Measurement of CSF core Alzheimer disease biomarkers for routine clinical diagnosis: do fresh vs frozen samples differ?

Background: Cerebrospinal fluid (CSF) amyloid-beta (Aβ) 42/40 ratio, threonine-181-phosphorylated-tau (p-tau), and total-tau (t-tau) represent core biomarkers of Alzheimer disease (AD). The recent availability of automated platforms has represented a significant achievement for reducing the pre-analytical variability of these determinations in clinical setting. With respect to classical manual ELISAs, these platforms give us also the possibility to measure any single sample and to get the result within approximately 30 min. So far, reference values have been calculated from measurements obtained in frozen samples. In this work, we wanted to check if the values obtained in fresh CSF samples differ from those obtained in frozen samples, since this issue is mandatory in routine diagnostic work.

Methods: Fifty-eight consecutive CSF samples have been analyzed immediately after lumbar puncture and after 1-month deep freezing (- 80 °C). As an automated platform, we used Lumipulse G600-II (Fujirebio Inc.). Both the fresh and the frozen aliquots were analyzed in their storage tubes.

Results: In fresh samples, a mean increase of Aβ40 (6%), Aβ42 (2%), p-tau (2%), and t-tau (4%) was observed as compared to frozen samples, whereas a slight decrease was observed for Aβ42/Aβ40 ratio (4%), due to the higher deviation of Aβ40 in fresh samples compared to Aβ42. These differences are significant for Aβ40, Aβ42/Aβ40 ratio, p-tau, and t-tau. Nevertheless, the Aβ42/Aβ40 ratio showed a lower variability (smaller standard deviation of relative differences) with respect to Aβ42. With respect to the AD profile according to the A/T/(N) criteria for AD diagnosis, no significant changes in classification were observed when comparing results obtained in fresh vs frozen samples.

Conclusions: Small but significant differences have been found for Aβ40, Aβ42/Aβ40 ratio, p-tau, and t-tau in fresh vs frozen samples. Importantly, these differences did not imply a modification in the A/T/(N) classification system. In order to know if different cutoffs for fresh and frozen samples are required, larger, multi-center investigations are needed.