Proposed comprehensive ototoxicity monitoring program for VA healthcare (COMP-VA)

Abstract

Prevention and rehabilitation of hearing loss and tinnitus, the two most commonly awarded service-connected disabilities, are high priority initiatives in the Department of Veterans Affairs (VA). At least 4,000 Veterans, most with significant hearing loss, will receive cisplatin this year, with more than half sustaining permanent hearing shift and nearly 40% developing new tinnitus. With improved survivability following cancer treatment, Veterans treated with cisplatin are approached with the dual goals of effective treatment and preserved quality of life. This article describes COMP-VA, a comprehensive ototoxicity monitoring program developed for VA patients receiving cisplatin. The program includes an individualized pretreatment prediction model that identifies the likelihood of hearing shift given cisplatin dose and patient factors. It supports both manual and automated hearing testing with a newly developed portable audiometer capable of performing the recommended procedures on the chemotherapy unit during treatment. It also includes objective methods for identifying outer hair cell changes and predicting audiogram changes using distortion-product otoacoustic emissions. We describe this program of evidence-based ototoxicity monitoring protocols using a case example to give the reader an understanding of how this program would be applied, along with a plan for future work to accomplish the final stages of program development.

Keywords: COMP-VA; DPOAE; OtoID; aural rehabilitation; chemotherapy; cisplatin; distortion-product otoacoustic emissions; hearing; ototoxicity monitoring; sensitive range for ototoxicity.

Figures

Figure 1

Veteran using OtoID in automated (self-test) mode. Veteran is alerted to upcoming listening interval. Test tone is either played or catch trial occurs in which no tone is played. Patient is then instructed to respond via touch screen whether or not tone was heard. Earphones shown are Sennheiser HDA 200. Reprinted from Dille et al. [17].

Figure 2

Pretreatment risk assessment audiograms using threshold information from case study 1. Series of prediction audiograms were generated using planned cisplatin dosing regimen (dashed lines) and patient’s actual baseline audiogram (solid line) in decibels hearing level (dB HL) shown as function of test frequency. Gray shading indicates “speech banana” with phonemes. This model of conventional frequency thresholds yielded overall accuracy of 4.9 to 8.0 dB prediction error.

Figure 3

Behavioral screening for early hearing changes using sensitive range for ototoxicity (SRO) protocol obtained from case study 1. SRO thresholds are provided in decibels sound pressure level (dB SPL) as function of frequency. Bold line indicates baseline evaluation, while dotted line indicates evaluation associated with 190 mg of cisplatin. Gray dotted line indicates monitoring result with American Speech-Language-Hearing Association significant threshold shifts, prompting examination of conventional audiometric frequency range. Otoscopy and tympanometry results are used to help rule out conductive component to loss.

Figure 4

Data from case study 1 in which distortion-product otoacoustic emission (DPOAE) level change (in decibels) from baseline is shown as function of f2 frequency measured in fine (1/48-octave) frequency steps. Test-retest reference limits obtained using same test protocol but on similarly aged subjects with no exposure to cisplatin are shown (gray fill). DPOAE level decrement was greater than test-retest at highest frequencies following initial cumulative cisplatin dose of 190 mg (black line). Note also large increment (amplitude increased) at around 3,000 Hz. At cumulative dose of 380 mg (dotted line), DPOAEs decreased over wide range of frequencies, with decrements at several frequencies extending beyond reference limits. Downward pointing triangles at bottom of graph indicate frequencies valid at baseline but were low amplitude response at monitoring visit (indicated by row). Upward pointing triangles show frequencies that gained response or showed increment beyond reference limits compared with baseline test. Otoscopy and tympanometry results are used to help rule out conductive component to loss.

Figure 5

Distortion-product otoacoustic emission (DPOAE) level plotted as function of L2 stimulus input level for f2 of 5,040 Hz from case study 1 stratified by cumulative drug dose. Measurements failing to meet criteria for valid response of +6 dB greater than combined noise and system distortion are indicated by “X” on figure corresponding to that input level. Gray line identifies monitor visit at which estimated American Speech-Language- Hearing Association-significant threshold shift will occur within sensitive range for ototoxicity using ototoxicity risk assessment. DPOAE level changes at cisplatin dose of 190 mg to inputs of 35–45 dB sound pressure level (SPL) were not clinically significant (≤5 dB). Otoscopic and tympanometric results were used to rule out conductive component to loss.

Figure 6

Screen failure follow-up testing done at same time as hearing change to determine extent that hearing changes include frequencies within conventional audiometric frequency range. This graph, similar to an audiogram, plots conventional frequency (≤8,000 Hz) as function of audiometric threshold using data from case study 1, stratified by cumulative drug dose. Comparison of Figure 6 with Figure 2, using same ear, reveals close correspondence between subject’s predicted and actual loss at 380 mg. HL = hearing level.

Figure 7

Maximum permissible ambient noise levels (MPANLs) (in decibels re: 20 μPa, American National Standards Institute S3.1–1999) for audiometric test room is shown as function of frequency when HDA 200 (circles) circumaural earphones and Etymotic ER2 (triangles) earphones are used. Hospital noise (asterisk) levels as function of frequency measured in 1/3-octave frequency bands (Gordon et al., 2005 [45]) are also shown. Data show that reliable behavioral hearing thresholds can be obtained for frequencies above 2,000 Hz in most circumstances. In addition, room noise using OtoID is measured just before presentation of each tone as extra measure that noise levels in room are well controlled. Using insert earphones for collecting distortion-product otoacoustic emissions, all f2 frequencies can be used.

Figure 8

Flowchart demonstrating how comprehensive ototoxicity monitoring program for Department of Veterans Affairs (COMP-VA) monitoring protocol is used for detecting hearing and outer hair cell function changes. Pretreatment tests include otoscopy, tympanometry, distortion-product otoacoustic emission (DPOAE) testing, and air-conduction testing in conventional and extended high frequencies, and sensitive range for ototoxicity (SRO) is determined for subsequent testing. From these tests, pretreatment risk calculation is done for educational and patient counseling purposes and may prove helpful when planning professional resources. Screenings minimally include SRO audiogram for responsive/reliable patients and DPOAE level plotted as a function of f2 frequency to detect early ototoxic changes. From DPOAE level plotted as a function of f2 frequency, DPOAE level changes are calculated relative to baseline measures and compared with our reference limits to determine whether changes are greater than those attributable to normal test-retest variability and DPOAE input-output functions (obtained from multiple levels) are reserved for patients who are too sick to provide reliable hearing test. Any screen failure requires follow-up testing, ideally done using behavioral hearing test.