Percutaneous coronary intervention for cardiogenic shock in the SHOCK Trial Registry

Interventional Cardiology

Percutaneous coronary intervention for cardiogenic shock in the SHOCK Trial Registry John G. Webb, MD,a Timothy A. Sanborn, MD,b Lynn A. Sleeper, ScD,c Ronald G. Carere, MD,a Christopher E. Buller, MD,d James N. Slater, MD,h Kenneth W. Baran, MD,f Patrick T. Koller, MD,f J. David Talley, MD,g Mark Porway, MD,e and Judith S. Hochman, MD,h for the SHOCK Investigators New York, NY, Watertown, Mass, Vancouver, British Columbia, Canada, Springfield, Mass, St Paul, Minn, and Little Rock, Ark

Background The SHOCK Registry prospectively enrolled patients with cardiogenic shock complicating acute myocardial infarction in 36 multinational centers.

Methods Cardiogenic shock was predominantly attributable to left ventricular pump failure in 884 patients. Of these, 276 underwent percutaneous coronary intervention (PCI) after shock onset and are the subject of this report.

Results The majority (78%) of patients undergoing angiography had multivessel disease. As the number of diseased arteries rose from 1 to 3, mortality rates rose from 34.2% to 51.2%. Patients who underwent PCI had lower in-hospital mortality rates than did patients treated medically (46.4% vs 78.0%, P < .001), even after adjustment for patient differences and survival bias (P = .037). Before PCI, the culprit artery was occluded (Thrombolysis In Myocardial Infarction grade 0 or 1 flow) in 76.3%. After PCI, the in-hospital mortality rate was 33.3% if reperfusion was complete (grade 3 flow), 50.0% with incomplete reperfusion (grade 2 flow), and 85.7% with absent reperfusion (grade 0 or 1 flow) (P < .001). Conclusions This prospective, multicenter registry of patients with acute myocardial infarction complicated by cardiogenic shock is consistent with a reduction in mortality rates as the result of percutaneous coronary revascularization. Coronary artery patency was an important predictor of outcome. Measures to promote early and rapid reperfusion appear critically important in improving the otherwise poor outcome associated with cardiogenic shock. (Am Heart J 2001;141:964-70.)

Cardiogenic shock is the major cause of death among patients admitted to the hospital with acute myocardial infarction (MI).1 The poor outcome with medical treatment alone has led to interest in percutaneous coronary intervention (PCI).2-4 The SHould we emergently revascularize Occluded coronaries for Cardiogenic shocK? (SHOCK) Registry was a prospective, multicenter registry of patients with known or suspected cardiogenic shock complicating acute MI who were not enrolled in From aSt Paul’s Hospital, Vancouver, British Columbia, Canada; bNew York Hospital–Cornell Medical Center, New York, NY; cNew England Research Institutes, Watertown, Mass; dVancouver General Hospital, Vancouver, British Columbia, Canada; eBaystate Medical Center, Springfield, Mass; fSt Paul Heart Clinic, St Paul, Minn; gUniversity of Arkansas, Little Rock; and hSt Luke’s–Roosevelt Medical Center, New York, NY. Supported by RO1 grants HL50020 and HL49970, 1994 to 1999, from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Md. Submitted August 11, 2000; accepted February 19, 2001. Reprint requests: John Webb, MD, Director Cardiac Catherization and Interventional Cardiology, St Paul’s Hospital, 1081 Burrard St, Vancouver, BC, Canada V6Z 1Y6. E-mail: [email protected] Copyright © 2001 by Mosby, Inc. 0002-8703/2001/$35.00 + 0 4/1/115294 doi:10.1067/mhj.2001.115294

the randomized SHOCK Trial. Through this registry, the role of PCI in the treatment of cardiogenic shock in a large and relatively unselected population was examined.

Methods SHOCK Registry Thirty-six multinational centers participated in the registry between April 1993 and August 1997. Local institutional review committee approval was obtained. These centers also participated in the prospective, randomized SHOCK Trial of early revascularization in the treatment of cardiogenic shock.5 Patients were prospectively identified by study coordinators for a clinical diagnosis of MI and suspected cardiogenic shock. Patients who failed to meet strict criteria for trial random assignment and those who met these criteria but were not randomly assigned were registered. In all, 1492 patients with suspected cardiogenic shock complicating acute MI were prospectively identified. We excluded 266 patients in whom shock was attributable to mechanical causes (ventricular septal rupture, cardiac tamponade, severe valvular disease), a catheterization laboratory complication, excess β-blockade, or calcium channel blockade; 40 patients who underwent PCI or coronary artery bypass graft surgery

American Heart Journal Volume 141, Number 6

Webb et al 965

Table I. Characteristics of SHOCK Registry patients with pump failure Angioplasty n All patients Age (y) <75 y Female White, non-Hispanic Transfer Diabetes Prior hypertension Kidney failure Peripheral vascular disease Prior congestive heart failure Prior CABG Prior PCI Prior MI Median time from MI onset to shock onset (h) MI to shock <6 h Location of infarction Anterior Inferior Posterior Lateral Apical Transmural infarction† Median highest creatine kinase (U/L) Left ventricular ejection fraction (%)‡ Systolic blood pressure (mm Hg)§ Diastolic blood pressure (mm Hg)§ Heart rate (beats/min)§ Cardiac index (L/min/m2)§ Pulmonary capillary wedge pressure (mm Hg)§

Medical therapy n

Bypass surgery

3-Group

n

P value

276 276 276 276 275 276 265 262 262 179 263 271 269 259 243 248

31.2% 63.8 ± 11.7 82.6% 39.9% 78.9% 50.4% 29.4% 50.0% 7.3% 11.7% 12.5% 5.9% 9.3% 32.4% 3.2 64.1%

499 499 499 499 492 499 487 476 471 293 474 488 477 481 430 459

56.5% 71.8 ± 11.4 56.9% 37.5% 82.7% 34.1% 36.3% 53.6% 12.7% 20.5% 25.1% 11.5% 4.8% 43.2% 7.5% 43.6%

109 109 109 109 109 109 109 108 106 82 106 108 103 108 98 106

12.3% 64.0 ± 11.4 83.5% 24.8% 91.7% 61.5% 27.5% 45.4% 12.3% 28.1% 13.2% 9.3% 3.9% 38.9% 9.2% 39.6%

<.001* <.001* .016 .005 <.001* .068 .262 .059 .004* <.001* .036* .039 .016* .001* <.001*

266 266 266 266 266 274 267 114 264 242 263 142 203

50.4% 51.5% 21.1% 32.7% 10.5% 88.0% 3126 32.9 ± 14.0 89.8 ± 22.4 55.0 ± 15.8 95.1 ± 24.6 2.1 ± 0.8 24.4 ± 8.9

430 429 428 430 430 497 451 125 468 392 469 201 252

57.4% 44.1% 16.8% 30.7% 9.5% 78.7% 1208 28.1 ± 13.0 86.7 ± 23.4 51.2 ± 17.8 93.4 ± 26.7 1.9 ± 0.7 23.5 ± 8.4

96 96 96 96 96 109 100 53 103 93 101 57 75

62.5% 42.7% 22.9% 34.4% 6.3% 67.9% 2027 30.0 ± 9.4 90.1 ± 22.7 51.6 ± 17.6 97.0 ± 25.2 2.2 ± 0.7 23.1 ± 8.6

.069 .122 .213 .724 .502 <.001* <.001* .014* .078 .080* .398 .007* .574

*P < .05 for comparison of angioplasty versus medical therapy. †ST elevation or new left bundle branch block. ‡Day of or after shock diagnosis. §Measurements obtained on support measures.

(CABG) before shock onset; and 302 patients randomly assigned in the SHOCK Trial.3 The remaining 884 registry patients whose cardiogenic shock was attributed predominantly to pump failure (left or right ventricular failure) constitute the basis of this report. Data were abstracted from the medical records by study coordinators.2 Cardiac catheterization reports were abstracted at the clinical coordinating center. Analyses were performed at the New England Research Institutes. If no Thombolysis in Myocardial Infarction (TIMI) flow grade6 was specified, an assessment was made on the basis of the following prespecified criteria: no antegrade flow, total occlusion, or 100% stenosis was set to TIMI grade 0 to 1; sluggish flow, subtotal occlusion, or 95% to 99% stenoses were set to TIMI grade 2; normal flow, patent, or <95% stenosis was set to TIMI grade 3. This method may systematically categorize some patients with TIMI 3 flow as TIMI 2.

teristics and outcomes of patients with PCI, medical therapy, and bypass surgery by using the Fisher exact test for categoric variables, the Kruskal-Wallis test for ordinal or nonnormally distributed variables, and analysis of variance for all other continuous variables. For variables with a 3-group P value <.05, Bonferroni adjustment was used for pairwise comparisons of the PCI and medical therapy groups. Association between death and ordered angiographic variables (coronary anatomy and TIMI flow grade) was assessed by means of the MantelHaenszel test for linear trend. To examine the association between transfer status and survival and between PCI versus medical therapy and survival, Cox proportional hazards modeling with a time-dependent covariate for the grouping variable was used. All analyses were conducted by means of the Statistical Analysis System (SAS) and S-PLUS for Windows.

Results Statistical analysis

Clinical characteristics

Continuous variables were summarized as mean ± SD and categoric variables as percentages. We compared the charac-

Registry patients undergoing PCI (n = 276, 31.2%) differed from those who were treated with medical ther-

American Heart Journal June 2001

966 Webb et al

Table II. Revascularization and in-hospital mortality rates of shock patients with pump failure undergoing angiography by disease severity Diseased vessels

Overall in-hospital mortality rate Angioplasty Mortality rate Bypass surgery Mortality rate Medical treatment Mortality rate

None (n = 5)

1 (n = 106)

2 (n = 108)

3 (n = 287)

60.0% 0% — 0% — 100% 60.0%

33.0% 75.9% 32.9% 5.4% 33.3% 19.8% 38.1%

40.7% 61.1% 42.2% 15.7% 17.7% 25.0% 55.6%

51.2% 41.1% 59.3% 34.2% 29.6% 30.0% 62.8%

apy (56.5%) or bypass surgery (12.3%), as shown in Table I. Patients with PCI, as compared with medically treated patients, were younger, had shock earlier, and were more likely to have been transferred from another institution. They had a higher left ventricular ejection fraction and cardiac index. Patients with PCI were more likely to have had a transmural MI and had a higher creatine kinase level. They were more likely to have a history of MI, heart failure, CABG, or peripheral vascular disease.

Survival after PCI The in-hospital mortality rate in patients undergoing PCI was 46.4%, as compared with 78.0% in medically treated patients and 23.9% in patients undergoing bypass surgery alone (P < .001). PCI survivors (n = 148) compared with nonsurvivors (n = 128) were younger (62.2 ± 12.4 vs 65.6 ± 10.5 years, P = .015) and had a higher left ventricular (LV) ejection fraction (36.1 ± 14.2, n = 77, vs 25.7 ± 10.6, n = 33; P < .001) and cardiac index (2.3 ± 0.8 vs 1.9 ± 0.7 L/min/m2, P < .001). MI location was more often inferior (52.2% vs 34.8%, P = .007) and less often anterior (44.0% vs 59.8%, P = .015). Survivors and nonsurvivors did not differ with respect to sex, diabetes, ST-elevation index MI, right heart catheterization, pulmonary artery wedge pressure, intra-aortic balloon support, stenting, administration of thrombolysis, glycoprotein IIb/IIIa antagonists, inotropes, or vasopressors, or a history of MI, PCI, or CABG. Survival modeling (time from shock onset to discharge) by means of a time-dependent grouping factor for PCI versus medical treatment and adjustment for patient differences (age, prior CABG, highest creatine kinase level, and diastolic blood pressure) demonstrated a PCI versus medical treatment relative risk for death of 0.65, with a 95% confidence interval of 0.43 to 0.97 (n = 562, P = .037).

Timing of interventions Median time from MI to hospital arrival was 4 hours, MI to PCI was 8.8 hours, and shock to PCI was 3.3

hours. PCI mortality rates did not differ significantly (P = .151) by timing category: 40.2% mortality rate for PCI within 6 hours of MI (n = 87), 50.9% if 6 to 12 hours after MI (n = 57), 60.5% if 12 to 24 hours after MI (n = 43), and 43.9% mortality rate if more than 24 hours after MI (n = 82), even after adjustment for transfer status (P = .087). Among the registry as a whole, 376 (42.5%) patients were transferred from a referral center to a SHOCK registry center, arriving a median 11.7 hours after MI onset. Transferred patients, compared with direct admissions to the SHOCK tertiary care centers, had longer delays from MI onset to PCI (median, 10.7 vs 6.3 hours, P < .001). Transferred patients had a higher crude survival rate than nontransferred patients (48.4% vs 31.3%, P < .001), perhaps because transfer patients by definition must survive long enough to be transferred. After adjustment for this bias, the association between transfer status and survival is not significant (P = .901). In 68% (n = 188) of patients undergoing PCI, the procedure was performed within 24 hours of MI. Mortality rates in patients undergoing PCI within 24 hours of MI were similar in the 39% of patients receiving prior thrombolysis (“rescue” angioplasty) and the 61% not receiving thrombolysis (45.2% vs 49.6%, respectively, P = .653).

Coronary anatomy The 513 registry patients who underwent angiography did so a median 14.2 hours after MI onset and 3.2 hours after shock onset. In this cohort, the number of diseased vessels (stenosis ≥50%) was zero in 1.0%, 1 in 20.9%, 2 in 21.3%, and 3 in 56.7%. Angioplasty survivors were more likely to have single-vessel disease than were nonsurvivors (41.2% vs 21.7%, P < .001) and less likely to have triple-vessel or left main disease (29.8% vs 55.7%, P < .001, and 10.0% vs 29.3%, P = .001). As the number of diseased arteries increased from 1 to 3, the in-hospital mortality rate increased from 33.0% to 51.2% (Table II). As the number of diseased vessels increased, the use of PCI declined, PCI-

American Heart Journal Volume 141, Number 6

Figure 1

Final TIMI flow grade after PCI and in-hospital mortality rates in SHOCK Registry patients with pump (left or right ventricular) failure (P < .001).

associated mortality rate increased, and CABG use increased. In-hospital mortality rate was similar after single-vessel procedures and single-stage, multivessel procedures (45.6% vs 53.9%, P = .422). Although PCI mortality rate increased as the number of diseased vessels increased, CABG mortality rates remained similar regardless of the number of diseased vessels (Table II). In 403 patients for whom a culprit artery was identified, this was most often the left anterior descending coronary artery (40.2%), followed by the right coronary artery (31.8%), the circumflex artery (14.4%), saphenous vein graft (6.9%), and the left main artery (6.7%). Among patients undergoing PCI, a critical stenosis (>90%) was present in 53.9% of the left anterior descending coronary, 53.9% of the right coronary, 33.9% of the circumflex, and 5% of the left main coronary arteries, as well as the majority (85.3%) of culprit arteries. The culprit artery was most often occluded, with TIMI grade 0 or 1 flow in 76.3%, TIMI grade 2 flow in 14.9%, and TIMI grade 3 flow in 8.8%.

Webb et al 967

Figure 2

Change in TIMI flow grade after PCI and in-hospital mortality rates in SHOCK Registry patients with pump (left or right ventricular) failure (P = .006).

Figure 3

Residual stenosis and in-hospital mortality rates in SHOCK Registry patients with pump (left or right ventricular) failure (P < .001).

Outcome after PCI Residual stenosis after PCI (n = 225) was <50% in 84.0% and ≥50% in 16.0%. TIMI flow grade (n = 170) was 0 or 1 in 20.6%, grade 2 in 14.1%, and grade 3 in 65.3%. Angioplasty was successful (defined as TIMI flow grade 2 or 3 with a residual stenosis <50%) in 74.8%. Although technically successful, TIMI flow remained suboptimal in 18.1% of patients in whom PCI successfully reduced stenosis to <50% (residual TIMI 0 or 1 flow in 2.5% and TIMI 2 flow in 15.6%). Patency of the infarct artery was a major predictor of in-hospital death after PCI, as shown in Figures 1 through 3. Both before and after PCI, coronary flow correlated with mortality rates. Baseline TIMI flow

grades of 0/1 (n = 174), 2 (n = 34), or 3 (n = 20) were associated with mortality rates of 44.8%, 64.7%, and 15.0%, respectively (Fisher exact test, P = .002). Normal flow after PCI was associated with the lowest mortality rate, TIMI grade 2 flow was associated with an intermediate mortality rate, and TIMI grade 0 or 1 flow had the highest mortality rate (P < .001) (Figure 1). An improvement in angiographic coronary flow with PCI also was associated with survival: lowest mortality rates in those with improved flow, intermediate mortality rates with no change in flow, and highest mortality rates in those with reduced flow (P = .006) (Figure 2). A residual stenosis <50% was associated with a lower

American Heart Journal June 2001

968 Webb et al

Figure 4

Angiographic success and in-hospital mortality rates in SHOCK Registry patients with pump (left or right ventricular) failure. Success is defined as residual stenosis <50% and final TIMI flow grade of 2 or 3 (P < .001).

in-hospital mortality rate (P < .001) (Figure 3), as was successful PCI (P < .001) (Figure 4).

Adjunctive interventions During the period of the study (from 1993 to 1997), the use of diagnostic coronary angiography increased from 45% to 67%. Although CABG rates remained relatively stable, PCI rates exhibited a trend upward from 24% to 36%, resulting in an increased overall rate of revascularization (from 36% to 48%). Over the same period, the use of stents increased from 0% to 32% and that of glycoprotein IIb/IIIa antagonists increased from 0% to 12%. Stent and abciximab use did not correlate with survival, but numbers were small. Coronary stents were placed in 24% of patients undergoing PCI. Stent deployment was successful in 94%. Stents used were the Palmaz-Schatz PS-153 (Johnson & Johnson Inc, Miami, Fla) in 34 patients, Gianturco-Roubin (Cook Cardiology Inc, Bloomington, Ind) in 9, Nir (Boston Scientific Inc, Galway, Ireland) in 5, Multilink (Guidant Inc, Temecula, Ga) in 4, and unknown in 13. A platelet glycoprotein IIb/IIIa antagonist, abciximab, was used in 7% of patients undergoing PCI. Intra-aortic balloon pump support was used in 80.1% of patients undergoing PCI. A total of 21 (7.6%) patients underwent CABG after PCI; the 12 (4.3%) patients who underwent CABG within 24 hours of PCI had an in-hospital mortality rate of 47.6%.

Discussion In the prospective SHOCK Registry, patients who underwent PCI had lower mortality rates than did patients treated medically. This improved outcome is in

part the result of selection bias. However, a favorable impact on survival remains apparent after adjustment for patient differences and survival bias with the use of a time-dependent covariate analysis. The favorable influence of revascularization is consistent with data from the randomized SHOCK Trial, which documented a 13% absolute reduction in 1-year mortality rates.7 Angiographic success rates and associated mortality rates were similar in the randomized trial and the larger cohort of patients in the registry who were clinically selected to undergo PCI.2,5 In the participating SHOCK Registry tertiary centers, almost two thirds of patients who had shock underwent angiography and one third underwent PCI. During the registry period (1993 to 1997) there was a trend toward more frequent coronary angiography and PCI.8 Given the beneficial effects on mortality rates, this trend seems desirable and likely to continue.

Timing of intervention Prior favorable reports of PCI in the setting of shock often include patients undergoing PCI relatively late in the course of infarction and shock.9-18 Early peri-infarct PCI may carry an inherently greater risk than a delayed procedure performed in a more stable survivor,4 and even late reperfusion may be beneficial.1,4,19 In this registry, in-hospital mortality rates were similar, regardless of the timing of PCI in relation to MI or shock onset. This time-independent effect may be related to patient selection in that more ill patients may not survive long enough to undergo late PCI or may be transferred to an institution with interventional facilities, although the survival difference was not significant, even after adjustment for transfer status. Early reperfusion therapy is desirable in the setting of acute MI,4,20 and the randomized SHOCK trial confirmed the long-term survival advantage of a strategy of “early” intervention (within 12 hours of shock onset and 48 hours of infarct onset) compared with initial medical stabilization and delayed revascularization.5,7

Coronary anatomy As expected, the most frequent single culprit artery was the left anterior descending coronary artery. However, the importance of other arteries in the pathogenesis of shock is evidenced by the observation that most patients had a culprit artery other than the left anterior descending. A critical (>90%) stenosis was present in approximately one half of left anterior descending and right coronary arteries; one third of circumflex arteries; and 1 in 20 left main coronary arteries. The majority of patients had multivessel disease, and the number of diseased vessels correlated with mortality rates. As the number of diseased arteries increased from 1 to 3, mortality rates increased from 33% to 51%. The high rate of multivessel disease (78%) and left main

American Heart Journal Volume 141, Number 6

disease (17%) argues for routine elective angiography and revascularization in patients with shock who were not undergoing a more urgent procedure. Angioplasty of the culprit lesion alone has been advocated in the treatment of unstable patients because of the risk of PCI of nonculprit lesions. However, the high rate of multivessel disease and left main disease argues for more complete revascularization. Recent data suggest that stenting may reduce the risk associated with multivessel procedures and increase the completeness of reperfusion.4,21-23 Although selection bias precludes direct comparison, our data suggest that CABG remains an excellent and often preferred option in patients with multivessel disease.

Webb et al 969

Conclusions This large, prospective, multicenter registry of patients with acute MI complicated by cardiogenic shock is consistent with a reduction in mortality rates with revascularization, although selection bias is evident. Angioplasty outcome is correlated with coronary anatomy and particularly with coronary patency. Measures to promote early and rapid reperfusion appear critically important in improving the otherwise poor outcome associated with cardiogenic shock. We thank Rasha Abboud, MS, for providing statistical programming support for this report.

Infarct artery patency

References

Although most patients undergoing angiography did so relatively late and many had received thrombolysis, the culprit artery remained occluded in most patients before PCI. This is consistent with studies of thrombolysis that have reported a lower rate of spontaneous and thrombolytic agent–induced reperfusion in patients with shock,1,4,24,25 in part because of reduced coronary flow and pressure.4,25,26 The outcome of patients who underwent rescue angioplasty after thrombolysis was similar to those who underwent direct angioplasty. The 75% angiographic success rate in the SHOCK Registry is consistent with the SHOCK Trial results and with prior reports in the setting of cardiogenic shock4 but much lower than expected compared with other settings. Early reperfusion is more often successful with PCI than thrombolysis and possibly may be improved further with stenting, intra-aortic balloon pumping, and newer antiplatelet regimens.4,21,27,28 Patency of the infarct artery, whether assessed by angiographic stenosis, flow, or a combination of the two, was a major predictor of survival after PCI. Of concern, a worsening of perfusion grade after PCI was associated with a high mortality rate. Importantly, coronary flow failed to normalize in 18% of patients with a successful reduction in stenosis. Recent reports suggest an important role for microvascular obstruction and impaired myocardial perfusion in myocardial infarction.29 Measures designed to improve coronary flow and myocardial perfusion may have an important impact on outcome.

1. Webb J, Hochman J. Pathophysiology and management of cardiogenic shock due to primary pump failure. In: Ghersh B, Rahimtoola S, editors. Acute myocardial infarction. New York: Chapman & Hall; 1997 p. 308-27. 2. Hochman J, Buller C, Sleeper L, et al. Cardiogenic shock complicating acute myocardial infarction: etiologies, management and outcome: overall findings of the SHOCK Trial Registry. J Am Coll Cardiol 2000;36:1063-70. 3. Hochman J, Sleeper L, Godfrey E, et al. Should we emergently revascularize occluded coronaries for cardiogenic shock: an international randomized trial of emergency PTCA/CABG: trial design. Am Heart J 1999;137:313-21. 4. Webb J. Interventional management of cardiogenic shock. Can J Cardiol 1998;14:233-44. 5. Hochman J, Sleeper L, Webb J, et al, for the SHOCK Investigators. Early revascularization in acute myocardial infarction complicated by cardiogenic shock. N Engl J Med 1999;341:625-34. 6. TIMI Study Group. The Thombolysis in Myocardial Infarction (TIMI) trial: phase I findings. N Engl J Med 1995;312:932-6. 7. Hochman J, Sleeper L, White H, et al. Early revascularization improves one-year survival for cardiogenic shock complicating acute myocardial infarction: results of the randomized SHOCK Trial. JAMA. 2001;285:190-2. 8. Carnendran L, Abboud R, Gurunathan R, et al. Trends in cardiogenic shock: a report from the SHOCK study. Eur Heart J. 2001; 22:472-8. 9. Himbert D, Juliard J, Steg PG, et al. Limits of reperfusion therapy for immediate cardiogenic shock complicating acute myocardial infarction. Am J Cardiol 1994;74:492-4. 10. Hibbard MD, Holmes DR Jr, Bailey KR, et al. Percutaneous transluminal coronary angioplasty in patients with cardiogenic shock. J Am Coll Cardiol 1992;19:639-46. 11. Lee L, Bates ER, Pitt B, et al. Percutaneous transluminal coronary angioplasty improves survival in acute myocardial infarction complicated by cardiogenic shock. Circulation 1988;78:1345-51. 12. Lee L, Erbel R, Brown TM, et al. Multicenter registry of angioplasty therapy of cardiogenic shock: initial and long-term survival. J Am Coll Cardiol 1991;17:599-603. 13. Gacioch GM, Ellis SG, Lee L, et al. Cardiogenic shock complicating acute myocardial infarction: the use of coronary angioplasty and the integration of the new support devices into patients management. J Am Coll Cardiol 1992;19:647-53. 14. Brown EJ, Swinford RD, Gadde P, et al. Acute effects of delayed reperfusion on infarct shape and left ventricular volume: a potential

Study limitations Interventional equipment, techniques, and treatment continue to evolve. This registry largely predates the advent of widespread stenting8,21,30-32 and use of glycoprotein IIb/IIIa antagonists.28 Lack of independent angiographic confirmation of locally reported or assumed TIMI flow grade are limitations of this registry analysis. Patient treatment bias remains a limitation.

American Heart Journal June 2001

970 Webb et al

15.

16.

17.

18.

19.

20.

21. 22.

23.

mechanism of additional benefits from thrombolysis therapy. J Am Coll Cardiol 1991;17:1641-50. Disler L, Haitas B, Benjamin J, et al. Cardiogenic shock in evolving myocardial infarction: treatment by angioplasty and streptokinase. Heart Lung 1987;16:649-52. Verna E, Repetoo S, Boscarina M, et al. Emergency coronary angioplasty in patients with severe left ventricular dysfunction of cardiogenic shock after acute myocardial infarction. Eur Heart J 1989; 10:958-66. Yamamoto H, Hayashi Y, Oka Y, et al. Efficacy of percutaneous transluminal coronary angioplasty in patients with acute myocardial infarction complicated by cardiogenic shock. Jpn Circ J 1992;56: 815-21. Seydoux C, Goy JJ, Beuret P, et al. Effectiveness of percutaneous transluminal coronary angioplasty in cardiogenic shock during acute myocardial infarction. Am J Cardiol 1992;69:968-9. LATE Study Group. Late assessment of thrombolytic efficacy (LATE) study with alteplase 6-24 hours after onset of acute myocardial infarction. Lancet 1993;342:759-66. Ellis SG, O’Neill WW, Bates ER, et al. Implications for triage from survival and left ventricular functional recovery analyses in 1500 patients treated with coronary angioplasty for acute myocardial infarction. J Am Coll Cardiol 1989;13:1251-9. Webb JG, Carere R, Hilton JD, et al. Usefulness of coronary stenting in cardiogenic shock. Am J Cardiol 1997;79:81-4. Garcia-Cantu E, Corcos T, Favereau X, et al. Multiple stenting in acute myocardial infarction with double vessel occlusion, complicated with cardiogenic shock. J Invasive Cardiol 1995;7:283-7. Antoniucci D, Valenti R, Santoro G, et al. Systematic direct angioplasty and stent-supported direct angioplasty therapy for cardio-

24.

25.

26.

27.

28.

29.

30.

31. 32.

genic shock complicating acute myocardial infarction: in-hospital and long-term survival. J Am Coll Cardiol 1998;31:294-300. Holmes D, Bates E, Kleiman E, et al. Contemporary reperfusion therapy for cardiogenic shock: the GUSTO-I trial experience. J Am Coll Cardiol 1995;26:668-74. Bates E, Topol E. Limitations of thrombolytic therapy for acute myocardial infarction complicated by congestive heart failure and cardiogenic shock. J Am Coll Cardiol 1991;18:1077-84. Prewitt RM, Gu S, Schick U, et al. Intraaortic balloon counterpulsation enhances coronary thrombolysis induced by intravenous administration of a thrombolytic agent. J Am Coll Cardiol 1994;23:794-8. Grines C. Aggressive intervention for myocardial infarction: angioplasty, stents, and intra-aortic balloon pumping. Am J Cardiol 1996;78:29-34. Hasdai D, Harrington RA, Hochman JS, et al. Platelet glycoprotein IIb/IIIa blockade and outcome of cardiogenic shock complicating acute coronary syndromes without persistent ST-segment elevation. J Am Coll Cardiol 2000;36:685-92. Gibson C, Cannon C, Murphy S, et al. Relationship of TIMI myocardial perfusion grade to mortality after administration of thrombolytic drugs. Circulation 2000;101:125-30. Holmes D, Hirshfield JJ, Faxon D, et al. ACC expert consensus document on coronary artery stents. J Am Coll Cardiol 1998;32:147182. Ahmad T, Webb J, Carere R, et al. Stenting for acute myocardial infarction. Am J Cardiol 1995;76:77-80. Grines C, Cox D, Stone G, et al. Coronary angioplasty with or without stent implantation for acute myocardial infarction. N Engl J Med 1999;341:1949-56.