Indeed, the failure of many recent clinical trials aimed at AD is commonly attributed to the exclusive enrollment of participants who already have mild or moderate dementia and concomitant neuron loss. Therefore, tools and strategies must be developed to diagnose and enroll individuals in the SCH772984 942183-80-4 pre-clinical phase of AD, when brain pathology is present but cognition remains intact. By definition, this phase is not reliably detected by clinical examination, so biomarkers will be required for diagnosis. Ideally, biomarkers should also estimate an individual��s risk of impending cognitive decline and even allow monitoring of pathological progression and response to treatment. Once such biomarkers are developed, clinical trials should become more efficient and effective treatments will be identified more quickly. Subsequently, once successful treatments are identified, these biomarkers are likely to remain useful in a clinical setting. Some progress has already been made in this direction. To date, leading modalities for such biomarkers include radiological imaging and cerebrospinal fluid analysis. Both techniques can detect amyloid deposits in the brain either directly, using amyloidbinding tracer compounds and positron emission tomography, or indirectly, by measuring low CSF beta-amyloid42 concentrations that correlate with amyloid deposition. Imaging and fluid biomarker studies have also shown potential to predict cognitive decline by measuring amyloid deposition, regional volumetric and metabolic changes in the brain, or specific changes in the CSF proteome. CSF SAR131675 analysis may even allow classification of disease stage and monitoring of acute changes in response to disease modifying therapies, as illustrated recently with gamma-secretase inhibitors. In spite of these advances, however, these techniques must still be improved. Additional biomarkers will be required to improve the sensitivity and specificity of pre-clinical AD diagnosis, increase the accuracy of prognosis, and expand the breadth of pathophysiological changes that can be monitored. CSF proteome analysis provides a favorable arena for such efforts. Indeed, many increasingly more powerful yet complementary proteomics technologies have been leveled at CSF biomarker discovery in the past decade, including: variations of 2D gel electrophoresis; SELDI-TOF-MS; offline LC �C MALDI-TOF; and LC-MS/MS with either isotopecoded affinity tags, Tandem Mass Tags or iTRAQ. These techniques have all been used successfully to identify candidate biomarkers because they provide accurate relative quantitative information between or among samples. In order to provide this information, they share a common requirement: proteins or peptides must be stained or labeled for precise and accurate quantification. This requirement necessarily increases procedural costs and also may introduce additional sources of error. These techniques also share a major limitation in clinical proteomics : they cannot readily be used to compare directly large numbers of samples, even with advances in multiplexing technologies and strategies. This second shortcoming is quite important because most CSF biomarkers, at least in the AD field, show relatively modest disease-associated quantitative changes, on the order of 30%; detecting such differences with statistical rigor in a cross-sectional study requires precise and accurate measurements with potentially hundreds of CSF samples. Label-free, quantitative proteomic methods have emerged that obviate the requirement for protein staining or peptide labeling. Many of these ��label-free�� approaches take advantage of the correlation between high-resolution LC/MS extracted ion currents and peptide abundances. Bioinformatics software tools have been developed that align LC elution times and accurate m/z values of the XIC��s across numerous samples.