Development and validation of a cisplatin dose-ototoxicity model

Area of Science:

• Ototoxicology
• Cancer Therapeutics
• Audiology

Background:

• Cisplatin chemotherapy is widely used but causes ototoxicity (hearing loss) in 50-60% of patients.
• Hearing shifts typically begin in the sensitive range for ototoxicity (SRO) and progress to lower frequencies.
• Predicting ototoxicity risk is crucial for patient counseling and treatment planning.

Purpose of the Study:

• To establish pretreatment risk curves for cisplatin-induced hearing loss within the SRO.
• To evaluate the predictive accuracy of these risk curves in a validation cohort of cancer patients.

Main Methods:

• Developed risk curves using a developmental sample (23 Veterans, 155 visits) analyzing cisplatin dose, radiation, concurrent medications, age, baseline hearing, and cancer characteristics.
• Modeled SRO threshold shifts using statistical analysis and reduced the model to significant variables.
• Validated predictions using ROC curve analysis on an independent sample (12 Veterans, 62 visits).

Main Results:

• Cumulative cisplatin dose and pre-exposure hearing were the only significant predictors of SRO hearing shifts.
• The developed dose-ototoxicity risk curves achieved an area under the ROC curve of 0.85 in the validation sample.

Conclusions:

• Validated dose-ototoxicity risk curves, incorporating cumulative cisplatin dose and pre-exposure hearing, can predict SRO hearing shifts.
• These curves aid in anticipating hearing loss, enhancing ototoxicity monitoring, patient counseling, and potential treatment adjustments.
• While not replacing serial audiometric testing, the tool significantly benefits ototoxicity management in cancer patients.