The crossover lung segment: congenital malformation associated with a variant of scimitar syndrome.

SIR,-We thank Drs Hyde and James for drawing our attention to two consecutive reports in the American Journal of Roentgenology on the association of horseshoe lung with scimitar syndrome. It is clear from the case description that many cases of so called horseshoe lung are, in fact, identical to our three cases, which we have described as crossover lung segments. The essential feature of horseshoe lung is partial fusion of right and left lungs posterior to the heart. We have clear evidence in our first patient, who underwent thoracotomy, that the lungs were not fused. There was merely a segment from the upper lobe of the hypoplastic bilobed right lung lying in the left hemithorax posterior to the heart. This was not an "isthmus of pulmonary parenchymal tissue arising from the right lung base, bridging the right and left lungs posterior . . .," as described by Frank et al. However, the cases described by Frank et al and, indeed, most of the other cases reviewed by them from the reports published worldwide did not have either in vivo or postmortem thoracotomies to establish the exact anatomy. The clinical and radiological features were identical to those of our cases, demonstrating a range of features that included hypoplastic right lung, anomalous venous drainage of the right lung. systemic arterial supply to the right lower lobe, and a segment of lung parenchyma in the left hemithorax supplied by a branch of the bronchial tree and pulmonary artery from the right. Interestingly, there is an overwhelming female predominance of this condition in the reports published worldwide (4:1), and in our patients the sex ratio was 2:1. All the other patients presented with major problems under 1 year of age and most died. Frank et al suggest aggressive surgical intervention with lung resection. Two of our patients have survived, one without any treatment and the other with only embolisation of the systemic arteries to the right lower lobe via a catheter. We suggest, contrary to the other publications, that, although the anatomy must be delineated, minimal surgical interference, at least in infancy, is most likely to be successful. Where there is an appreciable left to right shunt from systemic arteries to the right lower lobe via the scimitar vein, embolisation of the systemic arteries is indicated. Definitive correction may be attempted when the child is older. We also suggest that our nomenclature is more appropriate as it is clear that these patients do not have fused lungs. Indeed, conceptually, horseshoe lung is highly improbable from an embryological point of view.