Since its introduction as a further addition to the quinolone class in late 1999, gatifloxacin (Tequin, Bristol-Myers Squibb) has been relatively well tolerated. However, the prescribing information has been updated multiple times to highlight disturbances in glucose homeostasis reported to or by the manufacturer postmarketing. Current information includes a warning for careful monitoring of blood glucose when gatifloxacin is administered to patients with diabetes because of possible hypoglycemia and hyperglycemia. Several cases of significant hypoglycemia related to use of gatifloxacin have been published in the medical literature.1-3 Subsequent to our observation, 1 report also suggested possible gatifloxacin-induced hyperglycemia.3
Mr. J is a 70-year-old white male veteran with 26 years of service in the Marine Corps. He has a medical history that includes restrictive lung disease requiring home oxygen, obstructive sleep apnea on continuous positive airway pressure, ischemic cardiomyopathy with a known ejection fraction of 30%, status post coronary artery bypass graft and aortic valve replacement for aortic stenosis, status post abdominal aortic aneurysm repair, implantable cardiac defibrillator (ICD), anemia, and diabetes mellitus. His acute illness began 2 weeks prior to admission with complaints of a chest cold. He was initially thought to have decompensated heart failure in the setting of a viral illness, but symptoms failed to improve with adjustment of his diuretics. On the day of admission, his physical examination was significant for the following: blood pressure, 111/55; pulse, 90; respiratory rate, 20; temperature, 98.3°F; O2 saturation, 94% on 2 L O2 via nasal cannula; weight, 242 lb (110.0 kg; stable); height, 67 in (170.2 cm).
There was no evidence for decompensated heart failure; lung examination revealed only decreased breath sounds. Laboratory assessment revealed a white blood cell count of 27.4 K/cmm, with a predominance of polymorphonuclear cells, a hematocrit of 30.4 at his baseline, creatinine of 1.3 mg/dL, and glucose of 142 mg/dL. A chest x-ray was read as having possible lower lobe infiltrates. He was admitted that evening with a diagnosis of pneumonia, and intravenous gatifloxacin (400 mg/d) was added to his medical regimen of glipizide (5 mg/d), simvastatin, aspirin, furosemide, pseudophedrine, ascorbic acid, ferrous sulfate, rabeprazole, captopril, isosorbide mononitrate, metoprolol, spironolactone, digoxin, warfarin, citalopram, beclomethasone nasal inhaler, and eye drops for glaucoma.
Mr. J reported that his diabetes had been under very good control since beginning glipizide just 3 months prior to admission. He reported regular finger sticks of 100 mg/dL fasting and predinner blood glucose measurements of ~120 mg/dL. Confirming his report of good control were 4 outpatient serum values collected by our laboratory over the previous 3 months of 105, 121, 111, and 99 mg/dL, the most recent just 9 days prior to admission.
Over the course of his 4-day hospital stay, the patient's respiratory status and serum white blood cell counts steadily improved back toward his baseline. He was discharged to home to complete a 14-day course of gatifloxacin now on oral medication. Figure 1 denotes Mr. J's glucose measurements in the days before and after beginning gatifloxacin. As noted beginning on hospital day 3 and progressing, Mr. J had deteriorating glycemic control, despite his overall improved clinical status.
On the first day after discharge, Mr. J contacted his primary care provider to discuss his worsening glycemic control. He felt well but reported his home glucose readings were all around 400 mg/dL. Glipizide was increased to 10 mg bid.
Posthospital day 3, glycemic control remained very poor. Physical examination failed to reveal an intercurrent problem. Review of the patient record revealed normal urinalysis, except for glycosuria on the day of discharge. The question of gatifloxacin-induced hyperglycemia was raised and this medication was stopped.
Posthospital day 9, the patient was reevaluated. He experienced deteriorating respiratory status. On this same day, he noted his first fasting glucose under 200 mg/dL (162 mg/dL). Clinical assessment revealed slight increase in his white blood cell count and radiographic evidence of resolving retrocardiac infiltrate. He was treated with azithromycin and observed in hospital for 48 hours. He had a rapid and uneventful recovery, and despite his glipizide being reduced back to his original 5 mg/d dose, 2 fasting glucoses were recorded as 174 and 157 mg/dL in the hospital. A hemoglobin A1c of 7% was obtained during this period, confirming his generally good glycemic control.
In premarketing studies, no clinically significant changes in glucose tolerance and glucose homeostasis were observed in study subjects following the first intravenous or oral dose of gatifloxacin. Similarly, in diet-controlled diabetes, multiple doses of gatifloxacin did not have a significant effect on glucose homeostasis.4 However, in glyburide-controlled patients with type 2 diabetes, an ~30% to 40% decrease in serum insulin concentrations was seen along with modest increases in fasting glucose, which tended to occur after day 3 but did not reach statistical significance.5 In October 2001, the package insert was modified to reflect the occurrence of disturbances of blood glucose, including symptomatic hyperglycemia (including nonketotic hyperosmolar hyperglycemia), usually in patients with diabetes, but none of these cases are published.5 According to the prescribing information for Tequin, hyperglycemic episodes were found to occur typically after 4 to 10 days into gatifloxacin therapy (consistent with what we have described here) and persist for up to 28 days. Data on file with the manufacturer from an open-label postmarketing phase IV, drug safety trial reported hyperglycemia in 0.2%, but these data are from patients' own monitoring data or presenting with symptoms; no laboratory data were being routine collected. In this study of 15,635 patients, those with diabetes were over 100 times more likely to have noted hyperglycemia (2.65% vs. 0.02%). Although the mechanism of this observation is still unclear, an observation in some gatifloxacin-exposed animals shows decreased secretory granules in pancreatic B-cells and resultant decreased serum insulin concentrations.
Many of the patients requiring hospitalization for symptomatic and severe hyperglycemia reported to the manufacturer had possible confounders, which could have contributed to the glucose abnormalities observed. None of our patient's other medications are known to interact with gatifloxacin to increase its concentration, potentially augmenting toxicity. Our patient was admitted for the treatment of pneumonia, a metabolic stress known to worsen glycemic control in patients with diabetes. Yet, the development of hyperglycemia occurred after stabilization of his pneumonia as evidenced by improvement in his leukocyte count and subjective dyspnea. In addition, his glycemic control continued to improve, despite an exacerbation of his respiratory infection due to premature discontinuation of his antibiotics, suggesting that these stressors were unlikely the cause of his hyperglycemia.
Two early reports attributed hyperglycemia, seizures, and acidosis to nalidixic acid, the earliest quinolone.6,7 Yet, the frequency of hyperglycemia must be rare with this compound, if related at all, as no warning of alteration in glucose homeostasis accompanies current prescribing information for this medication.8
Gatifloxacin has also been associated with hypoglycemia occurring from initial dose through day 3. The manufacturer's postmarketing information reports a 0.08% frequency of low blood glucose that came to the attention of providers (glucose monitoring not required in the study). Three published reports include 6 patients who were all type 2 diabetes on differing diabetic regimens, who experienced severe hypoglycemia.1-3
Alteration of glucose homeostasis is noted by the manufacture of other more widely prescribed fluoroquinolones, ciprofloxacin9 and levofloxacin,10 as well as a newer addition to this class, moxifloxacin.11 Prescribing information for ciprofloxacin lists increase in glucose in <1% of patients and <0.1% for hypoglycemia in postmarketing reports, neither clearly associated with this medication. Severe hypoglycemia from ciprofloxacin in association with glyburide has been noted in 2 published reports.12,13 Levofloxacin prescribing information notes hypoglycemia and hyperglycemia occurring in <1% of patients in clinical trials and high measured glucoses in 2.2% of cases; a warning suggests that symptomatic glucose abnormalities have been noted as well. Moxifloxacin, a newer agent, reports <0.1% hyperglycemia of uncertain relationship to the compound and >2% of patients demonstrating lower serum glucose measures but without mention of symptoms.
The absence of many case reports in the literature attributing hyperglycemia to gatifloxacin or other fluoroquinolone administration may be due to several reasons. Drug-induced hyperglycemia may be underappreciated because infections themselves are associated with worsening glycemic control confounding the association. In addition, the biphasic effect on glucose homeostasis, with early lowering followed by potential increase in serum glucose, may stimulate changes in diabetic management, which are then considered the cause of the glycemic changes. At present, gatifloxacin has the strongest warning and most data included in its prescribing information relevant to alterations in glucose homeostasis. Clinicians should be aware of this manifestation as a potential problem when administering gatifloxacin to patients with diabetes.
1. Baker SE, Hangii MC. Possible gatifloxacin-induced hypoglycemia. Ann Pharmacother
2. Menzies DJ, Dorsainvil PA, Cunha BA, et al. Severe and persistent hypoglycemia due to gatifloxacin interaction with oral hypoglycemic agents. Am J Med
3. Briggs WS. Hypoglycemia and hyperglycemia associated with gatifloxacin use in elderly patients. J Am Board Fam Pract
4. Gajjar DA, LaCreta FP, Kollia GD, et al. Effect of multi-dose gatifloxacin or ciprofloxacin on glucose homeostasis and insulin production in patients with non-insulin-dependent diabetes mellitus maintained with diet and exercise. Pharmacotherapy
5. Tequin (gatifloxacin) [package insert] Princeton, NJ: Bristol-Myers Squibb Co; 2002.
6. Leslie PJ, Creegeen RJ, Proudfoot AT. Lactic acidosis, hyperglycaemia and convulsions following nalidixic acid overdosage. Hum Toxicol
7. Fraser AG, Harrower AD. Convulsions and hyperglycaemia associated with nalidixic acid. BMJ
8. NegGram (nalidixic acid). Prescribing information. Sanofi-Synthelabo, Inc. Revised November 2000. Available at: http://www.sanofi-synthelabo.us/index.html
. Accessed May 6, 2004.
9. Cipro (ciprofloaxacin). Prescribing information. West Haven, CT: Bayer Pharmaceuticals Corp. July 2002. Available at: http://www.univgraph.com/bayer/INSERTS/CIPROTAB.pdf
. Accessed May 7, 2003.
10. Levaquin (levofloxacin). Prescribing information. Raritan, NJ: Ortho-McNeil Pharmaceutical, Inc. Revised January 2003. Available at: http://www.ortho-mcneil.com/products/pi/pdfs/levatab.pdf
. Accessed May 7, 2003.
11. Avelox (moxifloxacin). Prescribing information. West Haven, CT: Bayer Pharmaceuticals Corp. Revised March 2003. Available at: http://www.univgraph.com/Bayer/inserts/Avelox.pdf
. Accessed May 7, 2003.
12. Roberge RJ, Kaplan R, Frank R, Fore C. Glyburide-ciprofloxacin interaction with resistant hypoglycemia. Ann Emerg Med
13. Whiteley M, Worlding J, Patel S, et al. Hypoglycemia in a diabetic patient, associated with ciprofloxacin therapy [case report]. Pract Diabetes