BASES TEÓRICAS

by guest on April 2, 2018

http://circ.ahajournals.org/

ed in-hospital mortality rate less than 7% in patients without cardiogenic shock. This would result in a risk-adjusted mor- tality rate with PCI comparable to that reported for fibri- nolytic therapy in fibrinolytic-eligible patients (40) and would be consistent with previously reported registry experi- ence (422-425). Otherwise, the focus of treatment should be the early use of fibrinolytic therapy (Figure 14) (40).

PCI appears to have its greatest mortality benefit in high- risk patients. In patients with cardiogenic shock, an absolute 9% reduction in 30-day mortality with coronary revascular- ization instead of immediate medical stabilization was reported in the SHOCK trial (301). In NRMI-II, patients with CHF had a 33% relative risk reduction with primary PCI compared with a 9% relative risk reduction with fibrinolytic therapy (302). Primary PCI in patients with anterior STEMI reduces mortality compared with fibrinolytic therapy, but there is no difference in patients with nonanterior STEMI (426,427).

Time from symptom onset to reperfusion is an important predictor of patient outcome. Two studies (294,295) have reported increasing mortality rates with increasing door-to- balloon times. Other studies have shown smaller infarct size, better LV function, and fewer complications when reperfu- sion occurs before PCI (345,346,428). An analysis of the ran- domized controlled trials that compared fibrinolysis with a fibrin-specific agent versus primary PCI suggests that the mortality benefit with PCI exists when treatment is delayed by no more than 60 minutes (Figure 15) (305). Mortality increases significantly with each 15-minute delay in the time between arrival and restoration of TIMI 3 flow (door- to–TIMI 3 flow time), further underscoring the importance of timely reperfusion in patients who undergo primary PCI (429). Importantly, after adjustment for baseline characteris- tics, time from symptom onset to balloon inflation is signifi- cantly correlated with 1-year mortality in patients undergo- ing primary PCI for STEMI (RR equals 1.08 for each 30- minute delay from symptom onset to balloon inflation, p equals 0.04) (Figure 21) (275). Given that the door-to-needle time goal is 30 minutes, this Writing Committee joins the Task Force on the Management of Acute Myocardial Infarction of the European Society of Cardiology in lowering the recommended medical contact–to-balloon or door-to-bal- loon time goal from 120 to 90 minutes in an attempt to max- imize the benefits for reperfusion by PCI (297) (Figure 22) (294).

If the expected door-to-balloon time exceeds the expected door-to-needle time by more than 60 minutes, fibrinolytic treatment with a fibrin-specific agent should be considered unless it is contraindicated. This is particularly important when symptom duration is less than 3 hours but is less important with longer symptom duration, when less ischemic myocardium can be salvaged. In both the CAPTIM trial (173), which showed lower mortality with prehospital fibri- nolysis than with primary PCI, and the PRAGUE-2 trial (177), which showed lower mortality with primary PCI after interhospital transfer than with on-site fibrinolysis, PCI was superior to fibrinolysis when symptom duration was greater STEMI if they are hemodynamically and electrically

stable. (Level of Evidence: C)

Primary PCI has been compared with fibrinolytic therapy in 22 randomized clinical trials (173,177,306,307,404-421). An additional trial, SHOCK (301), which compared medical stabilization with immediate revascularization for cardio- genic shock, was included along with the above 22 trials in an overview of primary PCI versus fibrinolysis (40). These investigations demonstrate that PCI-treated patients experi- ence lower short-term mortality rates (5.0% versus 7.0%, RR 0.70, 95% CI 0.58 to 0.85, p equals 0.0002), less nonfatal reinfarction (3.0% versus 7.0%, RR 0.35, 95% CI 0.27 to 0.45, p equals 0.0003), and less hemorrhagic stroke (0.05% versus 1.0%, RR 0.05, 95% CI 0.006 to 0.35, p equals 0.0001) than those treated by fibrinolysis but with an increased risk for major bleeding (7.0% versus 5.0%, RR 1.3, CI 1.02 to 1.65, p equals 0.032) (40). These results were achieved in medical centers with experienced providers and under circumstances in which PCI could be performed promptly after patient presentation (Figure 14) (40).

Additional considerations that affect the magnitude of the difference between PCI- and fibrinolysis-treated patients include the fact that UFH was used as the antithrombin with fibrinolytics as opposed to other antithrombins such as enoxaparin (see Section 6.3.1.6.8.1.1) or bivalirudin (see Section 6.3.1.6.8.1.2) that are associated with a reduction in the rate of recurrent MI after fibrinolysis; a smaller but sta- tistically significant advantage for PCI compared with a fib- rin-specific fibrinolytic versus streptokinase; and variation among the PCI arms as to whether a stent was implanted or GP IIb/IIIa antagonists were administered. Figure 14 shows the short-term and long-term outcomes of patients with STEMI treated by fibrinolysis versus PCI and the number of patients who need to be treated to prevent 1 event or cause 1 harmful complication when selecting PCI instead of fibrinol- ysis as the reperfusion strategy (Figure 14) (40). When pri- mary PCI is compared with tPA and the SHOCK trial is excluded, the mortality rate is 5.5% versus 6.7% (OR 0.81%, 95% CI 0.64 to 1.03, p equals 0.081) (421a).

There is serious and legitimate concern that a routine poli- cy of primary PCI for patients with STEMI will result in unacceptable delays in achieving reperfusion in a substantial number of cases and produce less than optimal outcomes if performed by less-experienced operators. The mean time delay for PCI instead of fibrinolysis in the randomized stud- ies was approximately 40 minutes (40). Strict performance criteria must be mandated for primary PCI programs so that long door-to-balloon times and performance by low-volume or poor-outcome operators/laboratories do not occur. Interventional cardiologists and centers should strive for out- comes to include 1) medical contact–to-balloon or door-to- balloon times less than 90 minutes, 2) TIMI 2/3 flow rates obtained in more than 90% of patients, 3) emergency CABG rate less than 2% among all patients undergoing the proce- dure, 4) actual performance of PCI in a high percentage of patients (85%) brought to the laboratory, and 5) risk-adjust-

by guest on April 2, 2018

http://circ.ahajournals.org/

adverse reactions to volume loading, contrast medium, and antithrombotic medications; technical complications; and reperfusion events. Reocclusion occurs in 10% to 15% of patients after PTCA but in fewer than 5% after stent implan- tation. Likewise, angiographic restenosis occurs in 30% to 40% of patients after PTCA but in 15% to 20% after stent implantation. The management of these complications is beyond the scope of this guideline (430-432).

than 2 to 3 hours but not when symptom duration was short- er (see Section 6.3.1.6.2.1). In the early hours of STEMI, prompt fibrinolytic therapy can decrease infarct size and the risk of developing cardiogenic shock (176).

6.3.1.6.4.2.1. Complications of primary PCI.

Potential complications of an invasive strategy for treating STEMI include problems with the arterial access site;

Figure 21.Symptom onset-balloon time and mortality in primary PCI for ST-elevation myocardial infarction. RCT = randomized con- trolled trial; PCI = percutaneous coronary intervention. The relationship between time-to-treatment and 1-year mortality, as continuous functions, was assessed using a quadratic regression model. The dotted lines represent 95% confidence intervals of the predicted mortality. Modified with permission from De Luca. Circulation 2004;109:1223-25 (275).

Figure 22.Relationship between 30-day mortality and time from study enrollment to first balloon inflation. Patients assigned to angio- plasty in whom angioplasty was not performed are also shown. PTCA = percutaneous transluminal coronary angioplasty. Reprinted with permission from Berger et al. Circulation 1999;100:14-20 (294).

by guest on April 2, 2018

http://circ.ahajournals.org/

6.3.1.6.4.2.2. Primary PCI in

fibrinolytic-ineligible patients.

Class I

Primary PCI should be performed in fibrinolytic- ineligible patients who present with STEMI within 12

hours of symptom onset. (Level of Evidence: C)

Class IIa

It is reasonable to perform primary PCI for fibri- nolytic-ineligible patients with onset of symptoms within the prior 12 to 24 hours and 1 or more of the following:

a. Severe CHF(Level of Evidence: C)

b. Hemodynamic or electrical instability (Level of

Evidence: C)

c. Persistent ischemic symptoms. (Level of Evidence:

C)

Randomized controlled trials evaluating the outcome of PCI for patients who present with STEMI but who are ineli- gible for fibrinolytic therapy have not been performed. Few data are available to characterize the value of primary PCI for this subset of patients with STEMI; however, the recom- mendations in Section 4.2 are applicable to these patients. Nevertheless, these patients are at increased risk for mortali- ty (433), and there is a general consensus that PCI is an appropriate means for achieving reperfusion in those who cannot receive fibrinolytics because of increased risk of bleeding (403,434-436).

6.3.1.6.4.2.3. Primary PCI without

on-site cardiac surgery.

Class IIb

Primary PCI might be considered in hospitals without on-site cardiac surgery, provided that a proven plan for rapid transport to a cardiac surgery operating room exists in a nearby hospital with appropriate hemodynamic support capability for transfer. The procedure should be limited to patients with STEMI or MI with new, or presumably new, LBBB on ECG, and should be done in a timely fashion (balloon infla- tion within 90 minutes of presentation) by persons skilled in the procedure (at least 75 PCIs per year) and at hospitals performing a minimum of 36 primary

PCI procedures per year. (Level of Evidence: B)

Class III

Primary PCI should not be performed in hospitals without on-site cardiac surgery capabilities and with- out a proven plan for rapid transport to a cardiac sur- gery operating room in a nearby hospital or without appropriate hemodynamic support capability for

transfer. (Level of Evidence: C)

Reports on emergency primary PCI from hospitals without established open heart surgery or elective PCI programs have demonstrated generally favorable results (307,437-450). PCI

in the early phase of an acute STEMI can be difficult and requires even more skill and experience than routine PCI in the stable patient. The need for an experienced operator and experienced laboratory technical support with availability of the broad range of catheters, guidewires, stents, and other devices (e.g., intra-aortic balloon pump [IABP]) required for optimum results in an acutely ill patient is of major impor- tance. Careful patient selection and continuous quality

Table 16. Criteria for Performance of Primary PCI at Hospitals Without Onsite Cardiac Surgery

The operators must be experienced interventionalists who regularly perform elective PCI at a surgical center (at least 75 cases/year). The catheterization laboratory must perform a minimum of 36 pri- mary PCI procedures per year.

The nursing and technical catheterization laboratory staff must be experienced in handling acutely ill patients and must be comfort- able with interventional equipment. They must have acquired expe- rience in dedicated interventional laboratories at a surgical center. They participate in a 24-hour, 365-day call schedule.

The catheterization laboratory itself must be well equipped, with optimal imaging systems, resuscitative equipment, and IABP sup- port, and must be well stocked with a broad array of interventional equipment.

The cardiac care unit nurses must be adept in hemodynamic monitor- ing and IABP management.

The hospital administration must fully support the program and enable the fulfillment of the above institutional requirements. There must be formalized written protocols in place for immediate

and efficient transfer of patients to the nearest cardiac surgical facility that are reviewed/tested on a regular (quarterly) basis. Primary PCI must be performed routinely as the treatment of choice

around the clock for a large proportion of patients with STEMI, to ensure streamlined care paths and increased case volumes. Case selection for the performance of primary PCI must be rigorous.

Criteria for the types of lesions appropriate for primary PCI and for the selection for transfer for emergent aortocoronary bypass sur- gery are shown in Table 17.

There must be an ongoing program of outcomes analysis and formal- ized periodic case review.

Institutions should participate in a 3- to 6-month-period of imple- mentation, during which time development of a formalized primary PCI program is instituted that includes establishment of standards, training of staff, detailed logistic development, and creation of a quality assessment and error management system.

PCI = percutaneous coronary intervention; IABP = intra-aortic balloon pump; STEMI = ST-elevation myocardial infarction.

Modified from Wharton et al. J Am Coll Cardiol 1999;33:1257-65 (445).

by guest on April 2, 2018

http://circ.ahajournals.org/

then the gains of early intervention may be lost because of increased complications. In such circumstances, transfer to a center that routinely performs complex PCI will often be a more effective and efficient course of action. Fibrinolysis is an acceptable form of therapy and is preferable to primary PCI by an inexperienced team.

Criteria have been suggested for the performance of pri- mary PCI at hospitals without on-site cardiac surgery (432,445) (Tables 16 and 17). Large-scale registries have shown an inverse relationship between the number of pri- mary PCI procedures performed and in-hospital mortality (295,303,304). The data suggest that both door-to-balloon time and in-hospital mortality are significantly lower in insti- tutions that perform a minimum of 36 primary PCI proce- dures per year (295). Suboptimal results may relate to oper- ator/staff inexperience and capabilities and to delays in per- forming PCI for logistical reasons. From clinical data and expert consensus, the Committee recommends that primary PCI for acute STEMI performed at hospitals without estab- lished elective PCI programs should be restricted to those institutions capable of performing a requisite minimum num- ber of primary PCI procedures (36 per year) with a proven plan for rapid and effective PCI and rapid access to cardiac surgery in a nearby hospital. The benefit of primary PCI is not well established for operators who perform fewer than 75

Table 17. Patient Selection for Primary PCI and Emergency Aorto- Coronary Bypass at Hospitals Without Onsite Cardiac Surgery (445) Avoid intervention in hemodynamically stable patients with:

Significant (greater than or equal to 60%) stenosis of an unpro- tected left main coronary artery upstream from an acute occlu- sion in the left coronary system that might be disrupted by the angioplasty catheter

Extremely long or angulated infarct-related lesions with TIMI grade 3 flow

Infarct-related lesions with TIMI grade 3 flow in stable patients with 3-vessel disease (432)

Infarct-related lesions of small or secondary vessels

Hemodynamically significant lesions in other than the infarct artery

Transfer for emergency aortocoronary bypass surgery patients: After primary PCI of occluded vessels if high-grade residual left

main or multivessel coronary disease with clinical or hemo- dynamic instability present

Preferably with IABP support

PCI = percutaneous coronary intervention; TIMI = Thrombolysis in Myocardial Infarction; IABP = intra-aortic balloon pump.

Figure 23.Comparison of elapsed time to fibrinolysis versus primary PCI. Time is presented as a continuous variable in minutes on the horizontal axis. For DANAMI-2, times reflect components of delay from symptom onset to randomization (vertical bar) and are fur- ther separated according to whether patients presented at community referral hospitals versus those equipped for primary PCI. For those patients randomized to PCI at a referral hospital, the 3 components of delay after randomization are related to duration of stay at referral hospital, time for transport to PCI hospital and delay from arrival at PCI hospital to balloon inflation. PCI = percutaneous coronary intervention; Rand = Randomization; SK = streptokinase; Transp = Transportation. PRAGUE-2 data modified from Widimsky et al. Eur Heart J 2003; 24:94-104.

improvement are critical components of a successful pro- gram. If these complex patients are treated by interventional- ists with limited experience at hospitals with low volume,

by guest on April 2, 2018

http://circ.ahajournals.org/

Mortality was reduced with PCI (6.8% versus 9.6%, RR 0.69, 95% CI 0.51 to 0.0.92, p equals 0.01), as was the com- bined end point of death, nonfatal reinfarction, and stroke (8.5% versus 15.5%, RR 0.51, 95% CI 0.39 to 0.65, p less than 0.0001). Importantly, mean time to treatment was delayed only 44 minutes in these studies (Figure 23) (177,306). In contrast, the time from presentation at the door of the first hospital to balloon inflation in the second hospi- tal, as recorded in 1346 patients in NRMI-4, was 185 minutes in the United States in 2002 (Figure 24) (458). Emergency transport in Europe is centrally organized and more efficient than in the United States (Table 18) (177,306,408,419,421,459) (Van de Werf; oral presentation, American College of Cardiology 52nd Annual Scientific Session, Chicago, IL, March 2003). Delays in door-to-bal- loon time versus door-to-needle time of more than 60 min- utes because of interhospital transfer might negate the poten- tial mortality benefit of transfer for primary PCI over imme- diate intravenous fibrinolysis with a fibrin-specific agent as shown in these trials (305). To achieve optimal results, time from the first hospital door to the balloon inflation in the sec- PCIs per year or in a hospital that performs fewer than 36 pri-

mary PCI procedures per year. In addition, the benefit of timely reperfusion of the infarct artery by primary PCI at sites without on-site surgery must be weighed against the small but finite risk of harm to the patient related to the time required to transfer the patient to a site with CABG surgery capabilities (452,453).

6.3.1.6.4.2.4. Interhospital transfer for primary PCI.

The enthusiasm for primary PCI has led to the concept of emergency interhospital transfer for catheter-based reperfu- sion rather than fibrinolytic therapy in the initial hospital (454-456). Complication rates are low during transport, but time to reperfusion is delayed, which results in larger infarct size and lower LVEF (457). However, as noted in Section 6.3.1.6.2.1, selection bias of patients enrolled in randomized trials likely resulted in an underestimation of the risk of inter- hospital transfer expected in routine practice. Five random- ized trials enrolled 2466 patients, with favorable results for PCI versus fibrinolytic therapy (177,306,408,419,421).

Figure 24.Door-to-balloon times: patients transferred in NRMI 4. Data are expressed in minutes as median time (25th percentile-75th percentile) Cath Lab = catheterization laboratory. Modified from NRMI-4 Investigators: The National Registry of Myocardial Infarction- 4 Quarterly Report. Genentech, South San Francisco, CA; March, 2003:2 (458).

Table 18. Transport of Patients With STEMI for Primary PCI

Time Between Randomization Study (Reference) No. Transported Distance, km and First Balloon Inflation, min

Vermeer et al. (419) 75 25-50 85*

PRAGUE-1 (408) 101 5-74 80*

AIR-PAMI (421) 71 51 plus or minus 58* 155†

PRAGUE-2 (177) 429 5-120 97* DANAMI-2 (306) 559 3-150 90† Total 1235 3-150 *Mean. †Median. by guest on April 2, 2018 http://circ.ahajournals.org/ Downloaded from

Preliminary reports suggest that compared with conven- tional bare metal stents, drug-eluting stents are not associat- ed with increased risk when used for primary PCI in patients with STEMI (468). Postprocedure vessel patency, biomarker release, and the incidence of short-term adverse events were similar in patients receiving sirolimus (n equals 186) or bare metal (n equals 183) stents. Thirty-day event rates of death, reinfarction, or revascularization were 7.5% versus 10.4%, respectively (p equals 0.4) (468).

6.3.1.6.4.4. Facilitated PCI