Pharmacotherapy 2002 May;22(5):663-8; discussion 668-72

Torsades de pointes associated with fluoroquinolones.

Owens RC Jr, Ambrose PG.

Department of Clinical Pharmacy, Maine Medical Center, Portland 04102, USA. owensr@mmc.org

Recent attention has been called to the interpretation of studies of antiinfective agents demonstrating effects on the QTc interval. It seems that the effects of many of these agents on the QTc interval are small, but in some patient populations, these drugs may cause morbidity and mortality related to TdP. It would be beneficial to researchers and clinicians alike for the FDA to standardize the types of studies designed to assess the QTc interval prolongation potential of a drug, methodologies, and interpretation criteria. To this end, it would increase the efficiency of the drug-approval process, give regulatory agencies and clinicians guidance, and increase patient safety. In summary we congratulate Dr. Frothingham for attempting to address the challenging issue of postmarketing safety surveillance. A critical review of his analysis of fluoroquinolone-associated TdP as well as other data on this potentially life-threatening adverse event support the following conclusions: Information from spontaneous reports is generally useful as an early warning system for excess adverse events, but reporting rates are not synonymous with incidence rates. The deficiencies of Dr. Frothingham's analysis lead to serious questions regarding the validity of both the numerators and denominators used in the incidence calculations (e.g., exclusion of European results, use ot extrapolated outpatient prescriptions, failure to account for inpatient versus outpatient utilization, failure to apply the appropriate statistical test to a rarely occurring, adverse event) and call into question conclusions about the relative risk of TdP with different fluoroquinolones. The association between which of the fluoroquinolones was administered to high-risk patients, which is important in the multiple-hit hypothesis, remains nebulous (e.g., failure to separate cases by route of drug administration and failure to identify which fluoroquinolones were given to patients with electrolyte abnormalities, concurrent QT interval-prolonging drugs, comorbid disease states). Preclinical and clinical trial data, as well as data from phase IV studies, indicate that levofloxacin, moxifloxacin, and gatifloxacin prolong the QTc interval, and the potential for TdP to develop as a result is rare and is influenced by many independent variables (e.g., concurrent drug administration of class Ia and III antiarrhythmic agents). These results should make clear that assessment of the cardiotoxicity of any new drug must take into account information (and its limitations) from several sources: preclinical studies that test effects on mechanisms underlying potential toxic reactions, controlled toxicodynamic studies in human volunteers safety results from controlled clinical trials, findings from phase IV studies, and postmarketing surveillance that includes spontaneously reported adverse events. One message that must not be lost in this discussion over the use of this reporting system to calculate incidences to incriminate certain agents is its overall importance, over time, in assisting governing bodies and clinicians alike in identifying compounds that may place certain patient populations at risk. It is imperative that clinicians not only submit adverse event reports to the FDA, but provide complete and accurate information. For moxifloxacin, levofloxacin, and gatifloxacin, the point must be clear that these agents should not be used in patients with risk factors predisposing them to TdP.