Reason for posting: Gatifloxacin (Tequin) is commonly used to treat respiratory infections, including community-acquired pneumonia, acute exacerbations of chronic bronchitis, sinusitis and urinary tract infections Health Canada has previously warned health care providers about cases of clinically significant hypoglycemia and hyperglycemia in patients taking the drug, usually those with diabetes.1 A recent case– control study confirmed the dysglycemic effects of the drug in patients with and without diabetes, and quantified its risks compared with other antibiotics.2 Health Canada has advised that the drug not be prescribed to patients with diabetes.3
The drug: Gatifloxacin is a third-generation, broad-spectrum fluoroquinolone with activity against gram-negative and gram-positive aerobic, anaerobic and atypical microorganisms. It undergoes minimal biotransformation and is excreted renally.
The recent case–control database study identified Ontario patients over 65 years of age seen in hospitals after they experienced either hypo-or hyperglycemia who had received an outpatient prescription for an oral fluoroquinolone, a second-generation cephalosporin or a macrolide during the preceding month.2 Because dysglycemia can result from infections or be caused by a hospital stay in itself, patients who had been taking macrolides (which are used for similar indications but which do not affect glycemic control) were chosen to form the control group. Each case of dysglycemia was matched (by the patient's age, sex, presence or absence of diabetes, and timing of the adverse event with respect to the initiation of the antibiotics) with up to 5 controls.
Patients who experienced hypoglycemia were in excess of 4 times as likely to have been treated with gatifloxicin rather than a macrolide, when compared with their matched controls (Table 1). The risks of both diabetic and nondiabetic patients were increased by use of gatifloxacin. Of the 788 patients treated for hypoglycemia, fewer than half (366) were admitted to hospital. Of these, 30 patients (8.1%) died before discharge. The median time from initiation of gatifloxicin therapy to hospital presentation was 6 days.
Patients who had experienced hyperglycemia, on the other hand, were almost 17 times as likely to have been treated with gatifloxacin rather than a macrolide (Table 1). Of the 470 patients with hyperglycemia identified, about half (237) were admitted and 39 (16.5% of those admitted) died in hospital. The median time from initiation of therapy to arrival at hospital was 5 days.
Levofloxacin was also associated with a slightly increased risk of hypoglycemia (Table 1) but not hyperglycemia. In supplementary analyses, the frequency of dysglycemia occurring within 30 days after antibiotic prescription was found to be highest for gatifloxacin (1.1%), followed by ciprofloxacin (0.3%), levofloxacin (0.3%), moxifloxacin and second-generation cephalosporins (both 0.2%), and least for macrolides (0.1%).2 These frequencies are probably underestimates, since visits to a hospital or emergency department were included in the database only when hypoglycemia or hyperglycemia was established.
Although the mechanism of these opposing adverse effects is unclear, gatifloxacin may cause hypoglycemia by promoting insulin release, and hyperglycemia by causing vacuolation of pancreatic beta cells, which would reduce insulin levels.2
What to do: Until now, risk factors for gatifloxacin's life-threatening dysglycemic effects were felt to include a patient's advanced age (> 75 yr), having diabetes or decreased renal function, and taking hypoglycemic medications (i.e., drugs to lower blood sugar) concurrently. The study's findings clearly indicate that even people without diabetes are at risk. People with diabetes should not be prescribed gatifloxacin; alternative antibiotics are preferable (levofloxacin should be used with caution). When gatifloxacin is prescribed, remember to warn the patient of the signs and symptoms of hypo-and hyperglycemia, and consider having the patient's blood glucose monitored during at least the first week of treatment with the drug.