Comprehensive Audiologic Analyses After Cisplatin-Based Chemotherapy

Abstract

Importance: Cisplatin is highly ototoxic but widely used. Evidence is lacking regarding cisplatin-related hearing loss (CRHL) in adult-onset cancer survivors with comprehensive audiologic assessments (eg, Words-in-Noise [WIN] tests, full-spectrum audiometry, and additional otologic measures), as well as the progression of CRHL considering comorbidities, modifiable factors associated with risk, and cumulative cisplatin dose.

Objective: To assess CRHL with comprehensive audiologic assessments, including the WIN, evaluate the longitudinal progression of CRHL, and identify factors associated with risk.

Design, setting, and participants: The Platinum Study is a longitudinal study of cisplatin-treated testicular cancer survivors (TCS) enrolled from 2012 to 2018 with follow-up ongoing. Longitudinal comprehensive audiologic assessments at Indiana University and Memorial Sloan Kettering Cancer Center included 100 participants without audiometrically defined profound hearing loss (HL) at baseline and at least 3.5 years from their first audiologic assessment. Data were analyzed from December 2013 to December 2022.

Exposures: Factors associated with risk included cumulative cisplatin dose, hypertension, hypercholesterolemia, diabetes, tobacco use, physical inactivity, body mass index, family history of HL, cognitive dysfunction, psychosocial symptoms, and tinnitus.

Main outcomes and measures: Main outcomes were audiometrically measured HL defined as combined-ears high-frequency pure-tone average (4-12 kHz) and speech-recognition in noise performance measured with WIN. Multivariable analyses evaluated factors associated with risk for WIN scores and progression of audiometrically defined HL.

Results: Median (range) age of 100 participants at evaluation was 48 (25-67) years; median (range) time since chemotherapy: 14 (4-31) years. At follow-up, 78 (78%) TCS had audiometrically defined HL; those self-reporting HL had 2-fold worse hearing than TCS without self-reported HL (48 vs 24 dB HL; P < .001). A total of 54 (54%) patients with self-reported HL showed clinically significant functional impairment on WIN testing. Poorer WIN performance was associated with hypercholesterolemia (β = 0.88; 95% CI, 0.08 to 1.69; P = .03), lower-education (F1 = 5.95; P = .004), and severity of audiometrically defined HL (β̂ = 0.07; 95% CI, 0.06 to 0.09; P < .001). CRHL progression was associated with hypercholesterolemia (β̂ = -4.38; 95% CI, -7.42 to -1.34; P = .01) and increasing age (β̂ = 0.33; 95% CI, 0.15 to 0.50; P < .001). Importantly, relative to age-matched male normative data, audiometrically defined CRHL progression significantly interacted with cumulative cisplatin dose (F1 = 5.98; P = .02); patients given 300 mg/m2 or less experienced significantly less progression, whereas greater temporal progression followed doses greater than 300 mg/m2.

Conclusions and relevance: Follow-up of cisplatin-treated cancer survivors should include strict hypercholesterolemia control and regular audiological assessments. Risk stratification through validated instruments should include querying hearing concerns. CRHL progression relative to age-matched norms is likely associated with cumulative cisplatin dose; investigation over longer follow-up is warranted.

Figure 1.. Speech-in-Noise Perception Measured With the Words in Noise Test (WIN) and Displayed by Self-Reported Hearing Loss (HL)

Blue designates patients with self-reported HL, while orange denotes those with no self-reported HL (see the eAppendix in Supplement 1 for detail). Mean performance (percentage correct) for each signal-to-noise ratio (24 to 0 dB SNR) is shown. The performance difference evaluated at the 50% threshold was statistically significant (P < .001) with those self-reporting HL having poorer speech-in-noise perception.

Figure 2.. Audiologic Performance Stratified by Self-Reported Hearing Loss (HL)

A, Blue designates patients with self-reported HL, while orange denotes those with no self-reported HL (see the eAppendix in Supplement 1 for detail). Change in hearing by age and time for each patient (displayed by self-reported vs no self-reported HL). B, Longitudinal audiograms for TPS-1 (initial assessment) and TPS-2 (follow-up assessment) for those with and without self-reported HL and relative to age- and sex-matched normative hearing (dashed lines). Both frames of B show the median audiometric hearing threshold (dB HL) across the tested frequency range (250-12 000 Hz). Dashed lines in each panel show the age- and sex-matched normative median results at both time points by frequency. TPS-1 indicates The Platinum Study-1; TPS-2, The Platinum Study-2; PTA, Pure Tone Average; dB HL, decibel reference to Hearing Level.