We present experimental evidence that, in Ossabaw miniature swine, selective surgical excision of adipose tissue in direct contiguity with one of the epicardial coronary arteries attenuated the progression of atherosclerosis, thus supporting the hypothesis that cEAT could contribute to underlying coronary atherogenesis . The findings are applicable to the early stages of CAD because of the relatively young age of the animals, the short duration of atherogenic diet feeding, and the lack of observed flow-limiting coronary stenosis typical of advanced clinical disease. We acknowledge the substantially high LDL cholesterol levels (>500 m/dL) in the obese group. We have conducted other studies in which the LDL was ~250 mg/dL (e.g. [10, 12]) for a longer duration, which yielded substantial coronary atherosclerosis. We predict a similar result of adipectomy. This very high LDL cholesterol level for a short duration is similar to LDLR−/− humans, who have substantial atherosclerosis. Horton and colleagues refer to this phenomenon as the effect of cumulative LDL cholesterol in g/dL-years .
We showed that in the middle LAD at the adipectomy site, progression of CAD was not significant compared to the significant increases in circumferential plaque in the adjacent, unperturbed proximal and distal segments. Thus, in the middle LAD segment, the increase in mean plaque length was 1.4-fold compared to 1.9- and 2.9-fold in the proximal and distal LAD, respectively. There were higher mean values and greater variability in baseline plaque length in the middle compared to the proximal and distal segments at the time of adipectomy, suggesting that the attenuated progression of atheroma could have been the consequence of higher initial plaque thickness at the inception of the protocol. However, a two-tailed, paired t-test did not indicate that this baseline difference was statistically significant (p = 0.26). The reason for the variability in the middle LAD plaque dimensions is not clear, since prior studies have indicated that mean plaque burden is greatest proximally and decreases distally in the LAD . One possibility is that adipectomy could have perturbed underlying plaque structure or vasomotor responses and interfered with IVUS measurements performed under heparin immediately after adipectomy. However, we have recently shown that cEAT has uniformly adverse effects on coronary artery function , in contrast to protective effect of PVAT in the aorta and other arteries. This explains why we observed no adverse mechanical changes after adipectomy.
We resected fat from one coronary artery for several reasons. Resection of as much fat as possible off the entire myocardium was not technically possible in vivo. cEAT on the LAD was the most accessible site within the operative field of exposure compared to either the circumflex or right coronary arteries, which would lessen both intraoperative trauma, morbidity and mortality, and the risk of post-operative complications. This strategy targeting one rather than two vessels was subsequently vindicated by the low intraoperative mortality (1 of 9 pigs) and no noticed post-operative complications. Finally, the longitudinal time course of the experiment enabled the pigs to act as their own controls. Immunoreactivity for T-cadherin was predominantly localized to the endothelium as previously established , but also to the neo-intima. It is considered to be a marker of active atherosclerosis  and a major receptor for adiponectin in the vasculature . T-cadherin and SRA, the scavenger receptor for LDL-cholesterol on plasma membranes of plaque macrophages and foam cells [24, 25], were reduced in the middle LAD intima-media implying less atherogenesis. Paradoxically, anti-inflammatory immunoreactive adiponectin in the same LAD segment was reduced significantly by 68%, despite a ~7-fold upregulation of adiponectin mRNA expression in overlying cEAT; thus, one might surmise that adiponectin released by epicardial adipocytes might not have been able to penetrate the coronary wall if such a mechanism exists. While a majority of the literature describes a decrease in adiponectin expression in CAD patients, we speculate that the 6.7-fold increase in obese pig cEAT mRNA was a compensatory response and that there must be some block in mRNA translation. We have previously reported that serum adiponectin is decreased in obese versus lean Ossabaw swine . On the other hand, there were no changes in vessel wall CD3 (a T lymphocyte cell marker), tryptase (a basophil marker), TNFα, MCP-1, and resistin.
The mechanisms whereby selective adipectomy might have attenuated the progression of CAD remain to be established, but hypothetically include (i) the removal of factors generated by cEAT that could contribute to atherogenesis by direct diffusion through the porous adventitia into the coronary intima-media , such as reactive oxygen species (ROS) and pro-inflammatory cytokines and/or (ii) the disruption of vasa vasorum in the adventitia and in the closely adjacent cEAT , which might interfere with vasocrine signaling of atherogenic adipokines  and restrict mononuclear cell access into the intima-media. The evidence for these suppositions includes the fact that relative to control cEAT, there was significant up-regulation in atherosclerotic cEAT of pro-inflammatory 11βHSD-1, CCL19, CCL21 and PGD2S; gp91phox [NADPH oxidase], a major source of vascular ROS; VEGF and VEGFGR1, which partake in vasa vasorum neogenesis and angiotensinogen, which promotes endothelial dysfunction and vasoconstriction. Similar to our current findings in pigs, we reported 3-fold up-regulation of 11βHSD-1 gene expression in human EAT adjacent to severe CAD over controls without CAD . Pharmacological inhibition of 11βHSD-1 prevented aortic plaque progression in a murine model of atherosclerosis, suggesting that 11βHSD-1 may play a role in atherogenesis . Lymphoid chemokines CCL19 and CCL21 are crucial for the recruitment of circulating naive T cells into lymph nodes. Increased levels of CCL19 and CCL21 have been reported within the atherosclerotic lesions of ApoE−/− mice, in human atherosclerotic carotid plaques, and in plasma of CAD patients . In mice, PGD2S deficiency induces obesity and facilitates aortic atherosclerosis , suggesting it is normally protective and anti-atherogenic in this species. The role of PGD2S in porcine atherosclerosis has not been reported, but -regulation of PGD2S in EAT in our study was associated with amelioration of atherosclerosis, suggesting the possibility that this pro-inflammatory prostanoid might otherwise be harmful either alone or in conjunction with other adipokines. No changes were observed in cEAT in the expression of TLR4 and acute phase cytokines TNFα, IL-1β, IL-6, IL-8 as well as the anti-oxidant genes eNOS and SOD. Finally, the enhanced expression of a selected array of pro-atherosclerotic genes in cEAT over paracardial fat attests to the relative specificity of the genomic response of cEAT associated with CAD that has been similarly documented in humans with CAD .
Epicardial fat volume determined by sequential CT scans significantly increased over 3 months of hypercaloric feeding under the same conditions previously shown to result in major expansion of visceral retroperitoneal and intraperitoneal as well as subcutaneous abdominal fat depots in the same animal model . Although our experimental design using pigs as their own controls in a longitudinal study has its strengths, there are also limitations in the lack of sham-adipectomies in age- and gender-matched pigs fed hypercaloric atherogenic diet. Future studies should employ sham-operated animals; randomization of experimental and control pigs; blinding of both the ultrasonographer and two independent IVUS data analysts to the pig group designation; reversing heparin anti-coagulation during IVUS with protamine sulphate so that IVUS can be done before surgery; CT scanning using radio-contrast dye to delineate the coronaries and to quantitate cEAT area around coronary plaques per se rather than EAT covering the entire myocardium; and immunohistochemical and mRNA analysis of cEAT and the coronary wall for expression of genes and proteins known to be important mediators of or participants in atherogenesis. To make these studies more translatable to clinical treatment, more clinical outcomes studies in the swine model would be helpful, for example, monitoring hard clinical endpoints such as MI, fatal events, etc. over a longer duration.
Our translational research project was done using experimental conditions prohibited in man by obvious ethical constraints and, with the provisos noted above, supports the human epidemiological evidence for a direct relationship between EAT and CAD. At this juncture and until more rigorous evidence is available one way or the other, it is premature to advocate an adipectomy as surgical treatment to attenuate the progression of CAD. Rather, it would be more appropriate from a clinical point of view if non-invasive methods were devised to shrink epicardial fat volume to serve the same beneficial purpose. In this regard, weight loss by caloric restriction and/or exercise in obese human subjects is associated with shrinkage of EAT volume by 9-32% of baseline in different reports [13, 31], but it is not known whether this is accompanied by decreases in chronic inflammation in EAT or reduction in CAD burden .