The rapid restoration of coronary blood flow via pharmacologic or mechanical recanalization of an occluded infarct-related artery limits the extent of myocardial necrosis and reduces subsequent morbidity and mortality. Both randomized controlled clinical trials and observational studies have demonstrated more effective, complete and durable coronary reperfusion following primary PCI compared with fibrinolysis, and thus, primary PCI has been embraced as the preferred and dominant strategy in clinical practice guidelines.
However, the relative advantage of primary PCI (vs. fibrinolysis) is influenced by the risk profile of the patient; the relative time delay for implementing therapy (primary PCI usually entails an obligate additional time delay); and the case volume experiences of both the operator (cardiologist) and hospital facility. The best clinical outcomes, and hence the greatest relative advantage of primary PCI, is obtained in higher-risk strata patients and with expedient primary PCI implementation by high-volume operators in high-volume institutions. As most hospital facilities in the United States do not provide primary PCI, efforts to treat more patients efficiently and effectively with primary PCI must focus on either the transport of STEMI patients to regional STEMI centers or the establishment of primary PCI capabilities in more remote community hospitals.
C-PORT E Outcomes
In this context, the C-PORT E trial was designed as a noninferiority comparison of clinical outcomes following PCI at centers with surgical backup (SOS) compared with PCI at centers without surgical backup (no SOS). The two primary endpoints of the trial were all-cause mortality at 6 weeks and MACE — a composite of target vessel revascularization, Q-wave MI and death — at 9 months.
Ernest L. Mazzaferri Jr.
Dean J. Kereiakes
Importantly, the fact that PCI in stable/elective CAD has never demonstrated a survival difference in comparison to optimal medical therapy alone raises concerns regarding the choice of a survival endpoint in this trial. Thus, the trial’s main finding that there was no significant difference in all-cause mortality to 6 weeks following PCI performed in centers with no SOS compared with centers with SOS among a very low clinical/angiographic risk cohort of patients is not surprising and might have been predicted. Nevertheless, this observation contributed to a recent update of the ACC/AHA/SCAI guidelines that give a Class IIB recommendation (may be considered; usefulness/efficacy is less well established by evidence/opinion) to elective PCI at no-SOS centers and a Class IIA recommendation (may be reasonable; weight of evidence/opinion is in favor of usefulness/efficacy) for primary PCI (STEMI) at these same centers.
A more global perspective of the Cardiovascular Patient Outcomes Research Team Non-Primary PCI (C-PORT E) trial identifies additional issues that must be considered before attempting to integrate the results into contemporary clinical practice:
PCI volume credentialing: Importantly, several procedural volume thresholds were required by all centers who participated in the trial:
- More than 200 PCIs per year — recently established programs were allowed to have performed more than 100 per year in their first year.
- Physician operators were required to perform at least 75 cases per year.
- Centers had to provide primary PCI “24/7,” 365 days a year.
In addition to procedural volume requirements, participating C-PORT E centers were required to implement a formal, robust angioplasty development program, which generally took 3 to 6 months to complete. Thus, in evaluating current and potential C-PORT hospitals in the United States, we would strongly encourage strict adherence to the rigorous standards imposed by the C-PORT E trial. Nevertheless, even among participating C-PORT centers, in-hospital mortality following primary PCI for STEMI has been demonstrated to be inversely correlated with institutional procedural volume. The hospitals in the lowest tertile (median volume, 45 PCIs/year) of primary PCI procedural volume had the highest in-hospital mortality rate, whereas hospitals in the highest volume tertile (median, 83 PCIs/year) had the lowest rates of mortality.
Patient selection: High-risk patients were excluded from participation in the trial. Indeed, of the 75,674 participants who consented for the trial, 56,807 (75%) were not randomized for multiple clinical and angiographic reasons. The most frequent reasons for exclusion included “high risk coronary anatomy” (23%); calcification (20%); tortuous anatomy (12%); bifurcation lesions (9%); left main coronary disease (9%) or a concomitant comorbidity (9%); and, finally, physician/operator discretion or “other” high-risk perception (18%). Furthermore, of the 18,568 randomly assigned participants, 371 (2%) either had no PCI performed or were not assigned to either SOS or no-SOS centers.
Thus, the final study cohort for analysis appears to be a very low-risk (both clinically and angiographically) population that was randomly assigned to PCI at either SOS or no-SOS facilities. Therefore, the results of C-PORT E that have been obtained in a low-risk patient cohort cannot be generalized or extended to apply to higher-risk strata patients. Although it is likely that PCI will continue in no-SOS facilities, it is imperative that the patient cohorts be similar to that which was enrolled into the C-PORT E trial (low clinical and angiographic risk), and adherence to this standard must be monitored and periodically audited to assure maintenance of the C-PORT E demographic/angiographic inclusion and exclusion criteria.
Procedural failure: Despite a low-risk clinical and angiographic profile of participants randomized in the C-PORT E trial, PCI procedural failure was significantly increased among those participants who had PCI at a no-SOS center (3.4% vs. 2.5% patient success; P=.0096). In addition, no-SOS sites were more likely to place bare-metal (vs. drug-eluting) stents, more likely to have additional unplanned catheterizations (4.41% vs. 3.35%; P=.002) and unplanned PCIs (2.11% vs. 1.32%; P=.001), and had trends toward more frequent bleeding events (3.41% vs. 3%; P=.18) and strokes (0.27% vs. 0.15%; P=.16). Despite the enrollment of a low clinical and angiographic risk strata cohort, PCI complications trended higher in no-SOS hospitals.
In this context, if the practice of PCI at no-SOS centers is continued, participating C-PORT hospitals should be required to continue to report data to the C-PORT registry with quarterly review of procedural and clinical outcomes by the C-PORT data and safety monitoring committee and potentially by individual state Departments of Health and Human Services. Such systematic monitoring is required to assure the safety of patients undergoing PCI in no-SOS facilities and to ascertain that the patients undergoing PCI remain in a low-risk strata.
Improving access: One of the motivations for performance of the C-PORT E trial was “to improve access to PCI services.” This motivation appears curious in the context of a documented progressive decline in national (US) PCI procedural volume that has occurred over the past 8 years. It was previously demonstrated that roughly 79% of the adult US population in 2000 lived within 60-minute ground transportation of a PCI facility. Nevertheless, between 2000 and 2006, 521 new PCI programs were introduced nationally (44% increase in PCI capacity), many of which were no-SOS C-PORT facilities. Although PCI capacity increased dramatically, PCI access (as defined to be within 60-minute ground transportation) increased by only 1% (to 80%). Importantly, the majority of these new, no-SOS PCI facilities were opened within 10 to 12 miles of an existing PCI center (with SOS). Furthermore, in Ohio, although 89% of the population lives within 30 minutes ground transportation of a PCI hospital with SOS, multiple C-PORT no-SOS PCI hospitals have been developed within close proximity to existing PCI facilities.
Thus, despite a decline in national PCI volumes, the proliferation of PCI centers (with no SOS) in close proximity to existing PCI facilities geographically exaggerates the relative decline in institutional PCI volumes with the potential to negatively impact clinical outcomes such as mortality (which has been inversely correlated with PCI volume). In addition, when the divergent strategies of PCI center proliferation (build more centers) and emergency medical services (EMS) integration using pre-hospital STEMI diagnosis and triage (utilize PCI centers more efficiently) have been modeled in comparison using the city of Dallas, EMS integration was the dominant cost-effective strategy. Therefore, should PCI in no-SOS centers continue, we would recommend that the distance from existing PCI facilities be seriously considered before allowing further PCI center proliferation.
Proliferation of PCI centers with no SOS has occurred with a consequent increase in PCI capacity by approximately 44%, despite the decline in PCI volumes noted nationally. Geographic disparity in the location of new PCI centers as suggested by the 1% increase in PCI access provided by the increased number of PCI centers may exaggerate the impact of reduction in regional institutional PCI volumes with the potential to negatively impact quality outcome measures such as mortality. Although PCI center proliferation has been accomplished under the auspice of “improving access,” the majority of new PCI centers have been in close proximity (within 12 miles) to existing PCI facilities, which suggests that a more likely motivation has been increasing market share rather than access. Recent data suggest that statewide programs focused on regional systems for reperfusion for STEMI via EMS integration with existing PCI centers utilizing pre-hospital diagnosis with early activation and triage is a more cost-effective strategy and achieves better clinical outcomes than the strategy of developing additional PCI centers.
Aversano T. Circulation. 2008:118:S-1076.
Aversano T. N Engl J Med. 2012;366:1792-1802.
Boden WE. N Engl J Med. 2007;356:1503-1516.
Buckley JW. Am Heart J. 2008;155:668-672.
Concannon TW. Circ Cardiovasc Qual Outcomes. 2010;3:506-513.
Concannon TW. Circ Cardvasc Qual Outcomes. 2012; 5:14-20.
Jollis JG. JAMA. 2007;298:2371-2380.
Katritsis DG. Circulation. 2005;111:2906-2912.
Levine GN. Circulation. 2011;124:e574-e651.
Nallamothu BK. Circulation. 2006;113:1189-1195.
Riley RF. Circ Cardiovasc Qual Outcomes. 2011;193-197.
Srinivas VS. J Am Coll Cardiol. 2009;53:574-579.
Ernest L. Mazzaferri Jr., MD, FACC, is the medical director at the Ross Heart Hospital at The Ohio State University Medical Center, Columbus; Dean J. Kereiakes, MD, FACC, FSCAI, is the medical director at The Christ Hospital Heart and Vascular Center/The Lindner Research Center, Cincinnati, and is a member of the Cardiology Today Intervention Editorial Board.
Disclosure: Dr. Kereiakes has received grant support from Abbott Vascular, Boston Scientific and Medtronic; Dr. Mazzaferri reports no relevant financial disclosures.