Trans-radial PCI. Is it safe? Is it feasible? Is it better, and to what extent ?
Daniel Obaid;Alexander Chase
British Heart Foundation Research Fellow. University of Cambridge Consultant Cardiologist. Morriston Cardiac Centre UK

Introduction
Although percutaneous coronary angioplasty has been performed via the trans-radial approach for over fifteen years1, uptake has remained limited. Its use is estimated at approximately 10% of PCI world wide, but the level in the US is much lower at around 1%2. This is mostly historical as the femoral route has been established for over two decades. However, it may also be due to the misconception that the radial route is associated with complications such as stroke or arterial access site complications. Another reason may be the radial approach is technically more challenging with a learning curve that is perceived to be associated with prolonged procedure time and increased procedure failure. This review will address these issues and discuss a growing collection of recent studies which suggest the trans-radial route is not only safe and feasible but also has proven advantages over the trans-femoral route.
Is it safe?
Cannulation of the radial artery is not a new technique. In this early study from 1983  prolonged cannulation of the radial artery occurred in 1699 patients in an intensive care setting. Despite there being temporary or persistant occlusion of the radial artery in up to 20% of cases, no significant ischaemic events were reported and there was no associated long term morbidity 3 . In more recent time when the radial artery has been the access site for angioplasty the rate of occlusion is reported at lower levels. In a meta analysis of Randomised Controlled Trials performed   by Agostoni et al (2004) the rate of radial artery occlusion detected by protocol driven-Doppler ultra sound was 3-6%. The rate of loss of radial pulse was 0-9% and none of these events were associated with clinical sequlae 4. Another study demonstrated that over half of radial arteries occluded after angioplasty have patency restored at one month. The rate of radial artery occlusion at 30 days was around 3% 5.
From the early days of trans-radial angioplasty it was noted from randomised controlled trial data that the rate of access site complications including haemorrhage, haematoma, arterio-venous fistula and pseudo aneurysm formation was much lower when compared to trans-femoral angioplasty 6. This results from the radial being smaller with a more superficial course than the femoral artery, leading to earlier recognition of bleeding and easier pressure haemostasis without a risk of concealed bleeding into the retroperitoneal space. Local bleeding can occur following arterial perforation (usually because of the guide wire exiting a small side branch) and forearm compartment syndrome is a very serious but extremely

rare complication of local haematoma. Calvino-Santos et al (2004)reported rates of perforation of around 1%, however they had no episodes of compartment syndrome or serious forearm haematoma despite continuing with the PCI with full doses of anti coagulation and anti platelet therapies 7. Dissection of the radial or brachial artery is also reported with variable frequency (1.5% by Hildick-Smith et al(1998)) 8 and its frequency may be increased by the use of hydrophilic wires. However as the dissection flap is contrary to the flow of arterial blood it is rarely clinically significant. A meta-analysis of 12 randomised controlled trials showed that overall entry site complications are encountered significantly less frequently when using radial access (0.3%) as compared to femoral access (2.8%) 4.There were initial fears that catheter and wire manipulation in proximity to the right internal carotid might increase the risk of peri procedural embolic stroke, fuelled by in vivo studies on cerebral micro emboli without clinical endpoints9. However these fears have been unsubstantiated in extensive clinical trials. A recent meta-analysis of 23 randomised controlled trials of coronary angiography and angioplasty performed between 1980 and 2008 in 7020 patients showed no significant difference in the rate of stroke between the radial and femoral route 10. In a comparison of higher stroke risk octogenerian patients there was no significant difference in the rate of stroke between trans-radial/ trans-femoral angioplasty11.
Another reason given to challenge adoption of the radial approach, not infrequently heard at leading intervention forums and by prominent physicians, is a perception of increased catheter induced coronary dissection. Even the most ardent radial enthusiasts will concede that many standard catheters, initially designed for use via the femoral artery, engage the coronaries with a less co-axial orientation12. Despite this, in an extensive search, we were unable to find any published evidence that radial access is associated with an increase in coronary dissection compared to femoral access. The truth behind this conflict likely lies somewhere in the middle. On one hand, experienced well trained and supervised radial operators who have contributed most to the radial verses femoral comparison literature have likely learned to negate this risk by careful adaptations of catheter selection and manipulation. On the other hand, for femoral trained operators using the radial route infrequently only by necessity or experimenting with changing access without the support of an experienced radialist colleague whilst on the learning curve, this threat may become reality and a powerful anecdotal dissuader from future pursuit of trans-radial PCI.

Is it feasible?

to accept a requirement to cross over to the femoral route in approximately one in twenty patients.  Eccleshall SC et al(2003) 15 demonstrated that it was feasible to implement a diagnostic and interventional trans-radial program, and with close supervision of trainees by an experienced operator achieved procedure and fluoroscopy time that were not significantly different between the radial and femoral route. We know from analysis of recent registry data (593,094 patients between 2004 and 2007) that radial operators now achieve procedural success rates that are not significantly different from those achieved  trans-femoraly  2. In the PREVAIL study there was no significant difference between radial or femoral access with respect to procedure duration, fluoroscopy time or contrast use 14. Trans-radial angioplasty was first safely implemented in an outpatient setting 16, but it has also been validated in the acute setting such as primary PCI in acute myocardial infarction 17, and in rescue angioplasty for failed thrombolysis 18. In the latter study trans-radial access was associated with reduced length of hospital stay and vascular complications when compared to trans-femoral angioplasty but with similar procedural success rates, procedure times and radiation doses 18.  Being used in a variety of settings the radial artery can be used for repeat procedures if required in the majority of cases. In a study of repeat radial artery cannulation in Japan, Sakai et al() found that it was possible to obtain repeat access in the same arm three to five times in most patients. A third trans-radial procedure was possible in 90% of men and 80% of women and a fifth procedure in 70% of men and 50% of women 19. Radial artery patency has probably been improved by advances in sheath and catheter technology, shorter compression time post procedure as well as the routine use of heparin in both diagnostic cases and angioplasty. A further interesting consideration and potential obstacle to wider adoption of the radial route is an impact upon potential arterial conduit patency for future coronary artery bypass grafting. In one small study of 22 patients using radial artery grafts after transradial catheterization, graft patency was reduced at one month angiographic follow-up without affecting clinical endpoints20. Further work on a much larger scale is required to more fully address this question.

A traditional limitation of  trans-radial has been the calibre of the radial artery. Historically there has been a tendency to assume that femoral route is required for more complex procedures due to the requirement of larger calibre access. In a study of 250 patients, Saito et al. 21 demonstrated that the radial artery diameter was smaller than the outer diameter of a 7 French Terumo (Terumo Co, Tokyo, Japan) introducer sheath in 28.5% of males and 59.7% of females.  However Mamas et al(2008) 22 have recently described the use of a sheathless guide catheter (Sheathless Eaucath, Asahi Intecc Co, Japan). These have an outer diameter of 2.49mm which is smaller than a conventional 6 French sheath (2.62mm). These catheters have an inner diameter which is equivalent to that of a 7.5 guide catheter. Using these guide catheters inserted directly into the radial artery without a sheath  it is feasible to perform complex interventions including crush stent bifurcations and rotablation entirely through the radial artery.
Is it better?
A prospective randomized control trial specifically designed to compare radial and femoral access adequately powered for mortality and morbidity is sadly lacking. The evidence base for the discussion that follows is therefore for the generation of hypothesis  rather than proof of concept. The trans-radial route has demonstrated significant advantages over the trans-femoral route. Because of the reduction in access site complications and facilitation of earlier ambulation, use of  trans-radial approach allows shorter hospital stays in both the elective 23 and acute settings 18. This translates to reduced costs, and the cost effectiveness of trans-radial access holds true even if the femoral puncture is closed with a closure device 24. The low rate of access site complications makes trans-radial angioplasty more suited to a day case setting. It has also been shown to be safe to perform angioplasty via the radial artery as a day case 25, in patients over the age of seventy five 26 and in those  traditionally considered  as high risk 27.  Not only does performing angioplasty as a day case have cost saving implications it has also been shown to be preferred by the majority of patients 28. In fact, radial access tends to score higher on both patient comfort and preference in comparison to femoral access 29.

REPLACE-2 34 where the rate of major bleeding was 3.2% in 6001 patients undergoing elective or urgent PCI, the 1 year cumulative mortality was 8.7% in the patients who  had major bleeds compared with 1.9% in those who did not have Patients in the Washington heart database 1991-2000 who had a major bleed had an in hospital mortality of 7.5% and one year mortality of 17.2% compared with 0.6% and 5.5% respectively in patients with no major bleeding 30.

non-bleeders or radial access and femoral access. The studies remain powerful however due to large patient numbers.
Analysing data from the national heart, lung, and blood institute (NHLB) dynamic registry Yatskar et al (2007)reported a degree of  significant  mortality in PCI patients who developed an access site haematoma requiring transfusion. Of the 6656 patients undergoing PCI, 1.8% developed an access site haematoma requiring a transfusion with an inpatient mortality of 9.3%, compared to 1.2% in those patients without access site haematoma requiring transfusion 42. In the MORTAL ( Mortality benefit of Reduced Transfusion After PCI via the Arm or Leg) retrospective study by Chase A, et al(2008) 43patients who had undergone PCI in the British Columbia registry and compared transfusion rates, 30 day mortality and 1 year mortality. Almost 39000 procedures were analysed and consistent with the numerous previous studies  there were twice as many transfusions (2.8%) in the femoral patients when compared with radial patients (1.4%). This remained significant after adjusting for all possible confounding variables that were recorded in the registry. Using propensity analysis,transfusion was found to be independently associated with a 5.7% increase in mortality. Using a multivariant regression model adjusting for all other risk factors captured in the registry, radial angioplasty  versus femoral was associated with a statistically significant reduction in 1 year mortality {odds ratio of 0.83 95% CI 0.71-0.98 , (p<0.001)}. Using a number needed to treat analysis calculated at prevention of 15 transfusion events to avoid one death, the authors postulated an impact of one death avoided per 1000 PCI if done radially rather than femorally. Other studies have also lent weight to the hypothesis that radial access might reduce ischaemic events. The RIVIERA study 44 looked at elective and primary angioplasty in 7962 consecutive patients at 144 hospitals. This mixed population was designed to identify treatment naïve patients and investigate variations in everyday practice that might impact outcome. Of all the variables recorded, only radial access was significantly associated with less bleeding. Also the use of radial access was independently associated with improved mortality along with the use of low molecular weight heparin and thienopyridine pre treatment, which  is mirrored in an analysis of the patients in the PRESTO ACS trial 45. A recent substudy of this trial looking at vascular access showed that radial access was associated with significantly less bleeding. This translated to a trend in improvement in the primary outcome of death or reinfarction at 30 days which was statistically significant at 1 year (4.9% in the radial group vs 8.3% in the femoral group; p = 0.05).

Intriguingly some recent data have raised the possibility that observed reduction in mortality using radial access may not just be down to reduced bleeding complications or transfusion alone. It remains possible that these findings simply reflect inadequate risk adjustment between radial and femoral cohorts in the absence of randomization but other explanations have also been proposed. In the PREVAIL 14 study 1052 patients undergoing PCI were recruited from 9 hospitals. In this study there were low levels of major bleeding in both the radial and femoral arms on intention to treat analysis, (although there was a decrease in bleeding in the radial patients if bleeds from crossovers to femoral access are included). Despite this there was a significant reduction in the secondary end point of death or MI in the radial group. The authors hypothesis that this may be due to increased glycoprotein  2b 3a inhibitor use (a class 1 indicated drug in unstable patients)46 in the radial access group. This can be called “the confidence hypothesis”. As the evidence for decreased bleeding complications in radial patients increases operators may be more willing to use glycoprotein 2b 3a inhibitors following radial puncture, or conversely be more reluctant to use them following femoral puncture. In the PREVAIL population the use of glycoprotein 2b 3a inhibitors was statistically higher with radial access (12%) compared to femoral access (8%) 14. This was also seen in PRESTO ACS (>50% usage radial verses 30% usage in femoral) 45. This higher use of glycoprotein 2b 3a inhibitors may go some way to explain the observed decrease in 1 year mortality given their previously proven survival benefit 47.  Another hypothesis proposed is radial access reduced renal injury, perhaps by less catheter related renal artery embolisation than the transfemoral approach although the supportive data are yet to be published.
            The vast majority of PCI’s performed world wide are trans-femoral. It may well be that it requires a prospective, randomised controlled trial powered specifically to look for the mortality benefit of radial versus femoral angioplasty to change practice. The CURRENT trial (Clopidogrel Optimal Loading Dose Usage to Reduce Recurrent Events) is recruiting and includes a vascular sub study of radial versus femoral access with 30 day outcome data as an endpoint. These data could prove less easy to dismiss if supportive of a trans-radial advantage.