To the Editor: A promising advance to complement clinical diagnosis at an early stage of Alzheimer's disease (AD) has been the measurement of biomarkers in cerebrospinal fluid (CSF).1 A decrease in CSF amyloid beta 42 (Aβ42) is a marker of AD pathology caused by the accumulation of amyloid-β in brain parenchyma,2 whereas increased total tau (t-tau) and hyperphosphorylated tau (p-tau) reflect neuronal degeneration and tangle pathology.3 AD is characterized by episodic memory loss, cognitive impairment, behavioral disorders, and finally, dementia.4 Individuals with mild cognitive impairment (MCI) have memory deficits but, in contrast to those with AD, are functionally intact and at higher risk of converting to AD than those without MCI.4 It is difficult to predict clinically which individuals with MCI possess AD brain pathology and will therefore progress to clinical AD. The aim of the present study was to evaluate CSF AD biomarkers in their capacity to discriminate AD from frontotemporal dementia (FTD) and to predict progression from MCI to AD. Data are presented on the first experience to the knowledge of the authors in Latin America of the use of CSF biomarkers in a clinical follow–up study. Individuals with MCI (n = 10), probable AD (n = 7), and FTD (n = 3) were recruited at the memory disorder clinic of an academic center in Argentina between 2005 and 2007. A trained neurologist performed a clinical evaluation, brain imaging, laboratory tests, and physical examination to exclude individuals with reversible causes of dementia. The Boston Naming Test,5 Rey Auditory and Visual Design Learning Verbal Test (RAVLT),6 Signoret's logical memory,7 Mini-Mental State Examination (MMSE),8 and Clinical Dementia Rating (CDR)9 were performed. Mean age was 69 for the groups with AD and MCI, whereas the group with FTD had a mean age of 60. Mean MMSE score was 28 for the group with MCI, 20 points for the group with AD, and 22 for the group with FTD. CDR was 0.5 for the group with MCI and 1 for the other groups. RAVLT mean results were 31 points for the group with MCI, 20 for the group with AD, and 15 for the group with FTD. Aβ42, t-tau, and p-tau were quantified in CSF using an enzyme-linked immunosorbent assay. Ratios of Aβ-42 to p-tau and CSF AD profile (Aβ42/(240 + [1.18 × t-tau]))10 were calculated. (A CSF ratio <1.3 was considered suggestive of AD pathology.) The Mann-Whitney one-tailed test was used to determine the difference between groups. Mean clinical follow-up was 4.7 years (range 1–8 years). As expected, functional status and overall cognitive tests deteriorated over time for individuals with AD and FTD. CDR was 2 for the groups with AD and FTD. For the group with MCI, participants were classified based on clinical and cognitive evolution into a group that progressed to AD (n = 5), with a mean MMSE score of 24 and CDR of 1, and a group that did not (n = 5), with MMSE and CDR scores that did not change from baseline. The mean value of biomarkers and the ratios were not significantly different in the three main groups (AD, MCI, FTD) because of the high dispersion observed in the MCI group. There were significant differences between the groups with AD and FTD in the biomarkers and ratios (Aβ42, P = .01; t-tau, P = .04; p-tau, P = .03). When the group with MCI was analyzed and divided according to clinical progression to AD over time, the mean value of Aβ42 was 355 pg/mL in those with progression, versus 800 pg/mL in those without AD progression, which was significantly different (Table 1). Regarding AD profile, a statistically significant difference was found in this subgroup analysis. A receiver operating characteristic curve analysis for each biomarker was performed between the groups with and without AD. The cutoff values for each biomarker were Aβ42, 532.5 pg/mL (sensitivity 100%, specificity 87.5%); t-tau 100 pg/mL (sensitivity 84.5%, specificity 87.5%); p-tau, 26.5 (sensitivity 69.2%, specificity 87.5%); Aβ42/p-tau, 20.5 (sensitivity 92.3%, specificity 87.5%); AD CSF profile 1.350 (sensitivity 100%, specificity 100%). Collectively, these results support the role of CSF biomarkers in the differential diagnosis of dementia and in the prediction of progression from MCI to AD. The conclusions of this study should be taken cautiously because of the small sample size and lack of confirmatory pathological examination, but active patient recruitment is underway to strengthen these observations. Overall, this first AD biomarker study in Latin America supports that combined analysis of all three core AD biomarkers represent a powerful tool in clinical setting. Conflict of Interest: This study was sponsored by the Fundación para la Lucha contra las Enfermedades Neurológicas de la Infancia. There is no conflict of interest for any of the authors. Author Contributions: Surace, Martinetto, Campos: study design; acquisition, analysis, and interpretation of data; preparation of manuscript. Martín, Smyth: acquisition of data. Cohen, Chrem Mendez, Russo, Amengual, Allegri, Leiguarda, Sevlever: study design, interpretation of results, preparation of manuscript. Sponsor's Role: The sponsor had no role in the design, methods, recruitment, data collection or analysis, or preparation of the paper.