Cost-effectiveness of cardiac resynchronization therapy

Implantable cardioverter defibrillators (ICD) represent the main tool for primary and secondary prevention of sudden cardiac death due to ventricular arrhythmias in selected patients affected from cardiac dysfunction; in fact, a relevant reduction in all-cause mortality was observed in numerous clinical trials on ICDs compared to optimal medical therapy^1–3. Despite their relevant contribution in survival improvement, ICDs per se have no impact on clinical outcomes related to congestive heart failure (CHF). Subjects with left ventricular conduction delay and consequent electromechanical ventricular dyssynchrony may develop ventricular dysfunction⁴. The advent of cardiac resynchronization therapy (CRT) ushered the era of device-based treatment of CHF on top of optimal medical regimen, considerably changing treatment and outcomes of patients suffering from CHF. In particular, CRT is a treatment of proven efficacy for patients with CHF, low left ventricular ejection fraction (LVEF <35%) and ventricular dyssynchrony due to intraventricular conduction delay, mainly left bundle branch block (LBBB, resulting in a QRS duration >130 ms). On top of optimal medical therapy, in such patients CRT may significantly improve New York Heart Association (NYHA) functional class, quality of life and survival; it promotes reverse left ventricular remodelling and reduces hospitalizations^5–8.

Current guidelines for CRT derive from large randomized clinical trials; in particular, the Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION)⁹ and the Cardiac Resynchronization-Heart Failure (CARE-HF)¹⁰ trial. Both studies demonstrated that biventricular stimulation reduced total mortality in this type of patients; subsequently, registry data and meta-analyses confirmed those results^11–13. Afterwards, CRT was tested on patients with mild CHF. In the Multicenter Automatic Defibrillator Implantation Trial with Cardiac Resynchronization Therapy (MADIT-CRT) CHF patients with LVEF of 30%, NYHA class I-II and a QRS duration >130 ms were enrolled, and total mortality in CRT group was 34% lower as compared to only medically treated patients¹⁴. Moreover, the Resynchronization-Defibrillation for Ambulatory Heart Failure (RAFT) trial reported that in patients with LVEF of 30%, a QRS duration >120 ms and NYHA class II-III, CRT reduced total mortality by 25%¹⁵. From MADIT-CRT¹⁴ and RAFT¹⁵ trials there were not sufficient data to recommend CRT in patients with NYHA class I or a QRS duration <120 ms; moreover, the EchoCRT trial¹⁶ demonstrated a trend to increased mortality in patients with narrow QRS complex, despite mechanical ventricular dyssynchrony on echocardiography. A similar conclusion was reported in the meta-analysis of Cunnington et al., that showed no benefit of CRT in non-LBBB patients¹². Atrial fibrillation (AF) was an exclusion criterion in the vast majority of CRT trials (except for the RAFT study), even though AF and CHF frequently coexist, and their association leads to a worse prognosis^17,18. Due to fast and irregular atrioventricular conduction, AF limits the benefit of CRT compared to sinus rhythm¹⁹; however, if AF conduction is limited by atrioventricular node ablation/modulation, the benefit of CRT was shown to be similar to patients without AF in non-randomized studies and meta-analysis^20,21. In general, one third of patients with the indication to CRT does not show a benefit over time (so-called “non-responders”)²²; however, there is no standard consensus on the definition of “responders” vs “non-responders”. Some authors consider clinical parameters (mortality, CHF hospitalizations and/or NYHA class)^4,23, others use echocardiographic measurements (increase in LVEF and decrease in end-systolic volume)²⁴, some other a combination of the two; even if there could be a discrepancy between clinical and echocardiographic response²⁵.

The decision to implant a CRT-device, and the choice between CRT-pacemaker (CRT-P) or CRT-defibrillator (CRT-D), may have important implications, either from an economic and a clinical point of view, in consideration of the increasing prevalence of CHF in general population. Similarly to ICDs, CRT-devices raise financial concerns as a consequence of high initial cost, that administrators of implanting centers have to face up when purchase the entire systems (devices and leads) from manufacturers. Despite the price reduction of CRT-D systems in the last few years, their relatively high cost in comparison with other medical devices (coronary stents, prosthetic cardiac valves, electrophysiology catheters, etc.) highlights the need to move from a purely financial perspective based merely on costs, to an economic perspective based on analysis of the relationship between costs and outcomes⁶. An early class II CRT-D implantation strategy totaled $95,292 compared with $91,511 for a late strategy. An early implantation strategy costs on average $3,781 for every year additional survival, resulting in an incremental cost-effectiveness ratio of $3,795/life year gained²⁶. The results of any cost-effectiveness analysis can be highly dependent on underlying assumptions of the model, and their applicability across different healthcare systems may be uncertain. Shah et al.²⁷ evaluated the cost-effectiveness of ICD, CRT-devices and optimal medical therapy in patients with CHF due to low LVEF. In their study the cost-effectiveness of ICDs was evaluated adapting the Mealing’s cost-effectiveness assessment²⁸ to United States real world evidences, considering baseline hospitalization risk and Medicare costs. Their analysis considered age, QRS duration, NYHA class, ischemic etiology, and the presence of LBBB. The authors found that CRT-D was the most cost-effective strategy in patients with CHF, low LVEF and LBBB ($100,000 per quality-adjusted life year willingness to pay). This observation fits well with current guidelines, that identify CRT-D as the main tool to counteract sudden cardiac death in patients with left ventricular dysfunction, wide QRS and optimized medical therapy. Notably, in clinical series analyzed by Shah et al.²⁷ the proportion of appropriately treated patients was high (70–85% on beta-blockers, 90% on ramipril, 75–80% on loop diuretic, 30% on spironolactone and eplerenone); this differs from what was observed in the REsynchronization reVErses Remodeling in Systolic left vEntricular dysfunction (REVERSE) trial²⁹, where the use of angiotensin converting enzyme inhibitors/angiotensin receptor blockers and beta-blockers (>95%) was high, but the proportion of patients receiving target dose of these medications was not reported.

The presence of combined conduction defects or need for permanent ventricular pacing may cause progression of left ventricular dysfunction over time: the biventricular vs. right ventricular pacing in patients with atrioventricular block heart failure (BLOCK-HF) ³⁰ showed that CRT-P resulted in better clinical and structural outcomes than right ventricular pacing in patients with atrioventricular block and left ventricular ejection fraction less than or equal to 50%; in these patients, CRT-P improved survival and reduced the progression of heart failure.

A cost-effectiveness analysis of CRT in these patients, performed by Chung et al.³¹, showed that patient survival was 6.78 years with right ventricular only stimulation and 7.52 years with CRT-P, with an increase in survival of 10.9%. In this setting, CRT-P allows an increase of 0.41 QALYs compared to right ventricular pacing alone, at an additional cost of $12,537. The Incremental Cost Effectiveness Ratio (ICER) was $30,860/QALY, which is below the acceptable threshold for the US ($50,000) and borderline for some European countries such as the UK (£20,000). The clinical and economic advantage was also evident in the group of patients with functional class NYHA I, for whom ICER observed was $43,687. From the data analysis of the BLOCK-AF study, Chen et al. (2019) predicted that patients who underwent CRT-D implantation will have greater benefits than patients who received a right ventricle-only stimulation system (0.84 vs. 0.49 years earned) in comparison to patients implanted with devices with only right ventricular pacing capabilities.

Even technical issues may lead to different outcomes in patients candidate to CRT implantation: there is strong evidence that coronary sinus (CS) lead implantation site can dramatically influence clinical response rate, and this may heavily influence the cost-effectiveness of CRT³². The rate of optimal CS lead implantation site in real-world practice is not yet well described, but it is likely lower than in rigorously conducted clinical trials, in which implantation procedures were conducted by experienced operators. Of consequence, CRT cost-effectiveness may be even better as optimal lead deployment will improve over time with advances in technology³².

To overcome this problem, a quadripolar catheter has been developed, which can provide more effective left ventricular stimulation over time, and prevent post-implantation complications such as an unmanageable increase in pacing threshold, phrenic nerve stimulation, or loss of capture caused by left ventricular pacing catheter displacement. Behar et al.³³ showed that patients with quadripolar implants had a lower rate of hospitalization than those with bipolar implants (42.6% vs. 55.4%; p = .002): this was due to the lower number of hospital readmissions for heart failure (16% for quadripolar implants vs. 26.1% for bipolar implants; p = .003) and replacement of generators (respectively 2.8% vs. 6.6%, p = .03). With lower clinical costs over time, despite a higher initial cost of the quadripolar catheter compared to the bipolar catheter, the incremental cost effectiveness ratio was favourable at £3,692 per QALYs earned (US $5,538).

Over the years, there was a considerable debate in the literature on whether patients who meet the criteria for CRT also need an ICD, or whether a CRT-P is an appropriate therapy. The Defibrillator Implantation in Patients with Nonischemic Systolic Heart Failure (DANISH) trial, in which it was observed that in patients with non-ischemic cardiomyopathy CRT-D was not superior to CRT-P (hazard ratio 0.91; p = .59), reinforced this concern³⁴. Similarly, Barra et al.³⁵ compared CRT-D to CRT-P in an extensive registry of 5,307 patients from Sweden, United Kingdom and France; evaluating the different effect of the two strategies in patients with ischemic vs. non-ischemic cardiac disease. In their study, the authors did not observe any additional benefit from ICD in patients with non-ischemic cardiomyopathy, whereas this was evident for ischemic patients. However, it should be noted that drug therapy and CRT-D reduced both CHF mortality and sudden cardiac mortality, while CRT-P reduced CHF deaths and deaths from other causes; in contrast, ICDs reduced only sudden cardiac mortality. Therefore, in selected patients with CHF combining optimal medical therapy with an appropriate cardiac implantable electronic device (CRT-P, ICD, or CRT-D) can reduce mortality by 60 to 80%³⁶. The addition of defibrillation capacity in a CRT device provides clinical and cost-effectiveness benefit only in categories of patients with an high risk of sudden death (>35%)^37,38, a sudden death rate ≥1.2%/year³⁹, and an annual mortality ≤25% (average survival of at least 3 years)^37,40–43.

Since cost effective data is needed to adjust investments in health care, the article by Saha et al. offers a real world view of interventions at high clinical impact. In line with the World Health Organization's suggestion⁴⁴, the economic thresholds applied in this analysis make these assessments applicable not only to the US, but also in other envirorments.

The development of new CRT technologies and their implementation may significantly impact its cost-effectiveness by both reducing costs and improving effectiveness. Technological innovations may include: implementation of device battery to prolong the longevity and to reduce the necessity of device replacements, dynamic pacing approaches (e.g. quadripolar leads and multisite pacing) that influence CRT clinical response, new implantation techniques for optimal lead positioning (e.g. active fixation CS leads), and novel device technologies (e.g. leadless devices) that reduce lead-related complications. Moreover, a more precise strategy to better identify CRT responders can orient the decision making to select patients who will derive the most benefit.

Transparency

Declaration of funding

There is no funding to declare for this manuscript.

Declaration of financial/other relationships

MB and MZ have received small fees as speakers for Medtronic, Boston Scientific and Biotronik.

JME peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Contribution statement

Authors contributed equally.

Acknowledgements

None reported.