Delayed Post-Myocardial Infarction Invasive Measures, Helpful or Harmful?

Delayed Post-Myocardial Infarction Invasive Measures, Helpful or Harmful?* A Subgroup Analysis Hon-Kan Yip, MD; Chiung-Jen Wu, MD; Cheng-Hsu Yang, MD; Hsueh-Wen Chang, PhD; Shyh-Ming Chen, MD; Wei-Chin Hung, MD; and Chi-Ling Hang, MD

Background: In patients who have experienced acute myocardial infarction (MI), primary percutaneous coronary intervention (PCI) has been shown to be of benefit in terms of clinical outcomes. However, the value of performing routine PCI in patients with early MI (ie, an MI occurring > 12 h to < 7 days before patient presentation) or recent MI (ie, an MI occurring > 8 days to < 30 days before patient presentation) has not been established. The purposes of this prospective observational study were to evaluate the impact of PCI on outcomes, and to delineate the predictors of lack of response to reperfusion and the prognostic determinants in patients with this clinical condition. Methods and results: A total of 377 consecutive unselected patients who had experienced early or recent MI underwent PCI. Successful reperfusion (ie, Thrombolysis in Myocardial Infarction flow grade 3 of the infarct-related artery [IRA]) was achieved in 90.2% of patients. By multiple stepwise logistic regression analysis, high-burden thrombus formation (odds ratio [OR], 15.53; 95% confidence interval [CI], 6.09 to 39.60; p < 0.0001) in the IRA, early PCI (ie, < 3 days) [OR, 4.10; 95% CI, 1.79 to 7.36; p ⴝ 0.0008], advanced congestive heart failure (CHF) [OR, 4.10; 95% CI, 1.70 to 9.91; p ⴝ 0.002], and diabetes (OR, 3.03; 95% CI, 1.03 to 7.06; p ⴝ 0.010) were independent predictors for lack of response to reperfusion. The 30-day mortality rate was 6.8%. The only variables that were independently related to the 30-day mortality rate were advanced CHF (OR, 29.85; 95% CI, 7.84 to 113.7; p < 0.0001), lack of response to reperfusion (OR, 7.57; 95% CI, 2.29 to 25.07; p ⴝ 0.0009), early PCI (OR, 4.81; 95% CI, 1.60 to 14.41; p ⴝ 0.005), and multivessel disease (OR, 9.22; 95% CI, 1.63 to 52.04; p ⴝ 0.0119). The surviving 351 patients were discharged from the hospital and followed-up for a mean (ⴞ SD) 38.9 ⴞ 14.2 months. Coronary angiographic follow-up was performed in 285 patients (81.2%). Restenosis of the IRA was found in 101 patients (35.4%). Reinterventions of the IRA were required in 69 patients (24.2%). Follow-up measurements of left ventricular ejection fraction (LVEF) showed significantly more improvement than the initial LVEF (59.3 ⴞ 13.8% vs 50.4 ⴞ 13%; p < 0.0001). The total cumulative mortality rate after hospital discharge was 6.5% for the entire group. Only advanced CHF (OR, 3.46; 95% CI, 1.26 to 9.52; p ⴝ 0.016) and old age (ie, > 70 years of age) [OR, 4.41; 95% CI, 1.59 to 12.24; p ⴝ 0.004] were independent predictors of long-term mortality. Conclusion: The performance of PCI on > day 4 in patients after they had experienced an MI was safe and had a high rate of success. The clinical benefits of a relative low mortality rate associated with successful PCI for patients with early and recent MI was maintained during the long-term follow-up. However, patients with advanced CHF along with old age continued to have a poor prognosis. (CHEST 2004; 126:38 – 46) Key words: early or recent myocardial infarction; percutaneous coronary intervention Abbreviations: AMI ⫽ acute myocardial infarction; CHF ⫽ congestive heart failure; CI ⫽ confidence interval; HBTF ⫽ high-burden thrombus formation; IRA ⫽ infarct-related artery; LVEF ⫽ left ventricular ejection fraction; MI ⫽ myocardial infarction; NYHA ⫽ New York Heart Association; OR ⫽ odds ratio; PCI ⫽ percutaneous coronary intervention; TIMI ⫽ Thrombolysis in Myocardial Infarction; TVR ⫽ target vessel revascularization

and complete restoration of the infarctE arly related artery (IRA) by either thrombolytic therapy or primary percutaneous coronary intervention (PCI) has been shown to improve the angiographic and clinical outcomes of patients who have experienced an acute myocardial infarction (AMI).1–3 However, many patients who experience an AMI are 38

not admitted to a hospital in time to receive either thrombolytic therapy or primary PCI. Therefore, there are still quite a few patients who may experience postinfarct angina as a result of either failed thrombolysis or delayed presentation (ie, arrival at the hospital after ⬎ 12 h) after the AMI and receive rescue PCI.4 –11 Although the clinical Clinical Investigations

value of rescue PCI early after failed thrombolysis has been debated extensively, it remains uncertain.5,6,10 One of the most important explanations For editorial comment see page 2 may be the relatively low rate of successful reperfusion after performing the rescue procedure, which in turn increases the likelihood of morbidity and mortality.4,5,7,9,11 Surprisingly, however, there is still no available information focusing on the short-term and long-term outcomes of patients with early or recent myocardial infarction (MI) undergoing routine PCI without preceding thrombolytic therapy. Therefore, the purposes of this prospective observational study were to test whether routine PCI provided additional benefits and improved shortterm and long-term outcomes in such consecutive and unselected patients. We also wanted to delineate the predictors of unsuccessful reperfusion and the prognostic determinants in patients with this clinical condition.

Materials and Methods Patient Population In our hospital, all patients who have had early or recent MIs were considered to be eligible for PCI under the protocol approved by the Institutional Ethics Committee of the Chang Gung Memorial Hospital. Informed consent was obtained from all patients prior to performing elective PCI. Patients were excluded if they had active upper GI bleeding, bleeding diathesis, or acute or recent intracranial hemorrhaging, or had undergone major surgery within the preceding 2 weeks. Procedure and Protocol Initially, stent-supported coronary angioplasty was the designated protocol for unsatisfactory angiographic results in this study. At a later stage, stent implantation was strongly encouraged unless the IRA had heavy calcification (ie, a reference lumen diameter of ⬍ 2.5 mm or after coronary angioplasty with stent-like results found on the treatment site). Therefore, ⬎ 60% of our patients in this study had received stent implantation. The stent type, including the Crown stent, Crossflex stent, and BX *From the Division of Cardiology (Drs. Yip, Wu, Yang, Chen, Hung, and Hang), Chang Gung Memorial Hospital, Kaohsiung, Taiwan; and the Department of Biological Sciences (Dr. Chang), National Sun Yat-Sen University, Kaohsiung, Taiwan, Republic of China. Manuscript received May 2, 2003; revision accepted February 26, 2004. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (e-mail: [email protected]). Correspondence to: Chi-Ling Hang, MD, Division of Cardiology, Department of Internal Medicine, Chang Gung Memorial Hospital, Kaohsiung, 123, Ta Pei Rd, Niao Sung Hsiang, Kaohsiung Hsien, 83301, Taiwan, ROC; e-mail: [email protected] www.chestjournal.org

velocity stent (Johnson & Johnson Interventional Systems; Warren, NJ), the NIR stent and Express stent (Boston Scientific; Natick, MA), the Wallstent (Schneider; Bulach, Switzerland), the GFX stent, S-670 stent, and S-7 stent (Medtronic Inc; Minneapolis, MN), and the Multilink stent and Penta stent (Guidant; Indianapolis, IN), was selected at the discretion of the primary operator. Glycoprotein IIb/IIIa receptor antagonists have been available in our country since August 2000. In our hospital, all patients who had experienced early or recent MI were considered to be eligible for PCI, and therapy with tirofiban was given after informed consent was obtained, unless there were contraindications for its use. Therefore, tirofiban was given to 55 patients in our study. The protocol-designated dosage was a bolus dose of 20 ␮g/kg body weight, which was administered to patients at least 30 min to 2 h before they underwent cardiac catheterization, followed by a maintenance infusion of 0.15 ␮g/min for 18 to 24 h. Heparin was administered as an initial bolus of 70 U/kg. If necessary, an additional bolus was administered to achieve an activated clotting time of ⱖ 250 s. Patients were treated with ticlopidine (250 mg bid) for at least 2 weeks if stent deployment was performed. Aspirin (100 mg po daily) was administered indefinitely to each patient. Angiographic Analysis The angiographic features of high-burden thrombus formation (HBTF) of the IRAs were morphologically classified as follows, based on quantitative and qualitative analyses12: (1) type II lesion (incomplete obstruction with an angiographic thrombus with the greatest linear dimension more than three times the reference lumen diameter); (2) cutoff pattern (lesion morphology with an abrupt cutoff at the obstructive level); (3) the presence of accumulated thrombus (ie, ⬎ 5 mm of linear dimension) proximal to the occlusion; (4) the presence of a floating thrombus proximal to the occlusion; (5) persistent dye stasis distal to the obstruction; and (6) an RLD of the IRA of ⱖ 4.0 mm. Definitions Early MI was defined as an AMI that had occurred ⬎ 12 h to ⱕ 7 days before patient presentation, and recent MI was defined as an AMI that had occurred ⱖ 8 days to ⬍ 30 days before patient presentation. The diagnosis of AMI was based on the following criteria: (1) a history of typical chest pain lasting for ⬎ 30 min with either an ST-segment elevation or a pathologic Q wave occurring in two consecutive inferior, lateral, or precordial leads; (2) typical chest pain lasting for ⬎ 30 min with a new onset of complete left bundle branch block; and (3) angiographic findings of an occluded coronary vessel with any one of the following criteria: an elevation of creatine kinase with a creatine kinase-MB fraction of ⬎ 4% or an elevated myoglobin or an elevated cardiac subfraction of lactate dehydrogenase levels on at least one occasion; new ECG changes suggestive of ischemia; or new wall motion abnormalities seen on an echocardiogram. Procedural success was defined as a reduction to a residual stenosis of ⬍ 40% by balloon angioplasty or successful stenting at the desired position with a residual stenosis of ⬍ 20% followed by Thrombolysis in Myocardial Infarction (TIMI)13 flow grade of 3 in the IRA. Advanced congestive heart failure (CHF) was defined as New York Heart Association (NYHA) classification of ⱖ 3. Multivessel disease was defined by a stenosis of ⬎ 50% in two or more major epicardial coronary arteries. Restenosis was defined as ⱖ 50% stenosis of a previous targeted lesion of the IRA. CHEST / 126 / 1 / JULY, 2004

39

Clinical Follow-up and Follow-up Methods A 5-year follow-up study (from January 1998 through March 2003) was conducted either through outpatient department contact (82.3%) or by telephone contact (7.1%) by personnel at individual clinic sites. Except for nine patients (2.6%) who were lost to follow-up, data are available for 5 years after patients underwent PCI. Statistical Analysis Continuous variables were compared using the Wilcoxon rank sum test. Categoric variables were compared using the ␹2 or Fisher exact test. Stepwise logistic regression analysis was used to determine independent predictors of unsuccessful reperfusion and 30-day mortality rate. Statistical analysis was performed using a statistical software package (SAS for Windows, version 8.2; SAS Institute; Cary, NC). A p value of ⬍ 0.05 was considered to be statistically significant.

Results From January 1998 through December 2001, cardiac catheterization was performed in 408 consecutive and unselected patients of any age who presented with either early or recent MI in our hospital. Thirteen of the 408 patients (3.2%) were treated conservatively due to ⬍ 60% stenosis of the culprit lesion with normal coronary blood flow. Of the remaining 395 patients, 8 patients (2.0%) with infarct-related mechanical complications (ie, severe mitral regurgitation, 3 patients; acquired ventricular septal defect, 5 patients) and 10 patients (2.5%) with significant left main artery disease and severe multivessel disease that required either urgent or emergency surgical intervention after diagnostic cardiac catheterization also were excluded. Seven of these 18 patients (including 1 patient with mitral valve regurgitation, 3 patients with ventricular septal defect, and 3 patients with left main artery and multivessel disease) died in the hospital after undergoing surgery. The remaining 377 patients were enrolled into this study. The baseline characteristics, angiographic results, and clinical outcomes of 377 patients are given in Tables 1 and 2. Of the 272 patients (72.1%) with early MI, 119 patients had an MI duration of ⱕ 3 days. Of 105 patients (27.9%) with recent MI, 64 patients had an MI duration of ⱖ 8 days and ⱕ 14 days. The procedural success rate was 90.2%. Reocclusion and repeat PCI were required in three patients. The 30-day mortality rate was 6.8%. Determinants of Unsuccessful Reperfusion and 30-Day Mortality Rate A univariate analysis of the baseline, clinical, angiographic, and procedural characteristics related to 40

Table 1—Baseline Characteristics of 377 Patients Patients Variables

No.

%

Age, yr Male gender Smoking Hypertension Diabetes mellitus Hypercholesterolemia Previous MI Previous coronary bypass surgery Anterior wall infarction Inferior wall infarction Lateral wall infarction Uncertain infarct location by ECG ST elevation MI Complete LBBB† Advanced CHF‡ Postinfarct angina Early MI Recent MI Intra-aortic balloon pump support Angiotensin-converting enzyme inhibitors§ ␤-Blockers§ Calcium-channel blockers§ Nitrates§ Diuretics§

377 306 218 169 126 168 18 2 184 139 5 49 370 7 88 212 272 105 53 288

61.0 ⫾ 11.4* 81.2 57.8 44.8 33.4 44.6 4.8 0.5 48.8 36.9 1.3 13.0 98.1 1.9 23.3 56.2 72.1 27.9 14.1 76.4

344 49 362 102

91.2 13.0 96.0 27.1

*Values given as mean ⫾ SD. †LBBB ⫽ left bundle branch block. ‡Defined as NYHA classification ⱖ 3. §Medication given in hospital and at discharge.

unsuccessful reperfusion was performed, and only the factors that were significantly associated with unsuccessful reperfusion are shown in Table 3. The

Table 2—Angiographic Results and 30-Day Clinical Outcomes of 377 Patients Patients Variables

No.

%

Left anterior descending artery Right coronary artery Left circumflex artery HBTF Multivessel disease Intercoronary collaterals Grade 1 ⱖ Grade 2 Pre-PCI TIMI flow grade ⱕ1 ⱖ2 Preceding thrombolytic therapy Adjunctive tirofiban therapy Stenting Procedural success Stage PCI to noninfarct vessel 30-d mortality

207 126 44 105 227 117 63 54

54.9 33.4 11.7 27.9 60.2 31.0 16.7 14.3

205 172 43 55 230 340 67 26

54.4 45.6 11.4 14.9 61.0 90.2 18.8 6.9

Clinical Investigations

Table 3—Univariate Logistic Regression Analysis of Baseline, Clinical, Angiographic, and Procedural Characteristics Related to Unsuccessful Reperfusion Variables

ⱕ TIMI Flow grade 2*

OR

95% CI

p Value

Diabetes vs nondiabetes Pre-TIMI flow grade ⱕ 1 vs ⱖ 2 CHF, NYHA classification ⱖ 3 vs ⱕ 2 With vs without HBTF Timely PCI ⱕ 3 d vs ⱖ 4 d

17.5 (22/126) vs 6.0 (15/250) 15.6 (32/205) vs 2.9 (5/172) 20.5 (18/88) vs 6.6 (19/289) 24.8 (26/105) vs 4.0 (11/272) 21.0 (25/119) vs 4.7 (12/258)

3.31 6.18 3.66 12.6 5.45

1.65–6.65 2.35–16.24 1.82–7.33 5.53–28.68 2.63–11.29

0.0007 0.0002 0.0003 ⬍ 0.0001 ⬍ 0.0001

*Values given as % (No. of patients with condition/total No. of patients with TIMI flow grade 2).

most significant factors among these variables were HBTF and timing (ie, ⱕ 3 days after presentation) of the PCI. Advanced CHF was significantly related to increased unsuccessful reperfusion. The incidence of HBTF was significantly higher in patients with advanced CHF than in patients without advanced CHF (37.5% [33 of 88 patients] vs 24.9% [72 of 289 patients], respectively; p ⫽ 0.043). A firm relation was also present for diabetes and unsuccessful reperfusion, whereas a pre-PCI TIMI flow grade of ⱖ 2 was associated with increased successful reperfusion. Using the ␹2 test, the incidences of HBTF and unsuccessful reperfusion were strongly correlated to the duration of MI (Fig 1). By multiple stepwise logistic regression analysis (Table 4), HBTF was an independent determinant for unsuccessful reperfusion, along with timing (ie, ⱕ 3 days after presentation) of PCI, advanced CHF, and diabetes.

A univariate analysis of baseline, clinical, angiographic, and procedural characteristics related to the 30-day mortality rate was performed, and only the factors significantly associated with the 30-day mortality rate are shown in Table 5. Of the known variables, including baseline, clinical, and angiographic variables, we found advanced CHF, failed reperfusion, and PCI time of ⱕ 3 days after presentation to be the most significant factors for the 30-day mortality rate. Strongly prognostic relations also were present for age ⱖ 70 years, diabetes, and multivessel disease, whereas a pre-PCI TIMI flow grade of ⱖ 2 was associated with improved survival. By multiple stepwise logistic regression analysis (Table 6), only advanced CHF, unsuccessful reperfusion, PCI time of ⱕ 3 days after presentation, and multivessel disease were significant independent predictors of an increased 30-day mortality rate.

Figure 1. Bar graphs showing the relationship between the duration of MI and the incidence of HBTF (dark bars) and unsuccessful reperfusion (light bars). * ⫽ p ⬍ 0.0062 for the incidence of HBTF at ⱕ 3 days vs ⱖ 4 to ⱖ 7 days and ⱖ 8 days; † ⫽ p ⬍ 0.0001 for the incidence of unsuccessful reperfusion at ⱖ 3 days vs ⱖ 4 to ⱕ 7 days and ⱖ 8 days. www.chestjournal.org

CHEST / 126 / 1 / JULY, 2004

41

Table 4 —Multiple Stepwise Logistic Regression Analysis of Predictors Related to Unsuccessful Reperfusion Variables

OR

95% CI

p Value

HBTF Timely PCI (ⱕ 3 d) Advance CHF (NYHA functional class ⬎ 3) Diabetes mellitus

15.53 4.10 4.10

6.09–39.60 1.79–7.36 1.70–9.91

⬍ 0.0001 0.0008 0.002

3.03

1.03–7.06

0.010

Furthermore, a subgroup analysis of advanced CHF demonstrated that patients who underwent successful reperfusion had a significantly lower 30-day mortality rate than did those who had undergone unsuccessful reperfusion (17.1% [12 of 70 patients] vs 61.1% [11 of 18 patients], respectively; p ⫽ 0.0004). Follow-up Angiographic Results and Long-term Outcomes At 6 months, follow-up angiographies were performed in 285 patients (81.2%). Follow-up measurements of left ventricular ejection fraction (LVEF) showed significantly more improvement than the initial LVEF (59.3 ⫾ 13.8% vs 50.4 ⫾ 13%, respectively; p ⬍ 0.0001). Angiographic restenosis of the IRA occurred in 101 patients (35.4%). Compared with balloon angioplasty, patients who underwent stenting had a significantly lower incidence of restenosis (30.7% [54 of 176 patients] vs 43.1% [47 of 109 patients], respectively; p ⫽ 0.033). Of the 101 patients, only 69 (24.2%) required repeated target vessel revascularization (TVR), including PCI in 64 patients and coronary artery bypass grafting in 5 patients. Total non-TVR was performed in 86 patients. Six of these 86 patients experienced AMI again due to another vessel infarction, and 3 patients received coronary artery bypass grafting due to left main artery and multivessel disease during the longterm follow-up. At a mean (⫾ SD) follow-up duration of 38.9 ⫾ 14.2 months (range, 15 to 64 months), the

Table 6 —Multiple Stepwise Logistic Regression Analysis of Predictors Related to 30-Day Mortality Variables

OR

95% CI

p Value

Advance CHF (NYHA functional class ⱖ 3) Unsuccessful reperfusion Timely PCI (ⱕ 3 d) Multivessel disease

29.85

7.84–113.70

⬍ 0.0001

7.57 4.81 9.22

2.29–25.07 1.60–14.41 1.63–52.04

0.0009 0.005 0.0119

total mean cumulative mortality rate was 12.5 ⫾ 1.96% (Kaplan-Meier analysis) [Fig 2]. The total cumulative mortality rate after hospital discharge was 6.5% (18 patients) for the entire group. Documented cardiac death occurred in 14 patients (4.0%). Four patients died of other causes, including one patient with lung cancer, two patients with rectal cancer, and one patient with hepatoma. The allcause cumulative mortality rates after hospital discharge were 2.6%, 3.6%, 4.9%, and 6.5%, respectively, after 1, 2, 3, and 4 years. Univariate analysis (Table 7) demonstrated that the significant risk factors for long-term mortality were old age (ie, ⱖ 70 years), multivessel disease, and advanced CHF. Multiple stepwise logistic regression analysis revealed that only advanced CHF (odds ratio [OR], 3.46; 95% confidence interval [CI], 1.26 to 9.52; p ⫽ 0.016) and old age (ⱖ 70 years) [OR, 4.41; 95% CI, 1.59 to 12.24; p ⫽ 0.004] were significant determinants of the long-term mortality rate after adjustment for multiple clinical variables. Discussion The Association of HBTF and Timely Reperfusion With Unsuccessful Reperfusion In the present study, one of the important findings was that the HBTF of the IRA was the most significant predictor of unsuccessful reperfusion. Interestingly, the link between distinguished angiographic morphologic features and unfavorable PCI

Table 5—Univariate Logistic Regression Analysis of Baseline, Clinical, Angiographic, and Procedural Characteristics Related to 30-Day Mortality Variables

Mortality*

OR

95% CI

p Value

Age ⱖ 70 vs ⬍ 70 yr Diabetes vs nondiabetes Multivessel vs single-vessel disease Pre-TIMI flow grade ⱕ 1 vs ⱖ 2 CHF, NYHA classification ⱖ 3 vs ⱕ 2 Failed vs successful PCI Timely PCI ⱕ 3 d vs ⱖ 4 d

11.8 (11/93) vs 5.3 (15/284) 14.3 (18/126) vs 3.2 (8/250) 10.6 (24/227) vs 1.3 (2/150) 9.8 (20/205) vs 3.5 (6/172) 26.1 (23/88) vs 10.4 (3/289) 35.1 (13/37) vs 3.8 (13/340) 15.1 (18/119) vs 3.1 (8/258)

2.41 5.04 8.75 2.29 33.7 13.6 5.57

1.06–5.44 2.13–11.9 2.04–37.6 1.17–7.63 9.83–115.7 5.69–32.62 2.35–13.22

0.035 0.0002 0.0035 0.022 ⬍ 0.0001 ⬍ 0.0001 ⬍ 0.0001

*Values given as % (No. of patients with condition/total No. of patients who died). 42

Clinical Investigations

Figure 2. Kaplan-Meier cumulative mortality curve in 377 patients for ⬎ 5 years.

results was also found in another of our studies12 in patients with AMI undergoing primary PCI. The same phenomenon in different clinical settings led us to a further understanding of the impact of HBTF on unsuccessful reperfusion. Another important finding in the present study was that PCI in patients with early MI (ie, ⬎ 12 h to

ⱕ 3 days after patient presentation) had a significantly higher incidence of unsuccessful reperfusion than did PCI in those patients with later MI (ie, ⱖ 4 days after patient presentation). The strong association between the duration of MI and failed reperfusion (Fig 1) further supported the results of our report,12 which showed that the combined slow-flow

Table 7—Univariate Logistic Regression Analysis of Predictors of Long-term Mortality Variables

Mortality*

OR

95% CI

p Value

Age ⱖ 70 vs ⬍ 70 yr Men vs women Diabetes vs nondiabetes Smoker vs nonsmoker With vs without hypertension With vs without hypercholesterolemia Multivessel vs single-vessel disease With vs without previous MI Anterior vs nonanterior MI Pre-TIMI flow grade ⱕ 1 vs ⱖ 2 With vs without tirofiban therapy With vs without thrombolytic therapy With vs without collaterals With vs without post-MI angina CHF, NYHA classification ⱖ 3 vs ⱕ 2 Failed vs successful PCI Timely PCI (ⱕ 3 d vs ⱖ 4 d) LVEF† ⬍ 50% vs ⱖ 50%

13.4 (11/82) vs 2.6 (7/269) 5.9 (17/288) vs 1.6 (1/63) 3.7 (4/108) vs 5.8 (14/243) 6.3 (13/207) vs 3.5 (5/144) 7.6 (12/157) vs 3.1 (6/194) 3.7 (6/161) vs 6.3 (12/190) 7.4 (15/203) vs 2.0 (3/148) 11.1 (2/18) vs 4.8 (16/333) 3.5 (6/173) vs 6.7 (12/178) 3.8 (7/185) vs 6.6 (11/166) 4.0 (2/50) vs 5.3 (16/301) 4.9 (2/41) vs 5.2 (16/310) 1.8 (2/110) vs 6.6 (16/241) 3.6 (7/197) vs 7.1 (11/154) 13.9 (9/65) vs 3.1 (9/286) 12.5 (3/24) vs 4.6 (15/327) 3.0 3/101) vs 6.0 (15/250) 7.4 (12/162) vs 3.2 (6/189)

5.80 3.89 0.63 1.85 2.59 0.57 3.86 2.48 0.50 0.55 0.74 0.94 0.26 0.48 4.95 2.97 0.48 2.44

2.17–15.50 0.51–29.78 0.02–1.95 0.64–5.31 0.95–7.07 0.21–1.57 1.10–13.57 0.52–11.71 0.18–1.36 0.21–1.46 0.17–3.33 0.21–4.26 0.06–1.15 0.18–1.27 1.88–13.02 0.80–11.08 0.14–1.69 0.90–6.66

0.0005 0.191 0.419 0.253 0.063 0.279 0.036 0.253 0.172 0.234 0.697 0.938 0.076 0.13 0.001 0.105 0.254 0.082

*Values given as % (No. of patients with condition/total No. of patients who died). †Measurements were immediately performed after the coronary angioplasty were recorded in the 30° right anterior oblique and the 60° left anterior oblique views. www.chestjournal.org

CHEST / 126 / 1 / JULY, 2004

43

and no-reflow phenomenon in patients in a clinical setting of AMI undergoing primary PCI was directly proportional to the time of reperfusion. Such a consistent clinical finding in different clinical situations may be explained because as time goes on after the AMI, the thrombus becomes organized and firmer.12 The organized thrombus is broken down into fragmented debris (ie, mixed macroemboli and microemboli) by mechanical devices such as balloons or stents during primary or elective PCI, and these debris cause embolization of the branch or distal vessels and can completely plug the microvasculature, resulting in the slow-flow or no-reflow phenomenon. This finding further explains why rescue PCI usually had a relatively low rate of successful reperfusion4,7 and a high rate of untoward clinical outcomes.5,11 Therefore, we suggest that there is no frantic rush to perform PCI immediately after an MI if the window of truly early intervention (ie, ⬍ 12 h) has closed. Another remarkable finding in the present study was that the incidence of HBTF was dramatically decreased later after AMI. Our findings increased our understanding that peculiarly vanishing HBFT was inversely proportional to the timing (ie, on the third day) after an AMI (Fig 1). This time course of the incidence of HBTF could explain why a significantly higher rate of unsuccessful reperfusion was observed during early PCI procedures in our patients and in previous rescue PCI studies.4,7 Taken together, we suggest that an early invasive approach in patients with this clinical condition would be hazardous, with the benefit seen only in those in whom invasive measures were carried on ⱖ 4 days after the MI. The Impact of Diabetes and Advanced CHF on the No-Reflow Phenomenon, and the Determinants of the 30-Day Untoward Clinical Outcomes The impact of diabetes on future angiographic and clinical outcomes after PCI has been debated extensively.14,15 More recently, the firm relationship between hyperglycemia and the no-reflow phenomenon in patients with AMI who were undergoing primary PCI has been found by Iwakura et al16 In the present study, we also found that diabetes was an independent predictor for unsuccessful reperfusion. Several possible mechanisms have been suggested to explain this situation. First, hyperglycemia increases the levels of intercellular adhesion molecule-117 or P-selectin,18 which augment the plugging of leukocytes in the capillaries. Second, leukocytes trapped in the coronary capillaries and venules early after coronary reperfusion are much more frequently observed in the diabetic rat heart than in the nondia44

betic rat heart.19 Third, the plugging of enhanced leukocytes in the microcirculation might further contribute to the no-reflow phenomenon.20 Surprisingly, advanced CHF was a strong predictive factor for unsuccessful reperfusion. We remain uncertain as to why advanced CHF also played a pivotal role for failed reperfusion. However, three reasonable explanations could account for its occurrence. First, the incidence of HBTF in the IRA was significantly higher in the patients with advanced CHF than in the patients without advanced CHF. Second, in the present study, we found that the incidence of advanced CHF was significantly higher in patients with diabetes than in patients without diabetes, which was strongly associated with unsuccessful reperfusion. This clinical observation was consistent with those of previous studies, which have found that patients with hyperglycemia usually had larger infarct sizes,21 a high incidence of CHF, and cardiogenic shock.22,23 Third, the strong positive correlation between patients with CHF and the hypercoagulable state that was found in the study by Gibbs et al24 also supported our finding. In the present study, we found that, without early and successful reperfusion in the IRA, MIs occurring in patients in the elective PCI group would lead to the development of advanced CHF in quite a few of the patients (23.3%). Furthermore, the initial lack of response to PCI in these patients directly translated into marked increases of in-hospital deaths. These findings explain why advanced CHF was another important factor related to the 30-day mortality rate in the present study. Our findings implied that performing PCI in patients with advanced CHF without a well-controlled situation was a formidable challenge and that the selection of an appropriate time for the performance of PCI should be taken into consideration. The striking impact of the lack of response to primary PCI on unfavorable outcome has been extensively debated in some studies,12,25 in which we found that the 30-day mortality rate was strongly related to unsuccessful reperfusion. Our findings are consistent with those of some studies12,25 and suggested that lack of response to PCI was harmful to the patients, even outweighing the benefit from a successful reperfusion. We have demonstrated26 that the presence of multivessel disease was a significant independent predictor of an increased 30-day mortality rate in patients with cardiogenic shock who were undergoing primary PCI. Therefore, it was not surprising that multivessel disease was also found to be a significant determinant for the 30-day mortality rate in the present study. Surprisingly, however, multivessel disease was no longer an independent preClinical Investigations

dictor for long-term mortality. This can be explained as the consequence of maintaining a high patency rate for the IRA after PCI and to more complete revascularization of other stenotic vessels. Follow-up Angiographic Results and Long-term Outcomes Repeated TVR in 22 to 24% of patients with AMI who were undergoing primary PCI has been reported by the Registry and Phone Patient Outcomes Research Trial27 and the Controlled Abciximab and Device Investigation to Lower Late Angioplasty Complications trial.28 In the present study, we found that ⬍ 25.0% of patients required further TVR. Our result was consistent with those of these two clinical trials.27,28 Of importance was the fact that this persistent high patency rate for the IRA directly translated into improved left ventricle function and a low long-term cardiac death rate of 4%. There were several potential limitations of this study. First, no patients were used as a control group for comparison. Therefore, our study did not determine exactly how PCI provided additional benefits to our patients. Second, as our study was not designed to discuss the basic mechanism of duration of the thrombin activity, we could not exclude active thrombin from being responsible for the high burden of thrombus. Third, although our study conveyed useful information for long-term outcomes in patients with this clinical condition, our study was a single-center experience and, therefore, might not completely reflect the real-world experience. Fourth, long-term survival also may be favorably influenced by certain pharmacologic treatments, such as ␤-blockers, angiotensin-converting enzyme inhibitors, and lipid-lowering agents. However, as our study was not designed to discuss these additional pharmacologic benefits, we could not provide evidence other than that for the impact of PCI on outcomes. In conclusion, this study provides intriguing evidence that PCI performed on ⱖ day 4 after an MI is probably a safe treatment alternative. The clinical benefit of successful PCI followed by complete revascularization of the other vessel diseases is maintained during long-term follow-up. Although the number of patients with advanced CHF and old age was small in this series, our results suggest that PCI was associated with poor prognosis in these patients.

References 1 Grines CL, Browne KF, Marco J, et al. A comparison of primary angioplasty with thrombolytic therapy for acute myocardial infarction. N Engl J Med 1993; 328:673– 679 www.chestjournal.org

2 The GUSTO Angiographic Investigators. The effects of tissue plasminogen activator, streptokinase, or both on coronaryartery patency, ventricular function, and survival after acute myocardial infarction. N Engl J Med 1993; 329:1615–1622 3 Gibbons RJ, Holmes DR, Reeder GS, et al. Immediate angioplasty compared with the administration of a thrombolytic agent followed by conservative treatment for myocardial infarction: the Mayo Coronary Care Unit and Catheterization Laboratory Groups. N Engl J Med 1993; 328:685– 691 4 Ellis SG, Van de Werf F, Ribeiro-daSilva E, et al. Present status of rescue coronary angioplasty: current polarization of opinion and randomized trials. J Am Coll Cardiol 1992; 19:681– 686 5 McKendall GR, Forman S, Sopko G, et al. Value of rescue percutaneous transluminal coronary angioplasty following unsuccessful thrombolytic therapy in patients with acute myocardial infarction. Am J Cardiol 1995; 76:1108 –1111 6 Vogt A, Neuhaus KL. Thrombolysis and mechanical intervention following myocardial infarction. Eur Heart J 1996; 17(suppl):49 –54 7 Gibson CM, Cannon CP, Greene RM, et al. Rescue angioplasty in the Thrombolysis in Myocardial Infarction (TIMI) 4 trial. Am J Cardiol 1997; 80:21–26 8 Nakagawa Y, Matsuo S, Kimura T, et al. Thrombectomy with angiojet catheter in native coronary arteries for patients with acute or recent myocardial infarction. Am J Cardiol 1999; 83:994 –999 9 Cafri C, Denktas AE, Crystal E, et al. Contribution of stenting to results of rescue PTCA. Catheter Cardiovasc Interv 1999; 47:411– 414 10 Gorfinkel HJ, Berger SM, Klaus AP, et al. Rescue angioplasty in failed thrombolysis in acute myocardial infarction: a community hospital experience. J Invasive Cardiol 1997; 9:83– 87 11 Topol E, ed. Textbook of interventional cardiology. 2nd ed. Philadelphia, PA: WB Saunders, 1994; 301–304 12 Yip HK, Cheng MC, Chang HW, et al. Angiographic morphologic features of infarct-related arteries and timely reperfusion in acute myocardial infarction: predictors of slow-flow and no-reflow phenomenon. Chest 2002; 122:1322–1332 13 TIMI Study Group. The Thrombolysis in Myocardial Infarction (TIMI) Trial: phase 1 findings. N Engl J Med 1985; 312:932–936 14 Kornowski R, Mintz GS, Kent KM, et al. Increased restenosis in diabetes mellitus after coronary interventions is due to exaggerated intimal hyperplasia: a serial intravascular ultrasound study. Circulation 1997; 95:1366 –1369 15 Moustapha A, Assali AR, Sdringola S, et al. Percutaneous and surgical interventions for in-stent restenosis: long-term outcomes and effect of diabetes mellitus. J Am Coll Cardiol 2001; 37:1877–1882 16 Iwakura K, Ito H, Ikushima M, et al. Association between hyperglycemia and the no-reflow phenomenon in patients with acute myocardial infarction. J Am Coll Cardiol 2003; 41:1–7 17 Marfella R, Esposito K, Giunta R, et al. Circulating adhesion molecules in humans: role of hyperglycemia and hyperinsulinemia. Circulation 2000; 101:2247–2251 18 Booth G, Stalker TJ, Lefer AM, et al. Elevated ambient glucose induces acute inflammatory events in the microvasculature: effects of insulin. Am J Physiol Endocrinol Metab 2001; 280:E848 –E856 19 Hokama JY, Ritter LS, David-Gorman G, et al. Diabetes enhances leukocyte accumulation in the coronary microcirculation early in reperfusion following ischemia. J Diabetes Complications 2000; 14:96 –107 CHEST / 126 / 1 / JULY, 2004

45

20 Engler RL, Dahlgren MD, Morris DD, et al. Role of leukocyte in response acute ischemia and reflow in dogs. Am J Physiol 1986; 251:H314 –H323 21 Tansey MJ, Opie LH. Plasma glucose on admission to hospital as a metabolic index of the severity of acute myocardial infarction. Can J Cardiol 1986; 2:326 –331 22 Capes SE, Hunt D, Malmberg K, et al. Stress hyperglycemia and increased risk of death after myocardial infarction in patients with and without diabetes: a systemic overview. Lancet 2000; 355:773–778 23 Oswald GA, Corcoran S, Yudkin JS. Prevalence and risks of hyperglycemia and undiagnosted diabetes in patients with acute myocardial infarction. Lancet 1984; 1:1264 –1267 24 Gibbs CR, Blann AD, Waston R, et al. Circulation 2001; 103:1746 –1751

46

25 Niederer AVW, Pfatterott C, Engel EL, et al. Direct percutaneous transluminal coronary angioplasty in acute myocardial infarction. Eur Heart J 1998; 19:917–921 26 Yip HK, Wu CJ, Chang HW, et al. Comparison of impact of primary percutaneous transluminal coronary angioplasty and primary stenting on short-term mortality in patients with cardiogenic shock and evaluation of prognostic determinants. Am J Cardiol 2001; 87:1184 –1188 27 Brener SJ, Barr LA, Burchenal JEB, et al. Randomized, placebo-control trial of platelet glycoprotein IIb/IIIa blockage with primary angioplasty for acute myocardial infarction. Circulation 1998; 98:734 –741 28 Stone GW, Grines CL, Cox DA, et al. Comparison of angioplasty with stenting, with or without abciximab, in acute myocardial infarction. N Engl J Med 2002; 346:957–966

Clinical Investigations