Minimizing the toxicity of aminoglycosides in cystic fibrosis

Antibiotic therapy in CF - benefits versus harms

By the late teenage years over half of cystic fibrosis (CF) patients in the UK will have chronic pulmonary infection with Pseudomonas aeruginosa. ¹ Regular treatment of this infection with intravenous antibiotics, at three monthly intervals, has been associated with increased survival in historical cohort studies. ² However, this increased survival may be achieved at the cost of cumulative side-effects of the antibiotics used and the more frequent occurrence of severe adverse-effects, such as acute renal failure. This article discusses how antibiotic treatment may harm patients, how such harm may be detected at an early stage and how it may be prevented.

DECLARATIONS

Competing interests

None declared

Funding

None

Ethical approval

Not applicable

Guarantor

ARS

Contributorship

ARS is the sole contributor

Acute renal failure in CF

The first reports of acute renal failure in CF were in 1998 in patients who had received both an aminoglycoside and a non-steroidal anti-inflammatory drug at the same time. ³ There are now over 50 cases of acute renal failure in CF in the world literature, including 24 cases identified in a UK national survey. ⁴ Children with CF are more than 100 times more likely to develop acute renal failure than the general population. ⁴ As renal damage is not believed to be intrinsic to the natural history of CF, it is likely that this increased risk is related to treatment. Patients with CF who have received an aminoglycoside have an 80-fold increase in their risk of acute renal failure. ⁵ In many cases aminoglycoside use may be unavoidable, due to the antibiotic sensitivity pattern of P. aeruginosa in that patient. However, known risk factors for renal impairment such as CF-related diabetes and concurrent non-steroidal or ciclosporin use also increase the risk of renal impairment. ⁵ Gentamicin use over a one-year period increases the risk of renal failure, whereas tobramycin use does not. ⁵ Almost half of isolates of P. aeruginosa from UK CF patients are resistant to gentamicin, compared to only 10% resistant to tobramycin. ⁶

Renal impairment

Patients with CF may have asymptomatic renal impairment. Earlier detection of this impairment may allow dose modification of drugs excreted by the kidney or a change in antibiotic regimen to minimize further toxicity. Renal impairment (as measured by reduced creatinine clearance - a measure of glomerular filtration) is present in between 31-42% of stable adult patients and is related to cumulative aminoglycoside use. ⁷ This may be potentiated by administration of intravenous colistin, though colistin alone does not seem to have an effect. ⁷ Other workers have found that prior intravenous colistin exposure is associated with cumulative renal tubular damage, as measured by the biomarker N-acetyl-beta-D glucosaminidase (NAG). ⁸ In children, there are no recent data, but a study using 24-hour creatinine clearance to measure renal impairment has suggested that the prevalence may be as high as 39%. ⁹ Further studies are needed, using measures of glomerular filtration rate which are more appropriate for children, such as chromium 51-ethylene diamine tetraacetate (Chromium⁵¹ EDTA). Future research will evaluate biomarkers of renal impairment and injury. Cystatin C, correlates closely with GFR, is independent of age and muscle mass and may allow estimation of GFR without the need for multiple blood samples or a radio-isotope. ¹⁰ Similarly neutrophil gelatinase associated lipocalin may allow the early detection of acute renal failure. ¹¹

Strategies to minimize nephrotoxicity

Clearly careful monitoring of aminoglycoside levels during intravenous therapy is essential - not simply after commencing therapy but also in the second and subsequent weeks of treatment. ¹² Doctors who care for CF patients can reduce the risk of renal failure by close monitoring of patients with known risk factors for renal impairment (such as known renal tract abnormality concurrent use of other nephrotoxic drugs and acute dehydration). Tobramycin should be used in preference to gentamicin. There is good evidence, from randomized controlled trials, that once daily tobramycin is associated with less nephrotoxicity (as measured by serum creatinine) in children and less proximal tubular damage (as measured by NAG) in adults and children. ¹³ There is also intriguing evidence to suggest that administering a once daily dose of tobramycin in the morning may be associated with more rapid clearance and possibly less toxicity. ¹⁴ This requires further study before adoption into clinical practice can be recommended.

Ototoxicity with aminoglycoside use in CF

Aminoglycosides are toxic to the hair cells of the cochlea and can cause irreversible sensorineural hearing loss. These drugs persist after systemic treatment and, in experimental animals, cumulative hair cell damage can be seen even when after an interval of 4 weeks between aminoglycoside courses. ¹⁵ This raises the possibility that ototoxicity could be reduced by increasing the interval between doses - for example, restricting aminoglycosides to alternate courses of antibiotics in patients who require frequent treatment. A study of 70 stable patients with CF identified sensorineural hearing loss, in keeping with cumulative aminoglycoside use, in 12 patients, including one child. ¹⁶ Those with hearing impairment had received a median of 20 courses of aminoglycoside treatment vs. a median of nine courses in the group with normal hearing. The per course risk of hearing impairment was estimated to be less than 2%. In a randomized trial of once vs. three times daily tobramycin, there was no change in pure tone audiogram from baseline compared to the end of treatment or follow-up and no difference between regimens. ¹⁷ There was also no difference in high frequency audiogram, suggesting the use of this technique may not allow early detection of hearing loss. ¹⁷ Vestibular toxicity has also been reported with aminoglycoside use in CF. ¹⁸

Genetic risk factors for ototoxicity

Individuals who have the mitochondrial DNA mutation m1555A>G are at risk of severe sensorineural hearing loss after a single dose of aminoglyocoside. ¹⁹ The prevalence of this mutation in the UK is around one in 500 ²⁰ and so many CF clinics will have no affected individuals. The mutation has, however, been reported in a CF patient with hearing loss. ²¹ Screening for this mutation in CF clinics has not so far been shown to be cost-effective. ²² However, this may change if other mitochondrial DNA mutations associated with hearing loss are identified and further research is needed.

Conclusions

CF patients receive frequent and prolonged courses of antibiotics, often at high doses. There is no comparable medical condition where so many individuals are exposed to the risk of cumulative antibiotic toxicity in this way. The CF team need a tool kit of strategies to allow safer antibiotic use in their patients. Some of the tools are available already. For example, clinicians should avoid the use of intravenous gentamicin in CF patients. Tobramycin should be used in preference and dosed once daily. Giving the once daily dose in the morning may enhance clearance and minimize toxicity, but further research is needed before a firm recommendation can be made. Some tools may be added to the tool kit in the next few years. Biomarkers of renal impairment or renal injury will be a valuable early warning of renal damage, if adequately qualified and validated for patients with CF. Further work is needed to determine whether limiting aminoglycosides to alternate courses will reduce ototoxicity. Further genetic risk factors for susceptibility to aminoglycoside-induced hearing impairment should be identified before screening is likely to be practical. Careful use of such tools will allow this valuable family of antibiotics to be used as effectively and safely as possible in CF patients.