Does the Presence of Mitral Regurgitation Strengthen or Weaken the Indication for Cardiac Resynchronization Therapy?

This article refers to ‘Prognostic implications of mitral regurgitation in patients after cardiac resynchronization therapy’ by M. Cipriani et al. published in this issue on pages 1060–1068. Functional mitral regurgitation (FMR) is common in patients with heart failure (HF) because of changes in left ventricular (LV) volume and shape, dyssynchronous LV contraction and mitral annulus dilation.1 Substantial MR leads to chronic haemodynamic overload, progressive symptoms and is associated with poor clinical outcomes.2, 3 Cardiac resynchronization therapy (CRT) is a well-established treatment for patients with systolic HF and a wide QRS complex.4-10 Successful CRT improves dyssynchrony and reduces LV volume.7, 9, 11 Reductions in LV volume have been strongly associated with improved clinical outcomes.12 These favourable changes frequently lead to more effective mitral valve closure and therefore rapidly detectable improvement of FMR.13 In this issue of the European Journal of Heart Failure, Cipriani et al. have published the results of a prospective analysis of an Italian registry of patients who received a CRT defibrillator (CRT-D).14 They used this registry to identify 1122 patients with FMR. These patients were divided into two groups based on the severity of FMR on baseline echocardiography: patients without haemodynamically significant baseline FMR (MR–) and patients with haemodynamically significant baseline FMR (MR+). The MR– patients included those with absence of FMR (degree 0) and mild FMR (degree 1) and comprised 45% of the cohort. The MR+ patients were patients with moderate FMR (degree 2–3) and severe FMR (degree 4) and comprised the remaining 55%. Of these 1122 patients, 60 died during the first year and 146 were lost to follow-up. The remaining 916 patients were followed up approximately 1 year later with another echo to determine if their FMR had improved following CRT device implantation. A total of 235 patients (26%) showed evidence of worsened FMR and 681 (74%) had improved FMR. At this 1-year echocardiography, 57.2% of patients were classified as echocardiographic CRT responders based on at least a 10% reduction in LV end-systolic volume or a 10% increase in LV ejection fraction (EF). A positive echo response following CRT implantation was significantly lower in the worsened group compared with the improved one (41.8% vs. 62.3%, P < 0.001). These 916 patients were then followed up further for the clinical outcomes of all-cause and cardiovascular (CV) mortality. Patients classified as MR+ at baseline had poor outcomes with increased all-cause and CV mortality compared with patients classified as MR– at baseline. The mortality annual rate in the MR– group was just over half of that of the MR+ group. Furthermore, the patients whose FMR did not improve with CRT also had significantly worse prognosis regarding all-cause and CV mortality compared with patients without evidence of improvement. The authors concluded that the patients most likely to respond favourably to CRT were those with mild FMR. Patients with severe FMR as well as those who did not improve in the extent of FMR following CRT should be considered for alternative valvular treatment options. There are a number of important limitations to this study. This was a prospective analysis of a database and not a randomized controlled trial. There is no control group without CRT and therefore the results must be interpreted with this in mind. The study is limited to CRT-D Medtronic devices. Echocardiographic examinations were performed and reported on site by the operator and not evaluated in a core laboratory. Mitral regurgitation and LV volumes can be difficult to assess and quantify and the results reported are therefore highly dependent on image quality. More uniform criteria for assessing the extent of FMR and volumes and rigid control of the echo quality would have strengthened the study. Furthermore, the repeat echo was only performed after approximately 1 year. The positive effects of CRT on FMR can be seen rapidly,13 so it would have been preferable if frequent, short-term longitudinal data had been collected but, as noted, the database was not constructed to address the authors' hypotheses. The results demonstrate substantial improvement in mean EF in the improved group at 1 year. However, the reductions in LV end-systolic volumes observed at 1 year in this group are even more impressive. It is important to emphasize that haemodynamically significant MR increases LV volumes and may exaggerate LVEF determination. The interactions between EF, volumes, and the extent of MR are complex and the improvement in EF following CRT implantation may be somewhat underestimated in patients demonstrating a reduction in MR. An interesting finding in this study is the overall low mortality rate observed following CRT implantation. Despite the fact that >70% of the patients included in the analysis were classified as New York Heart Association (NYHA) functional class III or IV, the overall mortality rate was only 18% after 38.6 months. Mortality data from trials in patients with moderate or severe HF are higher.15-17 In the COMPANION trial,6 the mortality of patients with severe HF and a CRT-D was as high as 12% after 12 months. In the CARE-HF trial7 the mortality of patients with a CRT pacemaker and severe HF was 20% at 29 months. The lower mortality rate shown in this registry data would suggest that either the patients were not truly as symptomatic as reported or that the device and improved therapy was very effective in reducing mortality. Furthermore, based on their echo data at 1 year, patients were divided into FMR improved and FMR worsened. However, 47% of these patients actually had unchanged MR and rather than placing these patients into a third ‘MR unchanged’ category, they have somewhat arbitrarily divided this group into the improved or worsened group depending on their baseline severity of FMR. The authors do not explain the rationale for placing patients into only improved and worsened groups. Furthermore, the analysis does not include hospitalization data, which is unfortunate, as it would have added value to this report to have information related to morbidity in these patients. Despite the limitations mentioned, the study clearly demonstrates that having severe FMR is a poor prognostic indicator and that, if this FMR does not improve following CRT implantation, alternative treatment strategies should be considered to improve long-term prognosis. Should the extent of MR be taken into account when considering a CRT device? The simplest answer to this question is yes, if patients are otherwise eligible for a CRT based on EF, NYHA class, QRS duration, and clinical status.18, 19 Improvement or progression of FMR should be detected early during follow-up after CRT implantation by echo and the programme and management strategy adjusted accordingly. The take-home message is that the presence of substantial FMR before CRT implantation is associated with a poor prognosis. However, improvement in FMR following CRT reduces this risk substantially. Further clinical research should focus on identifying the predictors of FMR improvement following CRT. Conflict of interest: none declared.