Although there have been case reports describing statin toxicity (rhabdomyolysis; acute kidney injury; hyperkalaemia) following co-prescription of clarithromycin or erythromycin (CYP3A4 inhibitors), no clinical or epidemiological studies have so far quantified the risk of serious toxicity from this drug-drug interaction. This was the subject of the current study, which sought to characterise the risk in a population of older adults in Canada.
The population-based retrospective study used linked healthcare databases in Ontario to establish a cohort of older adults with ongoing continuous prescriptions for statins metabolised by CYP3A4 (atorvastatin [most common; 73%]; simvastatin; lovastatin). New co-prescriptions of clarithromycin (n=72,591) or erythromycin (n=3,267) were then identified, and compared against a reference group of new users of azithromycin (n=68,478; similar indications and clinical use patterns, but does not inhibit CYP3A4). The index date was the first date of co-prescription of statin and study antibiotic. Patients receiving other potent CYP3A4 inhibitors within 30 days of the index date were excluded, as were those with end-stage renal disease prior to the index date, and those receiving any other antibiotics. The primary outcome was hospitalisation with rhabdomyolysis within 30 days following the index date.
Compared with azithromycin, co-prescription of a CYP3A4-metabolised statin with clarithromycin or erythromycin was associated with a higher risk for hospitalization with rhabdomyolysis (relative risk [RR] 2.17; 95% CI 1.04 to 4.53), acute kidney injury (RR 1.78; 95% CI 1.49 to 2.14) and 30-day overall mortality (RR 1.56; 95% CI 1.36 to 1.80).
Overall rates of these events were low and so the absolute increases were small (0.02% [0.01% to 0.03%] for rhabdomyolysis and 0.25% [0.17% to 0.33%] for 30-day mortality). Corresponding numbers needed to harm (NNH) were 5,870 (3068-67,758) and 399 (304-577). There was no statistically significant increase in risk for hospitalisation due to hyperkalaemia.
The above analyses were based on diagnostic codes. As the sensitivity of these is limited, the authors also examined a subpopulation for which laboratory values were available. In these patients, co-prescription of clarithromycin or erythromycin was associated with a higher risk of hospitalisation with acute kidney injury (RR 2.92; 1.47-5.79; absolute increase of 1.26%; NNH 79) and hyperkalaemia (RR 11.04; 1.48-82.58; absolute increase of 0.74%; NNH 135). Only six patients had evidence of hospitalisation with raised creatinine kinase and so meaningful comparisons were not possible.
The authors conclude that this drug-drug interaction is clinically important, given the high rate of co-prescription seen in their study, and that the combination should be avoided when possible. As database codes are insensitive to the diagnosis of rhabdomyolysis, they say that the absolute risk reported here is an underestimation. Further limitations include those inherent to observational studies, with retrospective data collection. Although the observed associations may not be causal, they are consistent with previous warnings about this interaction based on pharmacokinetic data and case reports.