Multi ‐ lead cephalic venous access and long ‐ term performance of high ‐ voltage leads

Background: Cardiac resynchronization therapy ‐ defibrillator (CRT ‐ D) implantation via the cephalic Conclusion: CRT ‐ D implantation via the cephalic route is not significantly associated with premature ICD lead failure. Female gender and BMI are predictors of lead failure. via usually by subclavian or axillary puncture and by cephalic vein cut ‐ down. There is no standard approach, but cephalic access feasible, effective, and safe implanted. 4 Traditional lateral puncture has been associated with a higher rate of lead failure than the use of cephalic venous access 5,6 but a recent report has suggested that multi ‐ lead defibrillator therapy utilizing the cephalic route is associated with early implantable cardioverter ‐ defibrillator (ICD) lead failure. were the cephalic, non ‐ cephalic routes. The procedure was also noted to be longer in the non ‐ cephalic procedures. there was no significant difference in the number of ICD lead failures between the two groups ( p =.29).

Conclusion: CRT-D implantation via the cephalic route is not significantly associated with premature ICD lead failure. Female gender and BMI are predictors of lead failure.

K E Y W O R D S
cephalic, CRT-D, implantable cardioverter defibrillator, lead failure, venous access

| INTRODUCTION
The cardiac resynchronization therapy-defibrillator (CRT-D) is an established treatment for heart failure which reduces morbidity and mortality. 1 The pacing leads are predominantly implanted via the transvenous approach, usually by subclavian or axillary puncture and sometimes by cephalic vein cut-down. There is no standard approach, but cephalic access is feasible, effective, and safe 2,3 irrespective of the number of leads being implanted. 4 Traditional lateral subclavian vein puncture has been associated with a higher rate of lead failure than the use of cephalic venous access 5,6 but a recent report has suggested that multi-lead defibrillator therapy utilizing the cephalic route is associated with early implantable cardioverter-defibrillator (ICD) lead failure. 7  Cephalic cut-down was preferred by most operators and has been reported previously. 2,4 Following isolation of the cephalic vein with blunt dissection close to the level of the coracoid process, the distal end of the vein was tied and a precise transverse venotomy was performed at this anatomical level. Two standard 50 cm 0.97 mm guidewires and a 150 cm angled 0.97 mm hydrophilic guidewire (Radiofocus RF*GA35153M, Terumo Corporation) were introduced and advanced towards the heart. Initially, the coronary sinus (CS) delivery system was advanced over the hydrophilic guidewire, to position the CS lead (with contrast venogram) in a suitable cardiac vein. Peel-away sheaths were then used to position an atrial lead (7 fr sheath) and an ICD ventricular lead (9 fr sheath). When this vessel was too small to allow access to all the leads, either the axillary or subclavian was used for the remainder. It was at the operator's discretion to decide which lead to place by other routes of access.

| METHODS
Pacing interrogation was performed within 24 h after implantation, at 6 weeks, and subsequently at 6-month intervals. Patients F I G U R E 1 Bar graph demonstrating the distribution of implant routes for other leads according to the implantation route of the highvoltage lead (A, B) and the distribution of access route for high voltage leads according to lead model (C). All systems included a left ventricular lead; 91% included an atrial lead. (A) When the high-voltage lead was implanted via the cephalic vein, the other leads followed the same route in 98.7% of cases. (B) When the ICD lead was non-cephalic, 78.6% of the other leads were also non-cephalic. (C) The Sprint Quattro (Medtronic) and the Endotak Reliance (Boston Scientific) leads were used in the majority of our population with a similar distribution of cephalic and non-cephalic access used across all lead models. ICD, implantable cardioverter-defibrillator with defibrillator leads implanted via the cephalic vein were categorized as the "cephalic" group and patients in whom the defibrillator lead was implanted utilizing the subclavian or axillary veins, were categorized as the "non-cephalic" group.

| Lead failure
Lead failure was defined as per the Heart Rhythm Society consensus. 8 High-voltage leads were considered to have failed if they exhibited: persistent oversensing of non-physiological rapid signals, the abnormal impedance in the pace/sense or the shock component, an increase in right-ventricular lead threshold, and/or decrease in sensing sufficient to make the lead unreliable. 8 All leads that met these criteria were extracted and replaced; all were inspected carefully before and after extraction. Lead extraction for infection and lead revision for displacement were considered separately. Radiological images from the time of implantation were inspected for all leads that subsequently failed.

| Statistical analysis
Continuous variables were conveyed as a mean ± standard deviation and median with interquartile range (IQR), whilst categorical variables were presented as a number and percentage. Statistical analysis was performed using a χ 2 test and an independent t test. A p < .05 was considered significant. Lead longevity was analyzed using the Kaplan-Meier model and risk factors were compared using a univariate, multi-variate, and Cox regression analysis.

| RESULTS
Over the study period, 633 patients underwent CRT-D therapy and were included in the analysis. In most cases (73.5%), the high-voltage lead was implanted via the cephalic vein (cephalic group) while the remainder were axillary (21.9%) and subclavian (4.6%) access; the majority (50.3%) of the implanted ICD leads were Sprint Quattro (Medtronic) (Figure 1).

| DISCUSSION
The current series is the largest to date evaluating the relationship between multi-lead defibrillator therapy delivered via the cephalic vein, and ICD lead failure. We found a very low This multi-center study reported a very low overall lead failure rate which is at odds with some previous series 7 but is validated by a previous large series (0.45%/year). 9 This low failure rate may reflect our conservative practice: a policy of concentrating on products with a track record of long-term safety and late adoption of less tested technology. The higher incidence of failure in the prior literature may represent a publication bias; it is reasonable to suppose that colleagues are more likely to report an unsatisfactory experience than to describe a lead performance that is in line with expectation.
This series demonstrates that cephalic vein access for multi-lead defibrillator therapy does not affect ICD lead longevity: the rate of lead failure was similarly low for cephalic and non-cephalic routes (0.36%/year vs. 0.13%/year; p = .12). This is in stark contrast to a recent report by Barbhaiya et al. 7 which found that cephalic access was significantly associated with a high rate of lead failure in multi-lead ICD therapy (11% per year for non-Linox and 19% per year for Linox leads).
There are significant differences between the two reports. Their method was to implant a maximum of two leads via the cephalic vein, whereas most of our patients received three leads by this route. Barbhaiya  We believe that many small technical 11 and methodological differences could play a role in lead durability: for example, our policy is to place all leads via peel-away sheaths to protect the tip from stress produced by passage through a tortuous cephalic vein. In our series, 75% of the operators would be considered as "cephalicoperators" with a similar well-honed technique and experience in accessing this vein, maintaining consistency, and minimizing error. It can be argued this may skew the findings, but it represents a multiinstitutional experience that is consistent and reproducible with favorable outcomes.
Our population included a low proportion of leads that are associated with a high rate of failure ( Table 4). The Sprint Fidelis (Medtronic) had been withdrawn before the study recruitment period; the Riata was seldom used in our centers. The Linox (Biotronik) which has been associated with an elevated failure rate in some series 12,13 but not all 7,14 demonstrated similar performance to other models, but the comparison is underpowered.
The Cox regression analysis found that the venous access route does not predict lead failure (p = .5), in keeping with prior reports. 15 Consistent with previous findings, we found leads implanted in women were much more likely to fail 16

| Limitations
This study was a retrospective analysis and therefore open to bias from confounding variables. Remote monitoring was available only for a minority of devices across the study period and may result in

| CONCLUSION
Cephalic vein access for multi-lead ICD therapy is not a significant risk factor for lead failure in the long-term. Our data confirm that female gender and BMI are predictors of lead failure.