Statistical identification of predictors for peripheral neuropathy associated with administration of bortezomib, taxanes, oxaliplatin or vincristine using ordered logistic regression analysis

Abstract

Chemotherapy-induced peripheral neuropathy (CIPN) is a major drug-induced adverse reaction that becomes a dose-limiting toxicity. However, effective strategies for preventing or treating CIPN are lacking. Accordingly, this study aimed to statistically identify predictors for CIPN. Retrospective analysis was carried out for 190 patients who had been treated with bortezomib (n=28), taxanes (paclitaxel or docetaxel; n=58), oxaliplatin (n=52) or vincristine (n=52) at our hospital between April 2005 and December 2008. The severity of CIPN was assessed at the time of chemotherapy completion, graded as grade 0-5 in accordance with the National Cancer Institute Common Terminology Criteria for Adverse Events v3.0. Multivariate ordered logistic regression analysis was used to investigate predictors for CIPN. Predictors for CIPN in patients that were administered bortezomib were no co-administration of dexamethasone [odds ratio (OR), 0.455; confidence interval (CI), 0.208-0.955; P=0.0376] and sex (male) (OR, 3.035; CI, 1.356-6.793; P=0.0069). For taxanes (paclitaxel or docetaxel), the predictor for CIPN was a large number of chemotherapy cycles (OR, 2.379; CI, 1.035-5.466; P=0.0412). For oxaliplatin, the predictors for CIPN were a large number of chemotherapy cycles (OR, 3.089; CI, 1.598-5.972; P=0.0008) and no co-administration of non-steroidal anti-inflammatory drugs (OR, 0.393; CI, 0.197-0.785; P=0.0082). For vincristine, predictors for CIPN were a large number of chemotherapy cycles (OR, 6.015; CI, 1.880-19.248; P=0.0025) and co-administration of an analgesic adjuvant (OR, 3.907; CI, 1.383-11.031; P=0.0101). In conclusion, our study indicates that CIPN will be alleviated by the co-administration of dexamethasone with bortezomib and non-steroidal anti-inflammatory drugs with oxaliplatin.