Patient-Reported Outcome Measures in Chemotherapy-Induced Peripheral Neurotoxicity: Defining Minimal and Clinically Important Changes

Background: Chemotherapy-induced peripheral neurotoxicity (CIPN) is a common complication of cancer treatment that produces func-tional disability. Increasingly, patient-reported outcome measures (PROMs) are used to assess CIPN, providing a broader symptom per-spective than clinician-graded scales. Understanding when a reported change in CIPN symptoms meets the threshold for clinical significance is challenging. This study aimed to provide interpretation guidelines for validated CIPN PROMs, and thereby enable estimation of thresh-olds to identify clinically relevant symptoms. Methods: Patients com-mencing neurotoxic cancer treatments were assessed at 3 timepoints: baseline, midtreatment, and end-of-treatment. Trajectory of CIPN development was assessed by means of CIPN PROMs, EORTC Quality of Life - Chemotherapy-Induced Peripheral Neuropathy questionnaire (QLQ-CIPN20), and Functional Assessment of Can-cer Therapy/Gynecologic Oncology Group - Neurotoxicity ques-tionnaire (FACT/GOG-NTX). Thresholds were estimated for CIPN PROMs using the NCI CTCAE sensory neuropathy scale as the clinical anchor by midtreatment and end-of-treatment. Patients were assigned to a clinical change group according to CIPN development: either no development; grade 1 neuropathy (minimally important dif-ference [MID]); or grade 2 neuropathy (clinically important difference). Distribution-based estimates (SD, 0.5) were also evaluated as support-ive evidence. Results: In total, 406 patients were recruited to the study, of whom 62% (n=199/320) developed CIPN by midtreatment and 80% (n=274/343) by end-of-treatment. Anchor-based MID estimates by midtreatment were 5.06 (95% CI, 4.26-5.86) for the QLQ-CIPN20 and 3.54 (95% CI, 2.87-4.20) for the FACT/GOG-NTX. End-of-treatment MIDs were estimated to be 7.32 (95% CI, 6.23-8.40) for the QLQ-CIPN20 and 4.84 (95% CI, 3.98-5.70) for the FACT/GOG-NTX. Distri-bution-based MID estimations yielded lower values than anchor-based methods, at 3.73 for the QLQ-CIPN20 and 2.64 for the FACT/GOG-NTX at midtreatment and 5.52 for the QLQ-CIPN20 and 3.64 for the FACT/GOG-NTX at end-of-treatment. Conclusions: Findings from the present series aid meaningful interpretation for commonly used vali-dated CIPN PROMs and provide thresholds that serve as guidance on how to interpret score changes, which will be useful for design and evaluation of clinical trials and clinical practice.

Automated summary

This study aims to provide interpretation guidelines for validated CIPN PROMs, allowing estimation of thresholds to identify clinically relevant symptoms.