Degeneration of the granular layer of the cerebellum.

The following investigation was initiated by the observation by one ofus (Meyer, 1944), of a selective degeneration of the granular layer of the cerebellum in a case of a Wernicke syndrome associated with carcinoma of the stomach. The possible relation to one or other type of nutritional deficiency (not necessarily that of vitamin B1) was discussed, but no consideration was given to the fact that there might be a more direct causal relation to the carcinoma. There has been a fair number of reports on cases of visceral carcinoma in which clinical and/or pathological signs of cerebellar degeneration (not associated with secondary deposits) were discovered. A causal relation between the carcinoma and the cerebellar lesion was by no means assumed in all of these communications. Indeed in the cases of Courville and Fried-man, and Buchanan and others, it was expressly rejected in view of the clinical history. In the opinion of Brouwer and Biemond (1938), however, the rapid development of a cerebellar lesion in cases of carcinoma was such as to suggest the influence of a carcino-toxic agent. The granular layer, if involved, was not always selectively affected in their cases, but in the most important observations published on this subject by Bertrand and Godet-Guillain (1942) the granular layer was indeed selectively involved. In fourteen out of sixteen cases of visceral carcinomata these authors described varying stages of degeneration of the granular layer, betraying itself even on macroscopic inspection by an appearance they termed " glace." These cerebellar changes were most marked in the dorsal part of the vermis and were usually accompanied by degenerative signs in the inferior olives, particularly in their dorsal portion. This predilec-tion for the dorsal aspects of the cerebellum recalls that seen in the delayed type of cerebellar atrophy (Foix and others, 1922), more recently known as parenchymatous cerebellar atrophy. This relationship, which is important in view of the usual age period of carcinoma, has in fact been discussed in the literature (Casper, 1929; Brouwer and Biemond, 1938). The latter authors concluded, however, that there was no relationship between the two conditions, as they found a diffuse distribution of the cerebellar changes in the carcinoma cases, as compared with the more focal distribution in the delayed atrophy. In the cases described by Bertrand and Godet-Guillain no clinical cerebellar signs were observed, and the authors considered, therefore, that the cerebellar lesions had developed rapidly during the …