2.1 Technical Considerations

Homogeneity in procedural aspects, techniques and materials used, were maintained over the study period. The operator team consisted of six physicians (three seniors, two juniors, and one trainee) throughout the duration of data acquisition. Seniors were classified as having performed > 1000 CAs/PCIs via forearm artery access prior to the study, juniors as having performed 250–1000 procedures and trainees as having performed < 250 procedures. All procedures were executed under the supervision of a senior physician. The choice of access site and forearm artery was determined by the operator based on preprocedural ultrasound assessment, considering vessel diameter and the presence of plaques or calcification at the intended puncture site. Arterial access sites included the proximal RA, the distal RA in the anatomical snuffbox, and the UA.

TRA was performed as described previously [5, 19]. Briefly, after positioning the forearm on a carbon fiber board (Starboard, Adept Medical Ltd., Auckland, New Zealand), the RA was punctured using a 0.8 × 50 mm G21 needle (B. Braun SE, Melsungen, Germany). The ultrasound probe (9L probe, Vivid T9, GE Healthcare, Boston, Massachusetts) was enveloped in a sterile cover (Flexasoft, Udo Heisig GmbH, Putzbrunn, Germany) and RA puncture was performed under out-of-plane ultrasound guidance. A 0.021-inch, 45-cm wire was advanced into the arterial lumen, followed by the placement of a 70-mm hydrophilic-coated introducer sheath (6 French Glidesheath Slender, Terumo Europe N.V., Leuven, Belgium). After sheath insertion, 2 mg of verapamil and 5000 IU of heparin were administered through the sheath. Subsequently, angiography of the distal upper arm and the forearm, extending to the wrist, was performed. With knowledge of the anatomy of the forearm arteries, a 0.035″ wire (Angiodyn, B. Braun SE) was advanced to the upper arm under continuous fluoroscopic guidance.

2.2 Modified Transulnar Access

On the wrist the UA shows a physiological kinking, which straightens after dorsiflexion of the hand. In its proximal course, the UA first follows the tendon of the flexor carpi ulnaris and then arches underneath its muscle belly into the depth. In this section the UA often shows calcifying atherosclerosis.

In our modified approach for TUA, those two anatomical obstacles are passed with an atraumatic hydrophilic 0.014″ PCI wire (Choice PT floppy, Boston Scientific, Marlborough, USA). Prior to this study, we observed that the 0.014″ wire is more successful in passing through atherosclerotically altered vessel segments than the standard sheath wire (Glidesheath Slender, Terumo Europe N.V., Leuven, Belgium).

TUA was obtained according to the above-described TRA: A 0.8 × 50 mm G21 needle was used for puncture, and in contrast to TRA, the 0.014″ wire was advanced into the lumen of the UA. Subsequently, a 70-mm hydrophilic coated introducer sheath was introduced. Upon successful sheath insertion, 2 mg of verapamil and 5000 IU of heparin were administered via the sheath. If the duration exceeded 20 min, an activated clotting time adjusted heparin therapy was initiated (> 200 s). With every procedure, a standardized forearm angiography, visualizing the mid-upper arm and the forearm up to the wrist, was performed as described previously (Figure 1) [19]. Under continuous fluoroscopic guidance a 0.035″ J wire was advanced to the upper arm via the sheath in analogy with the standard TRA as described previously [5]. The puncture site was closed with a RA compression device (TR Band, Terumo Europe N.V.).

2.3 Defining Difficult Vascular Anatomy

The classification of DVA used in our study provides a purely functional assessment, i.e., the ability to pass the forearm arteries with a guidewire and does not include visual aspects (e.g., the severity of vessel tortuosity) obtained during forearm angiography [5].

DVA of the forearm arteries (from the puncture site to the radio-ulnar-brachial region) was divided into three categories [5]. In short, no DVA (zeroth-degree DVA) was defined if the forearm could be passed with a standard 0.035″ guidewire. First-degree DVA was characterized by vessels exhibiting tortuosity, stenosis, or calcification requiring the use of a hydrophilic 0.035″ guidewire (Radifocus Guidewire M, Terumo Europe N.V.) and/or a 0.018″ guidewire (Advantage, Terumo Europe N.V., Figure 1b). Second-degree DVA encompassed severely tortuous, calcified, or occluded vessels that proved to be impassable with any guidewire, resulting in procedural failure and the need to switch to an alternative access site. After each procedure, the operating physician documented the severity of DVA at the forearm.

2.4 Statistical Analysis

For statistical analysis, Q–Q plots and the Shapiro–Wilk test were used to assess normality. Continuous variables were summarized as mean ± standard deviation (SD) and categorical variables as percentages (count). Student's t-test was used for continuous data and χ² test for categorical data.

Due to imbalances in potential confounding variables between primary TUA and primary TRA cases, we used k:1-nearest neighbor propensity score matching. Propensity scores were calculated by logistic regression with predefined variables (age, height, body mass index, sex, operator experience, clinical urgency, arterial hypertension, diabetes mellitus, left vs. right forearm artery access, and smoking status). The optimum matching ratio (k) was determined so that the maximum standardized mean difference of any confounder was minimized. This resulted in a matching ratio of 4:1. We applied the nearest-neighbor algorithm without replacement and used a caliper width of 0.25. The balance of covariates was considered satisfactory for a standardized mean difference of < 0.1.

After propensity score matching, we compared the prevalences of both outcomes through linear risk difference estimation and calculated standard errors using a cluster-robust technique [20]. To meet the assumption of independence, only the first access attempt was included to determine the prevalence of DVA, that is, if right TRA failed due to, for example, a radial loop and crossover to right TUA was necessary, observations from right TUA were not included in the analysis. All statistical analyses were carried out with Python 3.9 and R 4.0.3.

3.1 Unmatched Cohorts

The study population had a mean age of 69.6 ± 12.0 years and 824 (32.1%) patients were female. In the unmatched cohort, the TUA group did not differ significantly from the TRA group in terms of patient age (70.4 ± 10.5 vs. 69.5 ± 12.1 years, p = 0.381), height (171.7 ± 9.0 vs. 171.6 ± 9.3 cm, p = 0.864), body mass index (28.9 ± 5.8 vs. 28.3 ± 5.3 kg/m², p = 0.168), female gender (32.1 vs. 32.1%, p = 1.0), smoking history (26.5% vs. 28.0%, p = 0.764), and diabetes mellitus (23.5 vs. 22.1%, p = 0.770). The TUA group differed significantly from the TRA group in terms of arterial hypertension (100.0 vs. 92.3%, p < 0.001), left FAA (33.3 vs. 53.3%, p < 0.001), operator experience, clinical urgency and the year of intervention (Table 1). Zeroth-degree DVA was observed in 2171 (90.3%) TRA and 157 (96.9%) TUA cases, first-degree DVA in 172 (7.2%) TRA and 5 (3.1%) TUA cases, and second-degree DVA in 60 (2.5%) TRA and 0 (0%) TUA cases (p = 0.015, Table 2). Overall, in 232 (9.7%) TRA versus 5 (3.1%) TUA cases either first- or second-degree DVA was observed (p = 0.008, Table 2).

3.2 Matched Cohorts

After propensity score matching, the two groups were well balanced regarding all variables (Figure 3). In the matched groups, second-degree DVA was observed in 19 (2.9%) TRA cases and 0 (0%) TUA cases (p < 0.001). As second-degree DVA corresponds to procedural failure, we observed a crossover rate of 2.9% in the TRA group and 0% in the TUA group (p < 0.001). The combined outcome of either first- or second-degree DVA was observed in 52 (8.0%) TRA and 5 (3.1%) TUA cases (p = 0.005, Table 3).

3.3 Vascular Complications

We observed no ischemic or neurologic complications and no case of vasospasm during 162 TUA cases. In 17 (10.5%) TUA cases a local hematoma could be observed, which was treated conservatively.

4.1 Anatomical Differences in the Forearm Arteries Determine Procedural Success Rates

In both, unmatched and matched cohorts, the prevalence of DVA (first- and second-degree) was significantly lower in the TUA group compared to TRA. In the unmatched cohort, second-degree DVA—a common cause of access failure—was absent in TUA cases, whereas it was present in 2.5% of TRA cases. After matching, second-degree DVA persisted in the TRA group (2.9%), further demonstrating TUA's anatomical advantage. The lower DVA rate observed in TUA may be explained by anatomical differences between the RA and UA. Lo and colleagues describe the respective prevalences of vascular variants complicating TRA in their study. In this study, the authors report that failure of TRA was highest in cases of radial loops and radial tortuosity followed by atherosclerosis and high origin of the RA [9].

Emerging evidence in the literature suggests these vascular variants are less frequently found in the UA. A potential explanation for this could be that the UA represents the radiographic anatomical continuation of the brachial artery, with its—compared to the RA—larger diameter, straighter course and relatively flat branching angle from the brachial artery potentially making the UA less susceptible to these vascular variants [18, 21, 22]. These anatomical differences between the RA and UA may lead to higher procedural success rates highlighting TUA as a promising alternative.

4.2 Beyond the Guidelines There Is a Need for Alternative Access Routes

Current guidelines recommend TRA as a Class IA indication for routine CA/PCI, including cases of acute coronary syndrome [1, 3]. However, our findings underscore the advantages of TUA over TRA, especially when the efficacy of TRA is limited by DVA, like atherosclerosis, tortuosity, loops or a high origin of the RA. Additionally, TRA success rates may be compromised by difficulties in RA localization or a small RA diameter at the puncture site. Previously, TRA may have led to RA occlusion [23] or the RA may have been used or will be needed as a coronary artery bypass grafting or in arteriovenous fistula procedures [24]. In most cases, clinical judgment advocates for TFA; however, TFA shows higher bleeding and mortality rates [23, 25].

The lower prevalence of DVA in the UA suggests that TUA may be a viable alternative access route in patients with anatomical variations of the RA or in those with contraindications for TRA.

4.3 Ultrasound Guidance for TUA

In contrast to our results, Hahalis and colleagues described inferiority of TUA compared to TRA due to a higher puncture-related crossover rate. In this study, however, puncture of the UA was attempted even in cases without sufficient arterial palpability [26].

Another study showed results consistent with our findings, that is, no significant differences in procedural attempts between TUA and TRA. The high success rates in this study were primarily contributed to operator experience. However, TUA was only performed when the UA was easily palpable, which represents a selection bias [27, 28]. The more difficult palpability of the UA due to its anatomical course, represents a disadvantage for the UA as access site. This can be circumvented by the consistent use of ultrasound as was the case in our study. As puncture occurs in the three-dimensional space, ultrasound provides a real-time visualization of the target vessel and the needle's trajectory relative to the artery. Therefore, ultrasound guidance is the first key factor in enhancing the procedural reliability of TUA [27-30]. Additionally, damage to the ulnar nerve, which is located lateral to the UA, can be effectively circumvented [31].

4.4 Using a Hydrophilic 0.014″ PCI Wire to Straigthen the Ulnar Artery

The UA exhibits a physiological S-curve at the level of the wrist and continues with an arched course, running deeper beneath the forearm musculature. These anatomical peculiarities can potentially lead to the inability to advance the standard sheath guidewire or cause arterial perforation during sheath insertion.

In our modified approach for TUA, those two anatomical obstacles are passed with an atraumatic hydrophilic 0.014″ PCI wire which straightens the UA at the forearm and facilitates the advancement of the standard guidewire. Implementation of the hydrophilic 0.014″ wire is the second key factor in our modified TUA.

4.5 Standardized Forearm Angiography

In a previous study, we could show that a implementation of a standardized forearm angiography before guidewire advancement increases procedural success [19]. Despite the UA being less affected by DVA, it is reasonable to also perform a forearm angiography during TUA to allow for proactive guidewire adjustment, including an early switch to a hydrophilic guidewire in the presence of DVA.

4.6 Procedural Safety

Our secondary outcome analysis focused on intraprocedural complications associated with TUA. In our cohort, we observed no vasospasms, ischemic or neurologic complications—despite the UA's proximity to the ulnar nerve. Local hematoma formation occurred in 10.5% of cases and was treated conservatively, which aligns with previous studies on TUA, where the UA's deeper anatomical position compared to the RA makes it more difficult to achieve hemostasis [14]. With the first key factor of our modified TUA—ultrasound-guided puncture—the risk of local hematoma can be minimized which is consistent with the literature [14, 26, 32, 33].

Taken together, we demonstrate that TUA has a higher success rate due to a significantly lower DVA rate. Our modified TUA approach demonstrated high procedural success, attributed to ultrasound guidance, application of the 0.014″ wire to navigate the physiological S-shaped course of the UA at the wrist, and to accommodate the UA's naturally arcuate course along the forearm. Standardized forearm angiography enables proactive guidewire adjustment to the observed vascular variants and avoids misintubation of side branches. Furthermore, our modified TUA was associated with a favorable safety profile, as no major vascular complications were observed. Therefore, this study demonstrates the feasibility of TUA as an alternative to TRA for CA/PCI.