# Time-course based assessment of patient factors and their relationship to chemotherapy induced peripheral neuropathy

## TL;DR

This study finds that patients with breast cancer who develop chemotherapy-induced neuropathy often have higher depression, lower physical function, and higher BMI before treatment.

## Contribution

The study introduces a predictive model using patient-reported outcomes to identify CIPN risk factors over time.

## Key findings

- 75% of patients developed CIPN, with higher depression scores linked to CIPN across all time periods.
- Obese patients were 3.78 times more likely to develop CIPN compared to those with normal weight.
- A random forest model achieved 84% accuracy in predicting CIPN using BMI, physical function, and anxiety scores.

## Abstract

Patients commonly experience chemotherapy-induced peripheral neuropathy (CIPN) as an adverse effect from chemotherapies, such as taxanes. In some patients, CIPN symptoms are severe and significantly impact their quality of life. Gaining a deeper understanding of how patient factors change over time, and identifying predisposing patient factors for CIPN susceptibility, could provide opportunities to mitigate the risk of CIPN and improve patient outcomes.

A total of 229 patients with breast cancer receiving taxane chemotherapy completed study visits over 12 months. Data from patient reported outcomes (PROs) were collected from validated questionnaires to assess CIPN, anxiety, depression, weight, physical function, and sleep disturbance. Wilcoxon signed-rank tests were conducted to evaluate changes in PROs between timepoints (pre-treatment, on-treatment, post-treatment). Logistic regression was used to compare PROs between CIPN and non-CIPN groups at each time point, after adjustment for age, race, and pre-treatment CIPN score. Results were adjusted for multiple comparisons using a false discovery rate (FDR) procedure (FDR P < .05). A random forest model using 19 patient features was employed to build a CIPN predictive model.

Out of 229 patients, 75% developed CIPN. Higher depression scores were associated with CIPN development across the three time periods: pre-treatment (OR = 1.20, FDR P = 5.3 × 10−3), on-treatment (OR = 1.32, FDR P = 4.4 × 10−4), and post-treatment (OR = 1.24, FDR P = 4.3 × 10−3). Similarly, pre-treatment PROMIS physical function T scores were lower among patients who developed CIPN (FDR P = 1.30 × 10−2), on-treatment (FDR P = 5.78 × 10−6) and post-treatment (FDR P = 8.7 × 10−5). On treatment anxiety scores were higher on-treatment in patients experiencing CIPN (OR = 1.30, FDR P = 4.3 × 10−3). Patients with obesity were 3.78 times more likely to develop CIPN compared to those with normal weight (FDR P = 3 × 10−3). Patients with CIPN did not report experiencing higher levels of sleep disturbance (FDR P > .05). The random forest predictive model reached an accuracy of 84% with body mass index (BMI), physical function score and general anxiety score as the most important predictors.

These findings indicate that those experiencing CIPN were more likely to report reduced physical functioning, increased anxiety and depression, and have higher BMIs. As new multi-modal approaches are developed for CIPN, pre-treatment physical functioning, BMI and other patient reported measures may provide predictive ability.