The problem of restenosis is an important consideration any time angioplasty is contemplated. When the artery to be dilated is totally occluded, the restenosis rate is highest. As we showed in a large series of patients with chronically occluded arteries treated in the 1980s,1 although approximately half of arteries that were successfully dilated exhibited restenosis, fewer than 20% progressed to total reocclusion.
The study by Violaris and coworkers2 in this issue of Circulation reaffirms these findings in a large series with several new important characteristics: (1) there is a control group of patients with stenoses who are undergoing the same measurements; (2) the measurements performed are quantitative angiographic ones performed in a core laboratory; and (3) there is a high reangiogram rate due to the fact that the patients were enrolled in randomized clinical trials. The selection of these patients for clinical trials may have resulted in a somewhat lower restenosis rate than previously reported, since patients selected for the trials met certain clinical and anatomic entry criteria.
The authors offer several explanations for a higher reocclusion rate for totally occluded arteries and admit that all are speculative. These explanations are unmeasured hematologic factors; the morphology of the totally occluded lesions, which remains unknown; vasoconstriction occurring to a differential degree in postdilation patients; and differences in local flow dynamics due to differences in reference artery size and reference lesion diameter.
A major potential contributor to the reocclusion rate, however, was not discussed. It is assumed that the collateral circulation in the occluded arteries was substantially higher than collateral circulation in the stenotic arteries. It is also indicated that there was a higher previous myocardial infarction incidence and more myocardial fibrosis in the totally occluded arteries. Therefore, the difference in distal bed flow capacity probably exists and may be an important factor. If collateral vessels were well developed and supplied competing flow and if flow demand was reduced due to fibrosis resulting from previous myocardial infarctions, then the total demand for distal bed flow may have been less in the totally occluded cohort compared with in patients with stenotic arteries. Because such conditions likely existed, the phenomenon of progression to total occlusion would be expected to occur due to thrombotic causes brought on by slow flow rather than by fibrointimal hyperplasia or chronic vascular recoil. Thrombotic occlusions in such a setting would likely occur early after the angioplasty procedure, driven by the combination of an acutely injured artery and reduced flow. Although there is no direct way to determine when the total occlusions occurred, it would appear from Fig 1, which reflects time to clinical end points, that the differences appeared within the first week and then again after the 6-month angiographic follow-up. Because these clinical differences are entirely driven by repeat revascularization, one can see how an early thrombotic reocclusion could have led to a need for early revascularization within the first week or have led to a permanently occluded artery without requiring urgent reintervention, which would only have been discovered at the 6-month protocol angiogram.
Because the major difference in restenosis rates was driven by the total reocclusion phenomenon, the authors suggest this would be an area for major improvement in the long-term results of total occlusions. Whether thrombotic reocclusion in the early postangioplasty period can be reduced by other interventional means remains speculative. Among the therapies that may be of help, however, should be included the use of potent antiplatelets or antithrombins, which are undergoing clinical trials.3 Whether these agents can overcome the tendency to rethrombose, stimulated by decreased peripheral bed demand and consequent slow flow, remains to be seen. It is difficult to see how the distal flow demand can be artificially increased, at least not on a long-term basis.
Even though the major difference in restenosis rates between stenoses and occlusions was driven by this total reocclusion phenomenon, another major contributor is the late renarrowing process. To have an impact on the overall results of total occlusion angioplasty, an attack on the vessel geometry, such as that provided by intracoronary stenting, may be needed to reduce the restenosis rate.4 This strategy may produce new problems since stenting in the presence of diminished flow may itself predispose to a higher early thrombosis rate than would occur in stenting non–totally occluded vessels. Because reocclusion in totally occluded arteries appears to occur in approximately 20% of arteries, a higher thrombosis rate than occurs in stenosis stenting could be tolerated. It also appears to be intuitive that total reocclusion of previously totally occluded arteries will carry a lower risk for the patient.
There are a few problems with quantitative coronary arteriography when dealing with total occlusions. The quantitative angiographic method used to measure stenoses was not used in the group with totally occluded arteries.5 Therefore, all of the relative gain and relative loss comparisons are artificial since it is almost impossible for the CAAS system to read stenosis above 85% due to technical limitations, whereas all total occlusions were assigned a value of 100% stenosis. It is obvious that the gain-and-loss indexes will be quite different because of this difference in angiographic measurement methodology.
The fact that there were few differences between functional total occlusions and total occlusions is interesting. The success rate is influenced to a great degree by this feature,1 6 7 and in the present study, only successfully dilated arteries were considered. It is often unclear whether an artery is totally occluded or functionally totally occluded. If angioplasty is performed successfully with soft guide wires, it is often the case that a very tiny channel through the lesion existed, although flow may have been either TIMI 0 or TIMI I, depending on the nature of the collateral circulation and the force of the contrast media injection. The fact that these were all successfully dilated arteries may have blunted the difference between the total occlusion and the functional total occlusion group. Reocclusion rates remained higher, however, in the total occlusion group (23.9%) than in the functional occlusion group (15.9%), and the fact that this did not reach statistical significance may have been due to a type I error.
Increasingly, clinical end points are considered when comparing outcomes of angioplasty. Even though this survey was of prospective, randomized trial patients requiring follow-up angiography, that is becoming less frequent in clinical practice. Even considering all clinical end points, the significant statistical difference between total occlusions (28.5%) and stenoses (22.1%) may not have clinical relevance. Because the occurrence of death and myocardial infarction was not different (3.6% for occlusions and 3.0% for stenoses), this difference was entirely driven by revascularization (24.9% for occlusions versus 19.1% for stenoses). The difference in all clinical end points before the 6-month protocol angiograms was only 3% (22% for occlusions and 19% for stenoses). After the 6-month angiogram, the difference widened because of repeat interventions in the patients found to have restenosis. In the absence of routine follow-up angiograms, these differences in interventions performed may have been less.
The access of Violaris and coworkers to such a large, well-studied database significantly extends our understanding of the long-term results of total occlusion angioplasty. In their cohort of patients, the authors have shown that although statistically significant, the outcomes of patients with successfully dilated total occlusions are not sufficiently clinically different from those with stenotic lesions to preclude angioplasty attempts. This study provides a good basis for comparison with the results of new technology for this vexing problem. As with stenoses, however, I believe the future for keeping total occlusions open without restenosis is bright. Now, if we could only get more of them open.
The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association.
- Received February 9, 1995.
- Accepted February 9, 1995.
- Copyright © 1995 by American Heart Association
Ivanhoe RJ, Weintraub WS, Douglas JS Jr, Lembo NJ, Furman M, Gershony G, Cohen CL, King SB III. Percutaneous transluminal coronary angioplasty of chronic total occlusions: primary success, restenosis, and long-term clinical follow-up. Circulation. 1992;85:106-115.
Violaris AG, Melkert R, Serruys PW. Long-term luminal renarrowing after successful elective coronary angioplasty of total occlusions: a quantitative angiographic analysis. Circulation. 1995;91:2140-2150.
Almagor Y, Borrione M, Maiello L, Khlat B, Finci L, Colombo A. Coronary stenting after recanalization of chronic total coronary occlusions. Circulation. 1993;88(suppl I):I-504. Abstract.
Bell MR, Berger PB, Bresnahan JF, Reeder GS, Bailey KR, Holmes DR Jr. Initial and long-term outcome of 354 patients after coronary balloon angioplasty of total coronary artery occlusions. Circulation. 1992;85:1003-1011.