Serum nerve growth factor and early motor neuropathy in children receiving neurotoxic chemotherapy: a prospective single-center cohort study

Abstract Purpose No circulating biomarker identifies children at risk of chemotherapy-induced peripheral neuropathy before symptoms appear. We examined whether serum nerve growth factor (NGF) predicts incident motor neuropathy. Methods Twenty-eight children starting vincristine, cisplatin or carboplatin at one tertiary pediatric center were enrolled prospectively. Serum NGF was measured by ELISA before the course and on days 7 and 30, with standardized neurological examination at each point. Groups were compared by Mann–Whitney U test, within-patient change by Friedman and Wilcoxon tests, the group-by-time interaction by mixed-effects modelling, and discrimination by ROC analysis. Results Median age was 4.6 years (range 2.0–16.9); 24 children (85.7%) received vincristine. Motor neuropathy developed in 8 children (28.6%, 95% CI 15.3–47.1; 33.3% of vincristine recipients), resolving in seven of eight. Serum NGF did not differ between groups at any time point (p = 0.218–0.611) and did not discriminate neuropathy (all AUC ≤ 0.66). NGF rose significantly across the month in children without neuropathy (Friedman p = 0.024) but not in those with neuropathy (p = 0.417); the group-by-time interaction was not significant (p = 0.160). Lower cumulative vincristine dose in affected children (p = 0.040) depended on a single outlier and did not survive correction for multiplicity. Conclusion Serum NGF did not predict motor neuropathy and cannot be recommended as a biomarker; the trajectory divergence is hypothesis-generating only. One in three vincristine-treated children developed motor neuropathy detectable by bedside examination within a month of a single course, supporting systematic neurological assessment from the outset of vinca alkaloid therapy.