Circulatory support with attenuated pulse pressure alters human aortic wall morphology

Abstract

In biology “form follows function.” Hypertension causes the musculoelastic aortic wall to thicken. 1 The only situation characterized by chronic hypotension with diminished pulse pressure is in patients having long-term support with a left ventricular assist device (LVAD). In a sheep model of chronic nonpulsatile blood flow, the medial layer thinned with apoptosis of smooth muscle cells (SMCs) and fragmentation of elastic fibers. 2 We sought to determine the clinical implications of this finding for destination therapy patients with a rotary blood pump. Clinical Summary In 7 patients with end-stage heart failure, computed tomographic scans were taken before implantation of a Jarvik 2000 LVAD (Jarvik Heart, Inc, New York, NY) and then annually during survival. Preoperative and latest postoperative aortic diameters were compared. After 5 deaths at between 7 days and 34 months of support, we obtained multiple full-thickness sections of the descending thoracic aorta. At least two nonadjacent sections were processed for each patient. The samples were fixed and stained with hematoxylin and eosin (HE StatsDirect Ltd, Sale, Cheshire, United Kingdom) and presented as median SD. Analysis of variance was calculated using 2-way analysis of variance with Bonferroni comparison. None of the computed tomographic scans showed an increase in aortic diameter out to 6 years (Figure 1). The 7-day LVAD specimens were indistinguishable from controls, but distinct morphologic changes were evident in the 3-month samples. Medial thickness diminished during prolonged support (1.411 0.140 mm in controls vs 1.006 0.170 mm in LVAD patients; P .0001). Medial SMCs decreased in number (147 31 per field in controls vs 63 12 per field in LVAD patients; P .0001) (Figure 2). From the EVG-stained sections there was a concomitant decrease in the medial layer elastin content. The process of atrophy appeared to progress with time. Discussion These novel findings were predictable but have not previously been confirmed in humans. One study limitation is that the autopsy samples could not be compared with the same patients’ aorta obtained during LVAD implantation. However, we were initially unaware of the sheep findings from our Terumo ventricular assist device program. 2 It was this that triggered the human study. Matched controls were the alternative. Irrespective of this, the human changes closely parallel the laboratory findings. 3 Arteries remodel in response to pressure and flow to provide a constant level of shear stress. Remodeling follows a complex interaction between signaling pathways and altered gene expression for nitric oxide synthase, platelet-derived growth factor, and transforming growth factor 1. 4 Because the Jarvik 2000 LVAD is afterload sensitive, mean systemic arterial blood pressure (MAP)