|Home | About | Journals | Submit | Contact Us | Français|
To combine MRI, FDG-PET and CSF biomarkers in diagnostic classification and two year prognosis of Mild Cognitive Impairment (MCI) and Alzheimer's disease (AD), examining: 1) which measures are most sensitive to diagnostic status? 2) to what extent do the methods provide unique information in diagnostic classification 3) which measures are most predictive of clinical decline?
Alzheimer's Disease Neuroimaging Initiative baseline MR, FDG-PET and CSF data from 42 controls, 73 MCI and 38 AD patients and two year clinical follow up data for 36 controls, 51 MCI, and 25 AD patients were analyzed. The hippocampus, entorhinal, parahippocampal, retrosplenial, precuneus, inferior parietal, supramarginal, middle temporal, lateral and medial orbitofrontal cortices were used as regions of interest. CSF variables included aβ42, t-tau, p-tau, and ratios of t-tau/aβ42 and p-tau/aβ42. Regression analyses were performed to determine measures' sensitivity to diagnostic status as well as two year change in Clinical Dementia Rating sum of boxes (CDR-SB), Mini Mental Status Exam (MMSE), and delayed logical memory in MCI.
Hippocampal volume, retrosplenial thickness and t-tau/aβ42 uniquely predicted diagnostic group. Change in CDR-SB was best predicted by retrosplenial thickness, MMSE by retrosplenial metabolism and thickness, and delayed logical memory by hippocampal volume.
All biomarkers were sensitive to diagnostic group. Combining MR morphometry and CSF biomarkers improved diagnostic classification (controls vs. AD). MR morphometry and PET were largely overlapping in value for discrimination. Baseline MR and PET measures were more predictive of clinical change in MCI than were CSF measures.
Multiple biomarkers have proven sensitive to Alzheimer's disease (AD) and Mild Cognitive Impairment (MCI), a potential prodromal stage of AD. These include patterns of regional cerebral atrophy and hypometabolism detected by MRI and fluorodeoxyglucose (FDG) PET (1), and quantification of specific proteins in CSF, including tau protein and aβ42 (2). Tau is associated with axonal microtubules and is the main structural element of neurofibrils in AD. High CSF tau levels probably reflect axonal degeneration (3). Aβ42 is derived from cleavage of amyloid precursor protein, and CSF aβ42 levels are lowered early in AD, possibly due to sequestering of aβ42 in neuritic plaques (4). A full spectrum of imaging and CSF analysis methods is seldom employed, limiting knowledge on how they may best be combined. The Alzheimer's Disease Neuroimaging Initiative (ADNI), a large multi-site study, was launched to enable analyses of combinations of different candidate biomarkers for AD.
Recent findings indicate that MRI can be used to quantify regional atrophy in MCI distinguishing early and later preclinical stages of AD (5), and such measures are predictive of clinical decline across one year (6–8). A pattern of parietotemporal metabolic reductions in MCI and AD, and frontal metabolic reductions later in the disease, has been established through the last decades of research (1, 9, 10) and has recently been confirmed in ADNI PET data (11). The relative sensitivity of FDG-PET and MR morphometry to AD-related changes is, however, not well established. It has been assumed that metabolic changes associated with neocortical dysfunction may be detectable by FDG-PET before atrophy appears. Consistent with this, De Santi et al. (12) reported that metabolism reductions exceeded volume losses in MCI, and Mosconi et al (13) found the same in presymptomatic early-onset familial AD. However, Jagust et al. (14) found that cingulate hypometabolism was a significant risk factor in addition to MR measures of hippocampal atrophy, but the latter was a more statistically robust risk factor in a group of cognitively impaired but not demented (CIND) elderly (15). Different brain characteristics relevant for the understanding MCI and AD may be captured by FDG-PET and MR morphometry. For instance, a report based on ADNI data has indicated that FDG-PET and MRI measures may be complementary and differentially sensitive to memory in health and disease, with metabolism being the stronger predictor in normal controls, and morphometry most related to memory function in AD (16). As for CSF-MRI relations, recent reports (17–22) indicate that cerebral anatomical differences are related to tau and aβ42 and behavioral cognitive measures in AD and MCI. However, MRI and CSF biomarkers have not simultaneously been related and compared to information obtained by FDG-PET. It is important to test the specific sensitivity of all biomarkers simultaneously to be able to optimize the combination of measures in diagnosis and prognosis. We investigated 1) Which methods are the most sensitive to established AD-related pathology? 2) To what extent do the methods provide unique vs. overlapping information? 3) Which methods are the most predictive of clinical decline across two years?
The raw data used in the preparation of this article were obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database (www.loni.ucla.edu/ADNI). ADNI was launched in 2003 by the National Institute on Aging (NIA), the National Institute of Biomedical Imaging and Bioengineering (NIBIB), the Food and Drug Administration (FDA), private pharmaceutical companies and non-profit organizations. The primary goal of ADNI has been to test whether serial MRI, PET, other biological markers, and clinical and neuropsychological assessment can be combined to measure the progression of MCI and early AD. The Principal Investigator of this initiative is Michael W. Weiner, VA Medical Center and University of California – San Francisco. There are many co-investigators, and subjects have been recruited from over 50 sites across the U.S. and Canada. The ADNI has recruited 229 healthy elderly, 398 MCI and 192 AD patients to participate and be followed for 2–3 years. For up-to-date information see www.adni-info.org.
ADNI eligibility criteria are described at http://www.adni-info.org/index.php?option=com_content&task=view&id=9&Itemid=43. Briefly, participants are 55–90 years of age, had an informant providing an independent evaluation of functioning, and spoke English or Spanish. Subjects were willing and able to undergo test procedures including neuroimaging and longitudinal follow up, and all gave informed consent. Specific psychoactive medications are excluded. General inclusion/exclusion criteria of the ADNI study are as follows: 1. Normal subjects: Mini-Mental State Examination (MMSE) (23) scores between 24–30 inclusive (no person enrolled as an NC in the present sample had an MMSE score below 26), Clinical Dementia Rating (CDR) of 0, non-depressed, non-MCI, and nondemented. 2. MCI subjects: MMSE scores between 24–30, inclusive (exceptions made on a case by case basis, but no such exceptional cases were enrolled as MCI patients in the present sample), a memory complaint, objective memory loss measured by Wechsler Memory Scale Logical Memory II, CDR of 0.5, absence of significant levels of impairment in other cognitive domains, essentially preserved activities of daily living, and an absence of dementia. 3. Mild AD: MMSE scores between 20–26 (inclusive; exceptions made on a case by case basis), CDR of 0.5 or 1.0, and meets NINCDS/ADRDA criteria for probable AD. Only ADNI subjects for whom adequate processed and quality checked MRI, FDG-PET and CSF baseline data were available were included. This yielded a total of 153 participants. Demographics are shown in Table 1.
The protocol was approved by the institutional review boards of participating sites. Written informed consent was obtained from all subjects, or from guardians of patients.
All scans used here were from 1.5 T scanners. Data were collected across a variety of scanners with protocols individualized for each scanner, as defined at http://www.loni.ucla.edu/ADNI/Research/Cores/index.shtml, and processed as described elsewhere (5, 16). Briefly, raw DICOM MRI scans (including two T1-weighted volumes per case) were downloaded from the ADNI site (http://www.loni.ucla.edu/ADNI/Data/index.shtml), reviewed for quality, automatically corrected for spatial distortion due to gradient nonlinearity (24) and B1 field inhomogeneity (25), registered, and averaged to improve signal-to-noise. Scans were segmented as described by Fischl et al. (26), yielding volumetric data for the hippocampal formation (consisting of the dentate gyrus, CA fields, subiculum/ parasubiculum and the fimbria (27)). The procedure (26, 28) uses a probabilistic atlas and applies a Bayesian classification rule to assign a neuroanatomical label to each voxel. The cortical surface was reconstructed to measure thickness at each surface point using a semi-automated approach described elsewhere (29–34). Thickness measurements were obtained by reconstructing representations of the gray/white matter boundary (29, 30) and the pial surface and then calculating the distance between those surfaces at each point across the cortical mantle. The measurement technique used here has been validated via histological (35) as well as manual measurements (36). The entire cortical surface was parcellated into numerous cortical areas (28, 37). To limit multiple comparisons, candidate regions of interest (ROIs) were selected based on previous MRI and PET findings (1, 5, 7, 16, 38–43) indicating sensitivity to AD-related pathology: The hippocampi, entorhinal, parahippocampal, retrosplenial, precuneus, inferior parietal, supramarginal, middle temporal, lateral and medial orbitofrontal gyri. In the parcellation method used here (37), the entire cingulate cortex is defined and divided into four separate regions, including rostral and caudal anterior cingulate, posterior cingulate, and isthmus cingulate, the latter referred to here as retrosplenial cortex for consistency with other published studies (5, 16, 38). The retrosplenial region may also be referred to as the isthmus of the cingulate or caudal posterior cingulate area in other contexts. For a depiction of the exact ROIs used, see Figure 1.
Subjects were scanned after a 4 hour fast (water only). Plasma glucose had to be ≤ 180 mg/dL for FDG to be injected. An intravenous catheter was placed in one arm for injection of [18F]FDG. Imaging began at 30 min post injection, and the scan was acquired as six 5-min frames. For each subject, FDG-PET frames were averaged and registered to the corresponding distortion-corrected and intensity-normalized MRI volume. PET activity for each subject was sampled onto their reconstructed cortical surface, averaged within each ROI and normalized to activity within the pons (44).
CSF samples obtained by lumbar puncture were examined for t-tau, p-tau and aβ42 using an immunoassay method (45). The measurements were performed by Drs. L. Shaw and J. Trojanowski of the ADNI Biomarker Core at the University of Pennsylvania School of Medicine. The CSF biomarkers analysed in the present paper include beta amyloid 1–42 (Aβ42; M and SDs: 202 (56), 159 (51), 136 (39) pg/mL for NC, MCI and AD respectively), tau protein (t-tau; M and SDs: 68 (28), 100 (65), 125 (67) pg/mL for NC, MCI and AD respectively), and phosphorylated-tau protein 181 (p-tau; M and SDs: 26 (17), 36 (19), 45 (23) for NC, MCI and AD respectively). The ratio of tau and Aβ42 (tau/Aβ42; M and SDs: 37 (.21), .74 (.67), .98 (.56) for NC, MCI and AD respectively), and p-tau Aβ42 ratio (p-tau/Aβ42; M and SDs: .16 (.16), .26 (.19), .36 (.22) for NC, MCI and AD respectively) were also included. A one-way ANOVA on the residual CSF values after age and sex were regressed out showed significant (p < .001) main effects of group on all variables. Post hoc tests controlling for multiple comparisons showed significant (p <.05) differences between NC and MCI, NC and AD, and MCI and AD, with a few exceptions where trends (p < .10) were observed (difference in t-tau between MCI and AD, p-tau between NC and MCI, and t-tau/aβ42 between MCI and AD).
Change scores were calculated by subtracting baseline scores from scores obtained at the two-year follow up. In addition to CDR-SB (46) and MMSE (23), delayed recall on the Wechsler Logical Memory Test (47) was included. This test requires the subject to recall a story read by the examiner after a 30–40 minute delay, and is sensitive to the episodic memory deficits in MCI.
A repeated measures general linear model with the ten ROIs × hemisphere (left, right) × diagnostic group (NC, MCI, AD) with age and sex as covariates, showed no significant effect of hemisphere across ROIs (F[1,148] = 1.530, p = .218) and no interaction of hemisphere × diagnostic group (F[2,148] = .847, p = .431). Hence, values were averaged across hemispheres, effects of age and sex regressed out, and the standardized residuals were used in the analyses. Correlation analyses with MR, FDG-PET and CSF measures were run to assess their covariance. To select the measures yielding the most explained variance for each method, the values were entered in three separate logistic stepwise regressions using MR, PET and CSF measures respectively, predicting NC vs. AD. The selected MR, PET and CSF variables were then entered simultaneously in multi-method stepwise logistic regression analyses predicting NC vs. AD, and NC vs. MCI. Next, the variables identified by the NC vs. AD classification analysis were correlated with two year follow up CDR-SB, MMSE, and delayed logical memory change scores in the MCI group, and were entered as predictors in stepwise regression analyses with the respective behavioral change scores as the dependent variables.
Correlation analyses in the MCI group for morphometry and metabolism for the 10 ROIs and the five CSF variables showed no significant (p <.05, corrected for 10 ROI comparisons) correlations among CSF variables and morphometry or metabolism in any ROI, whereas moderate correlations were found between morphometric and metabolic measures for hippocampus, entorhinal, retrosplenial and inferior parietal regions (see Supplemental Table 1).
Table 2 shows results of the separate logistic stepwise regressions predicting NC vs. AD classification based on MRI, FDG-PET and CSF measures. For MRI, hippocampal volume, entorhinal and retrosplenial thickness were included in the final model, yielding an overall classification accuracy of 85.0%, and about 71 % explained variance (Nagelkerke R2). For FDG-PET, entorhinal, retrosplenial and lateral orbitofrontal metabolism were included in the final model, yielding an overall classification accuracy of 82.5%, and about 62 % explained variance. For CSF, the ratio of t-tau/aβ42 was the single unique predictor, yielding an overall classification accuracy of 81.2%, and about 52 % explained variance. Thus, hippocampal volume, entorhinal and retrosplenial thickness, entorhinal, retrosplenial and lateral orbitofrontal metabolism, and t-tau/aβ42 ratio were entered in a logistic regression analysis to classify NC vs. AD, and the results are shown in Table 3.
In the final model, hippocampal volume, retrosplenial thickness and t-tau/ aβ42-ratio were included as predictors, yielding an overall classification accuracy of 88.8%, and about 78 % explained variance. Figure 2 depicts the receiver operating characteristics (ROC) curves for these variables when using one (hippocampal volume), vs a combination of two (hippocampal volume and t-tau/ aβ42-ratio) and all three variables (hippocampal volume, t-tau/ aβ42-ratio and retrosplenial thicknes) shown to be unique predictors of normal control vs Alzheimer's disease classification. Predicted values from logistic regressions were used for calculation of the ROC curves. Statistical comparisons of the areas under the curve (AUC) of these classifiers were performed by using the method of Hanley and McNeil (48). This approach yielded a significant difference (p <. 05) between the AUC when using hippocampal volume alone vs when using hippocampal volume and t-tau/ aβ42-ratio in combination, and hippocampal volume, t-tau/ aβ42-ratio and retrosplenial thickness in combination, The difference of the AUCs when using hippocampal volume and t-tau/ aβ42-ratio vs hippocampal volume, t-tau/ aβ42-ratio and retrosplenial thickness in combination was clearly smaller and not significant (p > .05). Note however, that all meaningful differences in AUC e.g. in terms of sensitivity vs specificity, causing the curves to cross, may not necessarily be captured as statistically significant. The same set of predictor variables were entered in an analysis to predict diagnostic classification for NC and MCI, which revealed that hippocampal volume and t-tau/abeta42-ratio were unique predictors, yielding an overall classification accuracy of 79.1%, and about 40 % explained variance.
Table 4 shows correlations for each variable included in the regression models and the cognitive change scores (CDR-SB, MMSE, delayed logical memory). In MCI, baseline retrosplenial thickness correlated with two year change in CDR-SB and MMSE, where thicker cortex was associated with less CDR-SB elevation and less MMSE reduction. Retrosplenial and entorhinal metabolism correlated negatively with MMSE change. Hippocampal volume correlated positively with delayed logical memory. There were no significant correlations with clinical change measures for t-tau/aβ42 in MCI. T-tests of the Fischer z-transformed correlation coefficients showed that the T-tau/aβ42 ratio correlated significantly lower (p < .05) with CDR-SB and MMSE change than did retrosplenial thickness; and significantly lower with MMSE change than did entorhinal and retrosplenial metabolism. Lateral orbitofrontal metabolism also correlated significantly lower with change in CDR-SB and delayed logical memory than did retrosplenial thickness and hippocampal volume.
In the stepwise regression analysis predicting CDR-SB change, only retrosplenial cortical thickness was included as a unique predictor (y = 1.231 −0.731x, p = .002), explaining 18% of the variance. In predicting MMSE change, retrosplenial metabolism was included in the first step (y = −1.193 +1.534 x1, p = .002 for x1, R2 = .22), and retrosplenial thickness was added in the second (y = −1.197 +1.177x1+ 0.776 x2, p = .009 for x1 and .042 for x2, R2 = .29). Only hippocampal volume was included as a predictor of delayed logical memory change (y = 0.240 +1.669 x1, p = .003 for x1, R2 = .17). The regression plots for CDR-SB and MMSE change predicted from retrosplenial thickness and metabolism, and delayed logical memory predicted from hippocampal volume, are shown in Figure 3. There was one outlier for the MMSE change score, with a 13 point decline. Without this outlier, only retrosplenial metabolism was included in the model for predicting MMSE change (y = −1.023 + 1.091 x1, p = .004 for x1, R2 = .16), but a trend was observed for retrosplenial thickness (p = .079).
Morphometry, metabolism and CSF biomarkers were all sensitive to diagnostic status. The best classification accuracy of NC vs. AD was obtained by MRI morphometry measures (hippocampal volume, entorhinal and retrospenial cortical thickness). However, classification accuracies close to those obtained by MRI were also obtained by FDG-PET (entorhinal, retrosplenial and lateral orbitofrontal metabolism) and CSF measures (T-tau/ aβ42-ratio). In the multi-modal analysis, FDG-PET measures appeared to provide largely redundant information, whereas hippocampal volume, retrosplenial thickness and the t-tau/aβ42 ratio were unique predictors of diagnostic status. In particular, the inclusion of the CSF biomarker in addition to MR hippocampal volume did result in a significant improvement in classification in terms of area under the curve. Thus, the combination of MR morphometry and CSF biomarkers yielded the highest diagnostic classification accuracy. Contrary to this, in prediction of clinical change over two years, FDG-PET and MRI morphometry were the best predictors. However, with the exception of retrosplenial metabolism and thickness in the prediction of change in MMSE scores, the two measures were largely redundant. Thus, it seems that the benefits of including both MRI morphometry and FDG-PET are modest in predicting clinical decline in MCI.
Whereas CSF biomarkers added to diagnostic accuracy at baseline, they did not predict two year clinical decline in the current MCI group. This may be somewhat surprising since previous studies have found decreased CSF aβ42 and/or tau or tau/aβ42 levels to be predictive of future dementia in MCI patients (2). Several factors may have contributed to discrepancies. First, the ongoing ADNI study may have a more heterogeneous MCI group than some of the previously published CSF studies. As pointed out by Hansson and colleagues (49), participants included in CSF studies have generally been highly selected, for example, by inclusion of only MCI patients that progress to AD. In ADNI, the ultimate end-point is not known for many MCI patients. Further, studies have often employed dichotomized variables for CSF values and prognosis (49, 50). A stable/conversion dichotomization involves clinical judgment, which may vary from physician to physician, and demands long follow-up intervals impractical for clinical trials. It may be advantageous to identify other pre-selection criteria, biomarkers or clinical measures of decline than conversion. In light of this, it is important to relate the biomarkers to easily administered continuous behavioral measures. Interestingly, another study investigating continuous variables (51) did not find any association between MMSE change and change in CSF levels of either aβ42, tau or p-tau (r = .18, −.03, and −.07, respectively). This does not mean that CSF measures are not related to clinical change. CSF tau/aβ42 ratio did correlate in the expected negative direction with change in logical delayed memory in the present sample, but the effect size was too modest to reach significance. Select MR morphometry and FDG-PET measures at baseline were significantly more sensitive to two year change in CDR-SB and MMSE than were CSF measures. Both cortical thickness and metabolism of parietal ROIs served as unique predictors of clinical decline, indicating that even though FDG-PET did not contribute uniquely to diagnostic classification when MR morphometric variance was accounted for, some additional prognostic information can be obtained by combining the two imaging modalities.
While the present findings show that the different biomarkers all were sensitive to diagnostic group, a question of great interest is whether the findings regarding specific measures can be applied on an individual subject basis. McEvoy et al. (7) recently reported that semiautomated individually specific quantitative MR imaging methods identical to those employed here can be used to identify a pattern of atrophy in MCI that is predictive of conversion to AD after 1 year. Hence, in light of the present findings indicating also somewhat superior sensitivity of such MR morphometry measures compared to other biomarkers, it does seem that these measures are prime candidates to be used on an individual basis e.g. to enrich clinical trials. However, as seen from Figure 3, while the MR morphometry measures evaluated here do predict two year change in screening and memory parameters among MCI patients, the regression plots also show considerable scatter. Hence, while these measures can yield individual prognostic information, this will be associated with considerable uncertainty and at present any such estimate must be made with great caution.
The present results are limited by a number of factors: Participants were selected based on willingness and ability to undergo MR and PET scanning and lumbar puncture, and may thus not be fully comparable to other samples. However, imaging is an integral part of the ADNI protocol, so participants did enter with the intention of having brain scans performed and about half of the ADNI participants have also agreed to have CSF samples drawn (52). In terms of age, MMSE score, Aβ42, t-tau, p-tau and ratios of t-tau/Aβ42 and p-tau/Aβ42, the subgroups studied in the present paper do appear to be representative of the larger ADNI sample. The mean values for these indices in the present sample appear very similar to those reported by Shaw et al. (52) for 410 participants with CSF measures, and all the present mean values for age, MMSE, and CSF measures for NC, MCI and AD and deviate less than 1/5 of the standard deviations from the means reported by Shaw et al. (52) for the larger groups.
Still, the present sample may of course not be fully representative of the general population. Further, the multi-site design of the ADNI is likely to add some noise in data collection. Finally, the ADNI study is still ongoing, and the ultimate status of the current MCI group is unknown. That being said, the present study involving multiple-sites and two years of follow-up likely represents a more realistic model for current clinical trial designs than longer interval, single-site studies.
Each of the biomarkers demonstrated potential to inform diagnosis and/ or prognosis and enrich clinical trials. As a single classifier, MR morphometry (hippocampal volume) was the most sensitive to diagnostic group, but the inclusion of CSF biomarkers (t-tau/Aβ42) did result in significant improvement of classification (NC/AD). Still, both quantitative MRI morphometry and regional metabolism as assessed by co-registered FDG-PET data provided better prediction of clinical decline than did CSF biomarkers. MRI morphometry showed somewhat superior diagnostic and prognostic sensitivity and is the least invasive, less expensive and most widely available method. MRI scans are often routinely required as part of the diagnostic work-up, so a broader application of MRI morphometry may be feasible and useful.
This work was supported by grants from the National Center for Research Resources (U24 RR021382 - CFN, AMD, DJH) and the National Institute on Aging (R01AG22381 - CFN, LKM, AMD; K01AG029218 – LKM). The work of Walhovd and Fjell was supported by the Norwegian Research Council. Data collection and sharing for this project was funded by the Alzheimer's Disease Neuroimaging Initiative (ADNI; Principal Investigator: Michael Weiner; NIH grant U01 AG024904). ADNI is funded by the National Institute on Aging, the National Institute of Biomedical Imaging and Bioengineering (NIBIB), and through generous contributions from the following: Pfizer Inc., Wyeth Research, Bristol-Myers Squibb, Eli Lilly and Company, GlaxoSmithKline, Merck & Co. Inc., AstraZeneca AB, Novartis Pharmaceuticals Corporation, Alzheimer's Association, Eisai Global Clinical Development, Elan Corporation plc, Forest Laboratories, and the Institute for the Study of Aging, with participation from the U.S. Food and Drug Administration. Industry partnerships are coordinated through the Foundation for the National Institutes of Health. The grantee organization is the Northern California Institute for Research and Education, and the study is coordinated by the Alzheimer's Disease Cooperative Study at the University of California, San Diego. ADNI data are disseminated by the Laboratory of Neuro Imaging at the University of California, Los Angeles.