Frisoni 1. LENITEM (Laboratory of Epidemiology, Neuroimaging and Telemedicine) IRCCS S. Giovanni di Dio Fatebenefratelli Brescia, Italy;

Transcription

1 Frisoni 1 Imaging Markers for Alzheimer s Disease: Which versus How GB. Frisoni 1*, MD; M. Bocchetta 1, MS; G. Chételat 2,3,4,5, PhD; GD. Rabinovici 6, MD; MJ. de Leon 7, MD, PhD; J. Kaye 8, MD; EM. Reiman 9, MD; P. Scheltens 10, MD, PhD; F. Barkhof 10, MD, PhD; SE. Black 11, MD; DJ. Brooks 12, MD; MC. Carrillo 13, MD, PhD; NC. Fox 14, MD; K. Herholz 15, MD; A. Nordberg 16, MD, PhD; CR. Jack jr 17, MD; WJ. Jagust 18, MD; KA. Johnson 19, MD, PhD; CC. Rowe 20, MD, PhD; RA. Sperling 19, MD; W. Thies 13, PhD; L-O. Wahlund 21, MD, PhD; MW. Weiner 22, MD; P.Pasqualetti 23, PhD; Charles DeCarli 24, MD, for ISTAART s NeuroImaging Professional Interest Area. From: 1 LENITEM (Laboratory of Epidemiology, Neuroimaging and Telemedicine) IRCCS S. Giovanni di Dio Fatebenefratelli Brescia, Italy; 2 INSERM, U1077, Caen, France; 3 Université de Caen Basse-Normandie, UMR-S1077, Caen, France; 4 Ecole Pratique des Hautes Etudes, UMR-S1077, Caen, France; 5 CHU de Caen, U1077, Caen, France; 6 Memory and Aging Center, Department of Neurology, University of California San Francisco, CA; 7 New York University School of Medicine, Center for Brain Health, New York, NY; 8 Oregon Health and Science University, Portland, OR; 9 Banner Alzheimer's Institute, San Diego, CA; 10 Departments of Radiology, VU University Medical Center, Amsterdam, the Netherlands; 11 Department of Medicine, University of Toronto, Sunnybrook Research Institute, Toronto, Ontario, Canada; 12 Medical Research Council Clinical Sciences Center and Division of Neuroscience, Faculty of Medicine, Imperial College, London, United Kingdom; 13 Medical & Scientific Relations, Alzheimer's Association, Chicago, IL;

2 Frisoni 2 14 University College London Institute of Neurology, London, United Kingdom; 15 Wolfson Molecular Imaging Centre, University of Manchester, Manchester, United Kingdom; 16 Karolinska Institutet, Karolinska University Hospital Huddinge, Stockholm, Sweden; 17 Department of Diagnostic Radiology, Mayo Clinic and Foundation, Rochester, MN; 18 School of Public Health & Helen Wills Neuroscience Institute, University of California, Berkeley, CA; 19 Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Boston, MA; 20 Department of Nuclear Medicine, Centre for PET, Austin Health, Melbourne, VIC, Australia; 21 Division of Clinical Geriatrics, NVS-department, Karolinska Institutet, Stockholm, Sweden; 22 Northern California Institute of Research, San Francisco VA Medical Center, San Francisco, CA; 23 AFaR (Associazione Fatebenefratelli per la Ricerca), S. Giovanni Calibita - Fatebenefratelli, Rome, Italy; 24 University of California, San Francisco, San Francisco, CA. Corresponding author: Giovanni B. Frisoni, MD, Laboratory of Epidemiology Neuroimaging & Telemedicine, IRCCS Centro San Giovanni di Dio Fatebenefratelli, via Pilastroni 4, Brescia, Italy, Tel: , gfrisoni@fatebenefratelli.it Word count article: Word count abstract: 247. Characters count title: 57. Number of figures: 3. Number of tables: 3. Number of references: 35.

3 Frisoni 3 Supplemental Data: Suppl_Material_NIPIAreview_Neurology.pdf Key words: Alzheimer s disease; MCI (mild cognitive impairment); neuroimaging; biomarker; accuracy. Author Contributions: Drafting/revising the manuscript for content: Dr. Frisoni, dr. Bocchetta, dr. Pasqualetti, dr. Chetelat, dr. Rabinovici, dr. Herholz, dr. Kaye, dr. Jack, dr. Rowe, dr. Jagust, dr. Wahlund, dr. Brooks, dr. Nordberg, dr. Scheltens, dr. Reiman, dr. Weiner, dr. De Leon. Study concept or design: Dr. Frisoni, dr. DeCarli, dr. Barkhof, dr. Herholz, dr. Kaye, dr. Jack, dr Rowe, dr. Jagust, dr Wahlund, dr. Brooks, dr. Nordberg, dr. Scheltens, dr. Reiman, dr. Fox, dr. Black, dr. Sperling, dr. Johnson, dr. Weiner, dr. Carrillo, dr. Thies. Analysis or interpretation of data: dr. Frisoni, dr. Bocchetta, dr. Pasqualetti. Statistical analysis: dr. Bocchetta, dr. Pasqualetti. Study supervision or coordination: dr. Frisoni. Disclosure: The authors report no disclosures relevant to the manuscript. Full Disclosures: (Detailed disclosures are also reported in the Supplementary Material): Dr. Bocchetta, dr. Chetelat, dr. Wahlund, dr. Pasqualetti, dr. Carrillo, dr. Rabinovici, dr. De Leon, dr. Reiman, dr. Barkhof, dr. Black, dr. Brooks, dr. Fox, dr. Herholz, dr. Johnson, dr. Rowe, dr. Thies, and dr. DeCarli report no disclosures. Dr. Frisoni has served in advisory boards for Lilly, BMS, Bayer, Lundbeck, Elan, Astra Zeneca, Pfizer, Taurx, Wyeth, GE; he is member of the editorial boards of Lancet Neurology, Aging Clinical & Experimental Research, Alzheimer s Diseases & Associated Disorders, Neurodegenerative Diseases, and Imaging Section Editor of Neurobiology of Aging; he has received grants from Wyeth Int.l, Lilly Int.l, Lundbeck Italia, GE Int.l, Avid/Lilly, and the Alzheimer s Association. Dr. Nordberg has been PI for clinical trials sponsored by Torrey pines Therapeutics, GSK, Wyeth, Bayer Pharma served on the advisory board for Elan, Pfizer, GSK, Novartis, Lundbeck AB, Johnson & Johnson, GE Health Care, Avid; she received honorarium for lectures from Novartis, Jansen-Cilag, Pfizer, Merck and research grants from Novartis, Pfizer, GE Healthcare, Johnson & Johnson. Bayer Pharma; she owns no stocks and is member of the editorial Advisory Board for Current Alzheimer s research, Journal of Alzheimer s disease, Alzheimer s Research & Therapy. Dr. Sperling has served as a paid consultant for Bayer, Biogen-IDEC, Bristol-Myers- Squibb, Eisai, Janssen Alzheimer Immunotherapy, Pfizer, Merck, Roche, Satori, and as an upaid consultant to Avid; she is a site co-investigator for Avid, Bristol-Myers-Squibb, Pfizer, and Janssen Alzheimer Immunotherapy clinical trials. She has spoken at symposia sponsored by Eli Lilly, Pfizer, and Janssen Alzheimer Immunotherapy. Dr. Jagust has served as a consultant to Siemens, Genentech, TauRx, Janssen Alzheimer Immunotherapy; he receive research support from NIH (AG034570, AG025303).

4 Frisoni 4 Dr. Scheltens serves on the advisory boards of Genentech, Novartis, Pfizer, Roche, Danone, Nutricia, Jansen AI, Baxter and Lundbeck; he has been a speaker at symposia organised by Lundbeck, Lilly, Merz, Pfizer, Jansen AI, Danone, Novartis, Roche and Genentech; he serves on the editorial board of Alzheimer s Research & Therapy and Alzheimers Disease and Associated Disorders, is a member of the scientific advisory board of the EU Joint Programming Initiative and the French National Plan Alzheimer. The Alzheimer Center receives unrestricted funding from various sources through the VUmc Fonds; he receives no personal compensation for the activities mentioned above. Dr. Kaye received research support from the Department of Veterans Affairs and the National Institutes of Health (NIH); individuals that work in the research centers he directs received research support from Johnson & Johnson, Roche and Bristol Myers Squibb. Dr. Kaye was compensated for serving on a data monitoring committee for Eli Lilly; and as a paid advisor for Janssen Pharmaceutical; he received reimbursement through Medicare or commercial insurance plans for providing clinical assessment and care for patients; he has been salaried to see patients at the Portland VA Medical Center; he served as an unpaid Vice-Chair for the International Professional Interest Area Work Group of the International Society to Advance Alzheimer s Research and Treatments (ISTAART) and as an unpaid Commissioner for the Center for Aging Services and Technologies; he serves on the editorial advisory board of the journals, Alzheimer's & Dementia and Frontiers of Aging Neuroscience. Dr. Weiner served on the scientific advisory board for Lilly, Araclon and Institut Catala de Neurociencies Aplicades, Gulf War Veterans Illnesses Advisory Committee, VACO, Biogen Idec, Pfizer, BOLT Inter-national; he is a consultant for Astra Zeneca, Araclon, Medivation/Pfizer, Ipsen, TauRx Therapeutics LTD, Bayer Healthcare, Biogen Idec, Exonhit Therapeutics, SA, Servier, Synarc, Pfizer, Janssen, Harvard University and KLJ Associates; he received funds for travel from NeuroVigil, Inc., CHRU-Hopital Roger Salengro, Siemens, AstraZeneca, Geneva University Hospitals, Lilly, University of California, San Diego ADNI, Paris University, Institut Catala de Neurociencies Aplicades, University of New Mexico School of Medicine, Ipsen, CTAD (Clinical Trials on Alzheimer s Disease), Pfizer, AD PD meeting, Paul Sabatier University, Novartis, Tohoku University, Fundacio ACE, Travel edreams, Inc.; he is a member of the editorial advisory board of Alzheimer s & Dementia and MRI; he received honoraria from NeuroVigil, Inc., Insitut Catala de Neurociencies Aplicades, PMDA /Japanese Ministry of Health, Labour, and Welfare, Tohoku University and Alzheimer s Drug Discovery Foundation; he received research support from commercial (Merck and Avid) and from government entities (DOD and VA); he holds stock options from Synarc and Elan; he received funds from organizations contributing to the Foundation for NIH and thus to the NIA funded Alzheimer s Disease Neuroimaging Initiative: Abbott, Alzheimer's Association, Alzheimer's Drug Discovery Foundation, Anonymous Foundation, AstraZeneca, Bayer Healthcare, BioClinica, Inc. (ADNI 2), Bristol-Myers Squibb, Cure Alzheimer's Fund, Eisai, Elan, Gene Network Sciences, Genentech, GE Healthcare, GlaxoSmithKline, Innogenetics, Johnson & Johnson, Eli Lilly & Company, Medpace, Merck, Novartis, Pfizer Inc, Roche, Schering Plough, Synarc, Wyeth. Dr. Jack serves as a consultant for Janssen, Eisai, GE, Johnson and Johnson, and Lilly and received research support for clinical trials from Pfizer, Allon, and Baxter, Inc.

5 Frisoni 5 Dr. Reiman served as a scientific advisor to Sygnis, AstraZeneca, Bayer, Eisai, Elan, Eli Lilly, GlaxoSmithKline, Intellect, Link Medicine, Novartis, Siemens and Takeda; he has had research contracts with AstraZeneca and Avid/Eli Lilly; a patent pending for a biomarker strategy to evaluate pre-clinical AD treatments (through Banner Health); and research grants from the National Institute on Aging, Anonymous Foundation, Nomis Foundation, Banner Alzheimer s Foundation, and the State of Arizona.

6 Frisoni 6 ABSTRACT The NIA-AA revised and IWG criteria for Alzheimer s disease (AD) recognize a key role to imaging biomarkers for early diagnosis of AD. We reviewed diagnostic accuracy (likelihood ratios, LRs) of imaging biomarkers (amyloid imaging, 18F-FDG PET, SPECT, medial temporal atrophy on MRI) to separate AD from healthy and MCI converting to AD from non-converters. Accuracy was studied separately by metrics (i.e. how the marker is measured; e.g. manual segmentation, automated tool, etc). LR+ ranged from 9.4 to 4.4 and LR- from 0.08 to 0.25 for amyloid imaging and MRI, respectively. Prognostic LR+ ranged from 7.5 to 1.7 and LR- from 0.50 to 0.11 for 18F-FDG PET and amyloid imaging, respectively. Within individual metrics, LRs varied up to 100-fold (i.e. from 1 to 100 and from 0.01 to 1.00). Markers accounted for 11% and 18% of LR+ (diagnostic and prognostic) and for 24% and 16% of LR- variance. Across all markers, metrics accounted for an equal or larger amount of variance than markers: 13% of LR+ and 29% of LRdiagnostic, and 62% of LR+ and 18% of LR- prognostic variance. Within individual markers, the largest proportion of diagnostic LR+ variability was between 18F-FDG PET metrics, while the largest proportion of LR- variability was between MRI metrics. Accuracy of imaging biomarkers of AD is at least as dependent on how the biomarker is measured as the biomarker type itself. The development of standard operating procedures for biomarker measurement will be key to biomarker use in the clinical routine.

7 Frisoni 7 1. Introduction The prevailing understanding of the pathophysiology and natural history of AD has led researchers to recently propose alternatives to the traditional NINCDS-ADRDA diagnostic criteria 1-5. The IWG and NIA/AA criteria assign a key pathogenetic role for cerebral β- amyloidosis and neurodegeneration, hallmarked by senile plaques and neuronal tangles on microscopic examination. They further stipulate that positivity of one or more disease marker of brain amyloidosis (including increased binding of amyloid imaging agents with PET) and neuronal injury (including temporoparietal hypometabolism on 18F-FDG PET, temporoparietal hypoperfusion on SPECT, and medial temporal lobe atrophy on MRI) is associated with high likelihood that the patient s cognitive impairment is due to AD pathology. The view is largely shared that the criteria are not ready to be widely used in routine clinical practice 6-9, and none of the metrics used to quantify imaging biomarkers (e.g. hippocampal volume for medial temporal atrophy or standardized uptake value ratio for amyloid imaging) is currently certified as a diagnostic test by regulatory agencies or reimbursed by health care providers or third party payers. However, some clinical services are informally using biomarkers as adjuncts in the diagnostic process, supporting the practical urgency of quick progression on the track of criteria validation. In this context, the intrinsic test characteristics of biomarkers will represent a key factor for successful validation. A number of reviews are available on the diagnostic accuracy of imaging biomarkers. Reviews have generally focused on single modality markers, and only a few have addressed accuracy across different modalities. Fewer still have studied diagnostic accuracy across different operating procedures, and none have addressed diagnostic accuracy of imaging biomarkers across different modalities and operating procedures. The latter effort is relevant to appreciate the relative relevance of modality and operating procedure on diagnostic accuracy. This information will help design clinical research studies aimed to validate the new diagnostic criteria for AD, and contribute to the progression of imaging biomarkers from informal diagnostic adjuncts to fully validated biomarkers. This review is conceived by the Neuroimaging Professional Interest Area (NIPIA), a group of clinical imaging scientists born out of the Alzheimer s Imaging Consortium and the

8 Frisoni 8 specialist branch of the International Society to Advance Alzheimer s Research and Treatment (ISTAART) of the Alzheimer s Association (AA), in the context of its mission to promote the appropriate use of imaging in clinical and research contexts. The views expressed here are those of the authors and do not represent a formal position or endorsement by the AA.

9 Frisoni 9 2. Methods We performed a search on the PubMed database for literature published between 1989 and April 2012, using combined specific terms of AD, accuracy and biomarkers: condition AND marker AND submarker AND (accuracy OR sensitivity OR specificity), where conditions were Alzheimer s Disease and Mild Cognitive Impairment, markers were Amyloid PET, SPECT or SPET, 18F-FDG PET, magnetic resonance, while submarkers were 18F and 11C-PIB for Amyloid PET; hippocampus, amygdala, entorhinal cortex and temporal horn for MRI; 99m Tc-HMPAO and 99m Tc-ECD or 23I-IMP for SPECT. For instance for diagnostic performance of amygdala measured on MRI we used Alzheimer* AND (MRI OR magnetic resonance) AND amygdal* AND (accuracy OR sensitivity OR specificity). The Related Articles feature in PubMed for the selected research studies and references of retrieved articles was also screened to maximise the probability of finding additional relevant studies. The search was limited to articles involving human subjects and written in English. We selected studies that specified operating procedures (Table 1), and in particular: Amyloid imaging agents with PET: Visual read, the qualitative assessment of cortical ligand uptake for each individual image; Standardized Uptake Value Ratio (SUVR), the quantitative analysis of the ratio of cortical ligand uptake to a reference region for each individual image; Distribution Volume Ratio (DVR), the quantitative analysis of the ratio of cortical ligand distribution volume to the cerebellar uptake for each individual image; Temporoparietal hypometabolism on 18F-FDG PET: the computer-aided visual read (Neurostat/3D-SSP), which uses the 3-dimensional stereotactic surface projection (3D-SSP) technique through Neurostat automated image analysis procedure, comparing each individual 18F-FDG PET image on a pixel-by-pixel basis with a normative reference database, and producing parametric z-score images; t-sum/hci, the automated summary measures of AD-related hypometabolism based on the comparison of individual 18F-FDG PET images with a normative reference dataset in a pre-defined AD mask (t-sum score is computed as voxel-by-voxel sum of t-scores in a pre-defined AD-pattern mask, while HCI, hypometabolic convergence index, is calculated as the inner-product of the individual Z- map and a pre-defined AD Z-map); the computer-aided visual read using Statistical Parametric Mapping (SPM), computing a score as the average metabolism on a set of meta-

10 Frisoni 10 analytically derived regions-of-interest (ROI) reflecting AD hypometabolism pattern; Visual read, the qualitative assessment of cortical metabolism for each individual image; Temporoparietal hypoperfusion on SPECT or SPET: Visual Read, the qualitative assessment of cortical perfusion for each individual image; Quantitative/Semi-quantitative Assessment, the quantification of cortical perfusion for each individual image; Medial temporal atrophy on structural MRI: Visual Read, the qualitative assessment of structure atrophy using Likert scales; Manual Segmentation, the volumetric measurement through manual segmentation; Automated Volumetry measurement computed through automated segmentation algorithms (Freesurfer, that implements the subcortical segmentation by probabilistic segmentation based on a prior anatomical model; Adaboost- ACM, a machine learning method that learns features to guide segmentation; Brainvisa SASHA, the deformation constraint approach based on prior knowledge of anatomical features automatically retrieved from the MRI data); Linear Measure, the manual measurement of the medial temporal lobe and the temporal horn of the lateral ventricle. Table 2 shows that metrics are remarkably heterogeneous as to acquisition procedures, automation, stability, intensivity (in terms of time consume), availability of a normative population and threshold, and cost. We included studies reporting sensitivity and specificity for each single analytic method for each biomarker, the number and the diagnosis of subjects for each comparison group. Studies of AD versus other types of dementia were not considered. Regarding Mild Cognitive Impairment (MCI) patients, we included only studies that considered the sensitivity as the correct classification of MCI who subsequently converted to AD dementia versus stable MCI. We extracted single studies from reviews and meta-analyses 10-21, and addressed them individually. Those reviews in fact did not follow discrimination criteria as to specific tools and analytic methods, which was the aim of our investigation. We excluded studies if they did not: study AD or MCI patients; report figures for sensitivity and specificity; explicitly state procedures for marker measurement; assess the diagnostic performance of individual imaging biomarkers (e.g. accuracy of clinical diagnosis plus biomarkers, or a panel of biomarkers); disaggregate converters from stable MCI or different types of non AD dementias; provide information on group size.

11 Frisoni 11 We also checked that included studies did not include overlapping patient populations. In order to obtain a pooled measure of sensitivity and specificity, we used a classical Bayesian approach 22. Given the number of correctly (n CCS ) and incorrectly (n ICS ) classified subjects and by considering a binomial likelihood of parameter n CCS /n ICS and a beta(1,1) non informative conjugate prior, the posterior distribution of the proportion p CCS follows a beta distribution of parameters a=n CCS +1 and b=n ICS +1. The sampled beta distribution (1000 samples, R package v ) was therefore used to provide estimates for the pooled mean and 95% confidence intervals (CIs) of the ratio p CCS /p ICS =p CCS /(1-p CCS ). The estimation was repeated for each set of studies that investigated the same operational metrics on the same type of diagnostic groups. Positive and negative likelihood ratios (LR+s and LR-s) were computed as follows: LR+ = sensitivity/(100-specificity) and LR- = (100-sensitivity)/specificity. LR+ equal to or greater than 5 and a LR- equal to or lower than 0.2 are generally regarded as clinically meaningful, i.e. diagnostically useful. Analyses were carried out with SPSS using one-way ANOVA and nested ANOVA to test whether diagnostic and prognostic LR+ and LR- variability was due to differences between markers, metrics, and submarkers or due to variability among the metrics within markers, among the metrics within submarkers, or among the submarkers within markers. Statistical analyses and plots were restricted to metrics used by at least 3 studies. Through a linear regression analysis we investigated the effect of age, disease severity, group size and follow-up duration on sensitivity, specificity, and LR values.

12 Frisoni Results Table 3 shows that diagnostic accuracy was highest for amyloid imaging (sensitivity and specificity figures pooled across submarkers and metrics were 88%, 95% CI 84 to 91, and 85%, 81 to 88), and progressively lower for 18F-FDG PET (86%, 84 to 89; 84%, 81 to 87), SPECT (76%, 74 to 79; 84% 81 to 87), and MRI (75%, 73 to 77; 81%, 79 to 82). Prognostic accuracy had a similar pattern across markers, but was generally lower than diagnostic accuracy. Pooled sensitivity and specificity figures were 82% (74 to 88) and 56% (49 to 64) for amyloid imaging, 76% (70 to 82) and 74% (68 to 78) for 18F-FDG PET, 78% (72 to 85) and 64% (55 to 72) for SPECT, and 62% (58 to 66) and 73% (69 to 76) for MRI. Likelihood Ratio analysis, dementia stage The analysis of the LR+ (Figure 1A) mirrored the accuracy analysis pattern, being highest for amyloid imaging (9.4) and lowest for MRI (4.4). Considering amyloid imaging submarkers, LR+ values were 10.0 for 18-F ligands and 7.5 for 11-C PIB, while for MRI submarkers, they were 10.7 for temporal horn, 4.6 for amygdala, 4.4 for hippocampus and 4.0 for entorhinal cortex. At the metrics level, the variability of LR+ was often as high as between markers: LR+ of 18F-FDG PET ranged from 13.3 (automated assessment with Neurostat/3D-SSP) to 2.4 (visual read), and LR+ of MRI ranged from 10.7 (linear measures) to 4.2 (manual volumetry). The variability among metrics was lower for the other two markers, LR+ of amyloid imaging being 12.5 for DVR and 8.9 for SUVR, and of SPECT 6.7 for visual read and 6.7 for quantitative read. As to LR- (Figure 1B), values were highest for amyloid imaging (0.08) and lowest for MRI (0.25). LR- values across amyloid imaging submarkers were rather homogeneous (0.08 for 18-F ligands and 0.06 for 11-C PIB); little variation was also detected for MRI submarkers, (0.27 for temporal horn, 0.22 for amygdala, 0.26 for hippocampus, 0.22 for entorhinal cortex). The variability of the LR- of metrics was much higher: LR- of amyloid imaging was 0.01 for DVR and 0.10 for SUVR, and of 18F-FDG PET ranged between 0.05 (visual read) and 0.23 (SPM). Variability among metrics was lower for the other two markers: LR- of MRI

13 Frisoni 13 ranged from 0.21 (manual volumetry) to 0.32 (visual read), and of SPECT were 0.13 for quantitative read and 0.17 for visual read. For detailed information, see Supplementary Table e-1. Markers accounted for 11% (p=.005) of LR+ and 24% (p<.0005) of LR- variance and metrics for 13% (p=.009) and 29% (p<.0005), respectively (Figure 3A). When markers were divided into functional (i.e. 18F-FDG PET and SPECT) and structural (i.e. MRI), they accounted for 12% (p=.006) of LR+ variance. Of all metrics, those with the largest variability were 18F-FDG PET metrics (39%, p=.022) for LR+, and MRI metrics (37%, p<.0005) for LR-. The variance of LR- explained by metrics remained significant even when tested with the more stringent nested ANOVA within markers (17%, p=.008) and submarkers (15%, p=.027). When restricted to MRI, nested ANOVA analysis showed that metrics within MRI submarkers accounted for 41% (p<.0005) of diagnostic LR- variance. The variability of diagnostic LR+ within metrics was even greater than across markers and metrics. Many metrics spanned two orders of magnitude, LR+ ranging from a lowest between 1 and 3 up to between 70 to 100 (Figure 1A). The variability of diagnostic LRwithin metrics was similar, spanning two orders of magnitude from 0.01 to 1.00 (Figure 1B). Likelihood Ratio analysis, MCI stage Prognostic were generally lower than diagnostic LR+ figures, being above 5 only for 18F- FDG PET (7.5). The pattern was also different, the second highest LR+ values being that of MRI (2.6), followed by SPECT (2.2), and last by amyloid imaging (1.7). It should be noted, however, that the number of studies contributing to prognostic LR+ estimation was much lower than that of diagnostic LR+. Considering MRI submarkers, prognostic LR+ were 2.9 for hippocampus and 2.2 for entorhinal cortex (Figure 2A). In analogy with the pattern of diagnostic LR+, the variability across metrics was in some cases at least as large as that across markers. Prognostic LR+ of 18F-FDG PET metrics were 12.8 for SPM and 1.7 for t-sum/hci. The variability across MRI metrics was lower, being 3.2 for manual volumetry, 2.6 for visual read and 1.8 for Freesurfer.

14 Frisoni 14 Marker-wise, LR- values (Figure 2B) were highest for amyloid imaging (0.11), intermediate for SPECT (0.28), and lowest for MRI (0.49) and 18F-FDG PET (0.50). For MRI submarkers, LH- were 0.49 for hippocampus and 0.56 for entorhinal cortex. Again, LR- of 18F-FDG PET metrics were largely heterogeneous, (0.08 for SPM and 0.64 for t-sum/hci). The variability across MRI metrics was lower, being 0.46 for manual volumetry, 0.50 for visual read and 0.51 for Freesurfer. For detailed information, see Supplementary Table e-1. When compared to diagnostic LR+ variance, both markers and metrics accounted for a larger proportion of prognostic variance (18%, p=.06; 62%, p<.0005, respectively) (Figure 3B). Compared to diagnostic LR- variance, markers and metrics accounted on the contrary for a lower proportion of prognostic variance, respectively 16% (p=.09) and 18% (p=.07). Similarly to diagnostic metrics, those prognostic metrics with the largest LR+ variability were 18F-FDG PET metrics (82%, p=.019). Metrics accounted for 25% (p=.044) of prognostic LR+ variance of the MRI marker. When considered together, SPECT and 18F- FDG PET metrics accounted for 78% (p=.009) of the prognostic LR+ variance. The prognostic variance of metrics remained significant even when tested with nested ANOVA within markers (68%, p<.0005). When restricted to SPECT and 18F-FDG PET metrics, nested ANOVA analysis showed that these functional metrics accounted for 86% (p=.005) of prognostic variance. The variability of prognostic LR+ within metrics spanned one order of magnitude, with few exceptions spanning two orders of magnitude (from around 1 to around 10). The variability of LR- within metrics was similar, with a few exceptions spanning two orders of magnitude (18F-FDG PET with SPM and amyloid PET with SUVR) (Figure 2B). Effect of confounders on LR estimates Specific analysis regarding the effect of study group size, follow-up duration, age and disease severity on accuracy figures are reported in the Supplementary Material.

15 Frisoni Discussion We have estimated the diagnostic and prognostic accuracy of imaging markers of brain amyloidosis and neuronal injury and pertinent metrics. We have shown that the diagnostic and prognostic accuracy of AD imaging biomarkers is at least as dependent on how the biomarker is measured as the type of biomarker itself. While acknowledging that imaging biomarkers capture different neurobiological constructs (i.e. brain amyloidosis, neuronal injury at the molecular, and neuronal injury at the macroscopic level), this observation provides empirical support to current efforts aimed to develop standard operating procedures for AD biomarkers. Such efforts are key to the use of imaging biomarkers in the diagnostic routine and clinical trials. Diagnostic LRs were generally higher than prognostic LRs: the former were around 5 or above for all markers and metrics except visual read of 18F-FDG PET, and below 0.20, except for MRI metrics (between 0.20 and 0.35). This is expected in that biological changes in MCI who convert to AD are milder than in AD dementia patients The increasing awareness of AD and options for early diagnosis make biomarkers particular usefulness in clinical practice to predict MCI converters versus non converters. Here, LR+ were of 3 or below for all metrics and markers except 18-F FDG PET, where it was above 7; the pattern was similar for LR-, where all markers and the vast majority of metrics yielded LR- greater than 0.45 except amyloid imaging, with a LR- approaching 0.1. In the vast majority of conditions amyloid imaging metrics outperformed the other metrics. This is in line with current understanding of the pathophysiology of AD, positing that brain amyloidosis is a necessary condition for AD-related neurodegeneration to take place 4, 29. Although not investigated here, amyloid imaging is the most accurate marker for the differentiation between AD and non-amyloid related dementias such as the different moieties of frontotemporal degeneration 30, 31. It should be underlined however, that the number of studies on amyloid imaging is by far lower than those with MRI and, in the prognostic condition, also with 18F-FDG PET. More amyloid imaging studies will be needed to consolidate the pertinent estimates on LRs. The metrics with the largest variability were those of 18F-FDG PET: diagnostic and prognostic LR+ ranged between around 2 to 13, with highest diagnostic figures for Neurostat and lowest for visual rating. Interestingly, diagnostic LR+ for t-sum were higher

16 Frisoni 16 than SPM, and vice versa for prognostic LR+. This argues in favour of greatest benefits of standardization of 18F-FDG PET metrics. On the contrary, LR+ and LR- figures were poorest for MRI metrics in almost all conditions. This is expected in view of the little specificity of medial temporal atrophy, which is featured in AD as well as a proportion of cognitively healthy older persons and patients with non AD neurodegenerative conditions 32. Importantly, if the variability of diagnostic LRs across metrics varied by one order of magnitude (i.e. 10-fold), the variability within metric varied by as many as two orders of magnitude (i.e. 100-fold). The within metric variability of prognostic LRs varied also by one order of magnitude. This observation militates in favour of standardization of metrics, which should reduce this 100-fold variability to close to zero. Metrics vary for a number of features such as: dependency on a specific, sometimes nonroutine, image acquisition protocol; dependency on a human rater and automation; stability over time (test-retest reliability) and across raters (inter-rater reliability); feasibility in routine clinical settings, where human and technological resources are tailored to the use of routine tests; availability of rigorously standardized operational procedures for measurement; availability of a reference normative population; availability of reliable abnormality thresholds; and cost of the overall acquisition and measurement procedure. All of the above issues should be addressed by standardization efforts for metrics to be adopted in the clinical routine and to be used as the reference for validation of automated algorithms. This review has a number of limitations that should be underlined. Due to the low number of studies, we did not address the accuracy of imaging biomarkers for the differential diagnosis of dementia type (e.g. AD versus Lewy body dementia, AD versus frontotemporal degeneration, etc.). With the hopeful advent of drugs affecting specific core pathophysiologic substrates of AD this issue may become of greater relevance and need to be properly addressed. We accepted the definitions of AD and MCI adopted by the reviewed studies, including exclusion criteria (e.g. vascular disease, medications, etc.), thus accepting the inherent clinical heterogeneity. Data regarding the classification of AD dementia versus cognitively normal older adult subjects is only a necessary but not sufficient indicator of a test s value and does not reflect its diagnostic accuracy in clinical settings. Further tests of diagnostic value would be those that help in differential

17 Frisoni 17 diagnosis (e.g., among different dementia cases) in terms of predicting post-mortem neuropathology, in predicting a person s clinical course (as in the MCI analysis), and eventually, predicting response to treatment. Another limitation relates to the use of imaging markers for prognosis 33 and differential diagnosis 5 in the context of the new criteria. Here, imaging need to be used together with biological (CSF) markers, and the contribution of biological to LRs of imaging markers will need to be investigated in future studies with pathological confirmation. The definition we chose to classify metrics should be taken with due caution: hippocampal atrophy has largely been assessed using MRI, but data showed that it can also be usefully studied with FDG PET 34. Last, confounders had little effect on diagnostic and prognostic accuracy values, with the exception of the positive association of age with LR- and its negative association with specificity in AD, and negative association of age with LR+ and specificity in MCI stage. We believe that this observation is due to the relatively higher frequency of abnormal markers in older persons despite no disease 35.

18 Frisoni 18 Author Contributions Drafting/revising the manuscript for content: Dr. Frisoni, dr. Bocchetta, dr. Pasqualetti, dr. Chetelat, dr. Rabinovici, dr. Herholz, dr. Kaye, dr. Jack, dr. Rowe, dr. Jagust, dr. Wahlund, dr. Brooks, dr. Nordberg, dr. Scheltens, dr. Reiman, dr. Weiner, dr. De Leon. Study concept or design: Dr. Frisoni, dr. DeCarli, dr. Barkhof, dr. Herholz, dr. Kaye, dr. Jack, dr Rowe, dr. Jagust, dr Wahlund, dr. Brooks, dr. Nordberg, dr. Scheltens, dr. Reiman, dr. Fox, dr. Black, dr. Sperling, dr. Johnson, dr. Weiner, dr. Carrillo, dr. Thies. Analysis or interpretation of data: dr. Frisoni, dr. Bocchetta, dr. Pasqualetti. Statistical analysis: dr. Bocchetta, dr. Pasqualetti. Study supervision or coordination: dr. Frisoni. Acknowledgments The Alzheimer s Association funded the project: A Harmonized Protocol for Hippocampal Volumetry: an EADC-ADNI Effort, grant n. IIRG We would like to thank dr. Marco Lorenzi for his precious help on data analysis.

19 Frisoni 19 References 1. Dubois B, Feldman HH, Jacova C, et al. Research criteria for the diagnosis of alzheimer's disease: Revising the NINCDS-ADRDA criteria. Lancet neurology 2007;6: Dubois B, Feldman HH, Jacova C, et al. Revising the definition of alzheimer's disease: A new lexicon. Lancet neurology 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: Albert MS, DeKosky ST, Dickson D, et al. The diagnosis of mild cognitive impairment due to alzheimer's disease: Recommendations from the national institute on agingalzheimer's association workgroups on diagnostic guidelines for alzheimer's disease. Alzheimer's & dementia : the journal of the Alzheimer's Association 2011;7: McKhann GM, Knopman DS, Chertkow H, et al. The diagnosis of dementia due to 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: Khachaturian ZS. Revised criteria for diagnosis of alzheimer's disease: National institute on aging-alzheimer's association diagnostic guidelines for alzheimer's disease. Alzheimers Dement 2011;7: Frisoni GB, Hampel H, O'Brien JT, Ritchie K, Winblad B. Revised criteria for alzheimer's disease: What are the lessons for clinicians? Lancet Neurol 2011;10: Frisoni GB, Winblad B, O'Brien JT. Revised NIA-AA criteria for the diagnosis of alzheimer's disease: A step forward but not yet ready for widespread clinical use. Int Psychogeriatr 2011;23: Gauthier S, Patterson C, Gordon M, Soucy JP, Schubert F, Leuzy A. Commentary on "recommendations from the national institute on aging-alzheimer's association workgroups

20 Frisoni 20 on diagnostic guidelines for alzheimer's disease." A canadian perspective. Alzheimers Dement 2011;7: Bloudek LM, Spackman DE, Blankenburg M, Sullivan SD. Review and meta-analysis of biomarkers and diagnostic imaging in alzheimer's disease. J Alzheimers Dis 2011;26: Bohnen NI, Djang DS, Herholz K, Anzai Y, Minoshima S. Effectiveness and safety of 18F-FDG PET in the evaluation of dementia: A review of the recent literature. J Nucl Med 2012;53: Chetelat G, Baron JC. Early diagnosis of alzheimer's disease: Contribution of structural neuroimaging. Neuroimage 2003;18: Devous MD S. Functional brain imaging in the dementias: Role in early detection, differential diagnosis, and longitudinal studies. Eur J Nucl Med Mol Imaging 2002;29: Dougall NJ, Bruggink S, Ebmeier KP. Systematic review of the diagnostic accuracy of 99mTc-HMPAO-SPECT in dementia. Am J Geriatr Psychiatry 2004;12: Herholz K, Ebmeier K. Clinical amyloid imaging in alzheimer's disease. Lancet Neurol 2011;10: Jack CR,Jr. Alliance for aging research AD biomarkers work group: Structural MRI. Neurobiol Aging 2011;32 Suppl 1:S Laforce R,Jr, Rabinovici GD. Amyloid imaging in the differential diagnosis of dementia: Review and potential clinical applications. Alzheimers Res Ther 2011;3: Mosconi L. Brain glucose metabolism in the early and specific diagnosis of alzheimer's disease. FDG-PET studies in MCI and AD. Eur J Nucl Med Mol Imaging 2005;32: Patwardhan MB, McCrory DC, Matchar DB, Samsa GP, Rutschmann OT. Alzheimer disease: Operating characteristics of PET--a meta-analysis. Radiology 2004;231: Yuan Y, Gu ZX, Wei WS. Fluorodeoxyglucose-positron-emission tomography, singlephoton emission tomography, and structural MR imaging for prediction of rapid conversion

21 Frisoni 21 to alzheimer disease in patients with mild cognitive impairment: A meta-analysis. AJNR Am J Neuroradiol 2009;30: Zhang S, Han D, Tan X, Feng J, Guo Y, Ding Y. Diagnostic accuracy of (18) F-FDG and (11) C-PIB-PET for prediction of short-term conversion to alzheimer's disease in subjects with mild cognitive impairment. Int J Clin Pract 2012;66: Albert J. Bayesian Computation with R. Springer, Thurfjell L, Lotjonen J, Lundqvist R, et al. Combination of biomarkers: PET [F]flutemetamol imaging and structural MRI in dementia and mild cognitive impairment. Neurodegener Dis Morinaga A, Ono K, Ikeda T, et al. A comparison of the diagnostic sensitivity of MRI, CBF-SPECT, FDG-PET and cerebrospinal fluid biomarkers for detecting alzheimer's disease in a memory clinic. Dement Geriatr Cogn Disord 2010;30: Visser PJ, Scheltens P, Verhey FR, et al. Medial temporal lobe atrophy and memory dysfunction as predictors for dementia in subjects with mild cognitive impairment. J Neurol 1999;246: Fritzsche KH, Stieltjes B, Schlindwein S, van Bruggen T, Essig M, Meinzer HP. Automated MR morphometry to predict alzheimer's disease in mild cognitive impairment. Int J Comput Assist Radiol Surg 2010;5: Cuingnet R, Gerardin E, Tessieras J, et al. Automatic classification of patients with alzheimer's disease from structural MRI: A comparison of ten methods using the ADNI database. Neuroimage 2011;56: Ewers M, Walsh C, Trojanowski JQ, et al. Prediction of conversion from mild cognitive impairment to alzheimer's disease dementia based upon biomarkers and neuropsychological test performance. Neurobiol Aging Roberson ED, Mucke L. 100 years and counting: Prospects for defeating alzheimer's disease. Science 2006;314: Rowe CC, Ackerman U, Browne W, et al. Imaging of amyloid beta in alzheimer's disease with 18F-BAY , a novel PET tracer: Proof of mechanism. Lancet Neurol 2008;7:

22 Frisoni Rabinovici GD, Rosen HJ, Alkalay A, et al. Amyloid vs FDG-PET in the differential diagnosis of AD and FTLD. Neurology 2011;77: Frisoni GB, Redolfi A, Manset D, Rousseau ME, Toga A, Evans AC. Virtual imaging laboratories for marker discovery in neurodegenerative diseases. Nat Rev Neurol 2011;7: Albert MS, DeKosky ST, Dickson D, et al. The diagnosis of mild cognitive impairment due to alzheimer's disease: Recommendations from the national institute on agingalzheimer's association workgroups on diagnostic guidelines for alzheimer's disease. Alzheimers Dement 2011;7: Mosconi L, Tsui WH, De Santi S, et al. Reduced hippocampal metabolism in MCI and AD: Automated FDG-PET image analysis. Neurology 2005;64: Mattsson N, Rosen E, Hansson O, et al. Age and diagnostic performance of alzheimer disease CSF biomarkers. Neurology 2012;78:

23 Frisoni 23 Legends for figures Figure 1. Diagnostic (A) positive and (B) negative Likelihood Ratio (LR+ and LR-) for correct classification between AD patients and healthy subjects broken down by marker by metrics and by marker by submarker. Only metrics with at least 3 studies are shown. Boxplots denote median, 1st, and 3rd quartiles, and whiskers denote minimum and maximum values (excluding outliers). Dashed lines indicates the conventional thresholds of clinical relevance (5 for LR+ and 0.2 for LR-). Figure 2. Prognostic (A) Positive and (B) Negative Likelihood Ratio (LR+ and LR-) for correct classification between stable and converter MCI patients, broken down by marker by metrics and by marker by submarker. Only metrics with at least 3 studies are shown. Boxplots denote median, 1st, and 3rd quartiles, and whiskers denote minimum and maximum values (excluding outliers). Dashed lines indicates the conventional thresholds of clinical relevance (5 for LR+ and 0.2 for LR-). Figure 3. Proportion of explained variance and significance of LR+ and LR- for correct classification between (A) AD patients and healthy subjects, and (B) stable and converter MCI patients. Analysis included only values for metrics with at least 3 studies. Rectangles denote the markers or metrics whose LR variance is computed.

24 Frisoni 24 Table 1. Markers, submarkers, and metrics reviewed in the current study. MARKER INCREASED BINDING OF AMYLOID IMAGING LIGANDS WITH PET TEMPOROPARIETAL HYPOMETABOLISM ON 18F-FDG PET TEMPOROPARIETAL HYPOPERFUSION ON SPECT MEDIAL TEMPORAL ATROPHY ON MRI SUBMARKER 11C-PIB 18F ligands m Tc-ECD and 23I-IMP 99m Tc-HMPAO Amygdala Hippocampus Entorhinal Cortex Temporal Horn SUVR (Standardized Uptake Value Ratio) Computer-aided visual read (Neurostat/3D-SSP) Visual read Linear measure METRIC Visual read DVR (Distribution Volume Ratio) Automated TP hypometabolism (t-sum/hci) Computer-aided visual read (SPM) Visual read Quantitative/ Semiquantitative assessment Visual read Manual volumetry Automated volumetry

25 Frisoni 25 Table 2. Technical features of imaging metrics. Acquisition Automation Stability Intensivity Normative population Increased binding of amyloid imaging agents with PET Visual Read Any acquisition Fully subjective IRR:.85 compliant with subject current guidelines classification, SUVR Any acquisition compliant with current guidelines no automation Automated uptake quantification and subject classification DVR Any acquisition Automated compliant with uptake current guidelines quantification and subject classification Temporoparietal hypometabolism on 18F-FDG PET Visual read t-sum Any acquisition compliant with ESNM/SNM guidelines Any acquisition compliant with current guidelines Fully subjective, no automation Fully automated IRR:.56 - Software for image management required seconds p/t per scan Software required for semiquantitative uptake estimates (SPM) - Software required for semiquantitative uptake estimates (SPM) - No apparel required seconds p/t per scan - Commercial software required (PMOD software) - Digital images usually in DICOM format - 3min p/t per scan Abnormality threshold Cost Not required Not required Very low Not required medium Not required > 1.2 medium Notes Not required Not required Very low Poor stability healthy elders t-sum > Low HCI Any acquisition compliant with current guidelines Fully automated - Commercial software required (Matlab for SPM) - Digital images required in NIFTI or DICOM format - 3min p/t per scan healthy elders Z-score of 1.5 or greater Low Age-correction still needs to be implemented Neurostat/3D-SSP Any acquisition compliant with current guidelines Fully automated - Freely available software - Digital images required in Analyze format - 2min p/t per scan healthy elders Z-score of 1.5 or greater Low

26 Frisoni 26 SPM Any acquisition compliant with current guidelines Fully automated - Digital images required in NIFTI or DICOM format healthy elders Z-score of 1.5 or greater Low Temporoparietal hypoperfusion on SPECT Visual read Any acquisition compliant with guidelines Fully subjective, no automation Semi-Quantit. /Quantitative Any acquisition compliant with current guidelines IRR:.79 - No apparel required seconds p/t per scan Automated - Software for image management required Not required Not required Very low healthy elders Z-score of 1.5 or greater Low Medial temporal atrophy on MR Visual Read T1-weighted acquisition Fully subjective, no automation TRTR:.82 to.97 IRR:.82 to.86 - No apparel required seconds p/t per scan Not required 0-1/2+ best at separating AD from healthy elders Very low Poorly sensitive to the mildest degrees of atrophy Manual segmentation ADNI T1-weighted 3D Fully manual TRTR.85 to.99 IRR:.80 to.95 - Software for image management required - Digital images required in Analyzed or DICOM format - 1h p/t per scan ADNI healthy elders Conventional, 95th percentile of age-specific normal distribution Low once set up the segmentation apparel and procedure Normalization by head size not standardized Freesurfer Adaboost-ACM Brainvisa/SACHA Linear Measure ADNI T1-weighted 3D ADNI T1-weighted 3D ADNI T1-weighted 3D T1-weighted acquisition Fully automated Fully automated Fully automated Fully manual TRTR.73 to.75 TRTR.83 to.85 TRTR.91 to.98 TRTR and IRR.90 to.99 - freely available software required - Digital images required in Analyzed or NIFTI format - 10h p/t per scan - freely available software required - Digital images required in Analyzed or NIFTI format - 10min p/t per scan - Commercial software required (The Anatomist) - Digital images required in NIFTI format ADNI healthy elders ADNI healthy elders ADNI healthy elders Conventional, 95th percentile of age-specific normal distribution Conventional, 95th percentile of age-specific normal distribution Conventional, 95th percentile of age-specific normal distribution - No apparel required Not required Conventional, 95th percentile of age-specific normal distribution Low Low Low Very low

27 Frisoni 27 Abbreviations: Positron Emission Tomography (PET);Distribution Volume Ratio (DVR); Standardized Uptake Value Ratio (SUVR); Fluorodeoxyglucose (18F); Fluoro-2-deoxyglucose Positron Emission Tomography (FDG-PET); three-dimensional Stereotactic Surface Projection (3D-SSP); PMOD Alzheimer's discrimination analysis tool (t-sum); hypometabolic convergence index (HCI); Statistical Parametric Mapping (SPM); Single-Photon Emission Computed Tomography (SPECT); Magnetic Resonance Imaging (MRI); Alzheimer s Disease Neuroimaging Initiative (ADNI); Test-retest reliability (TRTR); ) inter-rater reliability (IRR); European School of Nuclear Medicine/Society of Nuclear Medicine (ESNM/SNM).

28 Frisoni 28 Table 3. Accuracy figures of imaging markers for AD at the dementia and MCI stages. Dementia stage denotes accuracy for the discrimination of AD dementia patients from healthy elders (diagnostic accuracy). MCI stage denotes accuracy for the discrimination of MCI who subsequently convert to AD from stable MCI patients (prognostic accuracy). Figures denote pooled mean values and 95% Confidence Intervals (CIs) weighted for group size. * denotes metrics with < 3 studies. : no published studies on this specific metric. List of e-references is reported in Supplementary Material. Specificity Sensitivity Dementia stage MCI stage Dementia stage MCI stage % (CI 95%) n of healthy elders % (CI 95%) n of stable MCI % (CI 95%) n of AD % (CI 95%) n of MCI converters ereference Increased binding of amyloid imaging agents with PET 11C-PIB Visual Read 85% (64 to 95) 20* 100% (84 to 100) 21* (e1) SUVR 83% (76 to 88) % (44 to 61) % (84 to 95) % (74 to 88) 87 (e2-e12) DVR 88% (77 to 95) 32 56% (34 to 75) 18* 93% (85 to 98) % (52 to 100) 5* (e13, e14) 18F ligands SUVR 86% (81 to 90) % (48 to 94) 10* 87% (83 to 91) % (46 to 93) 9* (e15-e23) Temporoparietal hypometabolism on 18F-FDG PET Neurostat/3D- SSP t-sum/hci 90% (85 to 94) 85% (79 to 90) % (60 to 89) % (47 to 64) 28* 88% (84 to 91) % (84 to 90) % (60 to 90) % (57 to 78) 22* (e24-e36) 65 (e37-e42) SPM 83% (77 to 88) % (81 to 97) 53 84% (78 to 90) % (61 to 84) 49 (e4, e13, e43-e49) Visual Read 68% (57 to 78) 66 74% (60 to 84) 52* 85% (78 to % (88 to 97) 105* (e1, e13, e35, e50- e53) Temporoparietal hypoperfusion on SPECT 99m Tc-ECD and 23I-IMP 100% Visual Read (71 to 100) Semi-Quantit. /Quantitative 99m Tc-HMPAO Visual Read 92% (66 to 98) 84% (78 to 88) 9* 13* 58% (46 to 70) % (45 to 88) 54 89% (84 to 92) 68% (63 to 72) 14* 231* 79% (71 to 91) 422 (e28) 121 (e30, e54-e57) (e58-e69) Semi Quantit. /Quantitative 83% (79 to 87) % (51 to 75) 83* 81% (78 to 84) % (65 to 88) 45* (e57, e60, e61, e70- e85)