Multicenter validation of CSF neurofilaments as diagnostic biomarkers for ALS
Oeckl P., Jardel C., Salachas F., Lamari F., Andersen PM., Bowser R., de Carvalho M., Costa J., van Damme P., Gray E., Grosskreutz J., Hernández-Barral M., Herukka SK., Huss A., Jeromin A., Kirby J., Kuzma-Kozakiewicz M., Amador MDM., Mora JS., Morelli C., Muckova P., Petri S., Poesen K., Rhode H., Rikardsson AK., Robberecht W., Rodríguez Mahillo AI., Shaw P., Silani V., Steinacker P., Turner MR., Tüzün E., Yetimler B., Ludolph AC., Otto M.
© 2016 World Federation of Neurology on behalf of the Research Group on Motor Neuron Diseases. OBJECTIVE: Neurofilaments are leading neurochemical biomarkers for amyotrophic lateral sclerosis (ALS). Here, we investigated the effect of preanalytical factors on neurofilament concentrations in cerebrospinal fluid (CSF) in a “reverse” round-robin with 15 centers across Europe/U.S. METHODS: Samples from ALS and control patients (5/5 each center, n = 150) were analyzed for phosphorylated neurofilament heavy chain (pNfH) and neurofilament light chain (NfL) at two laboratories. RESULTS: CSF pNfH was increased (p < 0.05) in ALS in 10 out of 15 centers and NfL in 5 out of 12 centers. The coefficient of variation (CV%) of pNfH measurements between laboratories was 18.7 ± 19.1%. We calculated a diagnostic cut-off of > 568.5 pg/mL for pNfH (sensitivity 78.7%, specificity 93.3%) and > 1,431pg/mL for NfL (sensitivity 79.0%, specificity 86.4%). CONCLUSION: Values in ALS patients are already comparable between most centers, supporting eventual implementation into clinical routine. However, continuous quality control programs will be necessary for inclusion in the diagnostic work-up.