Causes of death in heart failure could be sudden unexpected death due to arrhythmias leading to cardiac arrest or death due to progressive heart failure. In both situations, when drugs fail, devices come for rescue. Different types of devices are available for use in individuals with heart failure. They include implantable cardioverter defibrillator (ICD), heart failure device for biventricular pacing or cardiac resynchronization therapy (CRT-P), Combo devices incorporating an ICD with a CRT (CRT-D), ventricular assist device and finally a total artificial heart which is implantable. The current discussion will be focussing on ICDs and CRT-P/D.
Implantable cardioverter defibrillator (ICD)
Implantable cardioverter defibrillator is a device which automatically detects and treats life threatening cardiac arrhythmias. It has been well established that an ICD decreases mortality in survivors of cardiac arrest due to ventricular tachycardia/ventricular fibrillation (VT/VF). But only a minority of out-of-hospital arrests survive to reach hospital and still lesser number survive to hospital discharge. Hence the role of ICD in primary prevention of mortality due to cardiac arrest needs to be considered.
Even though the ICD is a life saver, it is an expensive device and not devoid of complications. Inappropriate shocks can occur in 21 – 25% of cases and is mostly due to supraventricular arrhythmia. Some of this can be reduced by appropriate programming and improving the sensing modalities. But these shocks can by psychologically quite traumatic, even to the extent that some individuals demand explanting of the device. Other procedure related complications like pneumothorax, hemothorax, infection, lead dislodgement and cardiac tamponade are fortunately much less common, with overall incidence less than 5%. To have a high cost benefit ratio, we need to evaluate the indication for ICD quite well.
Most important risk factor for sudden death in heart failure, is the extent of myocardial damage. The risk of cardiac arrest increases moderately with left ventricular ejection fraction (LVEF) below 40%. More significant increase is noted with LVEF < 30%. ICD implantation is justified in those with significant left ventricular dysfunction despite optimal medical therapy, for prevention of sudden cardiac death (SCD).
Various other predictors of risk of SCD in heart failure have been evaluated. They include the extent of heart failure, QRS width, neurohormone levels, inducibility of ventricular arrhythmias at electrophysiological (EP) testing, heart rate variability and T-wave alternans. But the positive and negative predictive values of these additional predictors have not been consistently good.
There have been many randomized controlled trials supporting the use of ICDs in primary prevention. They include MADIT I & II, MUSTT EP arm, DEFINITE and SCD-HeFT. The negative trials were CABG-PATCH and DINAMIT.
Cardiac Resynchronization Therapy
The other important device used in heart failure situation is the CRT device for cardiac resynchronization therapy. Conduction defects in heart failure cause decrease in contractile force due dyssynchrony of the ventricular conduction, development or prolongation of mitral regurgitation and wasted cardiac work due to mechanical asynchrony. The various levels of asynchrony could be intraventricular asynchrony with a regional contractile phase delay like that between the anteroseptal and the posterolateral left ventricle, interventricular asynchrony with a phase delay in contraction between right and left ventricles and atrioventricular asynchrony.
In cardiac resynchronization therapy (CRT), both the left and right ventricles are paced in an AV sequential manner. The right ventricle is paced by an endocardial lead while the left ventricle is paced using an epicardial lead introduced into a posterolateral vein through the coronary sinus. This leads to an improvement in the left ventricular regional wall motion and lowers left ventricular end-systolic volume. The device also optimizes atrioventricular delay. There is a decrease in mitral regurgitation and increase in the diastolic filling time with CRT. The ventricular pressure volume loop shows a shift to left with effective CRT.
Several randomized controlled trials have documented the efficacy of CRT. They include MUSTIC, MIRACLE, COMPANION and CARE-HF. These trials included those symptomatic despite optimal medical therapy and with QRS duration more than 120 ms, LVEF less than 0.35 and who were in New York Heart Association (NYHA) functional class III or stable class IV. All these trials showed improvement in 6-minute walk time, oxygen consumption and reduce frequency of hospitalizations for heart failure. Reduction in all-cause mortality was noted in CARE-HF trial involving 813 patients.
A non responder rate of around one third has been the disadvantage of CRT. It is often difficult to predict the non-responder before CRT implantation, in spite of various methods being available. It is also interesting that technological advances over a decade has hardly influenced the non responder rate with 67% responded in MIRACLE as well as in FREEDOM trial 8 years later. PROSPECT trial found a higher response rate of 76% in non-ischemic patients versus 64% of ischemic patients. The favourable factors which predict a good response include left bundle branch block (LBBB) with QRS duration of more than 150 milliseconds, presence of atrioventricular synchrony, moderately dilated left ventricle, mild to moderate mitral regurgitation, non ischemic cardiomyopathy and female sex. The less favourable factors include intraventricular conduction delay, right bundle branch block, QRS duration less than 150 milliseconds, permanent atrial fibrillation, massively dilated left ventricle, severe mitral regurgitation, ischemic cardiomyopathy with large anteroseptal or posterior scar and male sex. Stimulation from a scarred area may not be as effective as from a non scarred region. Various echocardiographic parameters have been described, to assess dyssynchrony and to choose the ideal site for biventricular pacing, with variable results. But in the PROSPECT trial, echo measures of ventricular dyssynchrony unable to distinguish responders from nonresponders. In RethinQ trial, CRT in patients with heart failure and narrow QRS with dyssynchrony on echo were evaluated. But no difference in quality of life (QOL), oxygen consumption, or heart failure events could be documented.
Studies on Combo device – CRT-D
Combo devices combine an ICD with CRT and they are designated as CRT-D. REVERSE and MADIT-CRT trials showed improvement in heart failure events and death combined, but not in death alone. RAFT study showed an absolute reduction in mortality of 6% over 5 yr (CRT-D vs ICD). Greater survival benefit was noted if patients had LBBB with QRS duration more than 150 ms.
Novel approaches for CRT
Several novel approaches for left ventricular pacing in those not suitable for conventional epicardial pacing through the coronary veins. Transseptal approach for left ventricular epicardial pacing has been tried. The main concern in left ventricular endocardial pacing is the potential for thromboembolism and the need for anticoagulation. In transseptal approach, septal puncture can be done using the conventional femoral route and the lead introduced from the subclavian route. Septal dilation using a special sheath is needed for introducing the lead into the left ventricle through the left atrium. Chance of mitral regurgitation in the long run is an additional concern with this approach.
Trans aortic left ventricular endocardial stimulation has been tried in experimental models. This has been suggested by the occasional cases of inadvertent left ventricular pacing going on without much complications. Only trivial aortic regurgitation occurred in the few experimental cases tried. Trans apical left ventricular endocardial stimulation has also been described in cases in which other approaches had failed. Following a limited thoracotomy left ventricular puncture was done with the Seldinger technique and the leads introduced and fixed under fluoroscopic technique using screw in leads. There is no risk of mitral regurgitation as there is no lead across the mitral valve. Larger studies are needed before we can assess the long term efficacy of this approach.
Single site and triple site pacing are the other approaches which have been tried. Left univentricular pacing (single site pacing) instead of biventricular pacing with the left ventricular lead placed in the region of most delayed activation has been found useful in small series. Some of the non responders to biventricular pacing had responded to left univentricular pacing and vice versa. Dual site pacing of the left ventricle along with right ventricular pacing (triple site pacing) is also another option which has been tried in those who failed dual site LV-RV biventricular pacing.