (HR) 2.27, 95% confidence interval (CI) 1.67–4.60, P < 0.01], early (HR 2.88, 95% CI 1.41–5.88, P< 0.01), and late mortality (HR 1.98, 95% CI 1.19–3.29, P < 0.01), while MVD without a CTO was a significant predictor only of early mortality (HR 1.75, 95% CI 1.00–3.06, P = 0.049). A significant interaction was observed between the severity of coronary artery disease and randomization to bivalirudin or heparin + a GPI with regard to 3-year mortality; the benefit of bivalirudin in decreasing 3-year mortality was confined to patients with SVD (Figure 3). There was no significant interaction between the extent of coronary artery disease and stent randomization on mortality at 3 years.

View this table:
Table 3

Independent predictors of early (0–30 days) and late (30 days–3 years) mortality

VariableHazard ratio (95% CI)P-value
Predictors of overall 3-year mortality
 History of congestive heart failure2.77 (1.67–4.60)<0.0001
 MVD with CTO vs. SVD2.27 (1.47–3.52)0.0002
 Killip class: 2–4 (vs. 0/1)1.99 (1.39–2.86)0.0002
 Age (per 10-year increase)1.88 (1.58–2.24)<0.0001
 History of prior MI1.56 (1.07–2.28)0.02
 Creatinine clearance <60 mL/min1.54 (1.05–2.22)0.03
 White blood cell count (per 1000 U increase)1.10 (1.06–1.14)<0.0001
 Final TIMI flow grade 30.57 (0.40–0.83)0.003
 One or more stents implanted0.53 (0.33–0.85)0.008
 MVD without CTO vs. SVDa1.33 (0.95–1.86)0.10
Predictors of early (0–30 days) mortality
 MVD with CTO vs. SVD2.88 (1.41–5.88)0.004
 Killip class: 2–4 (vs. 0/1)2.40 (1.39–4.14)0.002
 Age (per 10-year increase)2.11 (1.69–2.64)<0.0001
 MVD without CTO vs. SVD1.75 (1.00–3.06)0.0495
 White blood cell count (per 1000 U increase)1.14 (1.09–1.20)<0.0001
 Final TIMI flow grade 30.48 (0.27–0.83)0.009
 One or more stents implanted0.26 (0.15–0.47)<0.0001
Predictors of late (30 days–3 years) mortality
 History of congestive heart failure2.40 (1.30–4.45)0.005
 Age (per 10-year increase)2.28 (1.90–2.73)<0.0001
 History of prior MI2.06 (1.35–3.14)0.0009
 MVD with CTO vs. SVD1.98 (1.19–3.29)0.009
 Killip class: 2–4 (vs. 0/1)1.70 (1.08–2.68)0.02
 Male (vs. female)1.67 (1.08–2.61)0.02
 White blood cell count (per 1000 U increase)1.07 (1.02–1.12)0.008
 MVD without CTO vs. SVDa1.00 (0.67–1.49)0.99
  • CTO, chronic total occlusion; SVD, single-vessel disease; MVD, multivessel disease; TIMI, Thrombolysis in Myocardial Infarction.

  • aMVD without CTO vs. SVD was forced into the model.

Figure 3

Interaction between the severity of coronary artery disease and randomization to heparin plus a GPI vs. bivalirudin (A) and paclitaxel-eluting stents vs. bare-metal stents (B) on 3-year all-cause mortality. MVD, multivessel disease; CTO, chronic total occlusion; SVD, single-vessel disease; IRA, infarct-related artery; GPI, glycoprotein inhibitor; PES, paclitaxel-eluting stents; BMS, bare-metal stents.

Discussion

The present analysis from the large-scale, randomized HORIZONS-AMI trial shows that in patients with STEMI undergoing primary PCI, the presence of MVD with a CTO in a non-IRA is associated with impaired markers of reperfusion and increased early (0–30 days), late (30 days–3 years), and cumulative 3-year mortality. In contrast, MVD without a CTO was associated with increased early mortality, but not with late mortality.

The current analysis, drawn from a large international trial with careful data monitoring and event adjudication, confirms and extends previous reports about the poor prognosis associated with the presence of a CTO in a non-IRA in patients undergoing primary PCI for acute STEMI.4,6,7 Our finding that TIMI 3 flow is achieved less frequently, MBG is lower, and STR at 60min post-procedure is less common in patients with a CTO in a non-IRA is consistent with a subanalysis from the TAPAS (Thrombus Aspiration during Primary percutaneous coronary intervention in Acute ST-elevation myocardial infarction study) trial, in which patients with a CTO in a non-IRA had lower blush grades, less STR, and a higher frequency of persistent ST-segment deviation.7 Moreover, an observational study from the Academic Medical Center in Amsterdam, the Netherlands, reported that in patients with STEMI treated with primary PCI, MVD was associated with long-term mortality only if a CTO in a non-IRA was present.4 In this study, a CTO in a non-IRA was associated with lower LVEF after the index event and a further deterioration in LVEF during 1-year follow-up.

Prior studies have suggested that triple-vessel disease is associated with worse outcomes after primary PCI compared with double-vessel disease.15 Our analysis found that while mortality was increased in all patients with MVD compared with those with SVD, late mortality was higher in patients with triple- vs. double-vessel disease only in those in whom a CTO of the non-IRA was also present.

The mechanisms underlying the increased mortality (especially late) in patients with MVD and a CTO are likely multifactorial. Patients with a CTO in a non-IRA had a higher prevalence of cardiovascular risk factors and co-morbidities compared with SVD patients and MVD patients without a CTO. However, a CTO in a non-IRA remained an independent predictor of both early and late mortality after multivariable adjustment. The suboptimal epicardial and myocardial reperfusion in the IRA observed in the present study (resulting in less complete STR) may in part explain the impaired left ventricular function and subsequent higher mortality in patients with MVD and a CTO in a non-IRA. A possible mechanistic explanation for the lower MBG and less STR in the CTO group may be that microvascular ischaemia and reperfusion injury are more likely to occur in those myocardial regions that experienced the least collateral flow during ischaemia.16 It seems plausible that in the presence of a CTO in another territory, occlusion of the IRA may be accompanied by less collateral flow prior to PCI.

Another explanation for the high early and late mortality in patients with a CTO in a non-IRA could be that they are potentially at ‘double jeopardy’ from the acute MI; if the distal coronary bed of the CTO depends upon collateral blood flow from the IRA, the area of risk of the IRA would extend beyond its own supply territory to also include the myocardial distribution of the coronary artery in which the CTO is located, resulting in greater infarct size. In the present study, the peak creatine phosphokinase levels tended to be higher in the CTO group, supporting this hypothesis.

The finding that the presence of a CTO in a non-IRA adds incremental risk for adverse outcomes in patients with MVD may have important clinical implications. Would CTO revascularization improve prognosis after a recent STEMI? Several studies (including HORIZONS-AMI) have shown that treatment of non-culprit lesions during primary PCI for STEMI is associated with increased post-procedural morbidity without a survival benefit,1720 and the current ACC/AHA guidelines do not recommend non-culprit lesion intervention during primary PCI for STEMI without haemodynamic complications.21 Further studies are thus warranted to determine whether revascularization of a CTO in a non-IRA (either with PCI or CABG) as part of a staged revascularization strategy may improve late prognosis. The benefit of such revascularization may depend on the residual viability of the myocardium subtended by both the IRA and CTO vessel. The ongoing randomized EXPLORE (evaluating XIENCE V and left ventricular function in PCI on occlusions after STEMI) trial is investigating whether PCI of a CTO in a non-IRA within 1 week after primary PCI compared with optimal medical therapy has a beneficial effect on left ventricular dimensions and function.22

Limitations

The present study is a post hoc analysis from a large randomized clinical trial of patients with STEMI undergoing primary PCI and is limited by its observational nature. There were numerous differences in baseline clinical, angiographic, and procedural characteristics between the groups, and although multivariable Cox's proportional hazards analysis was performed, residual unmeasured confounders cannot be excluded. As such, the results of the present analysis should be considered hypothesis-generating. Similarly, the modest interaction between the extent of coronary artery disease and the 3-year survival benefit of treatment with bivalirudin compared with heparin plus a GPI (as previously described23) may be due to chance given the uncertain mechanism underlying this observation. Further studies are warranted to either confirm or refute this finding.

Conclusions

In patients with STEMI undergoing primary PCI, the presence of a CTO in the non-IRA is an independent determinate of early and late mortality.

Funding

The HORIZONS-AMI trial was supported by the Cardiovascular Research Foundation, with grant support from Boston Scientific and the Medicines Company.

Conflict of interest: G.W.S. is on the scientific advisory boards for and has received honoraria from Abbott Vascular and Boston Scientific, and has served as a consultant to the Medicines Company, Eli Lilly BMS/Sanofi and AstraZeneca. G.D.D. and has received speaker honoraria from Astra Zeneca, Bristol-Meiers Squibb, The Medicines Co, Sanofi Aventis, and Abbott Vascular. R.M. has received a research grant from Sanofi Aventis, and honoraria from The Medicines Company, Abbott Vascular, Sanofi Aventis, Bristol Meiers Squibb, Cordis, and Astra Zeneca. B.W. has received lecture honoraria from Boston Scientific and The Medicines Company. G.G. has served as a consultant to Boston Scientific and Volcano, and has received lecture honoraria from Boston Scientific, Medtronic, Lightlab, Labcoat, and received grant support from Medtronic and Boston Scientific. The other authors report no conflicts of interest.

References

View Abstract