Dement Geriatr Cogn Disord 2015;39:81–91 DOI: 10.1159/000366194 Accepted: July 22, 2014 Published online: October 24, 2014
© 2014 S. Karger AG, Basel 1420–8008/14/0392–0081$39.50/0 www.karger.com/dem
Original Research Article
The Use of MRI, CT and Lumbar Puncture in Dementia Diagnostics: Data from the SveDem Registry Farshad Falahati a Seyed-Mohammad Fereshtehnejad a Dorota Religa a–c Lars-Olof Wahlund a, c Eric Westman a Maria Eriksdotter a, c Divisions of a Clinical Geriatrics and b Neurogeriatrics, Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, and c Department of Geriatric Medicine, Karolinska University Hospital, Stockholm, Sweden
Key Words Alzheimer’s disease · Cerebrospinal fluid · Clinical practice pattern · Dementia and imaging · Differential diagnosis Abstract Background/Aims: The use of structural brain imaging [computed tomography (CT)/magnetic resonance imaging (MRI)] and the analysis of cerebrospinal fluid biomarkers are included in the guidelines for the diagnosis of dementia. The influence of variables such as age, gender and disease severity on the use of MRI, CT and lumbar puncture (LP) for the differential diagnosis of dementia and the consonance with the recommendations of the Swedish national guidelines were investigated. Methods: From the National Swedish Dementia Registry (SveDem), 17,057 newly diagnosed dementia patients were included in our study, with the majority from specialist care units (90%). Results: In the diagnostic workup, a CT was performed in 87%, MRI in 16% and LP in 40% of the cases. Age (p < 0.001) and cognitive status (p < 0.001) significantly influenced the use of MRI, CT or LP. Older patients with severe dementia were often investigated with CT. LP and MRI were used more often when less common dementia disorders were suspected. Conclusion: Our findings indicate that age, severity of cognitive impairment and the type of dementia disorder suspected are determinants for the choice of CT, MRI or LP. The majority of the dementia workups in specialist care units follow the recommendations of the Swedish national guidelines where CT is performed as a basic workup, and MRI and LP are chosen when extended workup is needed.
Farshad Falahati Division of Clinical Geriatrics, Department of Neurobiology Care Sciences and Society (NVS), Karolinska Institutet Novum, Plan 5, SE–141 86 Stockholm (Sweden) E-Mail farshad.falahati @ ki.se
Downloaded by: University of Utah 155.97.178.73 - 11/23/2014 3:32:27 PM
© 2014 S. Karger AG, Basel
82
Dement Geriatr Cogn Disord 2015;39:81–91 DOI: 10.1159/000366194
© 2014 S. Karger AG, Basel www.karger.com/dem
Falahati et al.: The Use of MRI, CT and Lumbar Puncture in Dementia Diagnostics: Data from the SveDem Registry
The diagnosis of dementia can be difficult, especially in the earliest stages. In order to confirm the presence of cognitive dysfunction and assess its severity, a review of the personal, family and medical history, a mental status assessment, physical and neurological exams, blood tests, cognitive tests, a functional assessment, and brain imaging are performed. Once the presence of cognitive dysfunction has been confirmed, the next step is to determine the subtype and cause of dementia. Accurate diagnosis of the type of the dementia disorder is of high importance to provide appropriate treatment. Structural brain imaging and the levels of certain proteins in cerebrospinal fluid (CSF) are potential biomarkers useful in the dementia workup process. Structural neuroimaging techniques including computed tomography (CT) and magnetic resonance imaging (MRI) provide valuable information on the brain structure and anatomy. Traditionally, structural neuroimaging was performed to exclude treatable and reversible causes of dementia such as brain tumors, subdural hematomas, cerebral infarcts, or hemorrhages. However, recent findings about the relationship between brain changes and neurodegenerative disorders have extended the role of structural imaging [1]. CSF occupies the brain cavities and directly interacts with the brain. Therefore, CSF can reflect pathological processes in the brain. Levels of amyloid-β as well as total and phosphorylated tau protein in the CSF have been shown to be valuable markers for the diagnosis of Alzheimer’s disease (AD) [2] and for differential diagnosis [3]. Structural neuroimaging and analysis of CSF are recommended in several national clinical guidelines for the diagnosis and management of dementia disorders such as the guidelines by the American Academy of Neurology [4], the European Federation of Neurological Societies [5, 6] and the Swedish National Board of Health and Welfare [7]. The Swedish guidelines were published in 2010 and state that CT should be performed in a basic dementia workup. MRI and CSF analysis should be used when an extended dementia investigation is needed [7]. A revision of the NINCDS-ADRDA criteria for the diagnosis of AD also highlights the importance of biomarkers [8]. Moreover, the use of CT, MRI and CSF analysis has been studied in other dementia disorders such as vascular dementia (VaD) [9, 10], dementia with Lewy bodies (DLB), Parkinson’s disease with dementia (PDD) [11–15], and frontotemporal dementia (FTD) [16, 17]. Although the use of structural imaging and CSF analysis is recommended by many national guidelines as discussed above, there are several factors that may influence the use in diagnostic workup, such as the clinical presentation, the physician’s choices and the local availability of the modalities. The aim of this study was to investigate which factors determine the selection of MRI, CT and CSF markers [obtained through lumbar puncture (LP)] in Sweden using the National Swedish Dementia Registry (SveDem) in dementia disorders as well as applying the actual modalities. Investigating which factors determine the choice of MRI, CT and LP in real clinical practice provides important knowledge for the implementation of clinical guidelines. SveDem is one of the largest registries of patients with dementia in the world and thus provides a unique opportunity to study the use of CT, MRI and LP in clinical practice. The specific aims were to investigate if baseline characteristics such as age, gender and disease severity [Mini-Mental State Examination (MMSE) score] influenced the choice of MRI, CT or LP and whether the use of these modalities differed due to clinical presentation. Moreover, the consonance of the use of CT, MRI and LP with the recommendations of the Swedish national guidelines was investigated.
Downloaded by: University of Utah 155.97.178.73 - 11/23/2014 3:32:27 PM
Introduction
83
Dement Geriatr Cogn Disord 2015;39:81–91 DOI: 10.1159/000366194
© 2014 S. Karger AG, Basel www.karger.com/dem
Falahati et al.: The Use of MRI, CT and Lumbar Puncture in Dementia Diagnostics: Data from the SveDem Registry
Materials and Methods Study Data The study data were obtained from SveDem (www.svedem.se), which is a national quality registry for dementia disorders that aims to improve the quality of diagnostics, treatment [18, 19] and care of patients with dementia disorders. The registry was launched in May 2007, and information on patients newly diagnosed with dementia and yearly follow-ups are recorded in a web-based system. Both primary and specialist care units are participating in the registry. SveDem data are regularly monitored by a coordinator. Registered patients are randomly selected (10%) from each unit and checked to verify if the data in SveDem correspond to the data in the medical records. Our dataset includes 17,057 patients diagnosed with incident dementia during the years 2007–2011. The majority of these patients (92.5%) were diagnosed in specialist care units. About 90% of the specialist units in Sweden participate in SveDem (56/62 units). The following variables were analyzed: age, gender, MMSE score at the time of the start of the workup, dementia workup investigations, and the resulting dementia diagnosis. Particularly, detailed diagnostic assessments such as the use of CT, MRI and LP tests (yes/no/do not know) were collected. There is no information on the results of the diagnostic assessments registered. The dementia diagnoses in SveDem are classified according to the ICD-10 [20] or defined criteria where appropriate. In the SveDem registry, nine dementia subtypes are defined: early-onset AD (EOAD), late-onset AD (LOAD), mixed AD, VaD, DLB [12], PDD [21], FTD [16], unspecified dementia, and other. The mixed AD subtype is a combination of patients with AD and VaD symptoms. Dementias with unknown etiology were categorized as unspecified dementia. Alcohol dementias and rare dementia disorders such as corticobasal degeneration were grouped as ‘other’ dementia. Subjects with mild cognitive impairment who do not fulfil diagnostic criteria for dementia are not registered in SveDem. Figure 1a shows the prevalence of the dementia subtypes from the study population. Ethics At the time of diagnosis, patients are informed orally and in writing about the SveDem registry, and they may decline participation. This study was approved by the regional ethics committee of Stockholm. More information about the SveDem registry is available on www.svedem.se. Statistical Analysis Frequency in percentages was reported to describe categorical variables (e.g., gender, diagnostic tests and dementia disorders), and means and standard deviations (SD) were used to describe quantitative variables (e.g., age). Baseline characteristics were evaluated with regard to the use of one or several of the three diagnostic modalities, i.e., CT, MRI and LP. Moreover, obtaining the selected diagnostic tests with respect to different dementia disorders was investigated. In univariate comparisons, the χ2 and the independent samples t test were used for categorical and quantitative variables, respectively. In multivariate analysis, binary logistic regression was used to calculate adjusted odds ratios (ORs) for the association between independent indicators and diagnostic modalities (CT, MRI and LP). A p value
© 2014 S. Karger AG, Basel 1420–8008/14/0392–0081$39.50/0 www.karger.com/dem
Original Research Article
The Use of MRI, CT and Lumbar Puncture in Dementia Diagnostics: Data from the SveDem Registry Farshad Falahati a Seyed-Mohammad Fereshtehnejad a Dorota Religa a–c Lars-Olof Wahlund a, c Eric Westman a Maria Eriksdotter a, c Divisions of a Clinical Geriatrics and b Neurogeriatrics, Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, and c Department of Geriatric Medicine, Karolinska University Hospital, Stockholm, Sweden
Key Words Alzheimer’s disease · Cerebrospinal fluid · Clinical practice pattern · Dementia and imaging · Differential diagnosis Abstract Background/Aims: The use of structural brain imaging [computed tomography (CT)/magnetic resonance imaging (MRI)] and the analysis of cerebrospinal fluid biomarkers are included in the guidelines for the diagnosis of dementia. The influence of variables such as age, gender and disease severity on the use of MRI, CT and lumbar puncture (LP) for the differential diagnosis of dementia and the consonance with the recommendations of the Swedish national guidelines were investigated. Methods: From the National Swedish Dementia Registry (SveDem), 17,057 newly diagnosed dementia patients were included in our study, with the majority from specialist care units (90%). Results: In the diagnostic workup, a CT was performed in 87%, MRI in 16% and LP in 40% of the cases. Age (p < 0.001) and cognitive status (p < 0.001) significantly influenced the use of MRI, CT or LP. Older patients with severe dementia were often investigated with CT. LP and MRI were used more often when less common dementia disorders were suspected. Conclusion: Our findings indicate that age, severity of cognitive impairment and the type of dementia disorder suspected are determinants for the choice of CT, MRI or LP. The majority of the dementia workups in specialist care units follow the recommendations of the Swedish national guidelines where CT is performed as a basic workup, and MRI and LP are chosen when extended workup is needed.
Farshad Falahati Division of Clinical Geriatrics, Department of Neurobiology Care Sciences and Society (NVS), Karolinska Institutet Novum, Plan 5, SE–141 86 Stockholm (Sweden) E-Mail farshad.falahati @ ki.se
Downloaded by: University of Utah 155.97.178.73 - 11/23/2014 3:32:27 PM
© 2014 S. Karger AG, Basel
82
Dement Geriatr Cogn Disord 2015;39:81–91 DOI: 10.1159/000366194
© 2014 S. Karger AG, Basel www.karger.com/dem
Falahati et al.: The Use of MRI, CT and Lumbar Puncture in Dementia Diagnostics: Data from the SveDem Registry
The diagnosis of dementia can be difficult, especially in the earliest stages. In order to confirm the presence of cognitive dysfunction and assess its severity, a review of the personal, family and medical history, a mental status assessment, physical and neurological exams, blood tests, cognitive tests, a functional assessment, and brain imaging are performed. Once the presence of cognitive dysfunction has been confirmed, the next step is to determine the subtype and cause of dementia. Accurate diagnosis of the type of the dementia disorder is of high importance to provide appropriate treatment. Structural brain imaging and the levels of certain proteins in cerebrospinal fluid (CSF) are potential biomarkers useful in the dementia workup process. Structural neuroimaging techniques including computed tomography (CT) and magnetic resonance imaging (MRI) provide valuable information on the brain structure and anatomy. Traditionally, structural neuroimaging was performed to exclude treatable and reversible causes of dementia such as brain tumors, subdural hematomas, cerebral infarcts, or hemorrhages. However, recent findings about the relationship between brain changes and neurodegenerative disorders have extended the role of structural imaging [1]. CSF occupies the brain cavities and directly interacts with the brain. Therefore, CSF can reflect pathological processes in the brain. Levels of amyloid-β as well as total and phosphorylated tau protein in the CSF have been shown to be valuable markers for the diagnosis of Alzheimer’s disease (AD) [2] and for differential diagnosis [3]. Structural neuroimaging and analysis of CSF are recommended in several national clinical guidelines for the diagnosis and management of dementia disorders such as the guidelines by the American Academy of Neurology [4], the European Federation of Neurological Societies [5, 6] and the Swedish National Board of Health and Welfare [7]. The Swedish guidelines were published in 2010 and state that CT should be performed in a basic dementia workup. MRI and CSF analysis should be used when an extended dementia investigation is needed [7]. A revision of the NINCDS-ADRDA criteria for the diagnosis of AD also highlights the importance of biomarkers [8]. Moreover, the use of CT, MRI and CSF analysis has been studied in other dementia disorders such as vascular dementia (VaD) [9, 10], dementia with Lewy bodies (DLB), Parkinson’s disease with dementia (PDD) [11–15], and frontotemporal dementia (FTD) [16, 17]. Although the use of structural imaging and CSF analysis is recommended by many national guidelines as discussed above, there are several factors that may influence the use in diagnostic workup, such as the clinical presentation, the physician’s choices and the local availability of the modalities. The aim of this study was to investigate which factors determine the selection of MRI, CT and CSF markers [obtained through lumbar puncture (LP)] in Sweden using the National Swedish Dementia Registry (SveDem) in dementia disorders as well as applying the actual modalities. Investigating which factors determine the choice of MRI, CT and LP in real clinical practice provides important knowledge for the implementation of clinical guidelines. SveDem is one of the largest registries of patients with dementia in the world and thus provides a unique opportunity to study the use of CT, MRI and LP in clinical practice. The specific aims were to investigate if baseline characteristics such as age, gender and disease severity [Mini-Mental State Examination (MMSE) score] influenced the choice of MRI, CT or LP and whether the use of these modalities differed due to clinical presentation. Moreover, the consonance of the use of CT, MRI and LP with the recommendations of the Swedish national guidelines was investigated.
Downloaded by: University of Utah 155.97.178.73 - 11/23/2014 3:32:27 PM
Introduction
83
Dement Geriatr Cogn Disord 2015;39:81–91 DOI: 10.1159/000366194
© 2014 S. Karger AG, Basel www.karger.com/dem
Falahati et al.: The Use of MRI, CT and Lumbar Puncture in Dementia Diagnostics: Data from the SveDem Registry
Materials and Methods Study Data The study data were obtained from SveDem (www.svedem.se), which is a national quality registry for dementia disorders that aims to improve the quality of diagnostics, treatment [18, 19] and care of patients with dementia disorders. The registry was launched in May 2007, and information on patients newly diagnosed with dementia and yearly follow-ups are recorded in a web-based system. Both primary and specialist care units are participating in the registry. SveDem data are regularly monitored by a coordinator. Registered patients are randomly selected (10%) from each unit and checked to verify if the data in SveDem correspond to the data in the medical records. Our dataset includes 17,057 patients diagnosed with incident dementia during the years 2007–2011. The majority of these patients (92.5%) were diagnosed in specialist care units. About 90% of the specialist units in Sweden participate in SveDem (56/62 units). The following variables were analyzed: age, gender, MMSE score at the time of the start of the workup, dementia workup investigations, and the resulting dementia diagnosis. Particularly, detailed diagnostic assessments such as the use of CT, MRI and LP tests (yes/no/do not know) were collected. There is no information on the results of the diagnostic assessments registered. The dementia diagnoses in SveDem are classified according to the ICD-10 [20] or defined criteria where appropriate. In the SveDem registry, nine dementia subtypes are defined: early-onset AD (EOAD), late-onset AD (LOAD), mixed AD, VaD, DLB [12], PDD [21], FTD [16], unspecified dementia, and other. The mixed AD subtype is a combination of patients with AD and VaD symptoms. Dementias with unknown etiology were categorized as unspecified dementia. Alcohol dementias and rare dementia disorders such as corticobasal degeneration were grouped as ‘other’ dementia. Subjects with mild cognitive impairment who do not fulfil diagnostic criteria for dementia are not registered in SveDem. Figure 1a shows the prevalence of the dementia subtypes from the study population. Ethics At the time of diagnosis, patients are informed orally and in writing about the SveDem registry, and they may decline participation. This study was approved by the regional ethics committee of Stockholm. More information about the SveDem registry is available on www.svedem.se. Statistical Analysis Frequency in percentages was reported to describe categorical variables (e.g., gender, diagnostic tests and dementia disorders), and means and standard deviations (SD) were used to describe quantitative variables (e.g., age). Baseline characteristics were evaluated with regard to the use of one or several of the three diagnostic modalities, i.e., CT, MRI and LP. Moreover, obtaining the selected diagnostic tests with respect to different dementia disorders was investigated. In univariate comparisons, the χ2 and the independent samples t test were used for categorical and quantitative variables, respectively. In multivariate analysis, binary logistic regression was used to calculate adjusted odds ratios (ORs) for the association between independent indicators and diagnostic modalities (CT, MRI and LP). A p value
