Clinical manifestations

Presentation and course

Patients can write fluently and spontaneously but cannot read what they have written. In milder cases, reading is slow and effortful, and words are read one letter at a time, creating a characteristic "word length effect," in which the time to read a word increases with the number of letters it contains (19; 26). This deficit is called “letter-by-letter reading” or “spelling dyslexia.” In severe cases, patients cannot read words or letters, a defect that may extend to numbers or other symbols, such as musical notation or map symbols (17; 51). Some authors suggest the term “visuographemic alexia," for this deficit, which is believed to be an inability to map the visual letter symbol to its abstract meaning (29).

In Japanese, which has two different writing systems, kana (phonetically based form) and kanji (non-phonetic, ideographic form), alexia without agraphia can selectively impair either form (50; 79) or impair both forms (78). Dissociations in bilingual patients can also occur, with the more recently acquired language being less affected (65), perhaps in keeping with the hypothesis that the right hemisphere is involved in new language acquisition. In patients with long-standing blindness, newly acquired alexia without agraphia may impair Braille reading in the absence of somatosensory deficits (45).

Patients with alexia without agraphia are able to recognize handwriting and font. This ability is impaired in patients with right fusiform gyrus lesions associated with prosopagnosia (72; 10). In a case report, a multilingual patient could easily identify the language of written words despite having difficulty reading them (34). Evidently some aspects of text processing are spared by left-sided lesions causing alexia without agraphia, suggesting hemispheric specialization for the type of information being extracted from written text. One study demonstrated alexia without agraphia to be associated with deficits in processing of facial features that led to difficulty in lip reading, a visual aspect of language. This deficit was more pronounced for line-drawn facial images than full content (grayscale) facial images (04).

Alexia without agraphia is often part of a triad that includes a right homonymous hemianopia and right hemiachromatopsia. The homonymous hemianopia is sometimes limited to the superior quadrants of visual field and may exist without the hemiachromatopsia (30). Alexia may also occur without homonymous hemianopia (35). In some cases, there is an inability to name colors (“color anomia”) even though the colors are visible in the normal left hemifield, and when the ability to match colors is preserved (30; 92; 47; 35).

Verbal memory deficits and visual object agnosia may also be present (30; 32), as well as an inability for the right hand to accurately target objects in the left visual field (“disconnection optic ataxia”) (30).

Prognosis and complications

In the presence of a stable lesion, some improvement in reading may occur. Among eight patients with pure alexia, five (63%) recovered, two (25%) died, and one (13%) did not recover after 2 years (97). A patient with a left fusiform gyrus lesion demonstrated improvement in perception of words but retained letter-by-letter reading. Based on functional MRI imaging, the authors attributed the recovery to reinforcement of parallel reading routes that bypassed the lesion to reach the occipital cortex (22).

Clinical vignette

A 58-year-old woman presented with a generalized seizure. Imaging revealed a tumor in the left occipital lobe, which on resection proved to be a glioblastoma multiforme. She had received radiation treatment for the lesion and was in remission when referred 9 months later. She reported profound problems with reading, especially of long words, but no other cognitive complaints. She displayed a dense macular-splitting homonymous hemianopia. She read with a letter-by-letter strategy, slowly and with great effort, taking more time to read long words. Repeat imaging at the time of examination, and after a 6-month interval, showed that the tumor and surrounding edema had extended to the anterior and medial left occipital lobe and splenium.

Biological basis

Etiology and pathogenesis

Alexia without agraphia is almost always caused by a lesion in the left hemisphere, most commonly in the medial and inferior occipitotemporal region (30; 17). Many cases have lesions that involve the splenium or callosal fibers, a feature critical to the disconnection hypothesis. However, some cases also have lesions involving the left mid-fusiform gyrus, which is the basis of the “word form agnosia” hypothesis (67). A single case report described a right occipital lesion in a right-handed person (75).

Most cases of alexia without agraphia are due to left posterior cerebral artery infarction. This condition can also rarely result from left carotid artery disease if the left posterior cerebral artery originates from the carotid artery rather than the basilar artery (“fetal origin”) (73). Alexia without agraphia may be a presentation of COVID-19 (68; 05). In that condition, patients are subject to arterial and venous infarctions due to a hypercoagulable state.

Other causes include primary and metastatic tumors (92), arteriovenous malformations (03; 18), lobar hemorrhages (47), nonconvulsive status epilepticus (56), multiple sclerosis (61), eclampsia (76), posterior reversible encephalopathy syndrome (20), herpes simplex encephalitis (35), Creutzfeldt-Jakob disease (02), cysticercosis (94), posterior cortical atrophy (90; 37), scrub typhus (43), and progressive multifocal leukoencephalopathy after treatment of chronic lymphocytic leukemia with monoclonal antibody (06).

A macular-splitting right homonymous hemianopia can slow reading (“hemianopic dyslexia”) and suggest alexia without agraphia. However, most patients with homonymous hemianopia do not have impaired reading. Moreover, alexia without agraphia can occur without homonymous hemianopia. Two studies have established that the word-length reading impairment in alexia without agraphia cannot be attributed to visual field loss (82; 07).

There are two principal pathophysiologic explanations of pure alexia: visuoverbal disconnection and word form visual agnosia.

Visuoverbal disconnection. The traditional explanation for alexia without agraphia is a disconnection of vision from language centers (33; 42). When there is a complete right homonymous hemianopia, intact vision in the left hemifield is processed by the right occipital cortex. However, the transfer of visual data to language centers in the left angular gyrus is interrupted when the occipital lesion also involves callosal fibers in the splenium, forceps major, or periventricular white matter surrounding the occipital horn of the lateral ventricle (30). Alexia without agraphia in the absence of right homonymous hemianopia is caused by a lesion of the white matter underlying the angular gyrus. Such “subangular” lesions cause an interruption of input just as it is reaching the angular gyrus. Most of the projection fibers travel ventral to the occipital horn (47). If the dorsal projections to the angular gyrus are spared, the alexia might be incomplete (“spelling alexia”) (17). Alexia limited to the right hemifield may be due to disruption of the intra-hemispheric occipital white matter tract (Barton el al 2022).

Support for the visuoverbal disconnection hypothesis comes from cases with multiple distinct lesions. Alexia without agraphia has been observed in these atypical cases, such as those involving a combination of a splenial lesion and a left lateral geniculate nuclear lesion causing right homonymous hemianopia (88; 89). These cases lack damage to the striate or extra-striate cortex, ruling out agnosia or simultanagnosia. Thus, they support the hypothesis that visuoverbal disconnection alone is sufficient to cause alexia without agraphia.

Word form agnosia. A second hypothesis is that alexia without agraphia is a specialized visual agnosia called “word form agnosia.” (95; 93). The alexic deficit is magnified with brief presentation of words and cursive script rather than print (95). In a single patient with alexia without agraphia, there was increased difficulty identifying words acquired at an older age compared with those acquired at a younger age (28). Along with the letter-by-letter reading strategy and the word length effect, these phenomena are ascribed to difficulty perceiving words as whole, unitary structures.

The visual agnosia may not be specific for words. For example, a patient with pure alexia without agraphia had difficulty perceiving complex textures, a problem with local pattern analysis that may be relevant to word perception (72). Others have shown that alexia without agraphia can be associated with subtle disturbances in visual recognition of nonverbal items (15; 85). This agnosia may be restricted to the visual domain. When tested on imagery of letters and words, a patient was able to read more correctly when allowed to trace letters manually than when asked to mentally manipulate visual images of letters (08). A visual word form agnosia may also explain why some patients with alexia without agraphia have subtle writing problems, consisting of a surface dysgraphia, in which their spelling relies on phoneme-to-grapheme rules. They have trouble reading words with irregular spelling, like “yacht” and “colonel” (70; 78) perhaps because they have trouble accessing an internal lexicon for writing purposes.

Studies with functional MRI suggest the existence of a visual word form area in the left fusiform gyrus (62), a component of a larger network involved in reading (71). Some authors speculate that the visual agnosia of alexia without agraphia represents disruption of this visual word form area (21). A study of patients with hemianopic dyslexia and those with alexia without agraphia suggested that damage to the region of the visual word form area was specific to the latter group (67). A survey of a large group of patients with left hemispheric strokes suggested that disruption of this fusiform region is specifically associated with computing grapheme sequences, that is, small letter strings that convey meaning when assembled into words (48). Additional support for the role of the visual word form area in alexia without agraphia comes from studies of patients with surgical lesions examined with functional neuroimaging. These studies show that in patients who become alexic postoperatively, the lesion destroys either the visual word form area (41) or the adjacent white and gray matter (24), possibly causing deafferentation of the word form area. Intraoperative cortical stimulation of a region corresponding to the left visual word form area was shown to cause a temporary alexia without agraphia (60).

Some authors have proposed a third explanation for alexia without agraphia as a type of “ventral simultanagnosia” in which patients can only process incomplete fragments of visual percepts (36). Although standard tests for simultanagnosia in some patients have been negative (95), there have been challenges to the adequacy of these tests in the presence of alexia (36). Supportive of this concept are findings in alexia without agraphia of reduced apprehension span for digits and numbers (86).

Differential diagnosis

Confusing conditions

Developmental dyslexia. Alexia without agraphia is an acquired disorder, distinguishing it from developmental dyslexia, a disorder of reading competency first noted when the child is learning to read. The origin of childhood dyslexia remains unsettled, although there is a suggestion of dysfunction in the rapid magnocellular stream of stimulus processing (69; 83). Surface dyslexia presents with an inability to pronounce irregular words like “yacht” and “colonel” due to loss of an internal dictionary (81; 27), whereas phonological dyslexia involves the loss of general pronunciation rules and hinderance in pronunciation of pseudo-words or new words (40; 38).

Detailed neuroimaging studies of subjects with developmental dyslexia display loss of gray matter in the fusiform regions bilaterally (55). In Japanese patients, alexia for kana (syllabograms) that spares reading for kanji is associated with similar fusiform lesions (80) as well as reduced activation on fMRI (74).

Visual disorders. Reading will be impaired nonspecifically in several visual disorders. Patients with bilateral reduction in visual acuity of ocular or cerebral origin will read poorly. To avoid confusion, their acuity should be tested with a nonletter optotypes, preferably gratings.

Visual field defects that do not affect visual acuity can also impair reading. In bitemporal hemianopia, the fact that no region of the visual field is represented in both eyes leads to unstable binocular alignment. As the eyes wander between esotropia and exotropia, words transiently double or disappear during reading (54). Homonymous hemianopias affecting the central five degrees cause hemianopic dyslexia (91). Reading left-to-right is slowed more by right than left homonymous hemianopia (31; 91). Patients with left homonymous hemianopia have trouble finding the beginning of lines because the left margin disappears into their blind hemifield as they scan rightwards (91). Right homonymous hemianopia prolongs reading times, with more fixations and smaller saccades (31; 91). However, the reduction in reading speed is not as severe as that due to the word-length effect in alexia without agraphia (67). Right hemianopic alexia may perhaps be improved with an optokinetic therapy aimed at encouraging rightward saccades (84).

Aphasia. Impaired reading is a component of many aphasic disorders. Testing auditory and oral linguistic skills will readily reveal these nonvisual elements and lead to the correct diagnosis.

Hemispatial neglect. Patients with left hemispatial neglect make left-sided reading errors known as “neglect dyslexia” (14). They omit the left side of lines and make left-sided omissions, additions, or substitutions with words. Vertical text is not affected (14). Rarely, it may occur without other signs of hemispatial neglect (66).

Eye movement disorders. Acquired ocular motor apraxia from bilateral frontal or parietal lesions can impair reading (49; 11). Inaccurate or delayed saccades to nonverbal targets are found. Accompanying signs include simultanagnosia and optic ataxia. Interruption of fixation by saccadic intrusions, such as ocular flutter or opsoclonus, will disrupt reading.

Diagnostic workup

A diagnosis of alexia without agraphia requires exclusion of significant visual or linguistic dysfunction. Visual acuity needs to be adequate, and its measurement sometimes requires testing with nonverbal optotypes. Standard tests of auditory comprehension and oral language output must establish the lack of aphasia. The patient must be asked to write a short paragraph to exclude agraphia.

If the patient is unable to read, and confounding disorders have been excluded, instruct the patient to read single letters or numbers. With spelling dyslexia, reading will be laborious. The examiner should listen to see if longer words take more time to read, revealing the characteristic "word-length" effect.

Formal neuropsychological assessment is invaluable in assessing partial or subtle reading dysfunction, particularly when the examiner is unclear about interpreting the patient's current reading ability in terms of premorbid linguistic competency (46).

Neuroimaging with MRI is important for lesion detection. FDG positron emission tomography has been reported to show hypometabolism when MRI appears normal (44). Functional imaging studies have implicated an area in the left lateral fusiform gyrus known as the “visual word form area” (23).

Stereoelectroencephalography (sEEG) presents a unique method for inducing “transient” and highly localized electrophysiological disruptions to investigate brain behavior during functional mapping. In one study, this approach was utilized to reproduce alexia without agraphia by stimulating the lateral fusiform gyrus at coordinates nearly identical to those previously established for the visual word form area (VWFA) using functional imaging techniques. This demonstration highlights the potential of sEEG in providing diagnostic insights by replicating symptomatology through precise electrophysiological manipulation (77).

Management

There is paucity of controlled trials regarding rehabilitation of alexia without agraphia (87). However, many imaginative ways have been applied, including phonological and orthographic training, reading-aloud training, and semantic feature analysis. Phonological training involves teaching patients to use phonological strategies to decode words, whereas orthographic training focuses on improving recognition of whole words (13). Reading-aloud training involves practicing reading aloud, enhancing oral articulation during reading (25), and repetitive oral reading of text (12). Semantic feature analysis involves identifying semantic features of words to aid in word retrieval (13). This approach includes altering text to highlight the spacing between words or phrases (12; 59), attempts to enhance implicit or covert processing of whole words (59; 39; 01), and finger tracing of letters (kinesthetic treatment) (59; 64).

Others have tried a “face font” that uses faces to represent phonemes and found that a patient was able to learn this nonalphabetic reading system, although not at the level of control subjects (63). Functional improvements in whole-word recognition were associated with changes in reading network connectivity as measured with magnetoencephalography following therapy designed to enhance processing of whole words (96). Enhancement of behavioral training and favorable changes on functional MRI followed transcranial direct current stimulation therapy (57). A controlled trial also found a small facilitatory effect of transcranial direct stimulation on word reading accuracy generalized to trained and untrained words (52).

The success of these approaches in improving both speed and accuracy requires further evaluation and will likely require tailoring to the specific reading defect (58). For example, one report described poor results of kinesthetic treatment in a patient found to have both visuoverbal and kinesthetic-verbal disconnections (53).