A single genetic variant inherited through the maternal line can turn one of the most widely used antibiotics in newborn intensive care into a cause of permanent hearing loss, according to an integrative review published in Frontiers in Pediatrics. The mutation, known as m.1555A>G, sits in the mitochondrial gene MT-RNR1 and alters a piece of cellular machinery so that it resembles the bacterial target of aminoglycoside drugs such as gentamicin. When carriers receive these antibiotics, the drugs can accumulate in the mitochondria of cochlear hair cells and trigger sensorineural hearing loss that is irreversible and dose-independent. In some individuals, a single therapeutic dose is enough to cause harm.
The review, led by P. E. Edison of Singapore General Hospital with colleagues at the All India Institute of Medical Sciences and the University of Southampton, evaluated the evidence linking m.1555A>G to aminoglycoside-induced hearing loss in neonates and examined whether pharmacogenetic prescribing could prevent it. The authors analysed the literature using the Wilson-Jungner framework of population-based preventive screening, asking whether the condition is sufficiently common, whether a reliable test exists, and whether an actionable intervention follows a positive result. Eighteen primary studies met their eligibility criteria, drawn from searches of Cochrane Library, PubMed, MEDLINE, Embase and ClinicalKey covering 2000 to August 2025.
The evidence base was entirely observational and methodologically heterogeneous, spanning cohort, cross-sectional, case-control and feasibility designs. Across population-based and epidemiological newborn cohorts, the study-level median frequency of m.1555A>G was 0.2 percent, with a range from 0 to 0.3 percent, corresponding to roughly one carrier in every 500 newborns in the populations studied. Regional and ethnic variation was pronounced. Small studies from South America reported no carriers at all, while several large population-based studies from China consistently reported carrier frequencies near 0.2 percent with narrow confidence intervals. Higher frequencies, up to 0.7 percent, appeared in selected high-risk cohorts such as low-birthweight infants in neonatal intensive care, though those estimates carried wider uncertainty because of smaller sample sizes.
The mechanistic story begins with the structure of the mutant ribosome. The 12S rRNA encoded by MT-RNR1, when mutated, structurally resembles the bacterial 16S ribosomal subunit, which is the binding site of aminoglycosides. Aminoglycosides kill bacteria by inducing ribosomal mistranslation and abnormal protein synthesis, and in carriers the same binding can occur in human cochlear mitochondria, disrupting protein synthesis and causing cytotoxic injury. Hair cells are especially vulnerable because of their high mitochondrial density. In preterm infants, the authors describe a double-hit model: loud neonatal intensive care unit noise overstimulates basal cochlear hair cells, opening mechanoelectrical transduction channels that increase intracellular accumulation of the drug, which then binds the mutant mitochondrial ribosomes and drives apoptosis of hair cells.
Clinical expression of the genotype is far from uniform. The review notes that m.1555A>G is not independently causative of hearing loss; the phenotype is modulated by environmental triggers such as aminoglycosides and noise, by heteroplasmy levels, by mitochondrial haplogroups and by nuclear modifier genes. Hearing loss can manifest as an immediate idiosyncratic reaction or as delayed-onset, bilateral, symmetric high-frequency impairment that is sometimes progressive. This delayed presentation matters for screening policy, because standard newborn hearing screens may miss it. AIHL primarily affects high frequencies, and conventional screening tools based on otoacoustic emissions and automated auditory brainstem response are designed for a frequency range of 1 to 4 kHz, so early stages of injury are likely underestimated.
The strongest quantitative signal in the review came from a multicentre prospective cohort study within the German Neonatal Network, which tested paired mother-infant buccal samples or umbilical cord tissue from 7,056 very-low-birthweight infants across 46 NICUs. Aminoglycoside treatment in m.1555A>G carriers was significantly but weakly associated with a failed newborn hearing screen, with an odds ratio of 1.26 (95 percent confidence interval 1.07 to 1.49, p = 0.0058). Low birth weight itself was the most important predictor of a failed screen. Maternal genotyping succeeded in 97.7 percent of samples and agreed completely with neonatal results. Three other studies found no association between aminoglycoside exposure and hearing screen outcome, underscoring how inconsistent the observational evidence remains.
A case-control study of 57 ex-very preterm infants with hearing loss and 180 matched controls illustrated the diagnostic gap. One control infant carried m.1555A>G, had been exposed to gentamicin during the neonatal period and passed the newborn hearing screen; hearing screening at two years of age was normal, yet pure-tone hearing impairment was detected at three-year follow-up. That same study found hearing loss independently associated with physiological instability during the neonatal period, at an odds ratio of 1.15 per day, and with cumulative exposure to ototoxic medications including gentamicin, vancomycin and furosemide, at an odds ratio of 1.23 per day of exposure.
On the testing side, the review documents rapid progress in genotyping technology. Early studies relied on low-throughput polymerase chain reaction and Sanger sequencing, later work used multiplex assays and microarray platforms, and recent studies applied high-throughput next-generation sequencing. The Pharmacogenetics to Avoid the Loss of Hearing trial, known as PALOH, demonstrated the analytical validity of a rapid point-of-care test using buccal samples containing a minimum of 16 cells; the assay achieved 100 percent sensitivity and 100 percent specificity with results available within 26 minutes, though recruitment challenges limited the study largely to a single centre. Pilot data also support the analytical validity of prenatal screening using buccal swabs or fetal cell-free DNA from maternal blood, with both methods showing complete agreement between maternal and neonatal genotypes. In one large Chinese NIPT study of 76,842 maternal samples, the positive predictive value for m.1555A>G was 90.8 percent.
Large-scale implementation experience comes mainly from China, where integrated programmes paired universal auditory screening with genetic screening for mutations in GJB2, GJB3, SLC26A4 and MT-RNR1. These programmes increased diagnostic potential for congenital hearing loss, shortened the time to etiological diagnosis, and enabled risk-mitigation measures such as antibiotic alert cards, genetic counselling and audiological surveillance for carriers. One programme reported a cost-benefit ratio of 1:7.27, though the authors caution that applicability to other healthcare systems is uncertain. A persistent weakness is follow-up attrition: long-term follow-up through two years of age was achieved in fewer than half of identified carriers, which matters because delayed-onset hearing loss is central to the rationale for screening and may go undetected without longitudinal data.
The review concludes that aminoglycoside-induced hearing loss fulfills key criteria for preventive screening: there is a latent preclinical period, genotyping tools are available, and aminoglycoside avoidance is an actionable intervention for carriers. The Clinical Pharmacogenetics Implementation Consortium recommends avoiding aminoglycosides in individuals with an MT-RNR1 mutation unless the severity of infection outweighs the risk and no safe alternatives exist. Yet the authors are explicit that clinical utility remains uncertain, since no study has directly compared universal and targeted screening approaches and long-term hearing outcomes are incompletely captured. They suggest genetic screening could be considered either as a universal component of newborn hearing screening programmes or as a targeted strategy informed by risk stratification based on family history, NICU admission and aminoglycoside predisposition, and they call for further research on clinical utility, cost-effectiveness and the ethical dimensions of neonatal genetic testing before policy can be settled.
Subject of Research: The association between the mitochondrial m.1555A>G mutation and aminoglycoside-induced hearing loss in neonates and the potential of pharmacogenetic screening
Article Title: Mitochondrial mutation m.1555A>G and aminoglycoside-induced hearing loss in neonates: an integrative review
Article References: Edison, P. E., Sankar, M. J., & Hall, M. (2026). Mitochondrial mutation m.1555A>G and aminoglycoside-induced hearing loss in neonates: an integrative review. Frontiers in Pediatrics, 14, Article 1973908. https://doi.org/10.3389/fped.2026.1973908
Image Credits: AI Generated
DOI: 10.3389/fped.2026.1973908
Keywords: aminoglycosides, m.1555A>G, MT-RNR1, hearing loss, neonates, newborn hearing screening, pharmacogenetics, mitochondrial DNA, gentamicin, point-of-care genotyping, NICU, ototoxicity
News Source: Harold Sullivan. (October 28, 2026). Common antibiotic can trigger permanent hearing loss in babies carrying one mitochondrial mutation. Scienmag.
