Primary progressive aphasia (PPA) leads to a gradual and relatively isolated dissolution of language function. The factors that determine the selectivity of the disease process remain unknown. We had speculated that PPA may occasionally arise as a tardive manifestation of genetic or acquired vulnerabilities involving the language network of the brain.
To explore predisposing factors for PPA.
In 2 patients, PPA developed with a background of mild left hemicranial hypoplasia.
In keeping with other observations of PPA in patients with dyslexia and childhood injury to the left temporal lobe, these 2 patients support the contention that some cases of PPA may arise in settings where the language network has become a locus of least resistance.
Primary progressive aphasia (PPA) is a clinical syndrome characterized by a relatively isolated dissolution of language function.1,2 The diagnosis of PPA is made when other mental faculties, such as memory for daily events, visuospatial skills, and attention remain relatively intact; when there are no major changes in personality or behavior; when language is the only area of progressive dysfunction for at least the first 2 years of the disease; and when structural brain imaging does not reveal a specific lesion, other than atrophy, that can account for the language deficit.3,4
Many patients with PPA display atrophy (indicative of neuronal loss), electroencephalographic slowing, decreased blood flow (measured by single-photon emission computed tomography [SPECT]), and decreased glucose utilization (measured by positron emission tomography) predominantly within the language-related areas of the left hemisphere.3,5- 8 The metabolic state of the contralateral right hemisphere can remain within the normal range, especially early in the disease.5,6 The reported neuropathologic changes in PPA have tended to be more pronounced in the left frontal, perisylvian, and temporal cortex than in the hippocampal or entorhinal areas most damaged by Alzheimer disease (AD).9- 11 Approximately 70% of the cases have displayed non-AD neuropathologic characteristics (neuronal loss and gliosis, occasionally accompanied by Pick bodies or other forms of taupathy), whereas the others have shown the plaques and tangles of AD, sometimes in an unusual distribution favoring language-related areas of the brain.
What accounts for the focal nature of the degeneration and cognitive deficits in PPA? The possibility was raised that PPA may occur in individuals with genetic or acquired vulnerabilities that affect language-related areas of the brain.3 The purpose of this report was to provide circumstantial support for this speculation with the help of 2 patients who developed PPA on a background of left hemisphere hypoplasia.
A 63-year-old ambidextrous woman developed difficulties naming objects, followed by trouble understanding spoken language. There was no prior history of aphasia. When she was seen 2 years later, she was still fully oriented and able to run her household, do the shopping, buy her clothes, travel to other cities, remember her appointments, and play bingo. Clinical examination displayed multiple word-finding hesitations. Naming by visual confrontation was impaired. Repetition of phrases was impaired. She was unable to obey complex commands. Reading and writing could not be appropriately assessed due to her low educational level. She had good memory for daily events. Results of routine laboratory studies were normal, and the apolipoprotein E genotype was 3/3. Left hemicranial hypoplasia, predominantly of the middle cranial fossa, and left perisylvian atrophy were noticed on neuroimaging studies (Figure 1). The cranial asymmetry was compatible with mild unilateral left coronal suture synostosis. Left hemisphere hypoperfusion with a perisylvian predominance was seen on SPECT (Figure 1). The aphasia slowly progressed, and at the age of 72 years, 9 years after onset, she could barely utter more than 2 to 3 words at a time, could not carry out simple commands, and needed help with most of her activities of daily living. By then, her right arm swing was slightly impaired when walking, there was a fixed facial expression, and both sucking and snout reflexes were elicited. Cerebral atrophy had progressed on magnetic resonance images, and hypoperfusion had become bilateral, but still with a left hemisphere predominance.
Patient 1. A, Axial computed tomographic scans. The middle and anterior fossas appear smaller on the left, with slightly thicker bony walls (asterisks). The skewed angle of the skull and facial structures, and the flat left frontal bone suggest mild unilateral coronal suture synostosis on the left. There is also left perisylvian atrophy (arrows). B, Axial single-photon emission computed tomographic scans obtained at the same time. There is left temporal and perisylvian hypoperfusion (arrows).
A 43-year-old right-handed man had 2 focal right-sided somatosensory epileptic events, followed a few days later by a grand mal seizure. Left temporal spikes were detected on an electroencephalogram. The epilepsy was controlled with carbamazepine. A computed tomographic scan obtained 5 years later (1980) noted cranial asymmetry based on a smaller left middle cranial fossa and a larger frontal ventricular horn on the left. No temporal or perisylvian cortical atrophy on the left was noted. Sixteen years later, at the age of 59 years, he complained of progressive word-finding difficulties as well as additional problems in writing and calculations. Electroencephalographic findings at that time were normal. He retired from his office job at age 63 years because of his language problems. At an examination at that time, 4 years after symptom onset, he was fully oriented, remembered daily events, and carried out all customary daily activities, including shopping and banking. He had a nonfluent aphasia. Conversational speech was effortful, with a slight stammering quality, mild dysarthria, loss of prosody, and circumlocutions. He also had acalculia, left-right disorientation, and ideomotor apraxia, although the aphasia was by far the most dominant finding and the chief cause of his limitations at work. Routine laboratory tests were unrevealing. Apolipoprotein E genotype was 3/3. Cranial computed tomographic scan and magnetic resonance imaging at that time (1996) showed the left hemicranial hypoplasia that had been seen before and additional left frontotemporal cerebral atrophy with perisylvian predominance (Figure 2). Left hemisphere hypoperfusion was reported on SPECT. At age 69 years, 10 years after the onset of aphasic problems, he was still able to choose his clothes, shave, dress, and travel without getting lost, even though he could barely speak. At that time, the cerebral atrophy had increased on magnetic resonance imaging and the SPECT hypoperfusion became more generalized, but still with a left perisylvian predominance.
Patient 2. A, Axial computed tomographic scans obtained at different times and acquired at different angles. All 3 scans show thicker bone in the middle cranial fossa on the left (asterisk) and a smaller left temporal pole (arrow). Other bony landmarks indicate that the head was adequately centered. B, Coronal T1-weighted magnetic resonance imaging (MRI). The larger size of the frontal ventricular horn on the left (double arrow) had been noted in 1980 before the onset of the aphasia. The slightly greater Sylvian cistern on the left and the mild atrophy of the left temporal lobe (single arrows) had not been noted in 1980 and may be of more recent origin. C, Axial single-photon emission computed tomographic scans, obtained at the same time as the MRI, showing left temporal and perisylvian hypoperfusion (arrows).
The patients in this report had 2 features in common. First, they had progressive neurologic impairment that initially exhibited a relatively isolated aphasic disturbance consistent with the diagnosis of typical PPA without any known history of childhood stuttering or learning disability. Second, they had radiologic evidence for a hypoplastic left hemisphere, predominantly within the middle cranial fossa, where the major components of the language network are located. The cranial asymmetry in these patients reflects a remote process that originated at the stage of craniocerebral development, decades before the emergence of the aphasia. At least in case 2, the cortical atrophy in the left hemisphere appears to represent a more recent event, perhaps temporally related to the onset of the progressive aphasia. The absence of a language deficit prior to the onset of the PPA indicates that the putative cognitive effect of this hypoplasia had been compensated. However, the focal seizures in patient 2 also indicate that the left hemisphere process was not entirely silent.
We had previously described a patient with a history of abscess removal from the left temporoparietal area at the age of 11 years. Although the surgery during childhood left no aphasic sequelae, he experienced a focal dissolution of language function almost 60 years later.3 There is also preliminary evidence that patients with PPA and their first-degree relatives have a greater incidence of developmental learning disabilities, especially dyslexia, when compared with patients with AD or normal controls.3 These relationships, together with the 2 cases reported here, support the possibility that PPA, at least in some patients, may reflect the tardive manifestation of a preexisting genetic or acquired vulnerability centered around the left hemisphere language network.
Premorbid dyslexic deficits in such patients could conceivably reflect the mild manifestations of a genetically determined left hemisphere dysfunction that eventually leads to PPA, perhaps in response to the additional biological stress of aging. An analogous phenomenon was identified in families with chromosome 17 mutations known to cause frontotemporal dementia. Carriers of such mutations displayed mild neuropsychologic evidence of frontal lobe dysfunction decades prior to the expected onset of the dementia.12
There are several settings in which acquired lesions provide a locus of least resistance for the clinical expression of subsequent disease processes. Women who recover from Sydenham chorea in childhood, for example, can experience chorea gravidarum during pregnancy in response to alterations of the hormonal milieu; patients who have recovered from poliomyelitis can develop, decades later, a progressive motor neuron disease in the previously affected muscles13; and patients who have recovered from childhood hemiplegia can develop a progressive hemiparkinsonism on the side of the recovered weakness later in life.14
The relationship between PPA and left hemicranial hypoplasia in the 2 patients described here may be coincidental. However, it is also possible that there is a causal link between the 2 events, supporting the contention that PPA may occasionally arise on a background of focal vulnerabilities affecting the language network of the brain. In some patients, this vulnerability may be visible in the form of hypoplasia. In others, it may elude detection by currently available imaging approaches. The clinical picture in such patients, and perhaps in many other focal degenerations, would thus reflect an interaction between the molecular aspects of the disease and host-specific regional vulnerabilities. According to this formulation, the incidence of hemicranial hypoplasia need not be elevated in PPA and, conversely, the incidence of PPA need not be elevated in hemicranial hypoplasia. The implication of the present report is that early injuries or developmental perturbations such as hemicranial hypoplasia may delineate the area of greatest involvement in some patients who later develop a disease that promotes cortical degeneration. It would be important to determine the frequency of this phenomenon in a larger sample of patients and to explore its biological mechanisms.
Corresponding author: Marsel Mesulam, MD, Cognitive Neurology and Alzheimer's Disease Center, 320 E Superior St, Chicago, IL 60610 (e-mail: email@example.com).
Accepted for publication August 12, 2003.
Author contributions: Study concept and design (Drs Alberca, Gil-Néciga, and Mesulam); acquisition of data (Drs Alberca and Montes); analysis and interpretation of data (Drs Russell and Mesulam); drafting of the manuscript (Drs Alberca, Montes, and Mesulam); critical revision of the manuscript for important intellectual content (Drs Alberca, Russell, Gil-Néciga, and Mesulam); administrative, technical, and material support (Drs Alberca, Montes, and Russell); study supervision (Drs Alberca and Mesulam).
This study was supported by grant AG138S4 from the National Institute on Aging, Bethesda, Md (Alzheimer's Disease Center, Northwestern University, Chicago, Ill).
Alberca R, Montes E, Russell E, Gil-Néciga E, Mesulam M. Left Hemicranial Hypoplasia in 2 Patients With Primary Progressive Aphasia. Arch Neurol. 2004;61(2):265-268. doi:10.1001/archneur.61.2.265