Thrombolytic therapy provides a beneficial treatment option for patients without contraindications who present with acute myocardial infarction (AMI) within 12 hours of symptom onset when the capability for primary percutaneous coronary intervention (PCI) within either 90 or 120 minutes of presentation is not possible (European Society of Cardiology [ESC], American College of Cardiology[ACC]/American Heart Association [AHA] Class Ia recommendation) . Although there is obvious concern for increased bleeding, current guidelines from the ACC/AHA and ESC further support a strategy of early invasive management for both unstable (Class Ia) and stable (Class IIa) patients treated with thrombolytic therapy with timely transfer for catheterization/PCI . Patients undergoing PCI for ST-elevation myocardial infarction (STEMI) are at particularly high risk of bleeding, and as such, ESC guidelines recommend transradial arterial access (TRA) as the preferred arterial access for qualified operators (Class IA) . The clinical benefit of TRA after pharmaco-invasive therapy for STEMI has been less well studied . In this issue of Cardiovascular Revascularization Medicine ( CRM ), Graham et al. attempt to address the impact of arterial access site selection for patients treated with PCI for this indication . To get further insight, the authors pooled individual data from patients enrolled in trials designed to study the potential benefit of TRA in facilitated thrombolysis.

Bleeding complications are an obvious concern associated with thrombolytic therapy. It has been well established that significant bleeding following PCI is associated with adverse outcomes. Over a decade ago, pooled data from the Gusto IIb, PURSUIT, and PARAGON B studies (n=24,122) demonstrated a significant increased risk of 30-day mortality and death/myocardial infarction in patients receiving transfusion with ≥1 U of blood . Early studies of adjunctive antiplatelet and antithrombotic drug therapies have shown a consistent association between major bleeding and adverse outcomes. The CURE and OASIS studies revealed a significant risk of mortality at 30 days that persisted out to 1 year . Ndrepepa et al. analyzed pooled data from four ISAR studies and reported a hazard ratio of 2.96 (95% CI: 1.96–4.48; p<0.001) for 1-year mortality following a major Thrombolysis In Myocardial Infarction post-PCI bleed in patients . Similarly, patients in ACUITY who suffered a major bleed demonstrated higher 30-day mortality (7.3% vs. 1.2%; p<0.0001) with multivariate analysis revealing major bleeding as a strong independent predictor of 30-day mortality (OR: 7.55; 95% CI: 4.68–12.18; p<0.0001) . Using pooled analysis from Replace II, ACUITY, and HORIZONS AMI, Verheugt et al. reported that both access site related (HR: 1.82; 95% CI: 1.17–2.83) and non-access site related (HR: 3.94; 95% CI: 3.07–7.15) bleeding are correlated with 1-year mortality .

Recognition of the significant clinical impact for post-PCI bleeding has subsequently led to: 1) Determination of those patients at highest risk for bleeding; 2) Identification of the sources of bleeding; 3) Quantification of clinical impact according to bleeding source and magnitude; and 4) Implementation and quantification regarding the benefit of bleeding avoidance strategies (BAS), such as TRA.

There is ample evidence that PCI performed for the indication of acute coronary syndromes (ACS), especially STEMI, is associated with increased risk for post-procedure major bleeding. This may be related to aggressive anticoagulation and antiplatelet therapy as well as patient-related factors. In their pooled analysis of three randomized bivalirudin trials, Mehran et al. reported that raised biomarkers and STEMI were independent predictors of major bleeding . A risk model for post-PCI bleeding derived from the National Cardiovascular Data Registry (NCDR) database also includes both urgent procedure and STEMI as predictors . The Society for Coronary Angiography and Interventions (SCAI) includes both urgent indication and bleeding risk within their PCI risk assessment application .

Identification of the sources of post-PCI major bleeding has been performed to help understand the potential impact for different BAS. The source of bleeding is related to patient presentation with a higher proportion of non–access site bleeding (NASB) in patients with PCI for ACS compared to PCI performed for more stable patients . The incidence of access site bleeding in pooled data from REPLACE II/ACUITY/HORIZONS-AMI was 38.6% compared to 61.4% for NASB . The SYNERGY study data reported that 37% of bleeding events were related to the access site . In RIVAL, 70% of non-coronary artery bypass grafting major bleeds were unrelated to the access site, and in MATRIX, 62.9% of Bleeding Academic Research Consortium 3 or 5 bleeds originated from non–access sites . Transradial arterial access as a BAS employed for PCI in patients with ACS would therefore be expected to impact roughly 30–40% of bleeding events.

The clinical impact of bleeding source has also been examined. Propensity adjusted analysis of over 3 million patients enrolled in the NCDR between 2004 and 2011 demonstrated a significant increase in the absolute risk of mortality overall for post-PCI major bleeding (3.39%; p<0.001), and further, while a significant absolute increase in mortality associated with both access site bleeding and non-access site bleeding (0.89% vs. 6.39%; p<0.001) was seen, the number needed to harm was significantly lower for non–access site bleeding (16 vs. 117) . A large meta-analysis of 25 PCI studies (n=2,400,645) also demonstrated an increased risk of peri-procedural mortality with access site bleeding (RR=1.71; 95% CI: 1.37–2.13) and even higher risk with non–access site bleeding (RR=4.06; 95% CI: 3.21–5.14) . The higher incidence and poor outcomes associated with non–access site bleeding in ACS patients may be related to more aggressive anticoagulation or may simply be related to a higher prevalence of co-morbid disease.

Large registry studies and multiple meta-analyses have demonstrated benefit with TRA compared to transfemoral access (TFA) for PCI in patients with access site bleeding . More important, large randomized studies examining the impact of arterial access site selection for PCI in ACS patients confirm this benefit – most convincingly in the STEMI patient population. RIFLE-STEACS (n=1001) revealed benefit with TRA in regard to 30-day net adverse clinical event (NACE) as well as cardiac mortality (5.2% vs. 9.2%; p=0.02) . STEMI-RADIAL showed a benefit with TRA with a reduction in NACE (4.6% vs. 11%; p=0.0028) and major bleeding (1.4% vs. 7.2%; p<0.0001) . Recently the MATRIX trial (n=8404) revealed a significant decrease in NACE with TRA compared to TFA (9.8% vs. 11.7%; p=0.0092) . There is very little published data regarding the benefit of TRA for PCI following thrombolytic therapy; however, a recent report of propensity matched patients undergoing rescue PCI (n=9494) from the NCDR demonstrated a lower rate of bleeding with TRA (OR: 0.67; 95% CI: 0.52–0.87; p=0.003) but no difference in mortality .

The study by Graham et al. in this issue of CRM uses patient-level data accumulated from seven studies to compare TRA to TFA in patients undergoing early PCI following thrombolytic therapy for AMI. Although there was no significant benefit in regard to the primary endpoint, lower absolute rates of 30-day major bleeding and mortality were seen for the TRA group. The authors correctly point out that statistical power is a significant limitation of this study.

When interpreting these results, several other issues must be highlighted. Four of the seven studies were not randomized, hence TRA and TFA patient groups were not well matched. As these studies were originally designed to compare immediate vs. delayed PCI following thrombolytic therapy, the timing of PCI post thrombolysis was inconsistent. Heterogeneity also existed in the patient population with mixing of both early and later (standard care) PCI groups. Furthermore, bleeding definitions were not consistent between studies. Operator experience also appears to be an important contributor to the benefit of TRA. In RIVAL, significant benefit for reduction of the primary endpoint and vascular complications was seen in the centers with the highest TRA volume . In MATRIX, a pronounced benefit in regard to the co-primary endpoints was observed for TRA in centers where >80% of procedures were performed radially . Benefit for TRA was seen in RIFLE-STEACS, which required each operator to perform ≥75 TRA PCI procedures annually, and STEMI-RADIAL, which required operators to perform ≥80% of their PCI volume using TRA . Graham et al. included studies for which a relatively small percentage of procedures was performed radially (18%). Further, there is no information regarding operator experience.

In summary, with the benefit of lower access site bleeding, the use of TRA is a sensible option for high-risk patients, including those with AMI treated with facilitated PCI. Even though the study by Graham et al. in this issue of CRM had several limitations in study design that precluded demonstration of significant superiority for TRA, there was non-significant reduction of major bleeding and mortality. This suggests that further investigation to more definitively establish the benefit of TRA for this application is warranted.