The aim of this study was to discuss the pathogenesis of PO and whether CEA could be an effective treatment for PO, based on the clinicopathological results. Our findings indicate that the PO plaque compositions may be one of the factors, causing reduction of the flood flow and collapsed distal portion. Alternatively, PO is a lesion treatable by CEA with minimum risk and a high success rate, and the plaque histology can be related to the surgical outcome. It is widely accepted that primary carotid artery stenotic lesions are caused by advanced atherosclerosis [8, 16]. The rupture of unstable plaques, characterized as having a large atheromatous core, thin fibrous cap, and infiltration of inflammatory cells, causes thrombotic events and is followed by plaque stabilization, including reorganization and fibrosis, over time [12]. In our group, plaque rupture and intraplaque hemorrhage were detected in approximately 50 % or more, which was compatible with other reports [8, 16], although fibrous or fibroatheromatous plaques were prevalent, and old organized thrombi were frequently found in the PO patients. These findings indicate that considerable time had passed since the initial thrombotic events. PO plaques had two subtypes of old organized thrombi, occlusive or non-occlusive of the original lumen; in other words, PO had two different histological features. Plaques of the total occlusion with recanalization (8 patients, 47.1 %) were composed of thrombotic total occlusion and lumen recanalization by large neovascular channels, whereas those of severe stenosis (9 patients, 52.9 %) were fibrous sclerotic plaque with severe stenosis of the original lumen. It is interesting to speculate as to why the PO falls into collapse and not into complete occlusion, despite severe stenosis. The abrupt increase in the degree of stenosis, reducing blood flow and collapse of the distal ICA, could be explained by either disrupted intraplaque hemorrhages with some connection to the arterial lumen or expanded intraplaque hemorrhages without plaque disruption or luminal extension [1, 16]. Plaques of the total occlusion with recanalization could be structured by the luminal thrombi arising from intraplaque hemorrhage with plaque disruption. On the other hand, plaques of severe stenosis might be formed by subluminal hemorrhages without plaque disruption or luminal extension. However, various clinicopathological features may occur, depending on the degree of stenosis, the site, and the plaque compositions [1, 13]. The angiographic appearance of PO may be caused not only by the severity of stenosis but also by plaque compositions. Changes in plaque compositions from atheromatous to fibrotic may reduce the intraluminal pressure and collapse the ICA distal portion. Symptomatology, acute occlusive thrombosis of the carotid arteries may be either asymptomatic or present with only mild symptoms if the Circle of Willis or other intracranial collateral circulation is adequate, as in our cases. This finding suggests that the adequate collateral flow plays an important role in maintaining the intraluminal blood pressure and delayed anterograde flow of the patent ICA. The natural history of PO of the ICA is obscure. A success rate of revascularization for PO was reported from 75.0 to 100 % in previous literature [5]. Several authors have described that CEA for PO is less beneficial. However, Morgenstern et al. reported that surgery for PO was indeed of benefit and resulted in a reduction of the stroke rate [10]. In our series, of all 17 patients, 16 (94.1 %) were successfully treated without perioperative complications. Our surgical results were also favorable compared with other previous reports. Of all 17 patients, 1 (5.9 %) showed occlusion of the ICA on DSA 2 weeks after surgery. Greiner et al. pointed out that frequent causes for failure of reconstruction are hypoplastic or fibrotic vessels, chronic subtotal thrombosis, or simply too long an interval between the initial diagnosis and intra-operative evaluation [5]. The present case involved a histologically fibrous plaque with severe stenosis, indicating that progression of atherosclerosis extended to the distal ICA and the fibrotic vessel wall could prevent satisfactory dilatation of the ICA. During the follow-up periods, restenosis was encountered in two patients (11.8 %), and they were successfully treated with CAS. Hellings et al. described that plaque composition is an independent predictor of restenosis after CEA and a lipid-rich, inflammatory plaque is associated with a reduced risk of restenosis [6]. In our study, two plaques with ICA restenosis were fibrous, indicating that the plaque compositions might be useful to identify a patient with restenosis after surgery. Recently, CAS has been developed as an alternative treatment to CEA in carotid atherosclerotic disease, and the advantages and disadvantages of each procedure are currently under discussion. Terada et al. described that CAS for PO under embolic protection was beneficial and resulted in a reduction of the stroke rate [15]. The indications for carotid artery revascularization have been basically determined by the degree of stenosis, but it is important to evaluate the plaque compositions. We demonstrated that PO plaques had two different histological features, including the thrombotic occlusion with recanalization or severe stenosis. Plaques of severe stenosis could be treatable with CEA or CAS; in contrast, those with total occlusion with recanalization, in the absence of true lumen, could be suitable for CEA because it seemed to be difficult for endovascular devices to pass through the arterial lumen.