Search for reliable diagnostic markers for Alzheimer’s disease

Abstract

The diagnosis of Alzheimer’s disease (AD) is rather difficult not at least in the earlier stages of the disease. The diagnostic accuracy has been reported to be around 65-90%. These figures however, come from academic centers with a special interest in the disease and are based on patients followed for several years. One can speculate that the accuracy is lower at the non-specialist level. This thesis deals with the problem of finding reliable biomarkers in the cerebrospinal fluid (CSF) with acceptable sensitivity and specificity which can be utilized as diagnostic tools not only for the specialist.

The defining lesions of AD are the neurofibrillary tangles (NFT) and senile plaques (SP), composed of hyperphosphorylated tau and ß-amyloid (Aß) respectively. Therefore, we focused on total CSF-tau and Aß42 as markers for AD, both individually and in combination with each other. Paper I demonstrates a high sensitivity of CSF-tau for the diagnosis of AD but a lower specificity against other dementias. Specificity for more limited cerebral lesions such as ALS and Parkinsons disease, depression and normal aging is good. Moreover, we demonstrate that the stability between baseline CSF-tau and follow-up is good.

Paper II demonstrates high sensitivity for CSF-Aß42 as a diagnostic marker for AD and a low intra-individual variation. Low CSF- Aß42 levels are found even early in the course of the disease.

In paper III we again focus on CSF-tau but this time in a large prospective sample of consecutive AD patients from a community based sample. CSF-tau was analyzed in routine clinical neurochemistry. We confirm the high sensitivity and specificity for depression and normal aging demonstrated in paper I. An increase in CSF-tau is seen very early in the course of the disease.

In paper IV we combine CSF-tau and CSF-Aß42, looking at all patients referred to the clinic during one year. Both assays were analyzed in routine clinical neurochemistry. Again we can demonstrate high sensitivity and specificity. We suggest the use of the combination of the two biomarkers in common clinical practice as an inexpensive and reliable tool in the diagnosis of AD.

In paper V we proceed with the early diagnosis. As we previously found high CSF-tau and low CSF- Aß42 very early in the clinical course of AD we wanted to examine the potential of the two markers in MCI and indeed we found the usefulness even here. Since the sample was relatively small, more studies are required to confirm our findings.

Our conclusion is that the combination of the biomarkers tau and Aß42 in the CSF have high sensitivity in the diagnosis of AD even in the very early course of the disease. The specificity for psychiatric disorders (e.g. depression) and some neurological disorders (e.g. Parkinson and ALS) is also good, but lower specificity is found for other dementias. Thus, we also stress that these biomarkers cannot be used alone, but should be part of the full clinical work-up.