Objectives: During the past decade, the initial popularity of metal-on-metal (MoM) hip implants has shown a progressive decline due to increasingly reported implant failure and revision surgeries. Local as well as systemic toxic side effects have been associated with excessive metal ion release from implants, in which cobalt (Co) plays an important role. The rare condition of systemic cobaltism seems to manifest as a clinical syndrome with cardiac, endocrine, and neurological symptoms, including hearing loss, tinnitus, and imbalance. In most cases described in the literature, revision surgery and the subsequent drop in blood Co level led to (partial) alleviation of the symptoms, suggesting a causal relationship with Co exposure. Moreover, the ototoxic potential of Co has recently been demonstrated in animal experiments. Since its ototoxic potential in humans is merely based on anecdotal case reports, the current study aimed to prospectively and objectively examine the auditory and vestibular function in patients implanted with a MoM hip prosthesis.
Design: Twenty patients (15 males and 5 females, aged between 33 and 65 years) implanted with a primary MoM hip prosthesis were matched for age, gender, and noise exposure to 20 non-implanted control subjects. Each participant was subjected to an extensive auditory (conventional and high-frequency pure tone audiometry, transient evoked and distortion product otoacoustic emissions [TEOAEs and DPOAEs], auditory brainstem responses [ABR]) and vestibular test battery (cervical and ocular vestibular evoked myogenic potentials [cVEMPs and oVEMPs], rotatory test, caloric test, video head impulse test [vHIT]), supplemented with a blood sample collection to determine the plasma Co concentration.
Results: The median [interquartile range] plasma Co concentration was 1.40 [0.70, 6.30] µg/L in the MoM patient group and 0.19 [0.09, 0.34] µg/L in the control group. Within the auditory test battery, a clear trend was observed toward higher audiometric thresholds (11.2 to 16 kHz), lower DPOAE (between 4 and 8 kHz), and total TEOAE (1 to 4 kHz) amplitudes, and a higher interaural latency difference for wave V of the ABR in the patient versus control group (0.01 ≤ p < 0.05). Within the vestibular test battery, considerably longer cVEMP P1 latencies, higher oVEMP amplitudes (0.01 ≤ p < 0.05), and lower asymmetry ratio of the vHIT gain (p < 0.01) were found in the MoM patients. In the patient group, no suggestive association was observed between the plasma Co level and the auditory or vestibular outcome parameters.
Conclusions: The auditory results seem to reflect signs of Co-induced damage to the hearing function in the high frequencies. This corresponds to previous findings on drug-induced ototoxicity and the recent animal experiments with Co, which identified the basal cochlear outer hair cells as primary targets and indicated that the cellular mechanisms underlying the toxicity might be similar. The vestibular outcomes of the current study are inconclusive and require further elaboration, especially with respect to animal studies. The lack of a clear dose-response relationship may question the clinical relevance of our results, but recent findings in MoM hip implant patients have confirmed that this relationship can be complicated by many patient-specific factors.