Right ventricular mechanical support for pulmonary arterial hypertension. The low flow concept.

Pulmonary arterial hypertension (PAH) is caused by a pathological increase in pulmonary vascular resistance (PVR) and ultimately leads to right ventricular (RV) failure. On modern therapy, there is still 15% mortality within 1 year and median survival remains limited to 5-6 years, with insufficient functional improvement in many survivors. Bilateral lung transplantation (bLTx) remains the only curative option. Unfortunately, most patients are unlikely to survive until transplantation.

We know that PAH-associated RV failure is potentially reversible because of the generally good recovery of the heart after bLTx. Therefore, the RV has become a potential therapeutic target. A medical PAH treatment strategy that both reduces PVR and improves RV function is likely difficult to develop because of contrasting priorities within the different cell populations of the heart and lungs. As such, surgical options to increase cardiac output (CO) and to reduce RV wall stress should be elaborated. Atrial septostomy (AS) decreases RV filling pressures and improves cardiac index, but has a procedural mortality up to 16% in PAH patients. A right ventricular assist device (RVAD) could offer the same advantages as an AS. Moreover, the fall in arterial oxygen saturation accompanying AS should not occur and the RV would be actively supported.

With this thesis, we aimed to investigate the potential role of an RVAD as a new surgical treatment strategy for PAH. Two major problems arose with this treatment strategy: 1) The RV demonstrates a high afterload sensitivity. Increased pulmonary artery pressures (PAPs) caused by the RVAD might cause further RV dilatation and failure. 2) Historically, pulsatile devices for end-stage RV failure secondary to PAH often resulted in intraparenchymal pulmonary hemorrhage, hemoptysis, and death. It is though feared that also continuous flow devices with relatively high flow rates might result in increased PAPs and lung injury. We hypothesized that smaller continuous flow devices with lower flow capacities might be more appropriate.

As a first step (chapter 4), we aimed to assess the feasibility of low flow RV mechanical support, and to describe the hemodynamic effects of low versus high flow support in an animal model of acute RV pressure overload. Therefore, a Synergy Micro-pump was implanted in seven sheep. Blood was withdrawn from the right atrium to the pulmonary artery. Hemodynamics and pressure-volume loops were recorded in baseline conditions, after banding the pulmonary artery, and after ligating the right coronary artery in these banded sheep. RV mechanical support improved arterial blood pressure (ABP) and CO, but intrinsic RV contractility was not significantly impacted. It provided RV diastolic unloading, but with a concomitant and RVAD flow-dependent increase of systolic afterload. These effects were most distinct in the pressure overloaded RV without profound ischemic damage. The low flow strategy avoided excessive increases in PAP, but sufficiently improved ABP and CO. Therefore, it was considered to be beneficial for the afterload sensitive RV. As it might also protect against the development of pulmonary hemorrhage and pulmonary edema, we advocated the use of this strategy for the acute pressure overloaded RV (e.g. heart transplant patients, acute lung injury, acute respiratory distress syndrome).

RV mechanical support is well described in cases of sudden increase in RV afterload, e.g. after heart transplantation. In cases of chronic RV pressure overload, e.g. PAH, it has rarely been described. Therefore, we aimed in a second step (chapter 5) to assess the hemodynamic effects of low flow mechanical support of the acute vs the chronic pressure overloaded RV. The pulmonary artery was banded in 18 sheep. In the acute group, we immediately implanted a Synergy Micro-pump, as described before. In the chronic group, this pump was implanted 8 weeks after the banding. Hemodynamics and pressure-volume loops were recorded before and 15 minutes after pump activation. Low flow RV mechanical support significantly improved ABP in both groups, but CO only in the acute group. Intrinsic RV contractility was not affected. The RV contribution to the total right sided cardiac output was 54 ± 8% in the acute group vs 10 ± 13% in the chronic group (p<1.10-5), indicating a more profound unloading in the latter. Diastolic unloading was successful in both groups, but systolic unloading only in the chronic group.

In a 3rd step we aimed to assess the hemodynamic and histological effects of long-term mechanical low flow support of the chronic pressure overloaded RV (chapter 6). Therefore, the pulmonary artery was banded in 20 sheep. Eight weeks later, a Synergy Micro-pump was inserted in 10 animals, as described before. Following magnetic resonance imaging, hemodynamics and RV pressure-volume loops were recorded. Another eight weeks later, RV function was assessed in the same way, followed by histological analysis of ventricular tissue. During support, RV volumes and central venous pressure significantly decreased while contractility increased. PAP increased modestly, mainly its diastolic component. The RV contribution to the total right sided CO increased from 12 ± 12 % towards 41 ± 9 % (p<1.10-4). After pump inactivation, and compared to eight weeks earlier, RV volumes had significantly decreased, tricuspid valve regurgitation had almost disappeared and RV contractility had significantly increased, resulting in a significantly increased RV forward power (0.25 ± 0.05 vs 0.16 ± 0.06 Watt, p<0.05). Also left ventricular (LV) dimensions and function had improved, despite the ongoing RV pressure overload. Fulton index and RV myocyte size were significantly smaller without changes in fibrosis, compared to control animals. Therefore, we concluded that long-term continuous low flow RV mechanical support significantly unloaded the chronic pressure overloaded RV and improved CO. After 8 weeks, the pressure overloaded RV showed clear signs of hemodynamic recovery and of reverse remodeling, without increase of fibrosis.

Temporary LV dysfunction after pulmonary endarterectomy (PEA) for chronic thromboembolic pulmonary hypertension (CTEPH) is well described. True LV failure has only been described after bLTx in PAH patients. Therefore, in chapter 7, we sought to identify factors that contribute to this LV failure and hypothesized that an AS before bLTx could prevent this LV failure. From our database, all PAH patients that underwent bLTx (n=24) and all CTEPH patients that underwent PEA, with a minimal reduction of 800 dynes.s.cm-5 (n=27), were selected. Perioperative demographic and echocardiographic data were analyzed.  Pulmonary hypertension was diagnosed at significant younger age and time between diagnosis and surgery was significantly longer in PAH patients. Before surgery, PAH patients had significant larger RV dimensions, a significant smaller LV wall thickness, but a similar LV diastolic dysfunction. Surgery caused significant decreases in RV dimensions, less extensive in CTEPH, and significant increases in LV dimensions. Preoperative AS for PAH caused increases in LV dimensions, stroke volume and cardiac index. Two PAH patients developed postoperative LV failure. Compared to other PAH patients, they were younger (< 12 years) at diagnosis, their time between diagnosis and surgery lasted 2.5 times longer, their LV mass was smaller and their pre-bLTx PVR was higher. Therefore, we concluded that a younger age at diagnosis and a longer duration of LV preload deprivation contribute to the development of a more undertrained LV compared to CTEPH patients. Together with a higher postoperative increase in preload and a lower postoperative residual PVR, this undertrained LV explains the occasional development of LV failure after bLTx. Preoperative AS, but also a low flow RVAD and other surgical strategies may train the LV before bLTx, by increasing its preload, and avoid postoperative LV failure.

The results of this thesis support the future clinical application of a low flow RVAD as a new treatment strategy for patients with RV failure due to PAH. This can be applied as bridge to bLTx or as destination therapy. The low flow RVAD will unload the RV and will lead to hemodynamic recovery, to RV reverse remodeling, and to an improvement of LV function with likely little change on damage to the lungs.