Ototoxicity risk assessment combining distortion product otoacoustic emissions with a cisplatin dose model

Marilyn F. Dille, Garnett P. McMillan, Kelly M. Reavis, Peter Jacobs, Stephen A. Fausti, Dawn Konrad-Martin

Research output: Contribution to journalArticle

19 Scopus citations

Abstract

An objective method for identifying ototoxic hearing loss among patients receiving cisplatin is necessary since the ability of patients to take a behavioral test may change over the course of treatment. Data from 56 monitoring visits by 19 Veterans taking cisplatin were used to identify combinations of distortion-product otoacoustic emission (DPOAE) metrics and ototoxicity risk factors that best identified ototoxic hearing loss. Models were tested that incorporated DPOAE metrics generated statistically using partial least-squares analysis. Models were also tested that incorporated a priori DPOAE change criteria, such as a minimum DPOAE level shift of 6 dB. Receiver Operating Characteristic analysis was used to compare the accuracy of these models. The best performing model incorporated weighted combinations of pre-treatment hearing, cumulative cisplatin dose and DPOAE metrics that were determined using partial least-squares and evaluated over a quarter octave range near each subjects' high frequency DPOAE limit. Using this model and the DPOAE recording methods described herein, the chance of ototoxic hearing change can be determined at any given observed change in DPOAE level. This approach appears to provide an accurate and rapid ototoxicity risk assessment (ORA) that once validated can be used clinically.

Original languageEnglish (US)
Pages (from-to)1163-1174
Number of pages12
JournalJournal of the Acoustical Society of America
Volume128
Issue number3
DOIs
StatePublished - Sep 1 2010

ASJC Scopus subject areas

  • Arts and Humanities (miscellaneous)
  • Acoustics and Ultrasonics

Fingerprint Dive into the research topics of 'Ototoxicity risk assessment combining distortion product otoacoustic emissions with a cisplatin dose model'. Together they form a unique fingerprint.

  • Cite this