Twelve years ago, clinicians gathered in Vancouver, Canada, for the first international working group meeting on posterior cortical atrophy. At the time, PCA was obscure. It lacked consensus on clinical diagnostic criteria, biomarkers, and even prevalence data. In PCA, amyloid plaques accumulate in the brain in a pattern that mimics Alzheimer’s disease, but neurofibrillary tangles sprout mainly in the occipital lobe, matching the posterior atrophy. PCA progressively hampers visuospatial processing, limiting perception and coordination. Yet memory and cognition tend to stay sharp, at least in the early years (for a detailed description of PCA, see Jul 2012 conference news). Back then, some neurologists considered PCA a rare form of Alzheimer’s disease; others did not.

  • A meta-analysis studied 1,092 people with PCA from 16 countries.
  • This rare form of young-onset dementia is almost always AD.
  • Structural MRI or FDG-PET are useful diagnostic biomarkers.

Now, after studying 1,092 people worldwide, the same clinicians, led by Gil Rabinovici at the University of California, San Francisco, together conclude that PCA, while rare, is almost always early onset AD. In the January 22 Lancet Neurology, they reported that PCA often begins around age 60, is driven by AD pathology in 94 percent of its sufferers, and affects women more often than men. ApoE4 increases risk, albeit less so than for amnestic AD. In fact, PCA symptoms indicate underlying AD pathology better than does mild cognitive impairment, the authors state.

"This paper is impressive. It stands to become the pivotal reference work for the PCA syndrome," wrote Marsel Mesulam, Northwestern University, Chicago (comment below).

Co-led by Sebastian Crutch at University College London, the Vancouver meeting led to classification criteria (Crutch et al., 2017), and the new paper represents the first large-scale test of those criteria.

“This work shows that even allowing centers to define PCA clinically using local criteria, these [diagnoses] closely match, and so validate, the core clinical features proposed in the PCA consensus classification framework for AD,” wrote co-author Jonathan Schott of UCL (comment below).

To characterize PCA in a large cohort that reflects the general population, co-first authors Marianne Chapleau and Renaud La Joie meta-analyzed published research studies on people clinically diagnosed with PCA who also had amyloid or tau PET scans, cerebrospinal fluid AD biomarker measurements, or brain autopsy data. They identified 1,092 people seen at 36 research centers in 16 countries across North and South America, Europe, Asia, and Australia. Six of 10 were women.

APOE4 seemed to up a person’s odds of developing PCA, as almost half of all participants carried one or more copies of the gene. APOE4 prevalence in cognitively normal people typically is below 25 percent but in AD ranges from 55 to 65 percent in Europe and North America; Chapleau and colleagues think APOE4 has a weaker link with PCA than with typical AD (reviewed by Crean et al., 2022).

On average, PCA symptoms began at age 59, people were diagnosed at 63, and survived for seven more years. The most common early symptom was difficulty copying a drawing. Half lacked depth perception, could no longer do simple math, or see more than one object at a time, a condition known as simultagnosia (image below). It means they literally cannot see the forest for the trees, or only notice a spoon on a fully set table. When quantified, these and other symptoms recapitulated the prevalence estimated in the consensus diagnosis criteria.

This Is PCA. The most common symptoms of posterior cortical atrophy span problems with visuospatial processing and motor coordination. [Courtesy of Chapleau et al., Lancet Neurology, 2024.]

By the time people with PCA were diagnosed, most had mild dementia. Sixty-two percent tallied a 1 or higher on the Clinical Dementia Rating, and the average Mini-Mental State Exam score was 20.7 out of 30. Women scored below men on the MMSE, again suggesting a sex difference.

It is notoriously difficult for people with PCA to get a correct diagnosis, as most clinicians tend not to suspect atypical AD in younger adults with intact memories. The average time for a diagnosis in this cohort, after symptom onset, was four years. By this time, half of them already had trouble learning and remembering new information, 40 percent struggled with executive function, and a third had speech or behavioral problems. The authors believe this reflects disease that has advanced from early stage visuospatial deficits to multidomain dementia.

People with PCA typically visit an ophthalmologist before an eye problem is ruled out, then see a neurologist, where they might still be misdiagnosed because their symptoms are so unusual. “Most memory clinics do not have the right tools to detect PCA at the early stages because neurologists tend to focus on memory tests and only ask in general terms if a person has problems seeing, reading, or writing,” co-author Yolande Pijnenburg of the Amsterdam University Medical Center told Alzforum.

“We hope this paper will raise awareness of PCA so eye-care specialists and neurologists will know to recognize and identify it early,” Rabinovici told Alzforum.

What biomarkers might point neurologists to PCA? Almost all participants poorly metabolized glucose in their posterior cerebral cortex, as seen on FDG-PET, and 85 percent had posterior atrophy on MRI. Why not 100 percent, given the disease etiology? Not everyone with PCA has frank neurodegeneration despite the disease name, explained Rabinovici. “In practice, this means that if the MRI is abnormal, there may be little to be gained by adding FDG-PET, but if the clinical syndrome suggests PCA and the MRI scan does not, then FDG-PET may provide additional evidence of posterior cortical dysfunction,” Schott said.

As for AD markers, almost every person in these studies was amyloid- and tau-PET-positive—94 and 97 percent, respectively. Fewer were positive for CSF biomarkers: 81 percent for Aβ42 and 65 percent for phospho-tau, mostly p-tau181, but isoform testing varied between clinics, Rabinovici noted. These data were reflected in autopsy findings, which were available for 145 people. Of these, 135 had a primary diagnosis of AD on autopsy, and 64 had Lewy bodies as a co-pathology.

Of the other 10 autopsy cases, four had a primary diagnosis of Lewy body disease (LBD with AD co-pathology), two had corticobasal degeneration (CBD), two had had strokes, one had Pick’s disease, and one had TDP-43 type A frontotemporal dementia due to a pathogenic GRN mutation. All but stroke cause atrophy of the posterior cortex and PCA symptoms in rare cases. Stroke precludes a PCA diagnosis, per the consensus criteria, as its onset is sudden.

The non-AD co-pathologies aligned with non-PCA symptoms some people reported. About 20 percent were diagnosed with PCA “plus.” They had symptoms and imaging indicative of PCA but on top of that they had symptoms of other neurodegenerative diseases, typically LBD or CBD. People with PCA plus were likelier to be amyloid-negative than those with pure PCA, hinting that non-AD diseases drive symptoms in the former.

Should people with PCA get Leqembi and related drugs? The authors believe that, for most of them, the answer might be yes. “Because PCA is strongly, but not always, associated with AD, it is important to determine if people with the syndrome have AD pathology to identify who might be candidates for disease-modifying therapies,” said Rabinovici. “Which AD biomarker to use will depend on whatever is accessible at each clinic.”

It is unclear how people with PCA might respond to Leqembi, the only disease-modifying AD therapy now available. The international PCA working group does not have a trial-ready cohort, but the Longitudinal Early Onset Alzheimer’s Disease Study (LEADS) recruits people with PCA (Aug 2023 conference news). LEADS plans to assemble a trial-ready cohort, following the model of the Dominantly Inherited Alzheimer Network. “It will include all early onset AD, not specifically PCA, but everyone will be well-characterized,” Rabinovici noted.

In the meantime, LEADS participants are allowed to take AD medication, including Leqembi, so the study will collect data on how disease-modifying therapy affects people with PCA. LEADS sites will open this year in the U.K., Netherlands, Sweden, Argentina, and Spain at clinics involved in this PCA meta-analysis.

This paper illustrates the power of clinician-researchers banding together to understand a rare disease. “A single center or two could have never found the richness of pathology and biomarker data seen in this paper, so we need the world to collaborate and study rare variants of Alzheimer’s disease,” Pijnenburg said.—Chelsea Weidman Burke


  1. This work is by far the largest study of PCA to date. It solidifies evidence from many smaller studies that PCA is a distinct syndrome, usually causes young-onset dementia, mean age at onset about 60 years, and is almost always underpinned by Alzheimer’s pathology. Patients with PCA are often mis-diagnosed, or diagnosed late, and so may not receive appropriate treatment, support, and information. This work shows that, even allowing centers to define PCA clinically using local criteria, these diagnoses closely match (and so validate) the core clinical features proposed in the PCA consensus classification framework for AD (Crutch et al, 2017).

    The very high probability that someone with PCA has underlying Alzheimer's pathology reinforces that such patients should be offered molecular diagnostic testing; and, especially as most patients are young and without other comorbidities, should not be denied access to new AD therapies and clinical trials.


    . Consensus classification of posterior cortical atrophy. Alzheimers Dement. 2017 Aug;13(8):870-884. Epub 2017 Mar 2 PubMed.

  2. This paper is impressive. It stands to become the pivotal reference work for the posterior cortical atrophy syndrome. It also reminds us that the mapping of syndrome to neuropathology is probabilistic rather than deterministic (Mesulam and Weintraub, 1992; Mesulam et al., 2014; Petersen et al., 2023). For example, a given neuropathologic entity (e.g., Alzheimer’s Disease Neuropathologic Change) can be associated with several syndromes, each association displaying a different probability and characteristic neuroanatomy. Conversely, each syndrome (e.g., PCA) can be associated with different neuropathologies, some more frequent (typical) than others. These two characteristics are summarized in the table below.

    A second point illustrated by this paper is relevant to the practicing clinician. While it is true that the vast majority of PCA is associated with AD, there are also definite cases of slow Creutzfeldt-Jacob, pure cortical Lewy body disease, and reportedly 4R tauopathy of the corticobasal type that can trigger the same syndrome. So, for the clinician who deals with individuals rather than groups, the value of careful differential diagnosis remains as important as ever.


    • Primarily amnestic AD (the typical form): Age- and ApoE4-related; neurofibrillary tangles (NFT) emerge and reach the highest densities in the hippocampal complex according to the Braak and Braak pattern. Known as amnestic MCI at early stages and dementia of the Alzheimer-type at later stages (DAT).
    • Primarily aphasic AD: Not age- or ApoE4-related; the ratio of language cortex to hippocampal NFT can be higher than in amnestic forms, in violation of the Braak and Braak pattern (Gefen et al., 2012). Known as primary progressive aphasia (PPA).
    • Primarily behavioral/executive AD: Relationship to age and ApoE4 to be determined; NFT can have higher density in frontal cortex than in amnestic forms (Johnson et al.,1999). Known as frontal-type dementia (FTD) or behavioral variant FTD (bvFTD).
    • Primarily visuospatial AD: Age-related but relationship to ApoE4 is not as high as in the amnestic form; NFT can have higher density in visual cortex and superior colliculus than in amnestic forms (Hof et al.,1993). Known as posterior cortical atrophy (PCA).


    . Spectrum of primary progressive aphasia. Baillieres Clin Neurol. 1992 Nov;1(3):583-609. PubMed.

    . Asymmetry and heterogeneity of Alzheimer's and frontotemporal pathology in primary progressive aphasia. Brain. 2014 Apr;137(Pt 4):1176-92. Epub 2014 Feb 25 PubMed.

    . A New Framework for Dementia Nomenclature. JAMA Neurol. 2023 Dec 1;80(12):1364-1370. PubMed.

    . Clinically concordant variations of Alzheimer pathology in aphasic versus amnestic dementia. Brain. 2012 May;135(Pt 5):1554-65. PubMed.

    . Clinical and pathological evidence for a frontal variant of Alzheimer disease. Arch Neurol. 1999 Oct;56(10):1233-9. PubMed.

    . Posterior cortical atrophy in Alzheimer's disease: analysis of a new case and re-evaluation of a historical report. Acta Neuropathol. 1993;86(3):215-23. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Researchers Join to Draw Posterior Cortical Atrophy Out of Shadows
  2. Cohort LEADS Toward Better Understanding of Sporadic Early Onset AD

Therapeutics Citations

  1. Leqembi

Paper Citations

  1. . Consensus classification of posterior cortical atrophy. Alzheimers Dement. 2017 Aug;13(8):870-884. Epub 2017 Mar 2 PubMed.
  2. . Apolipoprotein E ε4 prevalence in Alzheimer's disease patients varies across global populations: a systematic literature review and meta-analysis. Dement Geriatr Cogn Disord. 2011;31(1):20-30. PubMed.

External Citations

  1. Longitudinal Early Onset Alzheimer’s Disease Study

Further Reading

No Available Further Reading

Primary Papers

  1. . Demographic, clinical, biomarker, and neuropathological correlates of posterior cortical atrophy: an international cohort study and individual participant data meta-analysis. Lancet Neurol. 2024 Feb;23(2):168-177. PubMed.