* Avoid giving gatifloxacin to patients.

* Consider selecting an antibiotic other than a fluoroquinolone for an elderly patient with diabetes mellitus (especially those taking sulfonylureas), hepatic insufficiency, or renal insufficiency (A).

* Discontinue fluoroquinolone therapy if a patient experiences symptoms of hypoor hyperglycemia, or if blood glucose levels fall below 60 mg/dL or rise above 200 mg/dL (C).

When was the last time you checked a glucose level before prescribing a fluoroquinolone? Though most side effects of these drugs are mild and self-limited (nausea, anorexia, vomiting, abdominal pain, diarrhea, taste disturbance, dizziness, headache, and somnolence), dysglycemia–hypo- or hyperglycemia–is another side effect that is potentially fatal.

From their inception, fluoroquinolones were known to upset glucose metabolism. However, recent publication of several case reports of gatifloxacin-associated dysglycemia, and Bristol-Myers Squibb’s announcement of a contraindication for gatifloxacin in diabetic patients, brought the matter of potentially severe dysglycemia to the forefront. (1)

This article reviews the available literature on fluoroquinolone-associated dysglycemia–first describing the frequency of dysglycemic events and then discussing the possible causes of these complications. Fluoroquinolones are valued agents in treating several infections (TABLE 1), (1-4) and this article makes clinical recommendations, based on published evidence, to help you decrease the risk of dysglycemia.

* Scope of the problem

The most recent article published on this subject identified 178 episodes of dysglycemia requiring hospitalization within 30 days of treatment–this following 16,697 outpatient courses of gatifloxacin, for a rate of 1.1%. Compared with gatifloxacin, rates were substantially lower for ciprofloxacin (0.3%), levofloxacin (0.3%), and moxifloxacin (0.2%) and for comparison antibiotics (those not associated with glucose metabolism problems), such as second-generation cephalosporins (0.2%) and macrolides (0.1%). (5)

However, an earlier analysis of inpatients receiving gatifloxacin, levofloxacin, or ciprofloxacin reported similar dysglycemia rates of 1.01%, 0.93%, and 0%, respectively. (6) Postmarketing surveillance studies further confirm these rates. (7)

The 2 North American governmental drug regulatory agencies report the following.

Health Canada: Twenty-eight cases of gatifloxacin-associated dysglycemia, with 2 deaths, over a 2-year period–89% of affected patients had preexisting diabetes mellitus and 67% of dysglycemia cases were hypoglycemic events. (8)

US Food and Drug Administration: Fifty-six times as many reports of dysglycemia filed for gatifloxacin than for other fluoroquinolones; its data revealed rates of hypoglycemia ranging from 0.65% to 2%. (9)

* Hypoglycemia: Exact cause uncertain

Sulfonylurea-like action. The most plausible mechanism is a sulfonylurea-like action on pancreatic beta cells, thus increasing insulin secretion. (10) Drugs such as quinine and mefloquine share chemical structures with fluoroquinolones and work in a similar manner to release insulin. (11,12) Individual fluoroquinolones differ greatly in their affinity for pancreatic beta cells. Gatifloxacin and temafloxacin have greater affinity, and thus greater hypoglycemic effect than other fluoroquinolones such as ciprofloxacin and levofloxacin. (12)

Drug-drug interaction. Hypoglycemia may also be caused by a drug-drug interaction. Glyburide levels have been reported to increase in patients with hepatic or renal impairment taking ciprofloxacin. (13) Gatifloxacin has been shown to augment the activity of several oral hypoglycemics, namely repaglinide, glyburide, pioglitazone, and glimepiride. (14) Similar effects have been documented between glyburide and levofloxacin, moxifloxacin, and ciprofloxacin. (10)

P450 isoenzyme interaction may also be a factor in fluoroquinolone-associated dysglycemia. (16) This explanation is the least plausible given that, although hypoglycemia may be seen with all fluoroquinolones, ciprofloxacin is the only one with any appreciable hepatic P450 metabolism.

Frequency of hypoglycemic events

Clinical studies regarding the effect of fluoroquinolones on glucose homeostasis are difficult to interpret because they have yielded conflicting results. Differences in patient demographics, concomitant medications, existing medical conditions, and even the definitions of hypo-and hyperglycemia used may account for the varying results. The following is a synopsis of the available literature on fluoroquinolone-associated hypoglycemia. Only randomized controlled trials, case-control studies, and chart reviews are included.

Studies finding a hypoglycemic effect. The most comprehensive report on fluoroquinolone-associated dysglycemia screened the medical records of more than 1.4 million elderly patients from 2002 to 2004. During this case-control study, 788 patients were evaluated in an emergency department or admitted to a hospital for treatment of hypoglycemia within 30 days of receiving a fluoroquinolone, macrolide, or second-generation cephalosporin antibiotic. Ninety-two percent of patients suffering hypoglycemia were also being treated for diabetes. Gatifloxacin had the highest rate of hypoglycemic events, with an adjusted odds ratio of 4.3 vs a macrolide. Levofloxacin was also associated with a higher rate of hypoglycemia, with an adjusted odds ratio of 1.5. No increased risk was reported for moxifloxacin, ciprofloxacin, or second-generation cephalosporins.