Limbic encephalitis (LE) is often associated with lung, thymic, or testicular tumours and antibodies to Hu, CV2, or Ma2 (Ta) antigens. In these cases, it generally has a poor prognosis. Here we describe two patients with symptoms of LE, negative for typical paraneoplastic antibodies, in whom antibodies to voltage-gated potassium channels (VGKC) were detected retrospectively in serial serum samples. Patient 1 had a thymoma recurrence, but in patient 2 no tumour has been detected in the years following presentation. Plasma exchange was effective in reducing VGKC antibody levels, with substantial improvement in mental symptoms in patient 1. In patient 2, the VGKC antibodies fell spontaneously over two years, with almost complete recovery of mental function. Although neither patient had obvious neuromyotonia at presentation, both showed excessive secretions. We suggest that patients with limbic symptoms and excessive secretions should be tested for VGKC antibodies, and, if they are present, prompt and effective immunosuppressive treatment should be considered.

Two groups of university undergraduates differing in reading ability were tested on a number of reaction-time tasks designed to determine the speed of encoding visual information at several different levels. In addition, the subjects were given tests of sensory functions, verbal and quantitative reasoning ability, short-term auditory memory span, and ability to comprehend spoken text. The groups did not differ on the sensory tests. However, the faster reader group had faster reaction times on all of the reaction-time tasks, and the size of the fast-reader advantage increased with the mean reaction time. Faster readers also performed more accurately in verbal and quantitative reasoning, short-term auditory memory, and speech comprehension. Regression analyses suggested that the ability to comprehend spoken material and speed of accessing overlearned memory codes for visually presented letters represented two important independent correlates of reading ability in our sample of subjects. Two variables reflecting these abilities--the percentage of correct answers to a listening comprehension test and the reaction time for correct responses in a letter-matching task--accounted for nearly all of the variance in reading ability tapped by both of our reading tests. In a second experiment, no reaction-time difference was found between fast and average readers in a matching task requiring no long-term memory code access but considerable visual information processing as indexed by overall mean reaction time. The results supported the conclusion that one skill allowing fast readers to capture more information from each reading fixation is faster access to letter codes from print.

No abstract

No abstract