Recently, the safety of some members of the fluoroquinolones has been questioned, leading to a need to differentiate between these compounds in terms of adverse event profiles. One major concern is the potential for cardiotoxicity. Analysis of the structure-side effect relationship has revealed that cardiotoxicity is dependant upon the substituent added to position 5 of the fluorine ring.

In this situation, an amino group possessed by sparfloxacin and a methyl group for grepafloxacin increase toxicity, while the hydrogen possessed by other major fluoroquinolones is less likely to cause cardiac QTc prolongation.

Evidence supporting this relationship between QTc interval and risk of arrhythmia and risk of death has been published showing that the risk of sudden cardiac death with a QTc interval of 440 msec is 1.0. However, this rises to 1.4 with a QTc of 500msec, and then jumps to 2.8 at a QTc of 640 msec.

The risk of prolonged QTc increases as higher doses of drug are used, as these produce higher AUC levels. Females are also at greater risk than males, as are the elderly. Ranking the fluoroquinolones in terms of their risk of prolonging QTc shows sparfloxacin to be worse than grepafloxacin, followed by gatifloxacin and moxifloxacin. Much less risk is associated with trovafloxacin and levofloxacin.

A number of drugs have been recognized as causing prolonged QT interval. These include anaesthetics, antiarrhytmics, antidepressants, antihistaminics, antipsychotics, cholinergic agents and antibiotics. Drugs most commonly associated with torsades de pointes include sotalol, cisapride, amiodarone, erythromycin, terfenadine, quinidine, clarithromycin, haloperidol, digoxin, procainamide and furosemide.

The incidence of torsades de pointes associated with levofloxacin is less than 1 per million. In contrast, the rate is 14.5 per million for sparfloxacin, 3.8 per million for grepafloxacin and 2 per 1.29 million for gatifloxacin. This demonstrates a significant difference among the individual fluoroquinolones.

It is possible to directly correlate the strength of the fluoroquinolone QTc prolongation effect with their potassium antagonist action. This reveals that sparfloxacin and grepafloxacin have a high affinity for potassium current antagonists, potentially causing ventricular arrhythmia.

Research has demonstrated that QTc prolongation occurs with sparfloxacin and possibly moxifloxacin but not levofloxacin. In addition, this effect with sparfloxacin and moxifloxacin is compounded when they are prescribed with other anti-arrhythmic drugs. While QTc prolongation is probably a fluoroquinolone class effect, there is a 10-15 fold difference in ability to cause this among the individual fluoroquinolones. If the patient does not have an inborn prolonged QT interval syndrome, or does not receive antiarrhythmics causing QTc prolongation, levofloxacin appears to have no clinically relevant cardiotoxicity.