Abstract
Background
To evaluate factors that influence the successful cannulation of intracranial vessels using a transradial approach.
Methods
A total of 61 transradial diagnostic angiograms were evaluated in a tertiary care center from July 2020 to December 2021. We evaluated the learning curve and aortic arch vessel factors that may influence the cannulation of intracranial major vessels using a transradial approach.
Results
Learning curve for the procedure was established after 21 cases. We were successful in cannulating the supra-aortic arteries except in 4 cases where we were unable to cannulate the left VA (vertebral artery). Significant positive correlation was seen between time to Sim (Simmons curve) formation and aortic arch diameter (p = .002). Significant positive correlation was also seen between left VA take-off angle and time to cannulate left VA (p = .001) and negative correlation was noted between left CCA (common carotid artery) take-off angle and time to cannulate left CCA (p = .001).
Conclusion
Transradial approach is a feasible and safe approach for performing cerebral angiography. Multiple factors can influence the procedure time and successful cannulation of intracranial vessels. With the availability of radial specific hardware in the future, procedural success and time taken to complete the procedure may improve.
Keywords: Transradial approach, transradial access, diagnostic cerebral angiography, neurointervention, take-off angles, time to cannulate, fluoroscopy time
Introduction
Vascular access is a very important step for the success of any interventional procedure. Traditionally, the transfemoral route has been widely practiced by the interventional neuroradiologists owing to the familiarity with the approach. Transfemoral approach is associated with increased post procedure recovery time which is worrisome for ambulatory patients.1,2 This led to the introduction of novel yet practical and patient friendly transradial approach into the field of neurointervention. 3 Adaptability of the approach has significantly increased in the interventional neuroradiology as evidenced by the growing number of papers on the transradial technique in recent years.4–6 We sought to incorporate our own experience with this technique in this research, as well as discuss the numerous aspects that contribute to the transradial approach’s success.
Materials and methods
This was a prospective single center analytical study approved by the institutional ethics committee. We included all the patients >18 years who were indicated for diagnostic cerebral angiography according to our department protocol. Patients with radial artery diameter <1.8 mm, history of Raynaud’s phenomenon, severe atherosclerotic disease, and unhealthy puncture site were excluded. All procedures were carried out by single operator (VB) with 8 years of experience in field of neurointervention
Computed tomography angiography (CTA) head and neck was done prior to procedure for measurement of aorta metrics and take-off angles of the neck vessels. If CTA of the neck was not available prior to procedure, TOF MRA (time of flight magnetic resonance angiography) neck vessels was done post procedure.
Anatomy of the arch of aorta and neck vessels
Type of the arch, great vessel anomalies, diameter of aortic arch, take-off angles of great vessels and bilateral vertebral arteries were recorded by two neuroradiologists and mean values were taken. Aortic arch diameter was measured on all three planes with values averaged. Anteroposterior span and mediolateral span were considered as surrogate markers for aortic unwinding. Take-off angles of brachiocephalic trunk, bilateral CCAs, and bilateral subclavian arteries were measured using spine as the reference line as mentioned in previous literature.7,8 For bilateral vertebral arteries, angles were measured using a tangential line along the subclavian artery as reference with direction of flow marking the angle (Figure 1).
Figure 1.
Measurement technique employed to measure the angles of left CCA (a) and RVA (b).
Procedural details
Right sided radial artery was the principal access site in our study. Entire procedure was done under local anesthesia. Sedation and anxiolytics were used as per need in uncooperative patients and patients with painful radial spasm. 9 In most of the cases, radial access was achieved using ultrasound guidance and 21G needle with sheath inserted using Seldinger technique (Glide Slender sheath 5F, terumo, Japan). 10 Radial cocktail regimen used after sheath insertion was Inj. Diltiazem 2.5 mg, inj Nitroglycerin 200 micrograms and inj. Heparin 50 U/kg which was injected slowly over 30s. 11 Heparin dose was tapered according to procedure duration. However, there was no need of repeat heparinization in any of the cases. Pulse oximeter reading of ipsilateral thumb was recorded throughout the procedure to look for any compromised blood flow to the hand.
After cocktail injection, radial artery angiogram was taken visualizing the elbow. Radial artery anatomy and variations were assessed. Using roadmap angiography, arm vessel was traversed using over the wire technique (Glide catheter 5F- Simmons/sidewinder curve 2 & 0.035 Glidewire, Terumo intervention system, Japan).
Time durations for Sim formation and cannulation of each individual vessel from arch were measured with stop clock. Timings for cannulation of the artery were noted from the formation of Sim curve till the stabilization of the catheter in the concerned vessel. Crossover to alternate access was done in cases of failure to achieve the diagnostic indication.
Hemostasis
Post-procedure hemostasis was achieved by transradial compression band following patent hemostasis technique.12,13 Patient recovery duration, wellbeing, and mobility were assessed post procedure.
Follow-up with radial artery doppler was done at the time of discharge or after 3 days post procedure whichever was earlier.
Statistical analysis
Continuous data was mentioned as mean, range, and standard deviation. The normality of quantitative data was checked by measures of cannons and sparrows tests of Normality. Spearman correlation coefficient was calculated as the data was not normally distributed to see strength of the relation between take-off angles and time taken to cannulate the vessels and between aorta metrics and Sim curve formation time. Linear regression analysis was carried out for contribution of the various variables with significant correlation (i.e., arch of aorta diameter and time to Sim formation, take-off angles of left CCA and left VA) with time taken.
Correlation and regression analysis were carried out for the entire study population and part of the study population after the learning curve was achieved (i.e., After 21 cases).
All the statistical tests were two-sided and were performed at a significance level of α = 0.05. Analysis was conducted using IBM SPSS STATISTICS (version 22.0)
Results
There were a total of 61 patients out of which 70.4% were males (n = 43) and 29.6% patients were females (n = 18) with average age of 43.8 years (20–65 years). One included case was crossed over from transfemoral approach due to difficult arch. Left transradial and distal transradial approaches were excluded from the study (n = 4). 1 case was crossed over to transfemoral approach due to radial artery perforation, 2 cases due to severe radial artery spasm, and 1 case due to failed Sim formation in a case of aberrant subclavian artery.
Indications of cerebral angiography were broadly classified into 4 categories. Follow-up of vascular malformations (39%) included intracranial brain arterio-venous malformations, dural arterio-venous fistulas, and 1 case of facial malformation. CTA negative intracranial bleed population (34%) included both subarachnoid hemorrhage and intraparenchymal hemorrhage. There was also a case of CTA negative ptosis. Other groups were predominantly follow-up of aneurysm post coiling or clipping (13%) and angiographies done for pretherapeutic planning for aneurysms (14%).
Mean radial artery diameter was 2.06 mm (range 1.8–2.3 mm). Post cocktail average diameter was 2.14 mm (range 1.9–2.4). There was average 0.1 mm increase post cocktail diameter (p –.003). 55 cases were done with counter puncture technique and 6 cases were done with tactile technique.
Most common radial artery variation encountered was high origin of radial artery (n = 11) (18%). Tortuous course with loop was seen in 2 cases (3%)
Overall mean puncture time was 105 s, whereas after 21 cases, it had reduced to 68 s. Average number of attempts was 1.7, whereas after 21 cases, it improved to 1.2. (21 cases is the learning curve milestone in our study as described later).
We experienced one case of radial artery dissection and one case of radial artery perforation diagnosed by radial angiogram. These cases were conservatively managed with crossover to transfemoral and left transradial approach, respectively. Radial artery thrombosis in one of the cases was identified on follow-up ultrasound. Radial artery spasm was encountered in 23.8% (n = 14) patients identified by radial artery angiogram. Out of this, 22% (n = 12) had clinical spasm also. Majority of them were managed by mild sedation and intraarterial spasmolytics. Except 2 cases were crossover to transfemoral approach was required.
Learning curves
Learning curves throughout the course of study can be depicted by line graph drawn along the course of the study against the time taken to cannulate various arteries by the single operator (Figure 2). Learning curves were considered established if the fluctuations between time has reached a fairly plateau level with minimum standard deviation. Learning curve for Rt CCA and Rt VA were achieved within 10 and 12 cases, respectively. Lt CCA time normalized after 16 cases, whereas Lt VA achieved after 21 cases over the course of study. So, in our study, 21 cases were considered as the checkpoint of the learning curve.
Figure 2.
Line graph along the course of study depicting the time taken to cannulate various arteries (a–d).
Mean time to Sim formation was 24.25 s. Strong positive correlation was seen between arch of aorta diameter and time to Sim formation (Table 1). For every 1 mm increase in the arch of aorta diameter, time to Sim formation increases by 3.2 s.
Table 2.
Depicting the results of correlation of the various arteries and time to cannulate.
| Artery | Mean take-off angle (range) | Mean time to cannulate (range) | Spearman’s Rho correlation | p value |
|---|---|---|---|---|
| Rt CCA | 30.09° (11.4–55.6) | 40.03 s | −0.3 | .01 |
| Rt VA | 84.4° (30.7–140.5) | 23.9 s | 0.14 | .2 |
| Lt CCA | 21.2° (8.5–89) | 91.5 s (6–179) | −0.6 | <.0001 |
| Lt VA | 68.97° (6.6–128.2) | 90 s (8–215) | 0.8 | <.0001 |
The correlation of various arteries with time to successful cannulation is shown in Table 2. There was moderate negative correlation between Lt CCA take-off angle and time to cannulate. For every 1°, increase in angle time to cannulate decreased by 2.3 s (Figure 3(a)). Strong positive correlation was also seen between Lt VA take-off angle and time to cannulate with every 1° increase in angle time to cannulate increased by 4.4 s (Figure 3(b)). We were successful in achieving the goal of diagnostic cerebral angiography with 100% success rate in cannulating Rt CCA, Rt VA and Lt CCA. 10% failure rate in cannulating the left VA (n = 4). 3 of the patients had angle more than 90° and 1 had tortuous course of left subclavian artery along the clavicle not giving support for the catheter tracking over the wire along the left VA.
Table 1.
Depicting the correlation values of aorta metrics with time to Sim formation.
| Mean, mm | Spearman’s Rho | p value | |
|---|---|---|---|
| Arch of aorta diameter | 24.25 | 0.78 | .0002 |
| Anteroposterior span | 66.89 | 0.27 | .04 |
| Mediolateral span | 49.74 | 0.19 | .04 |
Figure 3.
Scatter plot depicting the correlation between various arteries take-off angles and time to hook the arteries.(a) Moderate negative correlation between Lt CCA take-off angle and time (p = .001), (b) strong positive correlation between Lt VA take-off angle and time (p = .001).
Our total fluoroscopy time for the cases was on an average of ∼15.2 mins (10.2–32 mins) with average dose area product 88000 cGy.cm2. After 21 cases, fluoroscopy time ranged from 10.2 to 18.8 mins.
Discussion
This study can be divided into two parts—the first section focuses on overcoming the neurointerventionist’s learning curve related with transradial access, while the second evaluates the many aspects affecting the transradial access’s safety and efficacy.
Evaluation of the safety was the important consideration of this study. We did not encounter any major intracranial complications in our study. There was one case of radial artery perforation in the initial part of the study which led to crossover to transfemoral access. Fortunately, the patient responded to conservative management without development of any major hematoma. We encountered one case of right radial artery dissection again in the initial part of the study with crossover to the left transradial approach. Patient was managed conservatively during the hospital stay. Follow-up USG doppler of patient’s right radial artery showed monophasic flow however with no evidence of compromised blood flow in the right hand.
Once the learning curve for puncture has reached a fair plateau, we did not encounter any major arterial complications like perforation or dissection. Most common complication encountered was radial artery spasm (n = 14). This is inclusive of both radiological and clinical spasm; however, management was done only for clinical spasm when patient complains of pain. Most of the cases, the spasm was managed conservatively with mild sedation and intraarterial spasmolytics, that is, NTG (nitroglycerin).
In two cases, there was severe vasospasm with risk of catheter entrapment. Access was abandoned with crossover to femoral approach. Moderate analgesia was given with transfemoral subclavian artery run showing resolution of the spasm.
Two cases of left transradial and 2 distal transradial access were excluded from the study.
Though many authors suggest to divide the procedure into various steps for analysis as beginners we encountered difficulty in following three steps.5,14
□ Radial artery puncture
□ Formation of the Simmons catheter
□ Cannulating the supra-aortic vessels.
These steps can be considered as important milestones in overcoming the learning curve of transradial access. In the first case, access was achieved after 4 attempts but reduced to single or two attempts in the later part of the study. Though incidence of radial artery spasm is partly associated with number of attempts, we did not find any significant correlation in our study mainly owing to small sample size. 8
Radial artery angiogram was usually taken after the access to look for variant anatomy. High origin of radial artery was the most common variant anatomy noted. Small radial loop (<360°) and tortuous radial artery course was seen in two of our cases; however, there was no statistically significant difference in the time taken from radial artery to arch. Data of variant anatomy was also limited in number to reach a conclusion.
Hydrophilic sheath with reduced lateral profile and hydrophilic Simmons 2 catheter were the workhorse of transradial access in our study.
Sim curve formation was done through various ways as previously described, most common being in the aortic arch, ascending aorta and reflecting off the aortic valve. Aorta metrics used in our study such as aortic arch diameter, anterolateral span, and mediolateral span served as surrogate markers of arch unwinding. Measurements of the aorta were done according to the method described by Knox et al. 7
Our study found significant positive correlation between time to Sim formation and aortic arch diameter. For every 1 mm increase in the arch of aorta diameter time to Sim formation increases by 3.2 s. This was in concordance with the study done by Khan et al. 8 However, the study did not perform regression analysis of individual Sim formation time and aortic arch diameter as in our study. Formation of Sim just by rotating the catheter in aortic arch is simplest and quickest way; however, it becomes difficult in aged patients due to unwinding of aorta. 15 As the aortic arch diameter increases, support to the catheter to form the curve in the arch is lost so the curve has to be formed either in the ascending aorta or reflecting off the aortic valve.
We had one case crossed over to transfemoral approach due to aberrant right subclavian artery leading to failed Sim curve formation.
Knox et al. 7 have shown significant correlation with take-off angles and reperfusion time is cases of mechanical thrombectomy via transfemoral approach. We utilized similar measurement techniques for take-off angles of right and left CCA as it was more reproducible. Some changes in bilateral vertebral artery angle measurements which were done according to the course of the subclavian artery as described by Khan et al. 8
Learning curve for left vertebral artery was steep when compared to rest of the supra-aortic vessels. Most significant correlation was seen with left vertebral artery take-off angle and time taken to cannulate it (Spearman = +0.72). More obtuse the take-off angles, more time was taken to hook the artery. All cases with take-off angle >90° required the usage of road-mapping increasing the time further. In our data, we were able to deduce that for each degree increase in angle time to cannulate increases by 4.4 s.
There was failure in cannulating the left vertebral arteries in 4 cases out of which 3 had obtuse take-off angle. Remaining one case had tortuous course of left subclavian artery with acute bend at origin of left vertebral artery. Despite the acute origin of vertebral artery, cannulation could not happen due to lack of support from the subclavian artery. None of the failed left VA cases were crossed over to transfemoral approach due to the fulfillment of the required indication.
Learning curve for left common carotid artery was moderately steep. We had moderate negative correlation of left CCA take-off angle with time taken to cannulate (Spearman = −0.65). The relationship further strengthened with addition of common and close origin of left common carotid artery with right brachiocephalic trunk. However, none of these measurements carried significance when considered alone except the take-off angle. For every 1° increase in take-off angle, the time taken to cannulate reduced by 2.2 s. Time was also paradoxically increased with >80°; however, it did not reach statistical significance in our study and were considered outliers (n = 2). One case of crossover from transfemoral approach due to failed left CCA catherization was seen in our study with left CCA take-off angle of 83°. This was successfully achieved with transradial approach; however, time taken was higher when compared to mean left CCA angle. This might require further data and analysis for accurate interpretation.
Learning curve of right vertebral artery and right common carotid artery were not difficult and showed no significant correlation with their respective take-off angles. Right vertebral artery was cannulated directly without the need of the formation of the Sim’s curve in all the cases.
When compared to earlier studies, our study’s learning curves appear to be completed sooner (21 vs 55). For novices, we recommend doing at least 50–100 cases on a continuous basis to obtain a firm grasp on the method, as there was significant diversity in learning curves between individuals.
Our study was in concordance with the results described by khan et al., except for the angle of left vertebral artery. We discovered difficult cannulation at the obtuse left VA angle by transradial approach, whereas Khan et al. described difficulty in the acute take-off angle of the left VA. In our study, however, we exclusively utilized the Sim 2 catheter for all procedures and did not transfer to a different catheter in the event of difficulty. The disparity between the Sim’s primary curve and obtuse VA angle explains our findings. In our investigation, the fastest fluoroscopy time was 10.2 min. In the first eight cases, it took 28–35 min, but after 21 cases, it took 10.2–18 min. In comparison to the earlier study by Kenawy et al., 16 this was slightly higher. However, as compared to their study, our sample size was substantially smaller.
On follow-up patients, satisfactory score was on an average of 7.6 out of 10. 6 patients gave score less than 5. Out of these 6 patients, 4 patients had pain due to moderate spasm which required NTG for control. One patient had follow-up radial artery thrombosis and other patient developed chronic mild pain in the forearm which persisted more than 3 months post procedure.
Limitations
This was a single center study. Reduced sample size is a limitation to our study. Our study may be prone to selection bias as all the consecutive patients referred for diagnostic angiogram were not included in the study.
We have shown these results using a single type of catheter only although dedicated radial hardware has yet to evolve and with greater availability may reduce the problems regarding successful cannulation of intracranial vessels using a transradial approach.
Conclusion
Transradial approach is feasible and bound to become a routine care in neurointerventions in the near future. Interventionists should overcome the learning curve and consider the factors that may influence the successful cannulation of intracranial arteries when a transradial approach is used as compared to transfemoral approach.
Footnotes
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
Author contributions: All authors of this work met ICMJE criteria for authorship and made substantial contributions to the conception and design, acquisition of data, analysis and interpretation of data, drafting, critical revising, and final approval of this manuscript
ORCID iD
Vikas Bhatia https:/orcid.org/0000-0003-3700-8025
References
- 1.Kok MM, Weernink MGM, von Birgelen C, et al. Patient preference for radial versus femoral vascular access for elective coronary procedures: the PREVAS study. Catheter Cardiovasc Interv 2018; 91(1): 17–24 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kaufmann TJ, Huston J, Mandrekar JN, et al. Complications of diagnostic cerebral angiography: evaluation of 19 826 consecutive patients. Radiology 2007; 243(3): 812–819. [DOI] [PubMed] [Google Scholar]
- 3.Gayed A, Yamada R, Bhatia S, et al. Society of interventional radiology quality improvement standards on radial artery access. J Vasc Interv Radiol 2021; 32(5): 761.e1–761.e21. [DOI] [PubMed] [Google Scholar]
- 4.Zussman BM, Tonetti DA, Stone J, et al. A prospective study of the transradial approach for diagnostic cerebral arteriography. J Neurointerv Surg 2019; 11(10): 1045–1049. [DOI] [PubMed] [Google Scholar]
- 5.Zussman BM, Tonetti DA, Stone J, et al. Maturing institutional experience with the transradial approach for diagnostic cerebral arteriography: overcoming the learning curve. J Neurointerv Surg 2019; 11: 1235–1238. [DOI] [PubMed] [Google Scholar]
- 6.Jo KW, Park SM, Kim SD, et al. Is transradial cerebral angiography feasible and safe? A single center’s experience. J Korean Neurosurg Soc 2010; 47(5): 332–337. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Knox JA, Alexander MD, McCoy DB, et al. Impact of aortic arch anatomy on technical performance and clinical outcomes in patients with acute ischemic stroke. Am J Neuroradiol 2020; 41(2): 268–273. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Khan NR, Peterson J, Dornbos D, III, et al. Predicting the degree of difficulty of the trans-radial approach in cerebral angiography. J Neurointerv Surg 2021; 13(6): 552–558. [DOI] [PubMed] [Google Scholar]
- 9.Deftereos S, Giannopoulos G, Raisakis K, et al. Moderate procedural sedation and opioid analgesia during transradial coronary interventions to prevent spasm: a prospective randomized study. JACC Cardiovasc Interv 2013; 6(3): 267–273. [DOI] [PubMed] [Google Scholar]
- 10.Seto AH, Roberts JS, Abu-Fadel MS, et al. Real-time ultrasound guidance facilitates transradial access: RAUST (Radial Artery Access with Ultrasound Trial). JACC Cardiovasc Interv 2015; 8(2): 283–291. [DOI] [PubMed] [Google Scholar]
- 11.Mason PJ, Shah B, Tamis-Holland JE, et al. An update on radial artery access and best practices for transradial coronary angiography and intervention in acute coronary syndrome: a scientific statement from the American heart association. Circ Cardiovasc Interv 2018; 11(9): e000035. [DOI] [PubMed] [Google Scholar]
- 12.Pancholy SB, Bernat I, Bertrand OF, et al. Prevention of radial artery occlusion after transradial catheterization: the PROPHET-II randomized trial. JACC Cardiovasc Interv 2016; 9(19): 1992–1999. [DOI] [PubMed] [Google Scholar]
- 13.Pancholy S, Coppola J, Patel T, et al. Prevention of radial artery occlusion - patent hemostasis evaluation trial (PROPHET study): a randomized comparison of traditional versus patency documented hemostasis after transradial catheterization. Catheter Cardiovasc Interv 2008; 72(3): 335–340. [DOI] [PubMed] [Google Scholar]
- 14.Nohara AM, Kallmes DF. Transradial cerebral angiography: technique and outcomes. Am J Neuroradiol 2003; 24(6): 1247–1250. [PMC free article] [PubMed] [Google Scholar]
- 15.Hadley C, Srinivasan V, Burkhardt J-K, et al. Forming the simmons catheter for cerebral angiography and neurointerventions via the transradial approach—techniques and operative videos. World Neurosurg 2021; 147: e351–e353. [DOI] [PubMed] [Google Scholar]
- 16.Kenawy AE, Tekle W, Hassan AE. Improved fluoroscopy and time efficiency with radial access for diagnostic cerebral angiography. J Neuroimaging 2021; 31(1): 67–70. [DOI] [PubMed] [Google Scholar]