Objective Measures of Ototoxicity
Total Page:16
File Type:pdf, Size:1020Kb
The following article appeared in the September 2005 issue (Vol. 9, No. 1, pp. 10-16) of the Division 6 publication Perspectives on Hearing and Hearing Disorders: Research and Diagnostics. To learn more about Division 6, contact the ASHA Action Center at 1-800-498-2071 or visit the division’s Web page at www.asha.org/about/membership-certification/divs/div_6.htm. Objective Measures of Ototoxicity Elizabeth Leigh-Paffenroth Veterans Affairs Rehabilitation Research and Development Service, National Center for Rehabilitative Auditory Research, VA Medical Center Portland, OR Department of Otolaryngology, Oregon Health & Science University Portland, OR Kelly M. Reavis, Jane S. Gordon, Kathleen T. Dunckley Veterans Affairs Rehabilitation Research and Development Service, National Center for Rehabilitative Auditory Research, VA Medical Center Portland, OR Stephen A. Fausti and Dawn Konrad-Martin Veterans Affairs Rehabilitation Research and Development Service, National Center for Rehabilitative Auditory Research, VA Medical Center Portland, OR Department of Otolaryngology, Oregon Health & Science University, Portland, OR A leading cause of preventable sensorineural hear- of patients who are unable to provide reliable re- ing loss is therapeutic treatment with medications that sponses; subsequently, many of these patients do not are toxic to inner ear tissues, including certain drugs receive monitoring for ototoxic-induced changes in used to fight cancer and life-threatening infectious dis- their hearing. The development of objective measures eases. Ototoxic-induced hearing loss typically begins that do not require patient cooperation is necessary to in the high frequencies and progresses to lower fre- monitor all patients receiving ototoxic drugs. quencies as drug administration continues (Campbell Two objective measures offer promise in their abil- & Durrant, 1993; Campbell et al., 2003; Macdonald, ity to detect and to monitor hearing changes caused by Harrison, Wake, Bliss, & Macdonald, 1994). It is im- ototoxicity: auditory brainstem responses (ABRs) and portant to detect ototoxicity before damage occurs to otoacoustic emissions (OAEs). ABRs are an objective, the region of hearing < 4 kHz, which is important for pre-behavioral measure of neural responses in the speech perception (De Paolis, Janota, & Frank, 1996). brainstem that reflect hearing function. OAEs are an Sensitive and time-efficient behavioral techniques have objective, pre-behavioral measure of cochlear mechani- been developed to monitor high-frequency (> 8 kHz) cal responses that reflect cochlear outer hair cell func- hearing to detect ototoxic-induced changes before dam- tion. After middle ear dysfunction has been ruled out, age has progressed to lower frequencies (Fausti et al., OAEs may be an excellent indicator of early ototoxic 1999). Hearing thresholds obtained through behavioral damage. Abnormal middle ear function and baseline audiometry are the current gold standard for detecting hearing loss greater than about 40 dB HL may pre- ototoxic-induced changes in hearing. However, behav- clude effective monitoring using OAEs. Use of ABR ioral techniques are not effective for a large population testing may be more appropriate in such cases. Hearing and Hearing Disorders: Research and Diagnostics 2 Changes in hearing can be identified by these tech- liest ototoxic changes, it is necessary to use high-fre- niques when monitoring patients before, during, and quency stimuli to elicit the ABR (Stapells & Oates, 1997). after drug treatment. The detection of hearing change at frequencies > 8 kHz can provide valuable information regarding individual A common use of ABRs and OAEs is the detection susceptibility to ototoxicity. This early detection of oto- of hearing loss in neonatal hearing screenings (Cone- toxicity makes it possible to prevent the progression of Wesson, Kurtzberg, & Vaughan, 1987; Hall, Smith, & hearing loss to lower frequencies where speech per- Popelka, 2004; Jacobson & Jacobson, 2004; Norton et ception can be greatly impaired. The key to detecting al., 2000; Sininger, Abdala, & Cone-Wesson, 1997). hearing changes in patients at risk for ototoxicity is These measures require no behavioral response and serial monitoring (i.e., before, during, and after drug can be recorded in sleeping patients. Assuming these treatment; Campbell et al., 2003). Thus, it is imperative measures correctly identify changes in hearing sensi- that the measure used to detect ototoxicity is reliable tivity, ABR and OAE testing is amenable to ototoxicity over time. monitoring of patients who cannot provide reliable re- sponses to behavioral hearing testing (e.g., infants and The reliability of ABRs elicited by high-frequency sick patients). Currently, there are no accepted clinical stimuli has been studied extensively by Fausti and col- protocols or criteria for ototoxic change using objec- leagues (Fausti, Frey, Henry, Olson, & Schaffer, 1992; tive measures of ototoxicity. A barrier to the widespread Mitchell, Fausti, & Frey, 1994; Fausti et al., 1995; Henry, use of ABR and OAE for ototoxicity monitoring is re- Fausti, Kempton, Trune, & Mitchell, 2000; Mitchell, lated to this lack of standardized monitoring proce- Ellingson, Henry, & Fausti, 2004). The following is a dures. Use of various testing protocols, different oto- brief review of that body of work. First, the use of single, toxic change cri-teria, and different patient popu- high-frequency tone bursts to elicit ABRs will be ex- lations in ototoxic studies has hindered the transla- amined, followed by the use of multiple, high-fre- tion of research results into clinical practice. The fol- quency tone bursts chained together, and the use of lowing review focuses on ABR and OAE test-retest vari- single, high-frequency broadband clicks will be dis- ability in subjects not receiving ototoxic drugs. Such cussed. studies provide the basis for developing objective pro- Fausti and colleagues (1991) first established the tocols for monitoring ototoxic damage. Further research test-retest reliability of ABRs elicited by high- is needed to validate these results in large groups of frequency tone bursts in listeners with normal hearing subjects receiving ototoxic drugs. using high-frequency tone bursts (8, 10, 12, and 14 kHz) Despite the lack of standardized protocols, ABRs presented a rate of 11.1/s. Each ABR was obtained are currently being used to monitor auditory brainstem twice to ensure within-session reliability. These ABRs activity before, during, and after drug treatment and were compared to both behavioral thresholds and offer the possibility of detecting early changes in oto- standard click-evoked ABRs. No significant differences toxicity for very sick patients. Changes in the ABR were within an individual subject across session were found concomitant with diminishing hearing in neonatal pa- for waves I and V. Absolute mean differences ranged tients receiving aminoglycoside antibiotics (Bernard, from < 0.01 ms to 0.05 ms across frequencies. Pechere, & Hebert, 1980) and adult patients receiving Once the test-retest reliability of high-frequency the chemotherapeutic agent cisplatin (De Lauretis, De tone burst ABRs was established, Fausti and colleagues Capua, Barbieri, Bellussi, & Passali, 1999). ABR (1992) applied the method to 34 patients receiving oto- changes have been characterized by a significant pro- toxic drugs. Again, tone burst ABRs were recorded at longation of wave latency or the disappearance of a 8, 10, 12, 14 kHz in addition to a standard click-evoked wave that was present previously. ABR. Patients underwent serial monitoring with the Use of the ABR to detect hearing changes caused decision criteria for ABR change defined as (a) a 0.3 by ototoxicity in humans has been limited to standard ms latency shift for wave I or wave V or (b) a scoreable (square wave) click-evoked responses (De Lauretis et response becoming unscoreable. The researchers al., 1999), time-consuming derived-band responses showed that the high-frequency tone burst ABRs iden- (Coupland, Ponton, Egger-mont, Bowen, & Grant, 1991) tified 87.1% (27/31) of ears demonstrating a behav- or primarily recorded in infants at risk for ototoxic- ioral change in hearing sensitivity. Only 28.6% of these induced hearing loss (de Hoog et al., 2003). Standard ears were detected by standard click-evoked ABRs. The click-evoked ABRs provide information regarding the results clearly showed the sensitivity of ABRs in de- normalcy of the peripheral auditory system in response tecting ototoxic change. to a broad-spectrum signal, but provide little informa- In a similar study (Fausti et al., 1995), ABRs elic- tion regarding high-frequency hearing where the ear- ited by a standard click and by tone bursts of 8, 10, 12, liest ototoxic changes occur. In order to detect the ear- Hearing and Hearing Disorders: Research and Diagnostics 3 and 14 kHz were obtained in 20 listeners with high- trains. Responses recorded from the multiple-stimu- frequency hearing loss. Hearing loss was limited to lus train were compared to responses recorded from thresholds of < 25 dB HL from 0.25 - 1 kHz and < 70 dB the same stimuli presented as single tone bursts. Al- HL from 3 - 6 kHz. ABRs were obtained in two sepa- though responses to individual tone bursts presented rate test sessions for each listener. ABR thresholds were in the train showed some adaptation compared to re- obtained for each individual at each tone burst fre- sponses to tone bursts presented singly, the reliability quency