Article Text

Download PDFPDF

Aminoglycoside-induced deafness during treatment of acute leukaemia
  1. M Bitner-Glindzicz1,2,
  2. V Osei-Lah3,4,
  3. I Colvin5,
  4. T Sirimanna3,
  5. D Lucas5,
  6. B Mac Ardle5,
  7. D Webb6,
  8. A Shankar7,
  9. J Kingston6,7,
  10. L Jenkins8,
  11. S Rahman1,9
  1. 1Clinical and Molecular Genetics Unit, UCL Institute of Child Health and Great Ormond Street Hospital NHS Trust, London, UK
  2. 2UCL Ear Institute, London, UK
  3. 3Department of Audiological Medicine, Great Ormond Street Hospital NHS Trust, London, UK
  4. 4Department of Audiological Medicine, Queen Alexandra Hospital, Portsmouth, UK
  5. 5Nuffield Hearing and Speech Centre, Royal National Throat Nose and Ear Hospital, London, UK
  6. 6Department of Haematology and Oncology, Great Ormond Street Hospital NHS Trust, London, UK
  7. 7Department of Paediatric Haematology and Oncology, The Royal London Hospital, Whitechapel, London, UK
  8. 8North Thames (East) Regional Clinical Molecular Genetics Laboratory, Great Ormond Street Hospital NHS Trust, London, UK
  9. 9MRC Centre for Neuromuscular Diseases, National Hospital for Neurology and Neurosurgery, Queen Square, London, UK
  1. Correspondence to Maria Bitner-Glindzicz, Clinical and Molecular Genetics Unit, UCL Institute of Child Health and Great Ormond Street Hospital NHS Trust, 30 Guilford St, London WC1N 1EH, UK; mbitnerg{at}


Three unrelated children from ethnically diverse backgrounds who were treated for acute leukaemia became profoundly and irreversibly deaf during treatment. Aminoglycoside levels were within the therapeutic range. Genetic testing showed all three to have a maternally inherited mutation of mitochondrial DNA, m.1555A>G, known to cause sensitivity to the ototoxic effects of aminoglycosides. One child has received a cochlear implant, and another will be implanted shortly. Children diagnosed with acute leukaemia should be tested for this mutation at diagnosis, and alternative antibiotics chosen for the treatment of sepsis. Consideration should be given to elective testing of other groups of patients likely to receive aminoglycosides.

View Full Text

Statistics from

Aminoglycosides have been widely used since the 1950s for the treatment of Gram negative sepsis in a variety of settings. They are commonly used to treat life-threatening infections, particularly in children, where they are used for the treatment of neonatal septicaemia, sepsis in children who are immunocompromised (paediatric leukaemias and immunodeficiencies), chronic lung infections especially in children with cystic fibrosis, and treatment of Gram negative endocarditis. They are also used electively in those with drug-resistant tuberculosis, beta-lactam allergic patients and for prophylaxis in children undergoing urological surgery. Their side-effects are well-documented and include ototoxicity, which is commonly hearing loss and/or vestibulotoxicity, and nephrotoxicity which can lead to acute renal failure. Adverse effects are minimised by maintaining aminoglycoside levels within the therapeutic range.

However, some genetically susceptible individuals can become deaf following short courses of aminoglycosides where levels remain within the therapeutic range. Such individuals have become profoundly and permanently deaf, sometimes following a single dose.1 We describe three children, all treated with aminoglycosides for septic episodes occurring during their treatment for acute leukaemia. We showed that all three carried a mitochondrial DNA mutation, m.1555A>G, a maternally inherited mutation of a ribosomal RNA gene, known to be associated with ototoxicity. All experienced rapid, profound hearing loss, treated in two cases by cochlear implantation.

Case report

Case 1

A 2-year-old Caucasian girl was diagnosed with standard risk precursor B cell acute lymphoblastic leukaemia (B ALL). She had no central nervous system disease and was treated on the UKALL 97 (revised 99) protocol. At the time of diagnosis of leukaemia, there was no history of hearing impairment or language delay. During her treatment she had five hospital admissions for neutropaenic sepsis and was treated with intravenous gentamicin and kanamycin with levels within recommended limits. Eleven months after diagnosis it was noted that she had virtually complete loss of spoken language and was not responding to sound. Audiological assessment confirmed a bilateral profound hearing loss (fig 1b,c). The child, mother and all siblings were positive for the m.1555A>G mutation.

Figure 1

(a) Family tree of case 1. A maternal great grandmother wore a hearing aid for hearing loss attributed to congenital infection, but the child’s mother and three of her half siblings have normal pure tone audiograms (the fourth has not been tested but reports no hearing difficulty at the age of 19 years). The child, her mother and her four maternal half-siblings are positive for the m.1555A.G mutation. (b) Pure tone audiogram of case 1, left ear and (c) right ear. (d) Family tree of case 2. In the family history there was a maternal uncle with hearing loss of unknown cause. Two maternal aunts and a maternal uncle have no history of hearing loss. (e) Pure tone audiogram of case 2, left ear and (f) right ear. (g) Family tree of case 3. Three siblings have normal hearing and the paternal grandmother experienced hearing loss in middle age, the cause of which was unknown. (h) Pure tone audiogram of case 3, left ear and (i) right ear.

Her leukaemia is in remission and following completion of her treatment she underwent cochlear implantation.

Case 2

A 5-year-old boy of British/Caribbean-British descent was diagnosed with precursor B ALL. He was treated on the MRC UKALL 97/99 protocol. He received several courses of aminoglycosides for neutropaenic sepsis and was referred to paediatric audiology services. There had been parental concerns about his hearing, which dated back to the age of 3, although assessment by the health visitor at that time gave no cause for concern. His pure tone audiogram at the age of 7 is shown in fig 1e,f. The child and his mother were positive for the m.1555A>G mutation.

He is in remission from his leukaemia.

Case 3

The third case is a 4-year-old boy of Asian descent, born to consanguineous parents, diagnosed with standard risk pre-B ALL. He was treated on the MRC UKALL 97/99 protocol. He was referred to the paediatric audiology clinic following a short history of progressive hearing loss. He had been admitted to hospital on several occasions for febrile neutropaenia and had received multiple courses of aminoglycosides. His hearing was documented to have deteriorated rapidly over a 5-week period (fig 1h,i). Genetic evaluation showed that the patient had inherited the m.1555A>G mitochondrial mutation from his mother. His father did not have this mutation.

He is in remission from his leukaemia and will receive a cochlear implant within the next few months.


The observation of three children with a similar history of deafness following treatment for acute leukaemia prompted us to determine the prevalence of the m.1555A.G mutation in the UK population. We have since shown, in a large birth cohort, that approximately 1 in 500 UK Caucasian children have this mutation and are at risk of lifelong deafness following aminoglycoside exposure.2 This is in keeping with the observation of three patients with leukaemia carrying m.1555A.Gina 2-year period (500 children are diagnosed with acute leukaemia every year within the UK and neutropaenic sepsis, a frequent complication of the therapy, is often treated with aminoglycosides). It is likely that other children with acute leukaemia have been treated with aminoglycosides and become deaf, but the cause may have been attributed to other factors, possibly toxic levels of aminoglycosides. None of the children described here had a significant family history of deafness, and from our studies, it appears that the mutation has very low penetrance in the absence of aminoglycosides.2 Screening of the ALSPAC birth cohort showed that at the ages of 7 and 9 years all of the children with the mutation had hearing levels within the clinically normal range. These children had not been treated with aminoglycosides to our knowledge.

The frequency of m.1555A .G, together with the normal hearing in children of the ALSPAC cohort who have the mutation, raises the question of whether it is truly pathogenic. The mutation occurs in the aminoacyl tRNA acceptor site of the small ribosomal subunit (A site) and alters the conformation of the encoded mitochondrial ribosome which facilitates aminoglycoside binding. Other mutations, including m.1494C.T, which also results in mutation at the A site, can cause aminoglycoside-induced deafness, but m.1555A.G is by far the most commonly reported.3 Recently, using artificially engineered hybrid ribosomes, it has been confirmed that the accuracy of correct amino acid incorporation into synthetic polypeptides is reduced sevenfold in the presence of m.1555A. G and was further reduced in the presence of aminoglycosides.4 Moreover, reduced fidelity of translation in mutant ribosomes was evident at 25-fold lower concentrations of aminoglycoside compared to that observed in non-mutant ribosomes, which resulted in reduced biological activity of the proteins assayed. As mitochondrial protein synthesis is known to be important for the oxidative phosphorylation (OXPHOS) apparatus in cells with high energy demands, such as those of the cochlea, it is likely that there is a subtle OXPHOS defect in these patients which is unmasked by aminoglycosides. Indeed, cells from patients with this mutation have reduced growth rates and reduced basal mitochondrial protein synthesis leading to reduced OXPHOS especially in the presence of aminoglycosides.

The children reported here are likely to be cured of their leukaemia but will now be deafened for life as a result of a preventable cause of hearing loss. One was deafened prelingually and although she received a cochlear implant, she has not developed spoken language and prefers to use sign language. Both of the other two suffered progression of hearing loss and one is due to be implanted within the next few months because he does not obtain satisfactory audition and speech perception using conventional amplification. Cochlear implantation is a significant undertaking in both emotional and educational as well as financial terms for families and health and education services.5 6 Children who have received cochlear implants remain deaf and can be at a significant disadvantage in a variety of developmental domains compared to normally hearing children.

The cost of testing all children with leukaemia for the mutation over 2 years needs to be balanced against the cost of cochlear implantation or lifelong deafness and is likely to be cost effective even based on current high costs.6 We suggest that genetic testing for inherited susceptibility to deafness caused by m.1555A.G should be performed when a diagnosis of leukaemia is made, so that an alternative antibiotic can be used in those who have the mutation. Consideration should also be given to elective testing of other patient groups who are likely to receive courses of aminoglycosides, such as those with severe immunodeficiencies, cystic fibrosis, those undergoing bone marrow transplantation, and cancer patients of any age in whom aminoglycosides may be used as first-line treatment for neutropaenic sepsis. All of these patient groups may also be at risk of hearing loss if they carry the m.1555A.G mutation. Special consideration needs to be given to neonates, especially those who are born prematurely, since they cannot be identified in advance of needing treatment for suspected sepsis, and no rapid bedside testing is currently available. However, the case for widespread population screening of pregnant women would require more data on penetrance of the mutation in the presence and absence of aminoglycosides and on false negatives and positives associated with testing.


Genetic testing for the m.1555A.G mutation should be performed at the onset of treatment for acute leukaemia, in order to prevent irreversible, profound hearing loss. Other patients who are also likely to receive aminoglycosides should be tested electively. However, the largest group receiving aminoglycosides, sick and premature neonates, pose a particular problem because testing cannot yet be performed in an acute setting. Consideration should be given to the feasibility and utility of genetic testing for m.1555A.G in pregnant women.


We would like to acknowledge the families for their consent to publish this report.


View Abstract


  • Funding MB-G and SR have received funding from SPARKS.

  • Competing interests MB-G and SR have been awarded funding from SPARKS to determine the prevalence of m.1555A>G in the UK.

  • Patient consent Parental consent obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.