In the field of cardiac resynchronization therapy is left ventricular pacing via the coronary sinus a mature technique

In the first reported case of cardiac resynchronization therapy (CRT) in 1994, left ventricular (LV) pacing was performed using an epicardial approach. 1 The use of a lead inserted into a coronary sinus (CS) tributary vein to pace the LV was initially proposed in patients in whom right ventricular (RV) pacing could not be performed by a transvenous approach and later by Blackburn and Gibbs 2 and Daubert et al.3 for cardiac resynchronization. The CS approach was used rather than the epicardial one because it was less invasive and did not require general anaesthesia in very delicate patients, but this new technique involved several difficulties: (i) to catheterize the CS in a usually marked dilated heart; (ii) to reach a CS tributary vein to pace the LV; (iii) to insure lead stability into the vein. The transvenous route aims to reach the target vein with an acceptable pacing threshold and without phrenic nerve stimulation in order to obtain the best mechanical resynchronization. During the first years of experience with LV pacing via the CS, implant success rates were low, ranging from 53% to 61%. Initially, conventional leads, designed for RV pacing, were used. Afterwards specific leads were designed to pace the LV via the CS. The development of new tools such as specific leads and guiding sheaths providing a better access to the CS and through the veins, helped to increase the success rate (82–98%). 3 , 4 Nowadays, in randomized trials, implant success rate ranges from 87% to 96%. 5 , 6 Most reported causes for implant failure are (i) the impossibility to catheterize the CS; (ii) the absence of a reachable vein; (iii) the lead instability; (iv) unacceptable electrical parameters and phrenic nerve stimulation. CS catheterization has been markedly facilitated by the development of guiding sheaths with various shapes to help the implanter according to the size of heart cavities and CS ostium position and orientation. Indeed, right atrium can sometimes be much dilated and CS ostium can be in an atypical position particularly in patients with a previous heart surgery. Otherwise, the CS anatomy is very different from one patient to the other with variable angulations of the veins with the CS and sometimes very tortuous veins. The development of ‘over the wire’ leads has made the entry of those veins easier. 7 , 8 Electrical obstacles such as high pacing thresholds and phrenic nerve stimulation may be solved by using bipolar leads and devices allowing different pacing vectors. The stability of the lead transvenously inserted remains an important issue. The lead dislodgment is not rare and can occur several months or years after implantation. One has to notice that a minor dislodgment may be responsible for a major increase in pacing thresholds or for the occurrence of phrenic nerve stimulation when it was initially absent. The need for re-interventions has decreased over time (25% at the beginning) but is still about 7–9%. In his paper Borleffs 9 reports a 7% need for CS lead re-intervention, similar to what is found in main trials and confirms that LV lead dislodgments could occur late after implant: 33% occurred after 6 months. He also questions about an important issue: can the problem be solved using an endovascular approach? In his series endovascular replacement using either the same lead or a new one was successful in 86% of the cases. The option of using the same lead may facilitate the procedure but the question is: will a lead that already dislodged be stable? In this paper the same lead was used in 30% of re-interventions with long-term stability (median 867 days). To alleviate the problem of lead stability pre-shaped and later, active fixation leads have been developed. But an important question worries implanters since the early use of these new leads: will it be possible to remove the LV lead from the CS particularly in the case of material infection? In older reports as well as in the one of Borleffs, no difficulties were met to remove LV leads even several months after implant but some describe the need for specific extraction tools. 4 , 9–11 Nevertheless, no data are available on extraction of active fixation LV leads. Beside those technical considerations, CRT aims to improve patients. In main trials there is a 20–30% rate of non-response to CRT. Non-response can be related to the selection of patients, the quality of resynchronization either possibly owing to the pacing sites or the lack of device optimization. Indeed, pacing sites are of major importance to ensure a good mechanical result. The mid-lateral wall is usually considered as the optimal LV pacing site but other authors suggest that optimal position may be different from one patient to the other. 12 , 13 Nevertheless, because of the technical difficulties we have described above, reaching the target site is not always feasible. Thus, despite great technical improvements, CRT using LV pacing via the CS is still often a compromise between theoretical optimal pacing sites and technical considerations. Further technical developments on LV pacing through the CS or towards new approaches are still necessary to improve CRT.