To determine the clinical correlation between the Scheltens scale, a measure of hippocampal atrophy, and the amnestic syndrome of medial temporal type.

Hippocampal atrophy, either volumetric or visually-rated is arguably the most widely available biomarker for the diagnosis of Alzheimer’s disease (NIA-AA criteria 2011). An association between hippocampal atrophy and this amnestic syndrome has been claimed, but the clinical validity of this concept remains to be established. This type of amnesia is one of the key elements in the diagnosis of the most common clinical syndrome of Alzheimer’s disease.

Retrospective study of 274 consecutive patients that were assessed in a tertiary care Memory Clinic between 01/2014 and 11/2018. Free and cued recall memory was assessed with a neuropsychological battery, using the FCRST. MoCA was also performed for all patients. Brain imaging performed within six months of the tests was also available.

Scheltens scale has a good correlation with FCRST free recall (Pearson correlation = -0,425, p<0,001) and low correlation with FCRST Cueing index (Pearson correlation = - 0,293, p< 0,001). However, 42% of our cohort (n=93) had an abnormal FCRST Cueing index (≤ 71%), but a normal Scheltens score. As for the MoCA-MIS, 26% of our cohort (n=57) had an abnormal score (≤ 7), but a normal Scheltens score.

Our results demonstrate the limits of the use of hippocampal atrophy as a biomarker in Alzheimer’s disease, given the large proportion of possible “false negatives”. Moreover, FCRST is a more sensitive test than the MoCA-MIS for detecting memory disorders. It remains to be determined what other factors, like atrophy of different cerebral regions or cognitive reserve, can explain why some patients have an amnestic syndrome that doesn’t improve with cueing, but have no hippocampal atrophy on routine brain imaging.