OBJECTIVE: To characterize reading impairments caused by lesions in the posterior occipital cortices. METHODS: We gave six patients with these lesions reading and writing tests and located a critical site for alexia using MRI and SPECT. RESULTS: The patients read three-character kana (Japanese syllabograms) nonwords, and five-character kana nonwords significantly or at a near significant level more poorly and slowly than normal subjects, whereas they read kanji (Japanese morphograms) almost correctly but more slowly. Letter-by-letter reading with a single-kana character identification impairment (in five patients), a word-length effect, kinesthetic facilitation, a lexicality effect, and minor to mild agraphia for kanji (in three patients) were observed. These deficits were characteristic of pure alexia. Alexia disappeared within a few months except in one patient who had extensive hypoperfusion in the left occipital lobe. A shared lesion was located in the left posterior fusiform/inferior occipital gyri (Area 18/19) on MRI, and there was blood flow reduction around this area on SPECT. This area coincided with the activation site for kana word covert reading in our previous study. CONCLUSIONS: These results suggest that pure alexia particularly for kana, or more generally pure alexia for letters, is caused by a lesion in the posterior inferior occipital cortex, characterized primarily by impaired kana character or letter identification, with relatively preserved kanji or word recognition.
OBJECTIVE: To characterize reading impairments caused by lesions in the posterior occipital cortices. METHODS: We gave six patients with these lesions reading and writing tests and located a critical site for alexia using MRI and SPECT. RESULTS: The patients read three-character kana (Japanese syllabograms) nonwords, and five-character kana nonwords significantly or at a near significant level more poorly and slowly than normal subjects, whereas they read kanji (Japanese morphograms) almost correctly but more slowly. Letter-by-letter reading with a single-kana character identification impairment (in five patients), a word-length effect, kinesthetic facilitation, a lexicality effect, and minor to mild agraphia for kanji (in three patients) were observed. These deficits were characteristic of pure alexia. Alexia disappeared within a few months except in one patient who had extensive hypoperfusion in the left occipital lobe. A shared lesion was located in the left posterior fusiform/inferior occipital gyri (Area 18/19) on MRI, and there was blood flow reduction around this area on SPECT. This area coincided with the activation site for kana word covert reading in our previous study. CONCLUSIONS: These results suggest that pure alexia particularly for kana, or more generally pure alexia for letters, is caused by a lesion in the posterior inferior occipital cortex, characterized primarily by impaired kana character or letter identification, with relatively preserved kanji or word recognition.