William D. Coats

Troglitazone inhibits vascular smooth muscle cell growth and intimal hyperplasia.

Vascular smooth muscle cell (VSMC) proliferation and migration are responses to arterial injury that are highly important to the processes of restenosis and atherosclerosis. In the arterial balloon injury model in the rat, platelet-derived growth factor (PDGF) and basic fibroblast growth factor (bFGF) are induced in the vessel wall and regulate these VSMC activities. Novel insulin sensitizing agents, thiazolidinediones, have been demonstrated to inhibit insulin and epidermal growth factor-induced growth of VSMCs. We hypothesized that these agents might also inhibit the effect of PDGF and bFGF on cultured VSMCs and intimal hyperplasia in vivo. Troglitazone (1 microM), a member of the thiazolidinedione class, produced a near complete inhibition of both bFGF-induced DNA synthesis as measured by bromodeoxyuridine incorporation (6.5+/-3.9 vs. 17.6+/-4.3% cells labeled, P < 0.05) and c-fos induction. This effect was associated with an inhibition (by 73+/-4%, P < 0.01) by troglitazone of the transactivation of the serum response element, which regulates c-fos expression. Inhibition of c-fos induction by troglitazone appeared to occur via a blockade of the MAP kinase pathway at a point downstream of MAP kinase activation by MAP kinase kinase. At this dose, troglitazone also inhibited PDGF-BB-directed migration of VSMC (by 70+/-6%, P < 0.01). These in vitro effects were operative in vivo. Quantitative image analysis revealed that troglitazone-treated rats had 62% (P < 0.001) less neointima/media area ratio 14 d after balloon injury of the aorta compared with injured rats that received no troglitazone. These results suggest troglitazone is a potent inhibitor of VSMC proliferation and migration and, thus, may be a useful agent to prevent restenosis and possibly atherosclerosis.

Remodeling of the coronary artery after vascular injury

Vascular injury, such as angioplasty, induces a complex healing process that is analogous to generalized wound healing. Following initial injury, platelet thrombi and an acute inflammatory response occur, followed by cellular inhibition by monocytes/macrophages, proliferation of vascular smooth muscle cells, and extracellular matrix deposition. The injury repair concludes with organization of the matrix and reendothelialization. Recent studies suggest that restenosis or renarrowing after vascular injury is due to two components: a cellular proliferative process leading to intimal hyperplasia and remodeling of the vessel with a change in vessel geometry. Animal and clinical studies have shown that at least 60% of restenosis can be explained by unfavorable remodeling with vessel constriction rather than intimal hyperplasia. Although a number of factors appear to be associated with remodeling, alterations in extracellular matrix metabolism may be most important.

Collagen Content Is Significantly Lower in Restenotic Versus Nonrestenotic Vessels After Balloon Angioplasty in the Atherosclerotic Rabbit Model

BACKGROUND It is recognized that restenosis is primarily due to alterations in geometric remodeling of the extracellular matrix rather than intimal hyperplasia. Prior studies have shown that angioplasty stimulates an increase in both synthesis and degradation of collagen in the atherosclerotic vessel. However, differences in collagen content and metabolism between restenotic and nonrestenotic vessels have not been examined. METHODS AND RESULTS Four weeks after angioplasty in an atherosclerotic rabbit model, collagen content in restenotic and nonrestenotic vessels was measured both biochemically by hydroxyproline quantitation and histologically by a digital subtraction method with the use of circularly polarized images of picrosirius red-stained sections. Collagenase and gelatinase activity also were measured in the same restenotic and nonrestenotic vessels by use of a radiosubstrate assay. Collagen content was found to be significantly lower in restenotic vessels than in nonrestenotic vessels both biochemically (127.0 +/- 32.6 versus 212.6 +/- 84.3 micrograms/mg tissue; n = 11 vessels; P < .05) and histologically (67.3 +/- 7.9% versus 76.3 +/- 11.8% area fraction; n = 20 sections from 6 vessels; P = .05). There was a significant inverse correlation between biochemically determined collagen content and gelatinase activity (P = .02) and a significant correlation between histologically determined lumen are and percent collagen content (P = .0071). CONCLUSIONS Collagen content is significantly decreased in restenotic versus nonrestenotic vessels after angioplasty in the atherosclerotic rabbit model. The increased collagen content in nonrestenotic vessels was associated with preserved lumen area and may play a role in geometric remodeling after angioplasty.

Coronary sinus occlusion pressure and its relation to intracardiac pressure

The hemodynamic components of coronary sinus (CS) occlusion pressure in humans have not been well described. If no other outflow for venous blood were present, then after acute occlusion of the coronary sinus the pressure would increase and equal aortic pressure. However, if thebesian vein drainage between the left ventricle and the coronary veins has an important role in humans, then CS occlusion pressure might reflect left ventricular (LV) pressure through transmitted LV pressure or intramyocardial pressure. To study this relation, 27 patients who underwent routine diagnostic cardiac catheterization were evaluated. Occlusion was accomplished by sudden inflation of a No. 7Fr balloon-tipped catheter placed into the CS. LV end-diastolic pressure and end-diastolic CS occlusion pressure were simultaneously recorded at rest. LV end-diastolic pressure (16.7 +/- 5.6 mm Hg) was not significantly different from end-diastolic CS occlusion pressure (15.9 +/- 5.4 mm Hg). LV end-diastolic and end-diastolic CS occlusion pressures were positively correlated (p less than 0.001) over the entire range of pressures (9 to 27 mm Hg). In contrast, systolic CS occlusion pressure was significantly lower than LV systolic pressure and unrelated to right-sided heart pressures. It is concluded that in humans, end-diastolic CS occlusion pressure closely parallels LV end-diastolic pressure, and measurement of CS occlusion pressure to assess LV end-diastolic pressure may have clinical use. These findings also suggest the existence of hemodynamically important thebesian vessel connections that may have implications for retroperfusion or pressure-controlled intermittent CS occlusion in humans.

Arterial Remodeling at the Reference Site After Angioplasty in the Atherosclerotic Rabbit Model

Recent studies suggest that arterial remodeling plays an important role in restenosis and that remodeling at the reference site may also occur. To assess the chronic effect of the reference site remodeling on angioplasty results, we evaluated reference site remodeling in an experimental atherosclerotic restenosis model. Histological sections of iliac stenoses and their associated proximal reference segments from 50 atherosclerotic rabbits killed 4 weeks after angioplasty were analyzed. Lumen area (LA), external elastic lamina area (EEL), and intimal plus medial areas (I+M) were measured at the lesion (L) and reference (R) sites. Angiography was performed preangioplasty, immediately postangioplasty, and 4 weeks postangioplasty. Restenosis was defined as an angiographic loss/gain ratio of greater than 50% at follow-up angiography. Twenty-three lesions were restenotic (R+) and 32 were not (R-). There was no difference in reference site diameters (RD) between these two groups at the time of angioplasty. However, RDs were significantly smaller in the R+ group than in the R- group (1.24+/-0.18 versus 1.52+/-0.28 mm, n=55, P<.01) at 4-week follow-up. Morphometric analysis also showed a smaller LA(R) in the R+ group (0.85+/-0.27 versus 1.06+/-0.37 mm2, n=55, P<.02), whereas there was no difference in I+M(R) between the two groups. EEL(R) significantly correlated with EEL(L), LA(R), and I+M(R) in both groups combined (r=.53, n=55, P<.0001; r=.62, n = 55, P < .0001; and r = .86, n = 55, P < .0001, respectively). Remodeling can favorably and unfavorably affect both the lesion and the reference sites and appears to occur in parallel and proportionately at both sites. These data suggest that angiographic measurement of late percent stenosis using reference site diameters may lead to an underestimation of the percent luminal narrowing in restenotic lesions because unfavorable remodeling occurs in both the lesion and reference sites in restenotic vessels.

Coronary sinus occlusion: Effect on ischemic left ventricular dysfunction and reactive hyperemia

Pressure-controlled intermittent coronary sinus occlusion (PICSO) has been shown to reduce experimental infarct size. To examine the role of PICSO in limiting the consequences of brief ischemia on left ventricular function, we studied the effect of PICSO in nine open-chest anesthetized dogs. PICSO was performed using a pump-inflated, balloon-tipped catheter in the coronary sinus until coronary sinus occlusion pressure reached a plateau (10 +/- 3 seconds). The balloon was then rapidly deflated (2 seconds) and the cycle was repeated. Regional left ventricular function in the ischemic zone was assessed by sonomicrometry. Coronary blood flow was measured with a flow probe around the left anterior descending artery (LAD) proximal to an occluding suture. Measurements were obtained at baseline, during a 3-minute LAD occlusion, and for 10 minutes of reperfusion. In an additional five dogs, this sequence was repeated during an infusion of adenosine at a dose that abolished reactive hyperemia following LAD occlusion. The addition of PICSO beginning 15 minutes prior to ischemia and continuing throughout LAD occlusion and reperfusion did not prevent, reduce, or shorten ischemic left ventricular dysfunction. PICSO uniformly blunted reactive hyperemia during reperfusion. However, PICSO also reduced coronary blood flow during maximal vasodilatation achieved by adenosine infusion prior to LAD occlusion. Therefore, it is likely that PICSO decreases reactive hyperemia due to mechanical factors arising from venous engorgement rather than by reducing the ischemic stimulus causing vasodilation.