Key PointsQuestion
Do study participants with biomarker-positive findings for preclinical Alzheimer disease have retinal microvascular alterations detectable by optical coherence tomographic angiography compared with control individuals with biomarker-negative findings?
Findings
In this single-center, case-control study, the foveal avascular zone was larger in participants with preclinical Alzheimer disease determined by the presence of β-amyloid biomarkers (mean [SD], 0.364 [0.095] mm2) compared with those without preclinical Alzheimer disease (mean [SD], 0.275 [0.060] mm2).
Meaning
Foveal avascular zone enlargement may offer a noninvasive, cost-efficient, and rapid screen to identify preclinical Alzheimer disease.
Importance
Biomarker testing for asymptomatic, preclinical Alzheimer disease (AD) is invasive and expensive. Optical coherence tomographic angiography (OCTA) is a noninvasive technique that allows analysis of retinal and microvascular anatomy, which is altered in early-stage AD.
Objective
To determine whether OCTA can detect early retinal alterations in cognitively normal study participants with preclinical AD diagnosed by criterion standard biomarker testing.
Design, Setting, and Participants
This case-control study included 32 participants recruited from the Charles F. and Joanne Knight Alzheimer Disease Research Center, Washington University in St Louis, St Louis, Missouri. Results of extensive neuropsychometric testing determined that all participants were cognitively normal. Participants underwent positron emission tomography and/or cerebral spinal fluid testing to determine biomarker status. Individuals with prior ophthalmic disease, media opacity, diabetes, or uncontrolled hypertension were excluded. Data were collected from July 1, 2016, through September 30, 2017, and analyzed from July 30, 2016, through December 31, 2017.
Main Outcomes and Measures
Automated measurements of retinal nerve fiber layer thickness, ganglion cell layer thickness, inner and outer foveal thickness, vascular density, macular volume, and foveal avascular zone were collected using an OCTA system from both eyes of all participants. Separate model III analyses of covariance were used to analyze individual data outcome.
Results
Fifty-eight eyes from 30 participants (53% female; mean [SD] age, 74.5 [5.6] years; age range, 62-92 years) were included in the analysis. One participant was African American and 29 were white. Fourteen participants had biomarkers positive for AD and thus a diagnosis of preclinical AD (mean [SD] age, 73.5 [4.7] years); 16 without biomarkers served as a control group (mean [SD] age, 75.4 [6.6] years). The foveal avascular zone was increased in the biomarker-positive group compared with controls (mean [SD], 0.364 [0.095] vs 0.275 [0.060] mm2; P = .002). Mean (SD) inner foveal thickness was decreased in the biomarker-positive group (66.0 [9.9] vs 75.4 [10.6] μm; P = .03).
Conclusions and Relevance
This study suggests that cognitively healthy individuals with preclinical AD have retinal microvascular abnormalities in addition to architectural alterations and that these changes occur at earlier stages of AD than has previously been demonstrated. Longitudinal studies in larger cohorts are needed to determine whether this finding has value in identifying preclinical AD.
Alzheimer disease (AD) is the most common form of dementia, affecting an estimated 5.4 million US residents.1 The pathophysiologic changes of AD involve loss of neurons, brain atrophy, extracellular deposition of β-amyloid (Aβ) plaques, and intracellular accumulation of neurofibrillary tangles.2,3 Unfortunately, the classic clinical symptoms of AD, including progressive memory loss and behavioral changes, are only apparent after massive, irreversible neuronal loss has occurred. Preclinical AD is a recently recognized period in which the key pathophysiologic changes are under way within the brain, but symptoms have not yet become apparent.4
Preclinical AD can be diagnosed based on the presence of clinically validated biomarkers measuring amyloid burden within the central nervous system. Carbon 11–labeled Pittsburgh Compound B (PiB) (N-methyl-[11C]2-(4′-methylaminophenyl)-6-hydroxybenzothiazole; not commercially available)5 and fluorine 18–labeled florbetapir (18F-AV-45; Amyvid) compounds bind amyloid protein within central nervous system tissue and can estimate disease burden when viewed by positron emission tomography (PET).5-9 In addition, levels of Aβ42 and τ protein in the cerebrospinal fluid (CSF) can be quantified in samples acquired by lumbar puncture.2,10,11
Both tests for biomarker status have especially high negative predictive value in assessing the risk of developing clinically detectable AD.10,12,13 In addition, both tests have been validated in long-term longitudinal studies to estimate onset of clinical dementia,14 such that positive findings for either test is considered diagnostic of preclinical AD.11 Although these methods are useful in assessing individuals at risk for AD, they are expensive, time-consuming, invasive, and difficult to implement in routine clinical screening and care.
Recent data have suggested that AD is also marked by vascular dysfunction, although whether the dysfunction is secondary or contributes to the Aβ accumulation is unclear.15 In the retina specifically, venous narrowing and reduced blood flow have been established in individuals with AD16-18 and mild cognitive impairment (MCI).19 A small study using optical coherence tomographic angiography (OCTA) to compare patients with MCI and those with advanced AD20 suggested decreased density of the deep vascular plexus specifically. However, determination of disease status was based on results of neuropsychiatric testing (eg, Mini-Mental State Examination) rather than objective biomarker status, which has been shown to be inaccurate in estimating conversion from MCI to dementia,21 because it is influenced by other factors such as socioeconomic status, level of education, and presence of confounding neuropsychiatric disorders such as depression and stroke.22
Because clinical trials are under way to evaluate new drugs designed to prevent neuronal loss, it is imperative to be able to identify which individuals with preclinical AD would benefit from potential therapy. Currently accepted testing methods are expensive and invasive. In this study, we evaluated whether OCTA technology has the potential to characterize early retinal architecture and vascular changes in individuals with preclinical AD.
Cognitively normal study participants were recruited from the Charles F. and Joanne Knight Alzheimer Disease Research Center (ADRC) of Washington University in St Louis, St Louis, Missouri. Study participants were volunteers in the Memory and Aging Project of the ADRC. The study design was approved by the institutional review board of Washington University in St Louis at the Human Research Protection Office and adhered to the tenets of the Declaration of Helsinki.23 Risks and benefits were discussed with each individual, and written informed consent was obtained before beginning the ophthalmologic examination.
Data were collected from July 1, 2016, through September 30, 2017. Inclusion criteria required a Clinical Dementia Rating classification of 0 (no evidence of dementia). The Clinical Dementia Rating is a 5-point scale used to characterize 6 domains of cognitive function and performance to evaluate Alzheimer type dementia, including memory, orientation, judgment and problem solving, community affairs, home and hobbies, and personal care, based on an extensive battery of neuropsychometric tests (eTable in the Supplement).
Additional inclusion criteria consisted of completion of PET imaging for PiB or 18F-AV-45 compound or CSF analysis of Aβ42 protein level within 1 year of recruitment; many participants underwent both tests. Biomarker status was kept by the ADRC during data collection stage so that testing and data gathering were performed in a masked manner. Additional data regarding age, sex, self-reported ethnicity, and medical history were collected from a review of the medical records. Information on family history or genetic testing (such as APOE4 allele status) was not collected in this study.
Exclusion criteria included previously diagnosed, clinically apparent AD. Additional exclusion criteria consisted of a known history of glaucoma or age-related macular degeneration; intraocular pressure of 22 mg Hg or higher; dense media opacity precluding measurement; history of ocular trauma or concomitant ocular diseases, including previous retinal disease; presence of significant refractive error (more than 5 diopters [D] of spherical equivalent refraction or 3 D of astigmatism); and previous retinal laser therapy. Additional medical exclusion criteria included diabetes and uncontrolled hypertension.
All participants received a complete neuro-ophthalmic examination, including standard assessment of Snellen visual acuity, color perception using Ishihara color plates, ocular motility, intraocular pressure, refractive status, and examination of the anterior segment and dilated fundus. Optical coherence tomographic imaging of the optic disc and macula and OCTA were performed using the Avanti Optovue OCTA system (Optovue, Inc). Measurements were automated using the manufacturer’s software (Optovue RTVue) from 6 OCT images per eye and thus collected in an objective manner. Although the software reproducibility has been substantiated in measuring central subfield thickness in diabetic macular edema,24 each data point was reviewed by two of us (B.E.O. and N.K.) to evaluate for potential confounding pathologic findings (eg, optic nerve head drusen) and subjective appropriateness of the measurements. Data outcomes collected included total and temporal retinal nerve fiber layer thickness; ganglion cell layer thickness; macular volume; inner, outer, and total foveal thickness; total macular, foveal, and parafoveal vascular density; and foveal avascular zone (FAZ) (Figure 1).
Data were analyzed from July 30, 2016, through December 31, 2017. Data outcomes measured on a ratio scale were analyzed using mixed-effects analysis of covariance, whereas data outcomes measured on a percentage scale were analyzed using mixed-effects generalized linear models (GLIMMIX in SAS software; SAS, Inc). Data points for each eye were treated as repeated measurements for the study participant. Separate analyses were run for CSF alone and PET alone and group analysis to compare all participants with at least 1 positive biomarker finding with participants without either biomarker. Age was included as a covariate in the models. Intraocular pressure was also included as a covariate in analyzing retinal nerve fiber layer and ganglion cell layer data. Two-sided P values were generated using SAS software (version 9.4), and these P values were not adjusted because comparisons were not made between the CSF group, PET group, or combined CSF-PET biomarker group.
A total of 32 study participants were recruited through the Washington University in St Louis ADRC. One patient was excluded owing to suspected undiagnosed normal tension glaucoma based on an increased cup-disc ratio; another was excluded owing to the presence of bilateral optic nerve head drusen. One eye was excluded owing to a full-thickness macular hole and another for vitreomacular traction causing distortion of the retinal architecture. Four images were excluded owing to motion artifact or segmentation error; an additional 6 images were excluded owing to poor automated mapping that did not accurately represent the optic nerve disc or FAZ. Data were collected from 58 eyes of 30 participants (16 women [53%] and 14 men [47%]; mean [SD] age, 74.5 [5.6] years; age range, 62-92 years) for inclusion in the analysis.
Mean (SD) age of participants with biomarker-positive status was 73.5 (4.7) years; of participants with biomarker-negative status, 75.2 (6.6) years. One participant was African American; the remainder reported white race. Among the biomarker-negative group, 10 of 16 (62%) were women; among the biomarker-positive group, 6 of 14 (43%) were women. Common comorbidities included medically controlled hypertension, hyperlipidemia, and depression.
Twenty-seven individuals completed PET scanning for PiB or 18F-AV-45 binding. Of these, 7 individuals had positive findings for preclinical AD. The mean (SD) age of the PET-negative group was 73.2 (4.6) years; of the PET-positive group, 76.4 (7.6) years old. Mean (SD) FAZ was larger in participants with PET-positive status (0.398 [0.066] mm2) compared with PET-negative controls (0.288 [0.0915] mm2; P < .001) (Figure 2A).
Twenty-eight individuals completed CSF sampling and analysis. Ten had Aβ42-positive findings and 18 had Aβ42-negative findings. Mean (SD) age of the Aβ42-positive group was 75.7 (7.2) years; of the Aβ42-negative group, 73.1 (4.4) years. Outer foveal measurements were thinner in the Aβ42-positive group (180.8 [8.8] μm) than the Aβ42-negative group (189.3 [10.0] μm; P = .03) (Figure 2B), as was total foveal thickness (245.9 [16.6] vs 263.0 [17.4] μm; P = .03) (Figure 2B).
All Biomarker-Positive Findings
Additional analysis was performed comparing individuals with results positive for the CSF or the PET biomarker compared with those with negative results. Inner foveal thickness was smaller in the biomarker-positive group (66.0 [9.9] μm) compared with the biomarker-negative group (75.4 [10.6] μm; P = .03) (Figure 2C). The FAZ was larger in participants with biomarker-positive findings (0.364 [0.095] mm2) compared with those with biomarker-negative findings (0.275 [0.060] mm2; P = .002) (Figure 2D).
A receiver operating characteristics (ROC) curve was generated for the FAZ in the all-biomarker analysis (Figure 3). The area under the curve was found to be 0.8007 (95% CI, 0.6647-0.9367). Given the limited sample size, single, lower 95% CI points were generated for the point along the ROC curve closest to the nondiscriminatory diagonal (50:50) line, assuming normal distribution and binomial distribution (0.26087 and 0.4166).
Our data suggest that individuals with biomarker-positive, preclinical AD might have retinal vascular and architectural alterations that are apparent before the onset of clinically detectable cognitive symptoms. This finding may be interpreted to imply that the retina undergoes neuronal loss and vascular modifications far earlier in disease progression than previously thought. A similar phenomenon is seen with AD-associated cerebral neuronal loss, which begins far in advance of symptom onset. However, these findings could be owing to confounding factors unrelated to the FAZ enlargement, and longitudinal studies in larger cohorts are needed to determine whether this finding has value in identifying preclinical AD.
Our findings of inner foveal thinning in participants with biomarker-positive test results are consistent with those of prior studies using traditional OCT technology and early autopsy studies.25-27 Unfortunately, although the difference between groups for disease status is statistically significant, the considerable overlap in distribution makes these findings of little use clinically.
We also observed dropout of vasculature specifically within the fovea, leading to enlargement of the FAZ in the biomarker-positive group. Since 2007, studies8,15,28-30 have reported that vascular dysfunction in individuals with MCI and AD leads to cerebral hypoperfusion during AD development. Older in vivo and autopsy data31-34 demonstrated that AD is associated with deposition of amyloid and collagen within the cerebral capillaries, resulting in cellular apoptosis and vessel dropout. Because retinal and cerebral vasculature are anatomically and physiologically homologous,30-33,35 the retinal vasculature may similarly be affected in AD progression; however, our study is observational and does not investigate causative mechanisms.
Another potential explanation for FAZ enlargement in individuals with preclinical AD may be secondary to retinal degeneration from Aβ accumulation within the retina itself. Several studies have demonstrated accumulation of Aβ plaques in the inner retina of postmortem tissue from individuals with AD36-39; although a few sources37,39 suggested that the accumulation is limited to the superior retinal tissue, most studies did not comment on location of the deposits. However, other studies in human tissues did not identify retinal Aβ,40 and still others suggest that τ accumulation may be more significant.41,42 A meta-analysis of the current literature published on retinal amyloid plaques ultimately concluded that “the limited number of eligible studies and their methodological heterogeneity make it impossible to come to a conclusion whether pathological retinal Aβ detection is an effective diagnostic tool for AD.”43
The difference in FAZ distribution between individuals with biomarker-positive and biomarker-negative findings (Figure 2D) provides a potentially clinically useful screening tool, if further studies confirm a false-positive rate of less than 40% as suggested by the ROC curve (Figure 3). Despite a promising area under the curve in the ROC with a lower 95% CI of greater than 0.5, larger, longitudinal studies may not validate our findings. If the final outcome confirms the lower 95% CI at the data points closest to the diagonal line, FAZ would prove to be a poor discriminatory marker in screening for preclinical AD.
Although we excluded participants with diabetes and uncontrolled hypertension from this study, we acknowledge that multiple other potential causes for an enlarged FAZ exist and that further assessment of OCTA in the general population is necessary. Despite this possible limitation, our data suggest that OCTA has the potential for rapid, noninvasive, and cost-effective identification of individuals who are likely to have preclinical AD unless these findings are owing to confounding factors unrelated to the FAZ enlargement. As noted, longitudinal studies in larger cohorts are be needed to determine whether this finding has value in identifying preclinical AD.
Strengths and Limitations
Strengths of our study include the use of biomarkers to identify individuals with preclinical AD. Previously published studies rely on the use of neurocognitive testing, namely, the Mini-Mental State Examination, to identify individuals with early dementia; however, a 2015 Cochrane review21 concluded that the Mini-Mental State Examination alone, without supporting testing or repetitive testing, could not accurately estimate conversion from MCI to dementia.
The PET and CSF biomarkers have been clinically validated and correlated with postmortem autopsy study findings4,6 and have been validated in longitudinal studies as an early diagnostic marker of individuals who will develop clinically significant Alzheimer-type dementia.14 In a comparative study, both biomarkers were found to be equally accurate in identifying early-stage AD44 with a relatively high concordance of approximately 80%.45 In our study, 5 participants underwent PET testing and lumbar puncture with conflicting results; in 4, PET findings were negative but CSF findings were positive; in 1, PET findings were positive but CSF findings were negative. Overall, the discordance rate was 15.6%. Although the participants with biomarker-negative findings who had only 1 biomarker available may have been misclassified, more likely these discrepancies are merely associated with the specificity of the individual test and are in line with a low rate of discordance.45-47 As such, any individual with a positive marker was considered to have biomarker-positive findings in the collective analysis.
A major limitation of our study is the small sample size, including a limited number of nonwhite individuals. An additional limitation is exclusion of individuals with known vascular disease from our study; we are therefore unable to determine whether these results are translatable to individuals who may have retinal microvascular changes due to other causes. Also, inclusion only of those with preclinical, biomarker-positive disease limits comparison to those with cognitive changes or advanced AD. Recruitment is under way to evaluate individuals with biomarker-positive MCI and more advanced AD and to follow up individuals with biomarker-positive findings over time for longitudinal evaluation of changes in retinal vasculature.
At present, preclinical AD is diagnosable only by invasive, expensive, and time-consuming PET or CSF testing. Our data suggest that OCTA may enable quick, inexpensive, and noninvasive screening for individuals with preclinical AD based on FAZ enlargement. However, these findings could be owing to confounding factors unrelated to the FAZ enlargement. Longitudinal studies in larger cohorts would be needed to determine whether this finding has value in identifying preclinical AD, so that these individuals may receive appropriate care.
Accepted for Publication: June 24, 2018.
Corresponding Authors: Rajendra S. Apte, MD, PhD (apte@wustl.edu), and Gregory P. Van Stavern, MD (vantaverng@wustl.edu), Department of Ophthalmology and Vision Science, Washington University in St Louis, 660 S Euclid Ave, St Louis, MO 63110.
Published Online: August 23, 2018. doi:10.1001/jamaophthalmol.2018.3556
Author Contributions: Drs Apte and Van Stavern were co–principal investigators. Drs O’Bryhim and Coble had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Apte, Van Stavern.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: O’Bryhim.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: O’Bryhim, Coble.
Obtained funding: Van Stavern.
Administrative, technical, or material support: O’Bryhim, Apte, Van Stavern.
Supervision: Apte, Van Stavern.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Funding/Support: This study was supported in part by an unrestricted grant from Research to Prevent Blindness, Inc (Department of Ophthalmology and Vision Sciences at Washington University in St Louis) and by an educational grant from Optovue, Inc.
Role of the Funder/Sponsor: The sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Meeting Presentation: This study was presented at the 4th Annual North American Neuro-Ophthalmology Society Meeting; March 6, 2018; Waikaloa Village, Hawaii; at the 41st Annual Macula Society Meeting; February 23, 2018; Beverly Hills, California; and the 70th Annual American Academy of Neurology meeting; April 23, 2017; Los Angeles, California.
Additional Information: Washington University in St Louis has filed intellectual property related to the discoveries described in this article.
4.Sperling
RA, Aisen
PS, Beckett
LA,
et al. Toward defining the preclinical stages of Alzheimer’s disease: recommendations from the National Institute on Aging–Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease.
Alzheimers Dement. 2011;7(3):280-292. doi:
10.1016/j.jalz.2011.03.003PubMedGoogle ScholarCrossref 8.Johnson
KA, Sperling
RA, Gidicsin
CM,
et al; AV45-A11 study group. Florbetapir (F18-AV-45) PET to assess amyloid burden in Alzheimer’s disease dementia, mild cognitive impairment, and normal aging.
Alzheimers Dement. 2013;9(5)(suppl):S72-S83. doi:
10.1016/j.jalz.2012.10.007PubMedGoogle ScholarCrossref 9.Clark
CM, Schneider
JA, Bedell
BJ,
et al; AV45-A07 Study Group. Use of florbetapir-PET for imaging beta-amyloid pathology [published correction appears in
JAMA. 2011;305(11):1096].
JAMA. 2011;305(3):275-283. doi:
10.1001/jama.2010.2008PubMedGoogle ScholarCrossref 12.Morris
E, Chalkidou
A, Hammers
A, Peacock
J, Summers
J, Keevil
S. Diagnostic accuracy of (18)F amyloid PET tracers for the diagnosis of Alzheimer’s disease: a systematic review and meta-analysis.
Eur J Nucl Med Mol Imaging. 2016;43(2):374-385. doi:
10.1007/s00259-015-3228-xPubMedGoogle ScholarCrossref 15.Di Marco
LY, Venneri
A, Farkas
E, Evans
PC, Marzo
A, Frangi
AF. Vascular dysfunction in the pathogenesis of Alzheimer’s disease: a review of endothelium-mediated mechanisms and ensuing vicious circles.
Neurobiol Dis. 2015;82:593-606. doi:
10.1016/j.nbd.2015.08.014PubMedGoogle ScholarCrossref 18.Golzan
SM, Goozee
K, Georgevsky
D,
et al. Retinal vascular and structural changes are associated with amyloid burden in the elderly: ophthalmic biomarkers of preclinical Alzheimer’s disease.
Alzheimers Res Ther. 2017;9(1):13. doi:
10.1186/s13195-017-0239-9PubMedGoogle ScholarCrossref 19.Feke
GT, Hyman
BT, Stern
RA, Pasquale
LR. Retinal blood flow in mild cognitive impairment and Alzheimer’s disease.
Alzheimers Dement (Amst). 2015;1(2):144-151.
PubMedGoogle Scholar 21.Arevalo-Rodriguez
I, Smailagic
N, Roqué I Figuls
M,
et al. Mini-Mental State Examination (MMSE) for the detection of Alzheimer’s disease and other dementias in people with mild cognitive impairment (MCI).
Cochrane Database Syst Rev. 2015;(3):CD010783.
PubMedGoogle Scholar 24.Bressler
SB, Edwards
AR, Andreoli
CM,
et al; for the Diabetic Retinopathy Clinical Research Network/Writing Committee. Reproducibility of Optovue RTVue optical coherence tomography retinal thickness measurements and conversion to equivalent Zeiss Stratus metrics in diabetic macular edema.
Transl Vis Sci Technol. 2015;4(1):5. doi:
10.1167/tvst.4.1.5PubMedGoogle ScholarCrossref 34.Farkas
E, De Jong
GI, de Vos
RA, Jansen Steur
EN, Luiten
PG. Pathological features of cerebral cortical capillaries are doubled in Alzheimer’s disease and Parkinson’s disease.
Acta Neuropathol. 2000;100(4):395-402. doi:
10.1007/s004010000195PubMedGoogle ScholarCrossref 35.Patton
N, Aslam
T, Macgillivray
T, Pattie
A, Deary
IJ, Dhillon
B. Retinal vascular image analysis as a potential screening tool for cerebrovascular disease: a rationale based on homology between cerebral and retinal microvasculatures.
J Anat. 2005;206(4):319-348. doi:
10.1111/j.1469-7580.2005.00395.xPubMedGoogle ScholarCrossref 36.Koronyo-Hamaoui
M, Koronyo
Y, Ljubimov
AV,
et al. Identification of amyloid plaques in retinas from Alzheimer’s patients and noninvasive in vivo optical imaging of retinal plaques in a mouse model.
NeuroImage. 2011;54S1:S204-S217. doi:
10.1016/j.neuroimage.2010.06.020Google Scholar 40.Ho
C-Y, Troncoso
JC, Knox
D, Stark
W, Eberhart
CG. Beta-amyloid, phospho-tau and alpha-synuclein deposits similar to those in the brain are not identified in the eyes of Alzheimer’s and Parkinson’s disease patients.
Brain Pathol. 2014;24(1):25-32. doi:
10.1111/bpa.12070PubMedGoogle ScholarCrossref 42.Chiasseu
M, Alarcon-Martinez
L, Belforte
N,
et al. Tau accumulation in the retina promotes early neuronal dysfunction and precedes brain pathology in a mouse model of Alzheimer’s disease.
Mol Neurodegener. 2017;12(1):58. doi:
10.1186/s13024-017-0199-3PubMedGoogle ScholarCrossref 44.Palmqvist
S, Zetterberg
H, Mattsson
N,
et al; Alzheimer’s Disease Neuroimaging Initiative; Swedish BioFINDER Study Group. Detailed comparison of amyloid PET and CSF biomarkers for identifying early Alzheimer disease.
Neurology. 2015;85(14):1240-1249. doi:
10.1212/WNL.0000000000001991PubMedGoogle ScholarCrossref 46.Forlenza
OV, Radanovic
M, Talib
LL,
et al. Cerebrospinal fluid biomarkers in Alzheimer’s disease: diagnostic accuracy and prediction of dementia.
Alzheimers Dement (Amst). 2015;1(4):455-463.
PubMedGoogle Scholar