Radial Artery Occlusion After Endovascular Procedure

Trial objectives and purposes

Our objective was to evaluate:

i. relationship between activated clotting time (ACT), measured before sheath removal and incidence of radial artery occlusion (RAO), assessed by continuous Doppler and confirmed by echoDoppler before discharge from the hospital.

ii. the relationship between RAO and other clinical and procedural characteristics, to assess independent predictors of RAO.

Trial design:

Observational cohort registry: prospective, multicenter independent evaluation of patients who undergo coronary angiography and/or PCI via radial artery. The purpose of this registry is to set the incidence of RAO according to ACT values, considered both as continuous and ordinal variable and its determinants. Coronary angiography and PCI will be performed according to usual practice.

Patient population:

All patients referred for coronary angiography and/or with planned radial access will be invited to participate.

If the radial puncture is not possible (precluding the use of radial artery as vascular access) the patient will be excluded from the registry.

Primary end-point:

incidence of RAO before discharge.

Secondary end-points:

i. incidence of symptoms related to vascular access ii. incidence of significant forearm hematoma iii. incidence of pseudoaneurysm at site of vascular access iv. incidence of procedural success v. MACCEs at 12 months

Definition:

i. RAO: no flow detectable at continous Doppler at the site of vascular access, confirmed by echoDoppler; ii. symptoms: hand pain or paresthesiae after radial cannulation; iii. hematoma: local bleeding extending > 15 cm2 on forearm iv. pseudoaneurysm: local ectasia of vessel wall not involving all the three layers at the site of puncture, diagnosed by 2D echoDoppler; v. procedural success: insertion of vascular sheath in the radial artery vi. Major cardiovascular cerebral events (MACCEs): death, cardiovascular death, myocardial infarction, any unplanned coronary revascularization, stroke/transient ischemic attack (TIA).

Subjects:

2168 consecutive patients in 5 centers (Divisione di Cardiologia, Ospedale Civile, Mirano (VE); Divisione di Cardiologia, Ospedale Civile, Conegliano (TV); Divisione clinicizzata di Cardiologia, Azienda ospedaliera-universitaria, Verona (VR); Divisione di Cardiologia, Ospedale Mater Salutis, Legnago (VR), Divisione di Cardiologia, Ospedale Santa Maria della Misericordia, Rovigo (RO), Istituto Clinico Humanitas, Milano) will be enrolled during a study period of 12 months. An interim analysis at 1084 patients will be performed. Inclusion and exclusion criteria will be applied to suitability for enrollment.

Evaluations:

Assessment of patency: whitin 24 hours from end of procedure or before discharge from the hospital, a physician blinded to ACT values will assess presence of flow along the radial artery with continous Doppler probe or with color Doppler probe (patency of radial artery will be considered if antegrade flow is present during contemporary compression of ulnar artery at the wrist); if a Doppler probe is used, a confirmation of occlusion with color Doppler must be obtained (see Appendix); in case of weak/biphasic flow with Doppler, an evaluation with color Doppler must be obtained to detect a segmental stenosis. Physicians will perform Doppler and echoDoppler evaluation after adequate training. Anonymized echoDoppler images will be recorded and send to referral center where the angiologist will review the images confirming the RAO.

Follow-up: if RAO occurs, 1 hour occlusion of homolateral ulnar artery could be attempted in order to increase flow and restore patency in occluded radial artery (25); if RAO persists, LMWH (100 u/Kg x 2/die) should be administered for 14-30 days (6). EchoDoppler follow-up at 1-3 months will be performed to check patency of occluded radial artery.

Long-term follow-up: a 12 months telephone call will be performed in order to assess MACCEs (death, cardiovascular death, myocardial infarction, stroke/TIA)

Statistics and data analysis:

The null hypothesis is that degree of anticoagulation (as assessed by ACT) has not relation with RAO. Alternative hypothesis is that incidence of RAO varies according to ACT values. Assuming a 6% of incidence of RAO at 24 hours, 95% confidence intervals would be 5% to 7% for a 2168-patient sample (110-151); PCIs must be at least 50% of the procedures, and enrollment should not stopped until at least 110 primary events (RAO) have occurred. An-interim analysis at 1084 patients enrolled will be performed. This sample size should allow evaluation of correlation of ACT with RAO both as continuous and ordinal variable. A statistical model (non-linear regression) will be fitted to the relation between RAO and ACT values. Receiver-operating characteristic (ROC) cut-off values will be obtained according to the best compromise between sensitivity and specificity, and a threshold effect will be investigated. To adjust for confounders, multivariate logistic regression analysis will be performed, considering the presence of RAO as dependent variable. In addition to ACT values (setting the group with lowest incidence as reference), other 4 variables known to be related with RAO will be entered into the model (heparin dosage, postprocedural compression time, patent hemostasis, sheath size) and odds ratios with 95% confidence intervals will be calculated. An exploratory analysis regarding possible role of other variables (sex, spasmolytic drugs, procedural time, multiple procedures, hemostasis device, pressure before sheath removal, antiGP IIb/IIIa drugs) will be performed. Pre-planned analysis of RAO incidence according to ACT values will be performed in these subgroups: patients who underwent coronary angiography vs patients who underwent PCI, patients in whom patent hemostasis was possible vs patients in whom patent hemostasis was not possible, patients with acute coronary syndrome vs without. Kaplan-Meier curves of MACCEs within 12 months after procedures according to RAO occurence will be plot and compared using log-rank statistics and Cox regression, adjusting for age, acute coronary syndrome and diabetes. Normality of data distribution will be assessed by skewness and kurtosis and by linear plot graph. Data will be reported as mean±standard deviation or median and interquartile range according to distribution. Chi-squared statistics will be used for categorical variables, and the t-test or Mann-Whitney U test for scale variables if data followed normal or non-normal distribution, respectively. Post-test Bonferroni correction and Dunn's test will be used for pairwise comparisons among groups. Statistical significance will be set at p<0.05 (two-tailed) for all tests. All data will be analysed using Prism (version 5, Graphpad, California) or SPSS software (version 20.0; SPSS Inc., Chicago, IL, USA).