The Transulnar Approach to Coronary Angiography and Intervention: Assessing the Anatomy of the Ulnar Artery Using Angiography

Background. The transulnar approach (TUA) has been proposed as a safe alternative to the more established transradial approach (TRA) for cardiac catheterization. However, no study has assessed the anatomy and variability of the ulnar artery using angiography. Methods. A retrospective analysis of patients who underwent transradial cardiac catheterization during routine clinical care was conducted. Both quantitative and qualitative measurements of artery diameter were collected. Results. Among 700 consecutive patients, mean distal ulnar artery diameter (UAD) was larger in men (3.2 ± 0.9 mm) compared with women (2.7 ± 0.7 mm; P<.001). UAD was larger than radial artery diameter (RAD) at all measured sites (distal ulnar, 3.0 ± 0.8 mm; distal radial, 2.9 ± 0.7 mm; P=.046). Compared with the radial artery, the ulnar artery had more atresia (4.3% ulnar vs 0% radial; P<.001), fewer loops (0.6% ulnar vs 2.4% radial; P<.01), and less spasm (2.7% ulnar vs 23.4% radial; P<.001). UAD had more variability (distal variance, 0.68) as compared with the RAD (distal variance, 0.53; P<.001). Conclusion. We found that the ulnar artery has a larger diameter, fewer loops, and less spasm, but more variance than the radial artery. Additionally, males have larger ulnar arteries than women. These findings have implications on the application of TUA either as an alternative to TRA or as the primary point of access.

J INVASIVE CARDIOL 2022;34(3):E164-E170. Epub 2022 February 18.

Key words: access approach, alternative access, transradial approach, transulnar, ulnar artery

The transradial approach (TRA) for percutaneous cardiac catheterization is a popular alternative to the more traditional transfemoral approach (TFA). A large meta-analysis found that TRA resulted in less major bleeding, vascular complications, major adverse cardiovascular events, and all-cause mortality when compared with TFA.¹ This is thought to be due to the radial artery’s superficial anatomy, easy compressibility, and early mobility post catheterization. However, the radial artery is not without limitations, including high prevalence of spasm and anatomic variations, such as loops or tortuosity.² Additionally, the smaller caliber of the radial artery limits the size of interventional equipment that can be used. This has led to increased interest in the transulnar approach (TUA) as a potential alternative to both TRA and TFA.^3-7 A recent meta-analysis of 6 trials showed noninferiority in major adverse cardiovascular events and vascular complications with TUA as compared with TRA.⁸ However, data on the anatomical characteristics of the ulnar artery during cardiac catheterization are lacking and have mostly been limited to ultrasound characterization. In the present study, we aim to characterize the ulnar artery anatomy in patients undergoing routine radial access percutaneous cardiac catheterization.

Between January 2012 and July 2017, consecutive adult patients who underwent TRA at a single, large, academic medical center during routine clinical care were retrospectively identified. Patients were eligible for inclusion in this retrospective study if they underwent proximal TRA with angiography of the radiobrachial vasculature and had visualization of the ulnar artery. For this analysis, we excluded patients with high radial take-offs, which often do not fully opacify the ulnar artery, those with cineangiograms with insufficient time to completely visualize the ulnar artery, and those with failed catheterizations (Figure 1). Angiographic measurements were made following sheath insertion and prior to the introduction of wires or catheters.

The electronic medical record was reviewed for pertinent demographic, clinical, and procedural information. Both sex and race were self reported. Information was stored in a secure electronic database. The institutional review board (IRB) at the University of Chicago reviewed and approved the study (IRB 16-1124).

Transradial access. Three high-volume transradial interventional operators performed the majority of cardiac catheterizations included in this study. Conscious sedation with intravenous fentanyl and/or midazolam and local anesthesia with 2% lidocaine were routinely administered prior to arterial access to ensure patient comfort and reduce radial vasospasm rates. Decisions related to patient selection, radial access side, use of ultrasound guidance, and micropuncture access were left to the judgment of the individual operators. Access was usually obtained with either a 20-gauge angiocath or 21-gauge, 4-cm needle, approximately 2-3 cm proximal to the styloid process at the point of maximal pulsation. Hydrophilic Terumo Glidesheaths (Terumo Interventional Systems) were used in all cases, the majority of which were 6-Fr standard sheaths, and a minority of which were Glidesheath 6/5-Fr Slender sheaths. The choice of sheath was left to the operator’s discretion.

Once radial artery access was secured and the sheath was inserted, a mixture of 50 U/kg bodyweight of unfractionated heparin and at least 200 µg of nitroglycerin ± 2.5 mg verapamil was administered to reduce radial spasm. After the administration of this spasmolytic “cocktail,” a 3 frame per second digital subtraction angiography of the radiobrachial and ulnar vasculature was obtained with a portion of the inserted sheath visible in the image (Figure 2). Cardiac catheterization then proceeded as per standard of care.

Measurements. Detailed methods for quantification of the radial artery diameter (RAD) measurements are described elsewhere.⁹ In brief, integrated quantitative angiography (QA) software (GE Centricity Cardiology CA1000 2.0; GE Healthcare) was utilized to make postcatheterization RA and UA diameter measurements. The outer borders of the radial sheath were identified, and point-to-point calibration was performed using manufacturer-provided specifications of the radial sheath diameter. Using the automatic border detection function, both the radial and ulnar arteries were measured at the point immediately proximal to the visible tip of the sheath (distal diameter), the mid segment of the artery (mid diameter), and the proximal origin of the artery, immediately distal to the take-off from the brachial artery (proximal diameter). The minimum and maximum artery diameters were also measured. Ulnar atresia was defined as having an ulnar branch that terminated early. Spasm was defined as a “beaded” appearance on angiography. Tortuosity was defined as >3 turns with >45° of angulation. All measurements were conducted by researchers not directly involved in the patients’ care.

Statistical analysis. All analyses and visualizations were performed with “tidyverse,” “tableone,” and “sjPlot” packages in R 4.0.2 (R Core Team, 2020). Differences between gender and artery type was compared using chi-square test for categorical variables and t test for continuous variables with normal distributions. Continuous variables with non-normal distributions were compared using the Mann-Whitney U test. Categorical variables with small cell counts were compared using the Fisher’s exact test. Variances were compared using the F test. Results were adjusted for gender, age, height, and body surface area in the multivariate analysis. All two-tailed P-values <.05 were considered statistically significant.

Between January 2012 and July 2017, a total of 1141 adult patients who underwent TRA were screened for inclusion and 700 patients were ultimately included for analysis (Figure 1). Baseline characteristics by gender are presented in Table 1. Median age was 64 years (range, 28-89 years), 283 (40.4%) were female, and 343 (49.0%) were Black. Males had significantly higher estimated glomerular filtration rates (P=.02), creatinine (P<.001), and hemoglobin values (P<.001) as compared with females in our cohort (Table 1). Males also had significantly lower estimated left ventricular ejection fraction (P=.03) and higher rates of coronary artery bypass surgery (P=.04). The rates of other comorbidities, such as myocardial infarction, congestive heart failure, hypertension, diabetes, chronic kidney disease, and never smokers, were similar between men and women (Table 1).

Females had smaller ulnar artery diameters (UADs) than males at distal (2.7 cm vs 3.2 cm), mid (3.1 cm vs 3.7 cm), proximal (4.5 cm vs 5.0 cm), minimum (2.6 cm vs 3.0 cm), and maximum (4.8 cm vs 5.4 cm) measurements (P<.001). Full results are listed in Table 1. Univariate analysis showed significant correlations between minimum UAD and gender (P<.001), age (P<.01), height (P<.001), and body surface area (P<.01). After multivariate adjustment, only gender (P<.001) and age (P<.01) showed significant correlations. Cumulative frequency distribution curves and density graphs showing variance for distal and minimum measurements between genders are demonstrated in Figure 3. Results were similar for a subanalysis of Black patients only, except that radial artery distal diameter was significantly larger than ulnar artery distal diameter and there was no difference observed in radial and ulnar artery minimum diameter (Supplemental Table S2).

UAD was significantly larger than RAD at distal (3.0 cm vs 2.9 cm; P=.046), mid (3.5 cm vs 3.1 cm; P<.001), proximal (4.8 cm vs 3.4 cm; P<.001), minimum (2.8 cm vs 2.7 cm; P<.001), and maximum sites (5.2 cm vs 3.5 cm; P<.001). UAD also had significantly more variability at distal (P<.001), mid (P<.001), proximal (P<.001), minimum (P=.03), and maximum measurements (P<.001). Results are listed in Table 2. Cumulative frequency distribution curves and density graphs showing variance for distal and minimum measurements between ulnar and radial arteries are demonstrated in Figure 4. Qualitatively, 4 ulnar arteries (0.6%) had loops, 19 (2.7%) had spasms, 126 (18.0%) had tortuosity, and 30 (4.3%) had atresia (Table 2).

The TRA to cardiac catheterization has been associated with decreased time to ambulation, reduced length of hospital stay, decreased costs, and enhanced patient experiences as compared with transfemoral approach.^10-13 In addition, the TRA has been associated with a lower risk of all-cause mortality (odds ratio [OR], 0.71; 95% confidence interval [CI], 0.59-0.87; P<.01), major adverse cardiac events (OR, 0.84; 95% CI, 0.75-0.94; P<.01), decreased risk of bleeding events (OR, 0.53; 95% CI, 0.42-0.65; P<.001), and decreased risk of major vascular complications (OR, 0.23; 95% CI, 0.16-0.35; P<.001).¹ However, the smaller caliber of the radial artery and its increased sensitivity to vasospasm have limited widespread adoption among many United States operators. Thus, interest in other access sites, such as the ulnar artery, have emerged. The AJmer Ulnar ARtery (AJULAR) Intervention Working Group study, which randomized more than 2500 patients to either TUA or TRA, found no differences in their composite primary endpoint of major adverse cardiac events during hospitalization, crossover, major vascular events, or vessel occlusion (risk ratio, 1.01; 95% CI, 0.83-1.2; P=.92).⁴ A recent meta-analysis by Fernandez et al, which included 6 randomized controlled trials, found no statistically significant difference in the incidence of major adverse cardiovascular events between TUA and TRA (OR, 0.90; 95% CI, 0.65-1.25).⁸

Despite interest in the ulnar approach as being a second-line or even primary access site for coronary angiography and interventions,¹⁴ the data on characteristics of the ulnar artery for catheterization remain relatively sparse. Previous data on ulnar artery quantification during coronary catherization have relied on ultrasound or cross-sectional imaging characterization and have not been definitive. A Chinese and a Cuban cohort used ultrasound quantification of ulnar and radial diameters and found that the two vessels were similar in size.^15,16 A group in Pakistan showed that the ulnar artery was larger than the radial artery while a US based group found that the ulnar artery had a smaller caliber than the radial artery.^17,18 A Turkish group using computed tomography angiography (CTA) found no differences between radial and ulnar artery diameters¹⁹ while a group in Korea also using 3-dimensional CTA found that 86% of their patients had ulnar artery diameters larger than radial artery diameters.²⁰  Angiography as compared with ultrasound allows for the direct measurement of vessel caliber during the administration of an irritant (ie, contrast) while also providing information regarding anatomic characteristics such as spasms, loops, or atresia. Our study showed that the ulnar artery was significantly larger than the radial artery at all measured sites. While statistically significant, the measurements of interest for most operators (the distal and minimum diameters) did not have significant clinical differences (distal ulnar, 3.0 mm; distal radial, 2.9 mm; minimum ulnar, 2.8 mm; minimum radial, 2.7 mm). Additionally, despite the mean UAD being larger than the RAD, there was significantly more variation in sizes among the ulnar artery as compared with the radial artery.

It is worth noting the percentage of measured diameters that would be able to accommodate standard hydrophilic Terumo Glidesheaths, which are the most used sheaths in our catheterization laboratory for coronary catheterizations. The manufacturer reported outer diameter sizes for the standard Glidesheath is 2.29 mm for the 5 Fr and 2.62 mm for the 6 Fr. These measurements are graphically represented in Figure 3 and Figure 4. In our study population, 82% of patients had distal (ie, the site of insertion) radial and ulnar artery diameters large enough to accommodate at least a 5-Fr sheath compared with ~66% of patients with vessel diameters large enough for a 6-Fr sheath. Interestingly, we also found that the ulnar artery had lower percentages of vessel loops (0.6%) and spasms (2.7%) compared with previously published reports on the incidence of radial artery loops (5.1%) and spasms (5.1%).⁹ However, these data should be interpreted carefully as comparisons between ulnar and radial anatomic variances in this cohort are difficult given the exclusion of patients with high radial takeoffs or incomplete ulnar visualizations.

Men had significantly larger ulnar diameters than women even after controlling for age, height, and body surface area, a finding similar to previously published angiographic quantification on radial arteries.⁹ Only 75% of women had distal ulnar diameters large enough for a 5 Fr sheath, compared to 86% of men. Note that body surface area was not a predictor of size after multivariate adjustment. We hypothesize that artery diameter is not physiologically correlated with weight and as a result body surface area did not correlate well with size, as body surface area includes weight in its calculation. It is also worth highlighting that there was less difference in radial and ulnar artery sizes in the subanalysis of Black patients, particularly in distal and minimum measurements. We suspect this may be due to demographic and comorbidity differences between Black and non-Black patients (Supplemental Table S1).

Study strength and limitations. There are several key strengths to this study. To our knowledge, this is one of the first angiographic quantification of the ulnar artery in a US cohort allowing for insight into not only caliber of the ulnar vessel but also its anatomic considerations. In addition, we examined a large number of patients (n = 700) with significant diversity (>40% female and nearly half Black), more representative of a contemporary United States population.

While this study does offer some useful insights into the ulnar artery, there are several limitations worth mentioning. First, although patients underwent prospective radiobrachial angiography, this study was conducted in a retrospective manner. As a result, patients with high radial takeoffs and those with incompletely visualized ulnar arteries either due to access issues or due to short cineangiograms were excluded, thus introducing possible selection bias. Additionally, the study selected only for patients who had TRA. Vasodilators and spasmolytic cocktails were introduced directly into the radial artery and not the ulnar artery, although we hypothesize that these agents would have also affected the ulnar artery. It is also worth noting that not all patients received verapamil but again this subset of patients is unlikely to have affected the results. The introducer sheath and wire were threaded through the radial artery and not the ulnar artery, which could have contributed to the differences in spasms observed between the two arteries. Conversely, atretic radial arteries would likely have been selected out due to technical difficulties and patient safety concerns. Finally, segment designation (distal, mid, and proximal) and standard calibration of measurements is dependent on the observer. We have attempted to minimize measurement errors by having multiple observers make measurements and by including a large number of patients.

In a contemporary United States cohort of patients undergoing TRA, we found that the ulnar artery had a larger diameter and less loops and spasm, but more variance than the radial artery. Additionally, the ulnar artery tended to be larger in men than women. These findings have implications on the application of TUA either as an alternative to TRA or as the primary point of access.

Data availability. The datasets generated and used during the current study are available from the corresponding author on reasonable request.

• Feasibility and Safety of Ipsilateral Ulnar Access in Cases of Impossibility or Failure of Radial Access for Coronary Angiography or Percutaneous Coronary Intervention
• Patient Specific Coronary Artery Bypass Graft 3D-Printing: Implications for Procedural Planning in Complex Percutaneous Coronary Interventions