Macrolides and TMP/SMZ aren’t the only antibiotics with potentially-harmful side effects. Next up: Fluoroquinolones.
This is the third in a series of columns dealing with non-allergic, potentially serious reactions caused by antibiotics. In the column dealing with macrolides, the increased risk of sudden death was noted. Clearly, the absolute numbers of deaths associated with macrolides has to be small, but it makes sense to take this fact into consideration when prescribing these antibiotics to those prone to arrhythmias (such as patients with congestive heart failure).
The fluoroquinolones have also been associated with an increased risk of sudden death. The following study used three different observational study designs comparing patients with ventricular dysrhythmias or cardiac arrest vs control patients and two of the three methods found a statistically significant association between recent fluoroquinolone use and these dysrhythmias.
EFFECT OF MACROLIDE AND FLUOROQUINOLONE ANTIBACTERIALS ON THE RISK OF VENTRICULAR ARRHYTHMIA AND CARDIAC ARREST
Zambon, A., et al, Drug Safety 32(2):159, February 2009
BACKGROUND: It has been reported that the effects of several non- cardiovascular drugs may trigger malignant arrhythmias.
METHODS: These Italian authors used three different observational study designs to evaluate the possible association between use of macrolide or fluoroquinolone antibiotics and the risk of ventricular arrhythmias and cardiac arrest. The analyses were performed in 1,275 adult case patients included in a provincial database who experienced a ventricular dysrhythmia or cardiac arrest, and 9,189 control patients matched for age, gender and general practitioner.
RESULTS: In a case-control analysis that adjusted for between-group differences in exposure to drugs or illnesses that possibly cause arrhythmias, adjusted odds ratios (ORs) for ventricular dysrhythmias or cardiac arrest were 2.13 with recent macrolide use and 3.58 with recent fluoroquinolone use (95% confidence intervals were above 1.0 for both of these comparisons). In a case-crossover design, in which case patients served as their own controls, adjusted ORs were 1.70 (95% CI crossed 1.0) and 1.98 (95% CI above 1.0) for recent macrolide and fluoroquinolone use, respectively. In a case-time-control analysis that controlled for exposure time trends, adjusted ORs were 1.62 with recent macrolide use and 1.59 with recent fluoroquinolone use (95% CIs crossed 1.0 for both comparisons).
- no reprints) 9/09 - #27
The next study points out the importance of assessing the multiple concomitant risk factors that patients may have (particularly the use of drugs that prolong the QT interval on the ECG) that may help precipitate ventricular dysrhythmias when a fluoroquinolone is prescribed.
TORSADES DE POINTES ASSOCIATED WITH FLUOROQUINOLONES: IMPORTANCE OF CONCOMITANT RISK FACTORS
Amankwa, K., et al, Clin Pharmacol Ther 75:242, March 2004
Increasing use of fluoroquinolone antibiotics has prompted concern about their proarrhythmic effects. Virtually all published cases of torsades de pointes attributed to fluoroquinolone antibiotics have occurred in patients with other risk factors for this arrhythmia. This report, from Wayne State University, describes a case of progressive QTc prolongation and torsades de pointes following initiation of treatment with IV levofloxacin in a woman with other risk factors. Reported risk factors for drug-induced torsades de pointes include a QTc interval exceeding 500ms or a prolongation of more than 60ms compared with pretreatment values, female sex, left ventricular dysfunction, hypokalemia, hypomagnesemia, advancing age, bradycardia, elevated plasma levels due to drug interactions or failure to adequately adjust doses for organ dysfunction, rapid IV infusion of torsades-inducing drugs, concomitant administration of more than one agent known to cause QTc prolongation or torsades de pointes, and possible genetic predisposition. In one study, nearly 10% of outpatients filling prescriptions for a drug that can prolong the QTc interval also filled prescriptions for at least one other drug than can be associated with QTc prolongation or inhibit clearance of the former agent. It has also been noted that at least one concomitant risk factor for torsades de pointes was present in 96% of published cases attributed to noncardiac drugs. The authors caution clinicians to consider risk factors for QTc prolongation and torsades de pointes when prescribing fluoroquinolone antibiotics or other drugs that can cause this complication, and to avoid such drugs in patients with multiple or noncorrectable risk factors. 29 references 8/04 - #32
The next study looks at the psychiatric and neurologic side effects of the quinolones and suggests that the incidence is 1-2%. Given the large number of prescriptions given for these antibiotics, this could represent a large number of people. Most physicians would not suspect an antibiotic could cause psychiatric or neurologic effects and may not consider them to be the cause if patients mention them. In addition, patients may not even consider some of these effects being caused by their antibiotic prescription and not even mention them to their physician. There are multiple reports on the internet of neuropsychiatric side-effects from the quinolones. The current package insert specifically advises, under the Warnings and Precautions section:
“Central nervous system effects, including convulsions, anxiety, confusion, depression and insomnia may occur after the first dose. Use with caution in patients with known or suspected disorders that may predispose them to seizures or lower the seizure threshold.”
“Peripheral neuropathy: discontinue if symptoms occur in order to prevent irreversibility.”
QUINOLONES: REVIEW OF PSYCHIATRIC AND NEUROLOGIC ADVERSE REACTIONS
Tome, A.M., et al, Drug Safety 34(6):465, June 2011
BACKGROUND: It has been reported that CNS events occur in 1-2% of patients taking quinolone antibiotics.
METHODS: These Portuguese authors reviewed 83 published articles (case reports or case series) describing 145 case reports involving 206 neurologic and/or psychiatric adverse drug events attributed to quinolone antibiotics.
RESULTS: Nearly half of the reports involved ciprofloxacin and nearly one-fourth involved use of ofloxacin. Adverse CNS events were considerably less common with levofloxacin. The authors suggest that widespread use of cirpofloxacin worldwide might account for overrepresentation of this agent in adverse event reporting. Just over half of the events (53.9%) involved psychiatric disorders, most commonly including mania, insomnia, acute psychosis, and delirium. Neurological events accounted for the remainder of the adverse CNS occurrences, most commonly including seizures, confusion, and myoclonus. Adverse CNS events generally developed within minutes to days after starting treatment. Some events were precipitated by concomitant treatment with a quinolone and theophylline, nonsteroidal antiinflammatory drugs or, less commonly, other agents (e.g., antivirals, antihistamines, antidepressants, and opioid analgesics). Because theophylline concentrations are increased, and clearance decreased, in the setting of quinolone treatment, a reduced theophylline dose is advised in this circumstance. Other factors that might contribute to the development of adverse CNS events with quinolone treatment include a history of seizures, electrolyte disturbances, renal and/or hepatic failure, older age, and excessive dosing. Dose adjustment is advised in patients with comorbidity.
- no reprints) (21585220 [PMID]) Copyright 2012 by Emergency Medical Abstracts 1/12 - #
There are also the known issues regarding tendinopathies. In July of 2008 a “black box” warning was put in the package insert of the quinolones. This side-effect, like the cardiovascular and neuropsychiatric effects, is also one that would not be intuited to be caused by an antibiotic and only on post-marketing surveillance would such effects likely be found.
Here is a summary of information from a variety of studies and reports regarding this problem:
- The risk of rupture is age related, one study found an odds ratio of 6.4 for patients aged 60-79 and a 20.4 ratio for older patients
- The risk is both related to the dose and specific quinolone prescribed. One study found the highest odds ratio (28.4) for ofloxacin and the lowest for ciprofloxacin (3.6)
- The Achilles tendon seems to be the most vulnerable with it being involved in 90% of cases in one study (rupture occurred in about 40%)
The effect on tendons seems to be long lasting. One study found the odds ratio for Achilles tendon rupture to be 4.3 with current use of a quinolone, 2.4 with exposure within six months and 1.4 for more remote exposure
- There appear to be multiple predisposing factors – steroid use (involved in about a third of cases in one report), renal dysfunction, rheumatic disease, diabetes, hyperparathyroidism, hypothyroidism and participation in sports
The first major tendinopathy lawsuit involved a 77-year-old patient given Levaquin and steroids for bronchitis. He ruptured both Achilles tendons within three months of being put on the drugs. In December of 2010 he won a $1.8 million award from Ortho-McNeil. The fundamental issue at the time was the perceived lack of adequate warning to prescribers by the manufacturer.
Multiple lawyer websites are currently seeking patients with tendinopathies. One report indicated that several thousand tendinopathy lawsuits have been filed. Pharmacists are now required to provide patients prescribed a quinolone with a “fact sheet” detailing the risk of tendinopathies with the specific content specified by the FDA.
The issue for clinicians currently is the fact that, since the initial lawsuit, manufacturers and the FDA have made substantial efforts to notify prescribers of this problem. So now the concern is that prescribers are at risk since there are multiple alternatives to the quinolones and they are “black boxed.” It would be difficult, in most cases, particularly on an outpatient basis, to claim that quinolones can generate an anticipated result that other antibiotics can’t.
Finally, it seems that all antibiotics (some more than others) have the potential to increase the effects of warfarin. Bowel flora are necessary to make vitamin K and that is essential in the liver’s manufacturing of selected coagulation factors (II, VII, IX and X in particular). One study demonstrated an almost 74% decrease in vitamin K production in people taking broad spectrum antibiotics. In addition, some antibiotics compete with warfarin for catabolism and can thus raise warfarin levels through this mechanism.
With regard to the quinolones in particular, one study found that levofloxacin raised the INR by 0.8 in patients on warfarin and that it pushed 39% above the therapeutic range. In the vast majority of patients this temporary rise and subsequent fall in INR levels is never appreciated clinically (no bleeding), yet one study noted a statistically significant increase in admissions for GI bleeding for patients on any of 10 antibiotics with a adjusted relative risk of 3-5 for doxycycline, amoxicillin (with or without clavulanic acid), ciprofloxacin, azithromycin and trimethoprim/sulfamethoxazole). The adjusted relative risk was 9 for tetracycline and 43 for cefradine and neomycin.
The bottom line – prescribers need to be aware of these less common, but potentially serious, side-effects of the quinolones and use them only when truly necessary and when potentially safer choices are not an option.