P14.02.B LONGITUDINAL COMPARATIVE ANALYSIS OF BLOOD NEUROFILAMENTS AS BIOMARKERS OF CHEMOTHERAPY-INDUCED PERIPHERAL NEUROTOXICITY

Abstract

Abstract BACKGROUND The identification of a biomarker of chemotherapy-induced peripheral neurotoxicity (CIPN) remains an unmet need. Increasing evidence supports that light-chain Neurofilament (NfL) could be useful to quantify the extend of peripheral nerve damage due to chemotherapy. The aim of the study is to ascertain differences in plasma NfL (pNfL) among agents to evaluate the clinical usefulness of pNfL. MATERIAL AND METHODS Prospective longitudinal comparative study including patients treated with paclitaxel (N=24) for breast cancer, brentuximab vedotin (BV) (n=20) for lymphoma and oxaliplatin (n=16) for gastrointestinal cancer. All patients were assessed using the Total Neuropathy Score-clinical, Common Toxicity Criteria for Adverse Events before (T1), at finishing (T2) and at next follow-up (T3). Serial pNfL were collected at the same timepoints and quantified using the SIMOA technique. The changes in pNfL were compared between group of agents (platinum (oxaliplatin) or microtubule disruptors (MT) (BV or paclitaxel) and correlated with clinical data, depending on the degree of CIPN. Clinically relevant (CR) was defined as > grade 2-CIPN. RESULTS 60 patients are included, mostly women (66,7%) with median age of 52 [22-88] years. pNfL at T1 was 16,51 ± 1,65 pg/mL, and correlated with age (r= 0.284, p < 0.036). pNfL increased at T2 (141,39 ± 18,89 pg/mL) and decreased at last follow-up (T3) (28,08 ± 2,52 pg/mL) in the whole series, at a median of 3 (1-5) months after finishing chemotherapy. Patients with CR-CIPN had higher pNfL at T2 (209,90 ± 40,09 vs 103.02 ± 96.74 pg/mL, p=0.005). One third of patients receiving MT disruptors (31,6%) or platinum agents (35,7%) developed CR-CIPN. Higher pNfL level at T2 were identified in patients receiving anti MT agents than oxaliplatin (160, 27 ± 22,75 vs 70,04 ± 11,24 pg/mL), p<0,001). Among the group of MT, pNfL at T2 were higher in those receiving paclitaxel compared to BV (233,90 ± 31.84 vs 62,09 ± 9.60 pg/mL, p<0.001), despite no differences in rate of CR-CIPN (26,6% BV and 34.7% Paclitaxel, p=0.7). At T3, no differences in pNfL were identified between BV (33,53 ± 4,67 pg/mL) and paclitaxel (29,20 ± 4,24 pg/mL) treated patients. A significant correlation between pNfL and TNSc was identified (r = 0.663, p < 0.001). CONCLUSION Variable change of pNfL according to the type of agent and mechanism of neurotoxicity can be observed regardless of the comparable degree of clinical CIPN severity. Further studies to better dissect cut-off values for each agent are needed before implementing pNfL as CIPN biomarker.