Right ventricular failure after surgical ventricular restoration: operation or myocardial protection problem?

The thoughtful report by Couperus et al. [1] indicates that surgical ventricular restoration (SVR) of failing dilated ventricles produces significant improvement in left ventricular (LV) function but causes right ventricular (RV) failure in 40% of patients—together with LV diastolic dysfunction. Their technical outcome was excellent, as SVR caused a 46% reduction in LV end-systolic volume (LVESVI). They question whether improvement in the left ventricle’s contribution to cardiac output could impair RV performance, because pulmonary artery pressure increased after interacting with new or worsened diastolic dysfunction. Alternatively, they wondered whether cardiac surgery itself could be responsible, despite its LV performance benefit. Any argument that blames LV success for RV failure is surprising, because the left ventricle and the right ventricle interacted normally in health, and the operation itself induced or worsened the diastolic dysfunction. To understand this disastrous complication, the mechanical determinants of RV function and how the SVR procedure impacts it must be analysed. The geometric factors responsible for RV performance share a common trait with the SVR procedure, since its changing of the spherical ventricle into elliptical form creates the architectural reasons for why LVESVI and New York Heart Association (NYHA) scores decreased, and why ejection fraction and survival increased. SVR can return twisting [2], which is the predominant force for RV performance [3]—a motion related to fibre orientation within the ventricular wall [4]. The helical ventricular myocardial band architecture clarifies the determinants of RV compression by its free wall and twisting due to its septum [3]. The RV free wall is formed by a wrap of transverse fibres that produce constriction or compression, whereas oblique helical fibres form the septum that will twist as they coil to shorten (Fig. 1). These muscle masses make vastly different functional contributions, because compression plays a minor role (20–30%) [5] and twisting accounts for 80% of RV performance [6]. ‘The septum is the lion of the RV’ [3], and its twisting action is quantified from longitudinal strain and tricuspid annular plane systolic excursion (TAPSE) measurements. Right ventricular structural components that include a circumferential wrap of transverse fibres around the free wall, a septum that comprised helical fibres. (A) Upper left is intact heart, (B) upper right is unfolding of free wall wrap and (C) lower image displays the helical muscle forming the septum. This dynamic balance is evident, as free wall exclusion by cauterization [7], excision and replacement [8] or ‘isolated’ ventricular fibrillation does not alter RV function [9], but RV failure follows pulmonary hypertension when the septum is damaged [10]. Their positive statistical correlations between fractional area change and RV failure only shows the secondary stretch of the free wall, as it expands due to septum dysfunction. The septum weighs 35–40% of the total ventricular weight and 50% of the LV weight [11], and its damage causes RV impairment [3] (Fig. 2). Clinical insight of its injury emerges from the reduced longitudinal strain (septum) and increased circumferential strain (from the wrap’s compression that compensates for impaired shortening) that follows coronary artery bypass grafting [12] and surgical aortic valve replacement [13]. This complication reduces twisting, ‘but does not occur if cardioplegia is not used’ [13]. This impediment sets the groundwork for using non-surgical transcatheter aortic valve replacement or mitral clips procedures, because ‘induced septum dysfunction does not occur because cardioplegia is not delivered’ [14]. (A) Separation of structures described in conventional architectural terms into RV, LV and septum. (B) Table shows that the septum weighs 35% of total ventricular weight, and it occupies 50% of LV weight. Reproduced with permission from Buckberg et al. [11]. LV: left ventricle; RV: right ventricle. Septum dysfunction is commonplace, and its high 1970s incidence [15] persists as —40% of 3292 cardiac patients developed paradoxical motion, rising to 60% in valve patients [16]. Prolonged aortic clamping was believed to be causative and the current authors approximately 130-min clamping interval fits into this category. Unfortunately, the extent of septum damage is unknown, because hypokinesia or akinesia is not reported. I believe that septum damage reflected reperfusion injury, and this concept was confirmed in 131 consecutive patients receiving integrated myocardial protection; their septum performance remained normal, despite up to 155 min of aortic clamping [11]. Cardioplegia strategy methods were critical when we compared integrated protection against patients receiving the same cardioplegic solution but underwent different delivery. Septum shortening (determined from longitudinal strain) ‘was augmented by the integrated strategy but became consistently impaired without it’ [17] (Fig. 3). Longitudinal ventricular strain before and after the use of integrated cardioplegia delivery or only antegrade delivery. All patients received the same solutions. Note (i) lower preoperative strain in patients receiving integrated myocardial protection, (ii) improved strain after integrated delivery and (iii) reduced strain when only antegrade delivery used. Reproduced with permission from Bhaya et al. [17]. MAS: mid-anteroseptal region; Preop: preoperatively; Postop: postoperatively. The commonplace event of septum damage is globally accepted, even though it reflects impaired myocardial protection. We addressed this dilemma in the article titled ‘Cardioplegia: Solutions or Strategies?’ [18], but this submission was rejected by the Journal of Thoracic and Cardiovascular Surgery and The Annals of Thoracic Surgery; yet, astoundingly, ‘not one of the 7 reviewers stated that their protection approach avoided septum damage’. Physicist Neil de Grasse Tyson stated that ‘the good thing about science is that it’s true whether or not you believe in it’. The European Journal of Cardiothoracic Surgery published our report [18]. This tale resembles a ‘sweeping it under the rug’ approach that hides a substantial septum injury, one that has a profound effect when pulmonary hypertension is present. This has enormous implications, because understanding the mechanisms of RV failure after SVR will ‘also’ explain how and why diastolic dysfunction interfaces with septum damage. Conventional anatomy is topographical. It correctly divides heart architecture into right ventricles, left ventricles and septum but has no functional interface. The helical ventricular myocardial band shows how the helix and circumferential wrap defines a heart architecture and furnishes a full explanation of function [3, 19–21]. Damage to its inner descending helical arm ‘equally affects’ the septum and LV free wall, because the same helix forms both regions (Fig. 4); their injury impairs the recoil needed for ventricular suction. The coherency of septum and diastolic function means that ‘the same disease is due to the same muscle and has same cause’; the measured dysfunction simply reflects ‘a symptom of inner descending helical arm injury’. Modification of anatomy in Fig. 1, with added ‘simulated left anterior descending artery’. Architectural image shows this vessel is a vascular highway perched on top of the helix that constructs the septum and the left ventricular (LV) free wall. (A and B) Upper images show only anatomy. (C and D) Lower ones show LV free wall on the vessel’s left side and septum to its right (but 3 dimensionally deeper). Note that the entire oblique muscle masses that construct the septum and LV free wall have the same architectural helical components. This combination exists clinically but remains unrecognized; approximately 50% incidence of septum damage [16] is matched by its sequel of diastolic dysfunction that exists in 44–75% of coronary grafting or aortic valve procedures [22]. This combined dilemma is especially powerful after SVR, since diastolic dysfunction increases pulmonary pressure, yet the injured septum prevents the development of twisting to overcome increased vascular resistance. Conversely, these predictable LV and RV consequences of inadequate protection of the descending helical arm are absent [11] if protection is adequate. Technical precision and absence of cardiac damage remain the 2 primary surgical goals. Satisfactory accomplishment of the first task is clear from their 46% reduction in LVESVI. This outcome dramatically differs from the flawed STICH trial—where only 19% LVESVI reduction occurred, scar was not documented in 72%, no volume was measured in 44%, SVR was done in 25% with ejection fraction >35% and no volume reduction occurred in 19% [23]. The aftermath of these grave problems in STICH’s scientific integrity has restricted the emergence of SVR for the huge ischaemic cardiomyopathy population. The authors only used SVR in LV aneurysm patients, yet aneurysm accounts for approximately 1% of dilated heart patients because of the universal use of angioplasty of thrombolysis. Acute revascularization has changed dyskinesia into akinesia, but the segment’s contractile function does not return if >50% of muscle is necrotic [24]. Consequently, ‘akinetic hearts will profit from SVR, provided there is no descending helical arm injury’. Insight into these benefits is evident in their Fig. 3 that shows nearly 80% 5-year survival when RV dysfunction is absent versus 50% survival with RV dysfunction. These salutary outcomes mirror Isomura et al.’s [25] report of 84% survival in 8 years when SVR rebuilt the elliptical form rather than just excluding the scar. These laudatory results are simply impossible to achieve following medical treatment of Class III/IV patients, which have high mortality ‘with or without RV failure’. The powerful importance of surgical rebuilding is appreciated by our cardiology colleagues who consistently follow our successful treatment pathways. This sequence started with angioplasty, was followed by mitral clips, and now takes hold with device-related LV rebuilding [26]. I believe septum and diastolic dysfunction injury is caused by impaired myocardial protection, and their induced presence will hamper our surgical approaches, because non-surgical approaches ‘without cardioplegia’ do not cause these complications. For this reason, my conclusion dramatically differs from the authors’ suggestion to abandon SVR when RV dysfunction is present; they favour transplantation or LV assist devices, whose limitations include uncommon donors and development of bleeding, clotting and organ-related problems. Anatomy is our friend, not our enemy, so that SVR is only contraindicated by total septal scarring, and dead muscle cannot restore twisting. Conversely, a stretched viable septum will regain contractile function after its helical shape is restored by SVR, if septum and LV free wall injury is avoided using effective myocardial protection strategies. I believe their last causative factor is correct—‘a problem with the surgery’. Their technical excellence is clear, but they must consider how to improve myocardial protection. They are well suited to test this, especially because their careful and meticulous database sets the stage for them using SVR together with a protection technique that avoids septum damage. Adequacy of protection can be easily confirmed in non-SVR patients undergoing coronary and/or valve disease. On confirming the absence of septum injury, SVR can also be done for akinesia. Their fresh data may help future decision making. Conflict of interest: none declared.