Products & Services

A Serum protein-based algorithm for the detection of Alzheimer's disease

Description
A Serum protein-based algorithm for the detection of Alzheimer's disease
Published
of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  A Serum Protein-Based Algorithm for the Detection ofAlzheimer's Disease Sid E. O'Bryant 1, Guanghua Xiao 2, Robert Barber 3, Joan Reisch 2, Rachelle Doody 4, Thomas Fairchild 5, Perrie Adams 6, Steven Waring 7, and Ramon Diaz-Arrastia 8  for theTexas Alzheimer's Research Consortium* 1 Texas Tech University Health Sciences Center, F. Marie Hall Institute for Rural & CommunityHealth, Department of Neurology, Lubbock, Texas, U.S.A. 2 University of Texas Southwestern Medical Center, Department of Clinical Sciences, Dallas,Texas, U.S.A. 3 University of North Texas Health Science Center, Department of Pharmacology andNeuroscience, Fort Worth, Texas, U.S.A. 4 Baylor College of Medicine, Department of Neurology, Houston, Texas, U.S.A. 5 University of North Texas Health Science Center, Office of Strategy and Measurement, FortWorth, Texas, U.S.A. 6 University of Texas Southwestern Medical Center, Department of Psychiatry, Dallas, Texas,U.S.A. 7 Marshfield Clinic Research Foundation, Epidemiology Research Center, Marshfield, WI, U.S.A. 8 University of Texas Southwestern Medical Center, Department of Neurology, Dallas, Texas,U.S.A. Abstract Background— Alzheimer's disease (AD) is the most common form of age-related dementia andone of the most serious health problems in the industrialized world. Biomarker approaches todiagnostics would be more time and cost effective and may also be useful for identifyingendophenotypes within AD patient populations. Methods— We analyzed serum protein-based multiplex biomarker data from 197 patientsdiagnosed with AD and 203 controls from a longitudinal study of Alzheimer's disease beingconducted by the Texas Alzheimer's Research Consortium to develop an algorithm that separatesAD from controls. The total sample was randomized equally into training and test sets and randomforest methods were applied to the training set to create a biomarker risk score. Findings— The biomarker risk score had a sensitivity and specificity of 0.80 and 0.91,respectively and an AUC of 0.91 in detecting AD. When age, gender, education, and APOE statuswere added to the algorithm, the sensitivity, specificity, and AUC were 0.94, 0.84, and 0.95,respectively. Address correspondence to: Sid E. O'Bryant, Ph.D., Texas Tech University Health Science Center, Department of Neurology, 36014th St. STOP 6232, Lubbock, TX 79430. Phone: (806) 743-1338 ext 271; Fax: (806) 743-4150; sid.obryant@ttuhsc.edu. Disclosure : A patent has been filed in conjunction with Rules Based Medicine for the algorithm contained within this manuscript. Thefollowing authors are named on the patent: SE O'Bryant, RC Barber, R Diaz-Arrastia, G Xiao, PM Adams, JS Reisch, RS Doody, andTJ Fairchild. NIH Public Access Author Manuscript  Arch Neurol . Author manuscript; available in PMC 2011 September 1. Published in final edited form as: Arch Neurol  . 2010 September ; 67(9): 10771081. doi:10.1001/archneurol.2010.215. N I  H -P A A  u t  h  or M an u s  c r i   p t  N I  H -P A A  u t  h  or M an u s  c r i   p t  N I  H -P A A  u t  h  or M an u s  c r i   p t    Interpretation— These initial data suggest that serum protein-based biomarkers can be combinedwith clinical information to accurately classify AD. Of note, a disproportionate number of inflammatory and vascular markers were weighted most heavily in analyses. Additionally, thesemarkers consistently distinguished cases from controls in SAM, logistic regression and Wilcoxonanalyses, suggesting the existence of an inflammatory-related endophenotype of AD that mayprovide targeted therapeutic opportunities for this subset of patients. Introduction There is clearly a need for reliable and valid diagnostic and prognostic biomarkers of Alzheimer's disease (AD) and, in recent years there has been an explosive increase of effortaimed at identifying such markers. It has been previously argued that, due to significantadvantages, the ideal biomarkers would be gleaned from peripheral blood 1 . Peripheral bloodcan be collected at any clinic (or in-home visit) whereas most clinics are not capable of conducting lumbar punctures. Furthermore, advanced neuroimaging techniques are typicallyonly available in large medical centers of heavily urbanized areas. A blood-based algorithmgreatly increases access to advanced detection and, while nearly all patients are willing toundergo venipuncture, fewer elderly patients agree to lumbar puncture and many are unableto undergo neuroimaging for a range of reasons (e.g. pacemakers).Even though there is a large literature demonstrating altered levels of a range of biomarkers(CSF, serum and plasma) in AD patients (as well as MCI patients) relative to controls,attempts to identify a single biomarker specific to AD have failed. In the highly publicizedRay et al 2  publication, a large set of plasma-based proteins was analyzed in an effort toidentify a biomarker profile indicative of AD. The overall classification accuracy for theiralgorithm was 90%; additionally, their algorithm accurately identified 81% of MCI patientswho would progress to AD within a 2-6 year follow-up period. To date, however, thesefindings have not been cross-validated, nor has an independent blood-based (particularlyserum-based) algorithm been published.In addition to offering more accessible, rapid, as well as cost- and time-effective methodsfor assessment, biomarkers (or panels of biomarkers) also hold great potential for theidentification of endophenotypes within AD populations associated with particular diseasemechanisms. Once identified, targeted therapeutics specifically tailored to endophenotypestatus could be tested. Drawing upon an example from cardiovascular disease, by identifyinga subset of patients where atherosclerosis is pathogenically related to hypercholesterolemia,plasma cholesterol is a useful biomarker in the management of coronary artery disease.Plasma cholesterol measurements are useful as indicators of efficacy of treatment withHMG-CoA reductase inhibitors. Translating this conceptual framework to AD would be amajor advancement in this field 3 . The identification of a pro-inflammatory endophenotypeof AD would have implications for targeted therapeutics for a subgroup of patients such thatthose with an over-expression of the pro-inflammatory biomarker profile may benefit fromtreatment with anti-inflammatory compounds while those patients with an under-expressionof this profile may get worse on such treatment.In the current study we sought to (1) determine if a serum-based biomarker algorithm wouldsignificantly predict AD status, (2) evaluate if inclusion of demographic variables directlyinto the algorithm would improve the overall classification accuracy and (3) determine if there was a predominance of inflammatory-related markers that were over- or under-expressed in AD, which would be an initial step towards the concept of an inflammatory-related AD endophenotype. O'Bryant et al.Page 2  Arch Neurol . Author manuscript; available in PMC 2011 September 1. N I  H -P A A  u t  h  or M an u s  c r i   p t  N I  H -P A A  u t  h  or M an u s  c r i   p t  N I  H -P A A  u t  h  or M an u s  c r i   p t    Methods Participants Participants included 400 individuals (197 AD subjects, 203 controls) enrolled in the TexasAlzheimer's Research Consortium (TARC). The methodology of the TARC project has beendescribed in detail elsewhere 4 ; each participant underwent a standardized annualexamination at the respective site that includes a medical evaluation, neuropsychologicaltesting, and interview. Each participant also provided blood for storage in the TARCbiobank. Diagnosis of AD status was based on NINCDS-ADRDA criteria5 and controlsperformed within normal limits on psychometric assessment. Institutional Review Boardapproval was obtained at each site and written informed consent was obtained for allparticipants. Assays— Non-fasting blood samples were collected in serum-separating tubes duringclinical evaluations, allowed to clot at room temperature for 30 minutes, centrifuged,aliquoted, and stored at -80°C in plastic vials. Batched specimens from either baseline oryear-one follow-up exams were sent frozen to Rules Based Medicine (RBM,www.rulesbasedmedicine.com, Austin, TX) where they were thawed for assay withoutadditional freeze-thaw cycles using their multiplexed immunoassay human Multi-AnalyteProfile (humanMAP). Multiple proteins were quantified though multiplex fluorescentimmunoassay utilizing colored microspheres with protein-specific antibodies. Informationregarding the least detectable dose (LDD), inter-run coefficient of variation, dynamic range,overall spiked standard recovery, and cross-reactivity with other human MAP analytes canbe readily obtained from Rules Based Medicine. As with all such technologies, rapidevolution is expected, therefore, the complete list of analytes utilized from the humanMAPat the time of the current analyses is provided in Appendix 1. Statistical Analyses— Analyses were performed using R (V 2.10) statistical software 6 .Fisher's exact and Mann Whitney U tests were used to compare case versus controls forcategorical variables (APOE ε 4 allele frequency, gender, race, or ethnicity) and continuousvariables (age and education). The biomarker data was log transformed and thenstandardized for each analyte. The random forest prediction model was performed using Rpackage randomForest   (V 4.5) 7 , with all software default settings. We used the method by  Bair et al 8  to de-correlate the RBM biomarker data and clinical variables. The ROC(receiver operation characteristic) curves were analyzed using R package AUC (area underthe curve) was calculated using R package  DiagnosisMed   (V 0.2.2.2). The significantanalysis of Microarray (SAM) was performed using R package samr   (V 1.27)9. The FDR(false discovery rate) was calculated to address the multiple comparison issues. The FDRfrom SAM analysis was determined by permutation and those from Wilcoxon test andlogistic regression model were determined by fitting the p values to Beta-uniform models10.The Beta uniform models were fitted using R package ClassComparison (V 2.5.0)(http://bioinformatics.mdanderson.org/Software/OOMPA/ ). Results Demographic characteristics of the study population are shown in Table 1. Alzheimer'spatients were significantly older (p<0.001), less educated (p < 0.001), and more likely tocarry at least one copy of the APOE ε 4 allele (p < 0.001) than control participants.Once randomized into a training set or a testing set via random number generator, a randomforest (RF) prediction model was built with the training set using all of the markers in theRBM human MAP. Using the training set as a guide, the random forest algorithm assigned arisk score to each subject in the test set that was reflective of the probability of being O'Bryant et al.Page 3  Arch Neurol . Author manuscript; available in PMC 2011 September 1. N I  H -P A A  u t  h  or M an u s  c r i   p t  N I  H -P A A  u t  h  or M an u s  c r i   p t  N I  H -P A A  u t  h  or M an u s  c r i   p t    diagnosed with AD. Using the humanMAP markers, when the cut-off for the risk score wasset at to optimize performance, 0.47 (i.e. patient's risk score > 0.47 = AD, ≤  0.47 = control),the area under the curve (AUC) for the biomarker algorithm was 0.91 (95% CI = 0.88 -0.95), the sensitivity and specificity were equal to 0.80 (95% CI = 0.71 - 0.87) and 0.0.91(95% CI = 0.81-0.94), respectively. Of note, when the non-optimized cut-off of 0.5 wasused, the results did not change significantly (AUC = 0.91, sensitivity = 0.73, specificity =0.91). To test the robustness against allocation to training and test sets, randomization wasalso done by TARC site, which yielded an AUC of 0.88 demonstrating the robustness of thealgorithm against choice of methodology. Figure 1 presents a variable importance plot of protein markers measured by the random-forest built from the training set.Next the biomarker data was de-correlated 8  from the clinical variables of age, gender,education, and APOE status and an additional random forest prediction model generated.Results from the multivariate logistic regression model (Table 2) demonstrate that thebiomarker risk score was a significant, independent predictor of case status. As can be seenin Table 3, clinical data alone accurately classified a large portion of the sample, which wascomparable to, though somewhat less accurate than the performance of the biomarker profilealone. However, a combined algorithm using biomarker and clinical data was superior toeither alone (see Table 3 and Figure 2). Using the non-optimized cut-off for the biomarkerrisk score did not change the findings for the algorithm using both clinical and biomarkerdata (AUC = 0.95, sensitivity = 0.90, specificity = 0.87).SAM analysis with a FDR of < 0.001 identified a total of 23 proteins with that were eitherdifferentially over (n=14) or under (n=9) expressed in AD relative to controls (see Table 4).There were 22 proteins identified by the Wilcoxon test with a FDR less than 0.0025 and 22by logistic regression with a FDR less than 0.01. Figure 3 demonstrates the consistencybetween methods utilized. Supporting our notion of a possible inflammatory-relatedendophenotype present with AD patients, 10 (MIP1, eotaxin 1, TNF α , fibrinogen, IL5, IL7,IL10, CRP, MCP1, and von Willebran Factor) of the total 30 markers identified in Figure 1were inflammatory in nature. Discussion In a recently highly publicized study, Ray et al 2  identified a subset of 18 plasma-basedproteins that yielded excellent classification accuracy in case versus controls and our serumprotein-based algorithm yielded comparable accuracy. It is noteworthy that the markersfrom our study (Figure 1) have only minimal overlap with those presented by Ray at al(ANG2 and TNF α ). It is likely that this differential signature profile resulted from differentsample mediums as well as different assay platforms. Our study has multiple distinctadvantages over that of Ray et al. First, our serum protein assays were conducted by RBM,who have developed high throughput methodologies for reliable assay of high volumes of samples and analytes. RBM is the leading biomarker company in the U.S., working withmultiple pharmaceutical companies, the Alzheimer's Disease Neuroimaging Initiative, aswell as several of the leading AD biomarker research labs both within and outside the U.S.Second, our sample size of controls and AD cases is more than twice as large of the sampleutilized by Ray and colleagues. Third, our study included demographic information in thepredictive algorithm (age, gender, education, APOE status) and we demonstrated that thecombination of biomarker and clinical information yields superior results to either alone.Finally, our study is unique in that we are the first group to present serum-based findings.In support of our theory of the existence of an inflammatory endophenotype, many of theproteins with the highest importance from the RF analyses were inflammatory in nature(Figure 1). Additionally, when SAM analyses were conducted, a large portion of the O'Bryant et al.Page 4  Arch Neurol . Author manuscript; available in PMC 2011 September 1. N I  H -P A A  u t  h  or M an u s  c r i   p t  N I  H -P A A  u t  h  or M an u s  c r i   p t  N I  H -P A A  u t  h  or M an u s  c r i   p t    proteins either identified as over- or under-expressed were inflammatory in nature. Takentogether, these data suggest the existence of an inflammatory endophenotype withinAlzheimer's disease cases, which could offer targeted therapeutic options for this subgroupof patients.Of note, it is possible that the algorithm identified in the current study is not AD-specific.The current findings are a preliminary in nature and follow-up is necessary to test the abilityof the algorithm to detect AD when mixed in with non-AD dementia samples. It is alsopossible that the inflammatory signature observed is not specific to AD, but rather is relatedto other co-morbid factors (e.g. cardiovascular disease). In fact, it is likely that a pro-inflammatory endophenotype exists within patients diagnosed with other dementiasyndromes. Such a finding would further support the utility of a pro-inflammatoryendophenotype, as it is likely to represent a common pathway for a wide array of diseases.The identification of blood-based biomarker profiles with good diagnostic accuracy wouldhave a profound impact worldwide and requires further validation. Additionally, theidentification of pathway-specific endophenotypes among AD patients would likewise haveimplications for targeted therapeutics as well as understanding differential progressionamong diagnosed cases. With the rapidly evolving technology and analytic techniquesavailable, Alzheimer's disease researchers now have the tools to simultaneously analyzeexponentially more information from a host of modalities, which is likely going to benecessary to understand this very complex disease. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments This study was made possible by the Texas Alzheimer's Research Consortium (TARC) funded by the state of Texasthrough the Texas Council on Alzheimer's Disease and Related Disorders. Investigators at the University of TexasSouthwestern Medical Center at Dallas also acknowledge support from the UTSW Alzheimer's Disease CenterNIH, NIA grant P30AG12300. The funders had no role in study design, data collection and analysis, decision topublish, or preparation of the manuscript.Investigators from the Texas Alzheimer's Research Consortium: Baylor College of Medicine: Eveleen Darby,Kinga Szigeti, Aline Hittle; Texas Tech University Health Science Center: Paula Grammas, Benjamin Williams,Andrew Dentino, Gregory Schrimsher, Parastoo Momeni, Larry Hill; University of North Texas Health ScienceCenter: Janice Knebl, James Hall, Lisa Alvarez, Douglas Mains; University of Texas Southwestern Medical Center:Roger Rosenberg, Ryan Huebinger, Janet Smith, Mechelle Murray, Tomequa Sears References 1. Graff-Radford NR, Crook JE, Lucas J, et al. Association of low plasma Abeta42/Abeta40 ratios withincreased imminent risk for mild cognitive impairment and Alzheimer disease. Archives of Neurology. 2007; 64(3):354–362. [PubMed: 17353377]2. Ray S, Britschgi M, Herbert C, et al. Classification and prediction of clinical Alzheimer's diagnosisbased on plasma signaling proteins. Nature Medicine. 2007; 13(11):1359–1362.3. Thal LJ, Kantarci K, Reiman EM, et al. The role of biomarkers in clinical trials for Alzheimerdisease. Alzheimer Disease & Associated Disorders. 2006; 20(1):6–15. [PubMed: 16493230]4. Waring S, O'Bryant SE, Reisch JS, Diaz-Arrastia R, Knebl J, Doody R, for the Texas Alzheimer'sResearch Consortium. The Texas Alzheimer's Research Consortium longitudinal research cohort:Study design and baseline characteristics. Texas Public Health Journal. 2008; 60(3):9–13.5. McKhann D, Drockman D, Folstein M, et al. Clinical diagnosis of Alzheimer's disease: Report of the NINCDS-ADRDA Work Group. Neurology. 1984; 34:939–944. [PubMed: 6610841] O'Bryant et al.Page 5  Arch Neurol . Author manuscript; available in PMC 2011 September 1. N I  H -P A A  u t  h  or M an u s  c r i   p t  N I  H -P A A  u t  h  or M an u s  c r i   p t  N I  H -P A A  u t  h  or M an u s  c r i   p t  
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks