1. IntroductionTo date, right heart failure (RHF) complicates more than 40% of left ventricular assist device (LVAD) implantations in patients on extracorporeal membrane oxygenation (ECMO) support [1], quadruplicating early mortality [2,3], especially when an unplanned right ventricular assist device (RVAD) is required [4]. The need for a postoperative RVAD affects even long-term prognosis dramatically, both in terms of survival and freedom from RHF recurrence [4]. Moreover, the persistence of RHF 3 months after LVAD implantation has been demonstrated to negatively influence the 2-year survival of patients in the recent analysis of the Society of Thoracic Surgeons/Interagency Registry for Mechanically Assisted Circulatory Support (STS/INTERMACS) Database [5].Several risk scores for predicting RHF after LVAD implantation have been proposed, but their reliability in discriminating patients with severe manifestations of RHF is still suboptimal [6]. However, even mild to moderate post-implantation RHF can significantly increase the risk of adverse postoperative events, including death [7]. The parameters that are commonly shared by available risk scores usually entail the need for inotropes to sustain the circulation, INTERMACS class I and II, and borderline hemodynamic and echocardiographic measures of right ventricular (RV) performance. Unfortunately, this vulnerable population currently represents almost half of patients who undergo primary LVAD implantation [8]. In addition, being on ECMO support limits the predictive power of echocardiographic measurements, and when fine RV-pulmonary circulation coupling indexes are considered [9]. Thus, anticipating the need for a RVAD after cardiopulmonary bypass (CPB) weaning represents the most significant challenge during LVAD implantation in patients on ECMO support.

In this work, we propose a novel protocol for planning and performing RV support, using a single percutaneous ProtekDuo cannula (Livanova, London, UK), in ECMO patients scheduled for a durable LVAD (d-LVAD) implantation. Our approach provides a unique solution for conducting CPB and for a rapid subsequent switch to a planned temporary RVAD. The preliminary results of the first six patients treated with this strategy are presented.

3. ResultsIn the study period, we used the ProtekDuo Cannula during and after 10 LVAD implantations in ECMO patients. Among these cases, six patients (mean age 56 [13] years, 100% male) underwent d-LVAD implantation from ECMO using the ProtekDuo cannula intraoperatively and then postoperatively to provide RV support, as a Planned Combo Strategy. The other four patients (mean age 53 [11] years, 75% male) developed an unexpected RHF after three extracorporeal and one i-LVAD implantations, requiring an unplanned RVAD. We herein present the results of the new Planned Combo Strategy.Left heart failure etiology was ischemic dilated cardiomyopathy in three patients (50%) and primitive dilated cardiomyopathy in three (50%). In two cases (33%), the patient had already undergone a cardiac surgical procedure (aortic valve and ascending aorta replacement in one and single coronary artery bypass graft and mitral valve repair in the other). The mean EUROMACS-RHF score was 4.9 (1.2) and in all cases the patient had at least one criterion of impaired RV performance at echocardiography. Demographic and preoperative clinical characteristics are summarized in Table 1.Four patients (67%) were implanted with HeartMate III and two (33%) with HVAD. In five of six implantations, a minimally invasive technique was adopted: bi-thoracotomy in four cases and left anterior mini-thoracotomy + mini-sternotomy in one (Table 2).In all cases, the ProtekDuo cannula was correctly positioned without complications and it provided satisfactory RV support (mean maximal flow of 4.2 [0.6] L/min) for a median of 8 (4–16) days. In one case (17%), an oxygenator was included in the RVAD circuit for respiratory support while the pulmonary function was recovering from cardiogenic pulmonary edema. During RVAD support, major bleeding occurred in one patient (17%): a bronchial bleeding requiring prolonged mechanical invasive ventilation and bronchial toilettes. The ProtekDuo cannula was easily removed at bedside without complications in all cases. During hospitalization, other major complications were: the need for temporary tracheostomy in three cases (50%), new-onset acute renal injury requiring renal replacement therapy in two cases (33%), sepsis in one case (17%), and major ventricular arrhythmias in one case (17%, see Table 2). Overall 30-day and 90-day mortality was 0% (0/6) and 17% (1/6), respectively. The patient who expired had developed septic shock and multi-organ failure 58 days after RVAD removal. None of the patients developed RHF after RVAD discontinuation. 4. DiscussionThe upgrade from a temporary ECMO support to a d-LVAD in patients with cardiogenic shock requires a meticulous clinical evaluation since the outcomes in this challenging population are historically expected to be poor, especially when multiple risk factors are present [1,15,16]. One of the most important determinants of the patient’s prognosis is the occurrence of RHF [1,2,3], which can significantly increase early and long-term mortality, even in mild-to-moderate forms [7]. Therefore, it is not surprising that the unexpected need for RVAD after LVAD implantation almost halves the chances of survival [4]. In the present study, we hypothesized that the early adoption of a planned RV mechanical support in selected patients at high risk of post-implantation RHF could represent a promising strategy for ECMO patients scheduled to have d-LVADs. Moreover, we opted for the utilization of the percutaneous ProtekDuo cannula to provide minimally invasive intra- and postoperative RV support.The actual scores designed for predicting the risk of RHF after LVAD implantation are still limited by suboptimal reliability. Sert et al., specifically compared the predictive power of available scores and found that the EUROMACS-RHF score displayed the highest discrimination power in their cohort [6]. Moreover, the authors demonstrated that the specificity and sensibility of the RHF scores can be enhanced by combining them with other hemodynamic and echocardiographic parameters [6]. However, by unloading the RV, the ECMO support alters RV preload and its measured performance, impeding an accurate evaluation of RV function [9]. For these reasons, we designed an institutional protocol for ECMO patients scheduled for d-LVAD implantation that considers both the EUROMACS-RHF score and a qualitative echocardiographic evaluation of the RV during a temporary low-flow ECMO support (Figure 1 and Figure 2). When a moderate-to-high risk of RHF was present, a Planned Combo Strategy was adopted. This approach provides an early and planned instauration of RVAD support from CPB weaning, which is accomplished by using the ProtekDuo cannula as a part of the CPB circuit itself.Traditional techniques for RV support require a sternotomy or a thoracotomy, adding further invasiveness to the LVAD implantation procedure, limiting early mobilization of the patient, and increasing the sources of bleeding and infection [1,2,5]. In addition, when RV support is no longer necessary, the removal of the device requires surgical re-entry, or, in the case of a tunneled prosthetic graft, it entails prolonged retention of foreign material in-situ [17]. Recently, the ProtekDuo cannula has emerged as an attractive option for RV support in patients with left ventricular failure developing temporary RHF in different clinical settings (see Table 3) [11,12,13,18,19,20]. This double-lumen cannula offers all the well-known advantages of a totally percutaneous approach, together with the capability to provide >4 L per minute of blood flow. We routinely adopted a third-generation magnetically levitated continuous flow pump, with proven hemocompatibility and safety for medium-term support [21], and implementability with an oxygenator.Due to its favorable characteristics of safeness and efficacy, the ProtekDuo cannula has been adopted to treat the occurrence of RHF after LVAD implantation mainly as an unplanned strategy [11,12,13]. Since a planned instauration of right mechanical support after LVAD implantation has demonstrated clear advantages [3,4], we hypothesized an intraoperative adoption of the ProtekDuo cannula in selected patients at moderate-high risk of RHF: the Planned Combo Strategy. Utilizing the ProtekDuo cannula since the establishment of CPB and subsequently providing postoperative RV mechanical support in a planned manner has several potential advantages. Firstly, this strategy offers a total peripheral CPB, upgraded with a pulmonary decompressing vent (Figure 3), which is extremely helpful in minimally invasive LVAD implantations and in reinterventions. In ECMO patients, intraoperative mechanical support can even be provided by the ECMO itself, without transitioning to CPB, which was recently proved to shorten operative times and reduce the risk of bleeding [22]. This intraoperative support can be rapidly switched to a temporary postoperative RVAD. Biventricular support is thus initiated, reducing the need for inotropes to sustain the hemodynamics after the LVAD implantation, ensuring a physiological circulation across the lungs and full-flow support, which we have shown to ameliorate patients’ outcomes with respect to the standard ECMO cannulation [23]. If combined with a sternal-sparing LVAD implantation, the ProtekDuo cannula allows extremely rapid patient extubation and mobilization, with improved early outcomes [24,25,26]. Moreover, in the case of pulmonary edema, an oxygenator can be included in the RVAD circuit, until pulmonary function recovers.During biventricular support, major bleedings may affect 20–40% of patients [27,28] and our preliminary findings showed a 17% rate of bleedings (non-fatal bronchial hemorrhage) that required intervention during RVAD + LVAD support. On the contrary, thrombotic events never occurred during biventricular support in our experience. As demonstrated in isolated LVAD implantation [24,29], we speculate that a minimally invasive approach may further reduce the rates of these complications, although further data are needed in the setting of the Planned Combo Strategy.During RVAD support, RV performance is monitored constantly via seriate echocardiography. By progressively reducing the RVAD flow, the RV is allowed to gently preload, potentially reducing the risk of unexpected RHF. When RV recovery occurs, satisfying device weaning criteria [11,12], decannulation can be accomplished bedside, with only one single deep hemostatic stitch. We suggest a careful post-removal echocardiographic assessment, since novel severe tricuspid regurgitation may affect up to one-third of patients [11], although this complication never occurred in our small series. As shown by our echocardiographic data after RVAD removal (Table 2), RV function usually improves, as well as the severity of tricuspid valve regurgitation in most cases.Early mortality of ECMO patients with LVAD implantation can be as high as 24–27% [1,22,30]. We have previously shown that in-hospital mortality following ECLS instauration ranges from 20 to 47%, depending on the acute vs. chronic etiology of heart failure [16], the percentage of flow required to sustain the circulation [23], and the type of device adopted (ECMO vs. paracorporeal LVAD) [23]. In addition, having a high risk of RHF further negatively affects the patient’s prognosis [2,3,4,5], contributing to the creation of an extremely challenging setting for d-LVAD implantation. In the present preliminary series, 30-day and 90-day mortalities were 0% (0/6) and 17% (1/6), respectively. In other words, none of the patients developed RHF after RVAD removal and the only death was not cardiovascular-related. These findings align with previous reports in which a planned RVAD support with the ProtekDuo cannula was scheduled for patients undergoing d-LVAD implantation (Table 3). Noticeably, weaning rates from the RVAD and patients’ outcomes seem to be inferior when the ProtekDuo cannula is used as a rescue strategy if postoperative RV failure occurs [13]. Our preliminary work aims to stimulate a prospective collection of patients to support these findings on a larger scale.