Ulnar Artery Interventions Non-Inferior to Radial Approach: AJmer Ulnar ARtery (AJULAR) Intervention Working Group Study Results
Abstract: Background. Percutaneous coronary procedures are undergoing a paradigm shift from femoral to forearm approach due to obvious advantages in terms of patient safety and comfort, and faster time to ambulation. Transradial access (TRA) has been established as a primary forearm access site. We have tried to use ulnar artery access as an alternative to radial route and found that transulnar access (TUA) is an excellent alternative and non-inferior to TRA if performed by an experienced operator. Methods. This was a prospective, single-center study involving 2532 patients who were randomized in a 1:1 manner to either TUA (n = 1270) or TRA (n = 1262). All cannulations were performed by operators who were experienced in radial artery (RA) cannulation and had performed a minimum of 50 ulnar artery (UA) cannulations. The primary endpoint was a composite of major adverse cardiac events during hospital stay, crossover to another arterial access route, major vascular events during hospital stay (large hematoma with hemoglobin drop of ≥3 g%) or vessel occlusion rate. To prove non-inferiority of TUA, a margin of 1.93 was derived by fixed-margin method (preserving 50% of difference of procedural failure rate [4.87%] between radial and femoral access from meta-analysis). Results. The composite primary endpoint occurred in 14.6% of TUA and 14.4% of TRA patients (risk ratio, 1.01; 95% confidence interval, 0.83-1.2; P=.92 at α=0.05). All assessed parameters (except large hematoma, for which non-inferiority could not be proved conclusively) were non-inferior in TUA when compared with TRA. Conclusions. TUA is non-inferior to TRA when performed by an experienced operator. The utilization of TUA as an access site option increases the chance of success with forearm access and reduces the need for crossover to femoral route.
J INVASIVE CARDIOL 2016;28(1):1-8
Key words: transradial access, ulnar artery, transulnar access
_______________________________________________
Forearm access for percutaneous coronary procedures has improved patient safety and comfort, as well as economic viability when compared with transfemoral access.1,2 Worldwide, transradial access (TRA) is the primary forearm access site, but the radial artery (RA) has its own shortcomings, including propensity for vasospasm, smaller caliber, and frequent anatomical variations, that result in failure rates as high as 11%.3,4 In addition, previous puncture and catheter manipulation of RA is associated with more intimal hyperplasia and reduced early graft patency.5 Transulnar access (TUA) may be a viable, alternative forearm access for percutaneous coronary procedures. Previous studies have attempted to compare complications with ulnar artery (UA) cannulation by operators naive to TUA against those with RA cannulation by operators with experience in TRA.6 Since the UA courses beneath the forearm tendons and is poorly palpable in comparison with the RA, cannulation of UA has its own learning curve. We believe that if a valid comparison between complications with the second access route has to be made, the operators have to attain a minimal level of experience in UA cannulation. Since operator experience has not been addressed in the past, we planned to conduct this study with the hypothesis that the use of TUA by an experienced operator is non-inferior to TRA.
Methods
This was a prospective, randomized, single-center, parallel-design study. Written informed consent was obtained from all patients. Analyses of initial data revealed the need of cannulating at least 50 ulnar arteries before events (large hematoma, spasm, crossover, and puncture times) could validly be compared with those that happened while cannulating RA by the same operator. Procedures in the current study were performed, therefore, by six operators who were default radialists (cannulation experience of >150 radial arteries/year) with experience cannulating at least 50 UAs. All operators involved in the study had at least level 2 competency (signifying an operator’s ability to perform all diagnostic and simple interventional procedures) as per Society for Cardiac Angiography and Interventions (SCAI) guidelines for forearm operators.1
Patient selection. A total of 2700 patients (Figure 1) requiring coronary angiography were screened for the current study. Percutaneous transluminal coronary angioplasty was performed as needed, per operator discretion. The preprocedural exclusion criteria were: inability to palpate one or both forearm vessels (radial or ulnar artery of either side); cardiogenic shock; chronic hemodialysis; vasospastic disease (Raynaud’s disease); severe skeletal forearm deformities; or history of prior coronary artery bypass graft surgery. Inability to cannulate the allotted artery within the first three attempts was considered a failure for the purpose of analysis; however, the decision to cross over to another access site was left to the operator’s discretion irrespective of number of attempts. The sequence to be followed while crossing over was predefined (Figure 2). Allen’s test was not performed routinely and it was assumed that palpable radial and ulnar pulses were normal before considering forearm approach.7
Randomization. Patients were randomized by using a table of random numbers. A table containing double-digit randomization codes (from 11-50) was generated using a computer program. Randomization codes were allotted to the enrolled patients starting at a random point in the table. Patients with codes from 11-30 were assigned to TUA and those with codes from 31-50 were assigned to TRA.
Procedure and vascular access. Five Fr and 6 Fr sheaths (Terumo Corporation) were used for diagnostic and interventional procedures, respectively. After the insertion of the sheath, a cocktail (composed of 50 µg nitroglycerin + 2.5 mg diltiazem + 2 mL 2% lignocaine without preservative) was given intraarterially as a rapid bolus. An intravenous bolus of unfractionated heparin (100 IU/kg) was given preprocedure and was repeated (60 IU/kg) if procedure time exceed 1 hour. Post-PTCA glycoprotein IIb/IIIa inhibitors were used if indicated. Arterial spasm during the procedure was managed by additional dose of cocktail. Ad hoc PCI was performed after upgrading the arterial sheaths from 5 Fr to 6 Fr. Arterial sheaths were removed immediately after the procedure regardless of anticoagulation and local hemostasis was achieved using a pressure bandage with sticky straps attached to the bandage to occlude the puncture site. These bandages were removed on the same day at 5 hours post procedure.
Follow-up and ulnar pulse evaluation. Ulnar and radial arteries were examined clinically on the day of the procedure and at scheduled 1-week follow-up. Clinically impalpable pulse at 1 week was labeled as artery occlusion. Detailed examination for ulnar nerve was also done at scheduled 1 week follow-up and ulnar nerve injury, if any, was noted.
Primary and secondary endpoints. The primary endpoint of the study was a composite of major adverse cardiac events (death, myocardial infarction, stroke, or urgent target-vessel revascularization), major vascular events during hospital stay, or crossover rate. Major vascular events comprised large hematoma with hemoglobin drop >3 gm% or need for blood transfusion and occlusion. Any bleed with requirement of blood transfusion was deemed a major bleed. Secondary endpoints were individual components of the primary endpoint, failed attempts (≥3 attempts prior to successful cannulation), total procedural and fluoroscopy times, contrast volume used, and vasospasm.
Data collection. Clinical characteristics of individual patients were noted down after the consensus of two investigators. In case of discrepancy between the two, the opinion of a third independent investigator was solicited and the decision of the majority prevailed. Periprocedural data were collected by the primary operator; however, postprocedure data regarding complications were noted during follow-up by investigators other than primary operator.
Clinical characteristics. Forearm vessels (radial and ulnar artery of either side) were examined clinically and data were collected for the following features: (1) volume of pulse (low-volume pulse if pulse pressure was <40 mm Hg); (2) palpability of pulse (easy vs difficult), ie, palpability of the UA was labeled “easy” if it was found to course in between forearm tendons or “difficult” if it was running beneath forearm tendons, RA palpability was labeled “difficult” if it followed a course that deviated from its natural course; (3) presence or absence of tortuosity; and (4) calcification of vessel wall, if present, was confirmed by forearm x-ray.
Periprocedural parameters. The following parameters were recorded for analysis purpose: (1) number of puncture attempts (number of times skin was entered by the puncture needle until successful arterial access); (2) arterial access time or puncture time; (3) crossover and final access site; (4) failed attempts (need for >3 attempts to cannulate artery); (5) total procedure time (time elapsed during coronary angiography after sheath insertion); (6) number of catheters used; (7) vasospasm (as per ease of catheter maneuverability and need for cocktail during procedure); (8) total fluoroscopy time; and (9) early and late postprocedural complications, ie, major and minor bleeds, Volkmann’s ischemia/contracture, forearm hematomas, pseudoaneurysms, arteriovenous fistula, acute closure of the ulnar artery, and gangrene of the upper limb.
Statistical analysis. The current study was designed to prove non-inferiority of TUA when compared with TRA in terms of primary endpoints if performed by experienced operators. To prove non-inferiority of TUA, a margin of 1.93% was derived by the fixed-margin method.8 The largest clinically acceptable difference (degree of inferiority) of the TUA compared with TRA was assumed to be 4.87%, which represents the difference in the procedural failure rate between radial and femoral access in a meta-analysis of 12 randomized trials.8 A preserved effect of 50% was used to determine non-inferiority margin as per draft Food and Drug Administration guidelines.9 Non-inferiority margin thereafter was calculated using natural logarithm: eln(1/M1) × (1 – preserved effect), where M1 is the upper bound of the 95% confidence interval (CI) of the pooled effect size (4.87% in this study).9 Non-inferiority of TUA was concluded if the upper bound of the 95% CI for the effect estimate remained smaller than the non-inferiority margin. The sample sizing assumed that the expected percentage events in either arm would be as in previous studies (14.3%).9 Considering the largest clinically acceptable difference (degree of inferiority) of 4.87% in previous studies, to achieve a statistical power of 90% with two-sided α=0.05, a sample size of 2172 patients (1086 patients in each treatment group) would be required to establish the primary hypothesis. Assuming a 10% drop-out rate, we planned to enroll ≥2386 patients. Intention-to-treat and per-protocol analyses were performed on the collected data. Non-prespecified subgroup analysis was performed using the Mantel–Haenszel fixed effects analysis model with 95% CI and risk ratio (RR) as effect measured. The stratified analysis for failed procedure (need for >3 attempts for cannulation) was performed on the following parameters: age <75 years or >75 years; male gender; diabetes; hypertension; renal dysfunction; palpability of vessel ease or difficulty; calcification of vessel wall; volume of pulse; tortuosity of vessel; and presence or absence of spasm during procedure. Statistical analyses were performed using MedCalc for Windows version 11.2.1.0, JMP Statistical Discovery version 11.2.0 (SAS, Inc), and Review Manager 5.3.
Results
Preprocedural parameters. The present study was conducted at a center dedicated for radial intervention. The demographic and baseline characteristics did not differ between the two groups (Table 1).
Periprocedural and postprocedural parameters. Total procedural time, fluoroscopy time, number of attempts, change of catheter, and amount of contrast used did not differ between the two groups (Table 2). Comparison of periprocedural and postprocedural parameters is shown in Table 3, Table 4, and Table 5. The differences in the incidence of crossover (4.4% in TUA vs 3.8% in TRA; P=.44) and large hematoma (1.0% in TUA vs 0.9% in TRA; P=.69) between the two groups remained statistically non-significant. Spasm occurred in 6.9% of the TUA group vs 8.7% of the TRA group (P=.09). In the intention-to-treat analysis, the composite primary endpoint did not show difference in the two study groups (14.6% in TUA vs 14.4% in TRA; RR,1.01; 95% upper CI, 1.2; P=.92) and proved non-inferiority of TUA compared with TRA. Using the non-inferiority margin of 1.93, analysis confirmed non-inferiority for all components of composite primary endpoints except large hematoma (Figure 3). Although the upper bound of the CI crossed the non-inferiority margin, the lower bound CI still remained within the non-inferiority zone, proving it inconclusive for large hematoma. Table 5 highlights the details of crossover rates and final access site following crossover in either arm. Subgroup analysis of predefined parameters (Figure 4) was performed so as to pinpoint the factors that can predict failure of the procedure 
(defined as need for >3 attempts to cannulate). Four parameters showed significant association with failed attempts: (1) low volume pulse (OR, 2.06; 95% CI, 1.25-3.41; P=.01); (2) inability to palpate ulnar artery with ease (OR, 2.15; 95% CI, 1.48-3.13; P<.001); (3) tortuosity of vessel (OR, 2.13; 95% CI, 1.26-3.59; P=.04); and (4) calcification vessel wall (OR, 2.15; 95% CI, 1.39-3.33; P=.01). Twelve patients in the TUA group had transient ulnar nerve parasthesia with no residual manifestation at the time of discharge, suggesting that damage to the ulnar nerve is uncommon and a reversible adverse event of UA cannulation.
Discussion
UA interventions are gaining increasing acceptance by experienced operators, thereby significantly reducing the need for crossover to femoral route. The recently published AURA of ARTEMIS study was terminated prematurely on first interim analysis with a conclusion that TUA strategy is inferior to TR strategy as a result of high crossover rates in the TUA arm.6 However, the comparison was between experienced transradialists performing RA cannulation vs naive transulnarists (no/minimal transulnar operator experience) performing UA cannulation. Also, attempts had been made to cannulate the UA even if pulse was absent or had low-volume quality. Because of its deep location and weak (but definite) palpability, UA cannulation has its own learning curve and cannulation of the UA is difficult for an operator naïve to transulnar cannulation. The current study, which is by far the largest to date on UA cannulation, tries to address both of these limitations: (1) operators in the current study were default radialists with experience of at least 50 UA cannulations, a level enough to make events in both groups 
comparable; and (2) the UA was considered for cannulation only if it was easily palpable.
Should UA access be a default strategy? Using the RA as the default route for coronary access should preferably be avoided for two reasons: (1) 6% of TRA patients develop vascular occlusion, thereby making repeat procedures difficult;10 and (2) prior puncture of the RA is associated with more intimal hyperplasia and reduced early graft patency.5 At 1-week follow-up in the current study, there was no significant difference in arterial occlusion rates between ulnar (6.1%) and radial routes (6.6%). There was absolute difference of 1.8% in incidence of spasm (6.9% in UA vs 8.7% in RA; P=.09). Furthermore, its relatively large size as compared 
with the RA can accommodate catheter sizes up to 7 Fr,11 and its straighter course, similar occlusion rates, and lower tendency to spasm (mainly due to fewer alpha-adrenergic receptors for epinephrine)12-15 makeTUA a viable alternative to TRA for both default and crossover strategies.
Will expertise in UA cannulation strengthen default forearm access approach? Non-ischemic complications (bleeding and vascular) with coronary intervention can be reduced by: (1) use of forearm vascular access (TRA or TUA) as the default strategy; (2) use of forearm vascular access (TRA or TUA) as the preferred crossover site;15 (3) minimizing use of the femoral route, thereby minimizing vascular access complication common with this approach;16,17 and (4) optimizing antiplatelet therapy. 18,19 By considering both the RA and UA for crossover, the need for cannulating the femoral artery was <0.5% in the current study, which is far less than the 7.7% in contemporary studies.20 The availability of UAs as an additional forearm vascular access site during both default and crossover strategies will markedly increase the chance that the access site will remain confined to the forearm and thereby decrease the incidence of non-ischemic complications.
Learning curve of UA cannulation. When performed by a default transradialist, we have reported a crossover rate of as high as 16% during cannulation of the initial 50 UAs. After attaining experience with as few as 50 procedures, however, the crossover rate of the UA group decreased to a level comparable with RA cannulation.7 The crossover rate with TUA in the current study was comparable with TRA (4.4% vs 3.8%, respectively; P=.44). The learning curve in TUA is slightly longer than with TRA. This may be due to the following reasons: (1) the UA courses along or beneath the forearm tendons, thus obscuring palpability; and (2) its proximity to the ulnar nerve and lack of posterior bony support make hemostasis more difficult. Further multiple attempts by an inexperienced operator while trying to cannulate the UA may cause tissue edema, hematoma, and vessel spasm, which may eventually lead to access failure. This suggests that the operator needs to attain a certain level of experience before attempting to cannulate the ulnar artery before events with ulnar cannulation can become comparable to those with radial cannulation, an issue that has not been addressed in the past.
Factors predicting failure of UA cannulation. Stratified analyses of data yielded significant association of five clinical parameters: (1) volume of pulse; (2) inability to palpate UA with ease; (3) tortuosity of vessel; (4) experience of operators; and (5) calcification of vessel wall, with requirement of >3 attempts to cannulate ulnar artery. Other than vasospasm as assessed by the operator during the procedure, the remaining parameters can simply be assessed on bedside clinical evaluation by palpation before proceeding to UA cannulation. Factors predicting failure to cannulate the UA can be expressed as “VInTEC parameters” which includes: Volume of pulse, Inability to palpate UA with ease, Tortuosity of vessel, Experience of operator, and Calcification of vessel wall.
Study limitations. This study had several limitations. First, recording of the clinical characteristics and events by the investigators themselves may be a source of bias. However, we attempted to address this by considering the majority decision of three operators for clinical characteristics and recording postprocedural events by investigators other than the primary operator. Second, there was no follow-up protocol for Doppler interrogation of cannulated vessel to assess occlusion rates; if the vessels were palpable, we concluded that no occlusion was present. Since we used a non-palpable pulse to define RA or UA occlusion, this may underestimate true occlusion since occluded arteries can still have a pulse from retrograde waveforms through the palmar arch. Delayed vessel occlusion due to intimal injury was not addressed. Although this mimics the real-world scenario where Doppler is not a routine investigation on follow-up, the incidence of occlusion may not be a true reflection of the vessel occlusion rate. No patient, however, had clinical evidence of hand ischemia, which rules out any clinically significant arterial occlusion. Third, operators in the current study were default radialists with experience in UA cannulation, so these results may be difficult to reproduce by physicians without such experience. Fourth, clinically impalpable UA was not considered for randomization; therefore, results do not hold true in such a clinical scenario. Fifth, patients with primary angioplasty and cardiogenic shock were not included in the study population, since time to intervene was of paramount importance in both of these situations. Lastly, although we have identified parameters to predict ulnar artery cannulation failure (VInTEC parameters), the significance of each individual parameter’s contribution in predicting failure to cannulate UA needs to be confirmed by trials addressing this issue specifically, so that a novel score can be established to predict preprocedural success rate of UA cannulation.
Conclusion
TUA is a viable and non-inferior alternative option for forearm access. In the event of RA cannulation failure, the UA should be considered the second favored crossover site. With increasing experience, TUA may even be considered a default route, thereby sparing the RA for future needs. Furthermore, careful assessment of the vessel prior to the procedure may help the experienced operator predict failure of UA cannulation and associated complications.
Acknowledgment. The authors would like to acknowledge Urmila Sharma, Fateh Singh, M.L. Bairwa, P. Gaur, and Khushvilla Parihar for their help with meticulous data collection.
References
1. Caputo RP, Tremmel JA, Rao S, et al. Transradial arterial access for coronary and peripheral procedures: executive summary by the transradial committee of the SCAI. Catheter Cardiovasc Interv. 2011;78:823-839.
2. Kedev S. Transradial and transulnar access for percutaneous coronary interventions. Arch Turk Soc Cardiol. 2011;39:332-340.
3. Campeau I. Percutaneous radial artery approach for coronary angiography. Cathet Cardiovasc Diagn. 1989;16:3-7.
4. Kiemeneij F, Laarmann G. Percutaneous transradial artery approach for coronary Palmaz-Schatz stent implantation. Am Heart J. 1994;128:167-174.
5. Kamiya H, Ushijima T, Kanamori T, et al. Use of the radial artery graft after transradial catheterization: is it suitable as a bypass conduit? Ann Thorac Surg. 2003;76:1505-1509.
6. Hahalis G, Tsigkas G, Xanthopoulou I, et al. Transulnar compared with transradial artery approach as a default strategy for coronary procedures: a randomized trial. The Transulnar or Transradial Instead of Coronary Transfemoral Angiographies Study (the AURA of ARTEMIS study). Circ Cardiovasc Interv. 2013;6:252-261.
7. De Andrade PB, Tebet M, Andrade M, Mattos L, Labrunie A. Performance of coronary procedures through the transulnar access without assessment of the integrity of the deep palmar arch. J Intervent Cardiol. 2008;21:562-565.
8. Agostoni P, Biondi-Zoccai GG, de Benedictis ML, et al. Radial versus femoral approach for percutaneous coronary diagnostic and interventional procedures: systematic overview and meta-analysis of randomized trials. J Am Coll Cardiol. 2004;44:349-356.
9. Wangge G, Roes KCB, Boer A, Hoes AW, Knol MJ. The challenges of determining noninferiority margins: a case study of noninferiority randomized controlled trials of novel oral anticoagulants. CMAJ. 2013;185:222-227. Epub 2012 Aug 20.
10. Stella PR, Kiemeneij F, Laarman GJ, Odekerken D, Slagboom T, van der Wieken R. Incidence and outcome of radial artery occlusion following transradial artery coronary angioplasty. Cathet Cardiovasc Diagn. 1997;40:156-158.
11. Pinto DS, Stone GW, Shi C, et al. Economic evaluation of bivalirudin with or without glycoprotein IIb/IIIa inhibition versus heparin with routine glycoprotein IIb/IIIa inhibition for early invasive management of acute coronary syndromes. J Am Coll Cardiol. 2008;52:1758-1768.
12. Yatskar L, Selzer F, Feit F, et al. Access site hematoma requiring blood transfusion predicts mortality in patients undergoing percutaneous coronary intervention: data from the National Heart, Lung, and Blood Institute Dynamic Registry. Catheter Cardiovasc Interv. 2007;69:961-966.
13. Lotan C, Hasin Y, Mosseri M, et al. Transradial approach for coronary angiography and angioplasty. Am J Cardiol. 1995;76:164-167.
14. Li Y, Zhao Y, Zhao Y, et al. Safety and efficacy of transulnar approach for coronary angiography and intervention. Chin Med J. 2010;123:1774-1779.
15. Petersen ED, Dai D, DeLong ER, et al. Contemporary mortality risk prediction for percutaneous coronary intervention: results from 588,398 procedures in the National Cardiovascular Data Registry. J Am Coll Cardiol. 2010;55:1923-1932.
16. Huggins CE, Gillespie MJ, Tan WA, et al. A prospective randomized clinical trial of the use of fluoroscopy in obtaining femoral arterial access. J Invasive Cardiol. 2009;21:105-109.
17. Sherev DA, Shaw RE, Brent BN. Angiographic predictors of femoral access site complications: implication for planned percutaneous coronary intervention. Catheter Cardiovasc Interv. 2005;65:196-202.
18. Mehta SR, Granger CB, Eikelboom JW, et al. Efficacy and safety of fondaparinux versus enoxaparin in patients with acute coronary syndromes undergoing percutaneous coronary intervention: results from the OASIS-5 trial. J Am Coll Cardiol. 2007;50:1742-1751.
19. Montalescot G, White HD, Gallo R, et al. Enoxaparin versus unfractionated heparin in elective percutaneous coronary intervention. N Engl J Med. 2006;355:1006-1017.
20. Hetherington SL, Adam Z, Morley R, et al. Primary percutaneous coronary intervention for acute ST-segment elevation myocardial infarction: changing patterns of vascular access, radial versus femoral artery. Heart. 2009;95:1612-1618.
___________________________________________
From the 1Department of Cardiology and JLN Medical College, Ajmer, Rajasthan, India; and 2Mayo Clinic, Rochester, Minnesota.
An abstract of the current study was accepted and presented at the 2015 American College of Cardiology Scientific Sessions as a “late breaking clinical trial in featured clinical section.”
Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.
Manuscript submitted June 3, 2015, provisional acceptance given July 2, 2015, final version accepted November 10, 2015.
Address for correspondence: Rajendra Gokhroo, DM, Department of Cardiology, JLN MC, Ajmer, Rajasthan, India. Email: drrajendragokhroo@yahoo.in
