Pathways of Cerebral Collateral Circulation

Despite Considerable anatomical proof to the contrary, the myth of “end-artery” circulation to the brain has been perpetuated in textbooks for years. This interpretation of Cohnheim's original theory concerning infarction and his concept of end-arteries, published in 1872, has led to the erroneous assumption that cerebral collateral circulation is sparse and inefficient, if not actually nonexistent (1). The presence of a cerebral collateral circulation and its significance have become especially evident with our continuing angiographic evaluation of patients with stroke syndromes during the past three years (2). Our studies have prompted an evaluation of the importance of collateral circulation in these and other cases and an investigation of the pathways by which this occurs, as demonstrated on the cerebral angiogram. Normal Cerebral Blood Flow As far back as 1912, Kramer (3) demonstrated by means of dye injection in living animals that each internal carotid artery normally supplies blood only to the distribution of its anterior and middle cerebral branches, and the vertebral arteries, via the basilar, only to the distribution of the cerebellar and posterior cerebral branches. This was confirmed by vascular injection studies in the human cadaver by Fay in 1925 (4). In 1948 Shenkin et al. (5) corroborated these observations in man by injecting Evans blue into the internal carotid arteries and recovering the dye from the internal jugular vein on the injected side only. These concepts are borne out by everyday experience with normal internal carotid and vertebral arteriograms. Thus, it is seen that the circle of Willis acts as a shunt mechanism and does not, under normal circumstances, allow mixing of blood coming from each of the two internal carotid arteries and the basilar artery (6). Not uncommonly on internal carotid arteriography filling of the posterior cerebral artery is observed. This may be because the posterior cerebral artery originates as a branch of the internal carotid and not of the basilar artery. Such an origin was reported by Riggs and Griffiths as occurring in approximately 26 per cent of autopsies on persons with nervous and mental disorders (7) although Fawcett and Blackford found it in less than 2 per cent of autopsies (8). Despite statements that filling of the posterior cerebral artery may be seen in carotid arteriography in 15 to 25 per cent of cases, an evaluation of 564 normal carotid arteriograms showed, in our hands, posterior cerebral artery filling on one or both sides in only 11.5 per cent (Table 1). Collateral filling of the opposite anterior cerebral artery on internal carotid arteriography is also generally considered a normal phenomenon. However, a study of 1,083 normal carotid angiograms does not seem to bear out this contention.