Potential role of multiple rectal biopsies in the diagnosis of equine grass sickness.

Two samples were taken postmortem from the rectum of each of 14 horses with grass sickness and 10 control horses, and four sections stained with haematoxylin and eosin were examined. By using as a criterion of grass sickness the presence of three chromatolytic neurons, 10 of the 14 cases were positive and none of the control horses was positive, giving a sensitivity of 71 per cent and a specificity of 100 per cent. No other histological features appeared to be of diagnostic value, and staining for Nissl substance in neurons with cresyl fast violet and methyl green-pyronin did not improve the sensitivity of the test.