Pitfalls in grading severity of chemotherapy-induced peripheral neuropathy

Abstract

Background: Reliable reporting of chemotherapy-induced neurotoxicity is important. The objectives of the current study were to evaluate the differences in the peripheral neurotoxicity sections of several widely used chemotherapy-related toxicity grading systems, and the differences in the way in which observers interpret these scales.

Patients and methods: Two neurologists independently rated the severity of chemotherapy-induced peripheral neuropathy, according to WHO, ECOG, Ajani, and NCIC-CTC criteria in 37 patients.

Results: The highest percentage grade 1, grade 2 and grade 3 peripheral neurotoxicity was noted when employing the WHO, Ajani and NCIC-CTC scales, respectively. Percentage inter-observer agreement across all grades of severity ranged from 45.9 (NCIC-CTC) to 83.8 (WHO). The degree of agreement varied from 'poor to fair' to 'substantial'. Percentage inter-observer agreement for the dichotomy grade < or = 2 and grade 3 ranged from 81.1 (NCIC-CTC) to 94.6 (Ajani and WHO), however, exact agreement on grade 3 peripheral neurotoxicity ranged from 0 (Ajani and WHO) to 42% (NCIC-CTC). Percentage interscale agreement for the dichotomy grade < or = 2 and grade 3 varied from 67.6 (WHO and NCIC-CTC) to 100 (WHO and ECOG). Interobserver disagreement of severity grading was partly due to different interpretation of scale parameters.

Conclusions: Our results suggest that caution should be used in interpreting results across studies using different scales for neurotoxicity grading in chemotherapy-related peripheral neuropathy. When (multicentre) trials are to be undertaken with potential neurotoxic or neuroprotective agents, consensus should be sought regarding the toxicity rating scale used, and its interpretation by participating physicians.