Unilateral rib-notching from the collateral circulation after division of the subclavian artery.

The operation of subclavian-pulmonary anastomosis has provided another cause for unilateral rib-notching-the collateral circulation to the arm after division of its subclavian artery. The flow is, of course, in the reverse direction to the flow in the collateral arteries that by-pass a coarctation of the aorta. We are reporting three cases in which there was striking unilateral rib-notching after a subclavian-pulmonary anastomosis, in one with other signs of a collateral circulation in the chest wall. My attention was drawn to the first of these patients by Dr. Curtis Bain as long ago as 1949 (Fig. IA), but I did not then see the explanation. I thought there might be some new collateral circulation to the lungs, for small arteries from the mediastinum had often been seen entering the hila in patients with Fallot's tetralogy. Barrett and Daley (1949) also drew attention to the arteries they had seen at this site coming from the mediastinal, phrenic, and internal mammary arteries. It was not until several years later when this boy was re-examined and two further patients were seen, all of whom had rib-notching on the side where the subclavian artery had been divided, that the correct explanation became apparent. The point at which the subclavian is divided is not far from the usual site of a coarctation, and anastomotic vessels that can carry blood from some branches of the subclavian and axillary arteries to parts below the coarctation can do so equally well in the opposite direction, from the lower part of the chest wall to the axillary artery and the arm. When discussing this conclusion with my radiological colleagues, Dr. Hills and Dr. Dow, I found that unilateral rib-notching after division of a subclavian artery had been described by Kent (1953). He was, I think, the first to link these two features and to suggest the right explanation. He reported two cases, both with rib-notching on the side of the operation. The first had a complex lesion, probably dextrocardia and tricuspid atresia, and the second did not have an anastomosis completed: but, as he says, this does not affect the argument, because the subclavian artery was divided in both-for a successful anastomosis in the first and for an attempted anastomosis in the second. Since his paper this syndrome has been observed by several radiologists: it does not, however, seem to be so widely known among physicians. As will be seen later, the condition is much more common than has been supposed, though generally in a less striking form than in these three cases. Sometimes there are also signs of the collateral circulation in the chest wall. Thus in Case 1, there was a prominent pulsating artery passing over the right clavicle very similar to the one shown by Kent (1953, his Fig. 10) as well as arteries in the chest wall over the lower ribs.