**5. Discussion**

We have reported the results of our surgical treatment of severe patients with ICM and functional MR and described the details of our surgical strategy. Three kinds of SVR technique effectively reduced LV dimension and changed the spherical shape of the LV into an elliptical shape. Concomitant mitral valve surgery decreased the severity of MR during SVR. This combined surgery would contribute to better surgical outcomes for these patients. The final goal of SVR for ICM with functional MR is re-establishment of the geometric balance of the remodeled LV to increase the forward flow by obtaining concentric contraction and decreasing the extent of MR. We detected the akinetic region of the LV with various techniques and excluded it with three kinds of SVR based on the location of the region. Subsequently, the contractile myocardium was connected by the elliptical patch placed on the "contractility trail". Simultaneously, for patients with a dilated posterior LV wall between two papillary muscles, it was approximated during SVR to restore subvalvular geometry beneath the mitral valve. Although there is no gold standard technique for patients with ICM and functional MR, our combined surgery appears to achieve the final goal at this moment.

For patients with ischemic heart disease, SVR has yielded beneficial short-term effects on functional status, exercise performance, long-term results, and quality of life48,49). However, concomitant SVR is still controversial during CABG for these patients48,49,50). Recently, the Surgical Treatment for Ischemic Heart Failure (STICH) trial addressed this question and demonstrated that anatomical change by SVR was not associated with a greater improvement in symptoms or exercise tolerance or with a reduction in the rate of death or hospitalization for cardiac causes50). Patient selection issues and hemodynamic effects of LV volume reduction have been proposed to explain these contradictory results50). Thus, it would be very difficult to conclude anything about the efficacies associated with SVR, even though a large, multicenter, randomized trial such as STICH has been done. Especially for a small number of patients with ICM and functional MR, the same would be true.

More recently, we have suggested the effectiveness of SVR for patients with ICM51). According to our results, SVR is most effective when a >33% volume reduction rate achieves an LVESVI of <90 ml/m2. No long-term benefits occur when SVR induces an LV volume reduction of <15%, leaving a residual LVESVI >90 ml/m2. Although the results also contradict the STICH trial findings, long-term prognosis in ICM would be determined by the relationships between accurate methods for measuring ventricular volume and the extent of SVR volume reduction.

Due to the diverse patient population, it is very difficult to compare the surgical outcomes among clinical studies and trials. Although details of patients' background were disregarded, the cumulative survival rate was assessed by a systematic review of the literature associated with SVR in ischemic heart disease48). According to the review, the weighted average early mortality (defined as in-hospital or 30-day mortality) was 6.9%, and the cumulative 1-year and 5-year survivals were 88.5% and 71.5%, respectively. Although our surgical outcome did not reach the cumulative value, the extent of LV dysfunction with coexisting MR secondary to ischemia was much more severe in our series. More than 50% of patients had a large antero-septal akinetic region of the LV requiring the SAVE procedure, and all of them were classified as NYHA functional class III and IV. In fact, the remodeled hearts presented with severe LV dysfunction (EF <20%) with a dilated LV (LVESVI > 140 ml/m2). Moreover, more than half of the patients had concomitant severe MR (grade III and IV) in the present series. Earlier clinical reports demonstrated that the mortality risk is related to the degree of functional MR in patients with ICM52,53). Thus, our early and late surgical results would be acceptable in patients with such severe backgrounds.

Although SVR improved cardiac function and functional status for patients with ICM, it was reported that potential determinants of hospital mortality included preoperative advanced heart failure status, postoperative large LV volume (LVESVI > 60 ml/m2, LVESV > 80 ml), coexisting MR, and need for mitral valve surgery53,54). Many potential risks were involved in this series, and baseline LVESV would be much larger in a patient population with ICM and functional MR. In the present series, preoperative LVESVI (140±50 ml/m2) was larger than in other reports, and thus, postoperative LVESVI (104±42 ml/m2) was not included in the smaller LV volume category with low mortality. Although more exclusions to reduce LVESV would result in better surgical results, we believe that excessive exclusions involving contractile myocardium should be avoided for such ICM patients with severely dilated LV accompanying MR. Accordingly, prediction of the exclusion area of non-functional scar or myocardium is very important to perform effective SVR for these patients.

As one of the additional surgical adjuncts, we performed papillary muscle approximation to reduce LV volume for patients with a severely dilated LV requiring the SAVE procedure. The SAVE procedure effectively excludes a broad akinetic region of the antero-septo-apical wall, and papillary muscle approximation shortens the posterior wall between both papillary muscles. Thus, these combined procedures achieve further reduction of the LVESV. Although the volume reduction rate was increased by papillary muscle approximation, the early surgical effect on functional MR was almost the same, irrespective of papillary muscle approximation. Although the long-term effect on the LV dimension has not been elucidated, it may contribute to prevention of MR due to re-dilation of the LV.
