Hypoglycemia Associated with the Use of Levofloxacin

Seth M. Garber; Melanie W. Pound; Susan M. Miller


Am J Health Syst Pharm. 2009;66(11):1014-1019. 

In This Article


The proposed mechanism by which the fluoroquinolones induce glycemic abnormalities is not clearly understood. Reports of fluoroquinolone-induced hypoglycemia are abundant, but reports of hyperglycemia induced by fluoroquinolones are also available.[15–19]

The primary theory of fluoroquinolone-associated hypoglycemia is twofold, consisting of both pharmacokinetic and pharmacodynamic effects. The pharmacokinetic mechanism involves drug–drug interactions, while the pharmacodynamic mechanism comprises the possibility of enhanced pancreatic β-cell stimulation and subsequent increased insulin release.[20]

Pharmacokinetic Mechanism

The theory of a possible drug–drug interaction has been considered among many case reports and case series involving fluoroquinolones and antihyperglycemic medications.[4,14,21–25] The fluoroquinolones involved in these reports were gatifloxacin (seven cases),[1–3,21,22] ciprofloxacin (two cases),[24,25] and levofloxacin (three cases).[4,14,23] Glyburide was the most often implicated antihyperglycemic drug, accounting for five of these cases.[4,14,22,24,25] The remaining interactions involved glimepiride, pioglitazone (with glyburide), and repaglinide.

Kirchheiner et al.[26] investigated genetic polymorphisms of the cytochrome P-450 (CYP) isoenzyme system and their effect on the activity of oral antihyperglycemic medications. They found that CYP2C9 is the primary isoenzyme pathway responsible for metabolizing glyburide, glimepiride, and glipizide, though other pathways may play a minor role in the metabolism of these drugs. Other antihyperglycemic medications metabolized by CYP isoenzyme systems include nateglinide (CYP2C9), repaglinide (CYP2C8), pioglitazone (CYP2C8 and CYP3A4), and rosiglitazone (CYP2C9 and CYP2C8).[26] Genetic polymorphisms in CYP2C8 and CYP2C9 may alter the clearance of these drugs from the body in certain individuals with altered genotypes of these isoenzymes. Theoretically, concomitant use of a drug that is a substrate or an inhibitor of a CYP pathway by someone with an altered genetic polymorphism of that same CYP isoenzyme could result in increased serum drug concentrations, resulting in an enhanced hypoglycemic effect of the oral antihyperglycemic drug.

Like the sulfonylureas, certain fluoroquinolones interact with CYP isoenzymes to some extent. In one study, ciprofloxacin inhibited the metabolism of theophylline via the CYP1A2 pathway.[27] Ciprofloxacin has also been reported to be an inhibitor of the CYP3A4 pathway.[25] Nakamura et al.[28] described a levofloxacin–theophylline–clarithromycin interaction believed to be mediated by CYP1A2 and CYP3A4. According to this case report, the clearance of theophylline decreased because its metabolism was inhibited by levofloxacin and clarithromycin through the CYP1A2 and CYP3A4 pathways. According to the manufacturer, neither gatifloxacin nor moxifloxacin is thought to be an inhibitor, an inducer, or a substrate of the CYP pathways based on in vitro studies; therefore, a drug–drug interaction would have been unlikely.[29,30]

Although both the sulfonylureas and certain fluoroquinolones interact with the CYP isoenzyme system, each proceeds through a different primary pathway. Thus, a drug–drug interaction via CYP seems highly unlikely, indicating that other mechanisms must be involved.

Pharmacodynamic Mechanism

The proposed pharmacodynamic mechanism by which the fluoroquinolones induce glycemic abnormalities is not completely understood. Augmentation of insulin release from the islet cells of the pancreas has been reported as the most likely mechanism for fluoroquinolone-induced hypoglycemia.[31,32] Adenosine triphosphate (ATP)-sensitive potassium channels are involved in insulin secretion. When these channels are blocked, the membrane of the β-cells is depolarized, allowing calcium to enter the cell through the voltage-dependent calcium channels. Insulin granules then exit the β-cells, and blood glucose is reduced. The ATP-sensitive potassium channels of the islet cells are inhibited by the fluoroquinolones. Due to this inhibition, insulin secretion is increased, and hypoglycemia can ensue.[31]

Saraya et al.[31] studied the effect of levofloxacin, gatifloxacin, and temafloxacin on insulin secretion and ATP-sensitive potassium-channel activity in rat pancreatic islet cells. Only small increases in insulin secretion occurred with levofloxacin, while insulin secretion increased significantly with gatifloxacin and temafloxacin. Levofloxacin slightly reduced potassium-channel activity, while gatifloxacin and temafloxacin markedly inhibited potassium-channel activity.

On a cellular level, eight subunits (four Kir6.2 and four SUR1) comprise the ATP-sensitive potassium channel of the pancreatic β-cell. Saraya et al.[31] found that levofloxacin, gatifloxacin, and temafloxacin inhibit the Kir6.2 subunits of pancreatic β-cells. Gatifloxacin and temafloxacin appeared to have greater inhibitory potential than did levofloxacin on the Kir6.2 subunit, which may explain why more cases of hypoglycemia have been reported with gatifloxacin than with levofloxacin. Temafloxacin was voluntarily recalled from the U.S. market in 1992 due to reports of its association with hemolytic anemia and several cases of severe hypoglycemia.[11]

Under normal conditions, the body can compensate for a decrease in blood glucose levels through physiological mechanisms.[33] Normally, a decrease in blood glucose levels causes the pancreas to decrease its insulin secretion and glycogenolysis to increase in the liver. Glucose is produced endogenously from lactate, glycerol, and amino acids.[34] Malnourished patients, such as the elderly, may not have sufficient glycogen reserves to mobilize in response to the hypoglycemia caused by fluoroquinolones.[35] This inability to appropriately compensate, along with decreases in renal function in elderly patients, may cause higher drug levels or decreased drug clearance. This may explain why fluoroquinolone-induced hypoglycemia is most frequently described in older patients.

Use of the Naranjo et al.[36] adverse-drug-reaction probability scale revealed that levofloxacin was possibly the cause of our patient's hypoglycemia (score = 4). Our patient had many of the risk factors often cited for fluoroquinolone-induced hypoglycemia, including renal insufficiency, diabetes, and sulfonylurea use. His age may have also been a contributing risk factor.

Several case reports cited improper adjustments in fluoroquinolone dosage in the setting of reduced creatinine clearance as a possible reason for hypoglycemia.[1,2,7] Our patient had a serum creatinine concentration of 1.5 mg/dL and a calculated creatinine clearance of 42 mL/min. His levofloxacin was dosed at 750 mg orally every 48 hours, as recommended by the manufacturer for patients with a creatinine clearance of < 50 mL/min. The patient's serum glucose concentration decreased very slightly after the first dose of levofloxacin, from 150 to 131 mg/dL. The second dose was given 65 hours, instead of 48 hours, after the initial dose due to a delay in medication administration. It is therefore unlikely that improper dosing of levofloxacin in the context of renal insufficiency was responsible for the patient's hypoglycemic episode.

One important difference between the first and second doses of levofloxacin was that glipizide 10 mg was administered simultaneously with the second dose of levofloxacin. After the concomitant administration of these two medications, the patient's serum glucose concentration dropped to 82 mg/dL, the last measured value before he was discharged. Upon his readmission to the hospital, his serum glucose concentration was 20 mg/dL. Of interest, the patient had no significant hypoglycemia associated with his first dose of levofloxacin and only experienced severe hypoglycemia when it was administered concomitantly with glipizide. This supports the theory that a drug–drug interaction may be responsible for the severe and resistant hypoglycemia noted in this and previous case reports.

Since the patient had been previously stable on a "sugar pill" at home, subsequently identified through the community pharmacy as glimepiride (2 mg orally daily), administration of the sulfonylurea was not the sole cause of the severe hypoglycemia. Per computerized refill records at the community pharmacy this patient utilized, the patient had been compliant with his glimepiride during February–May 2006, the time period during which these hospitalizations occurred. Furthermore, per the community pharmacist, the patient had subsequent glimepiride prescriptions filled after these hospital admissions, yet the patient has not been hospitalized at our institution with hypoglycemia since this incident. Based on the pharmacokinetics of glimepiride and glipizide in a patient with renal dysfunction, glimepiride would be more likely to induce hypoglycemia than would glipizide due to its metabolism to an active metabolite, whose excretion would be impaired in this patient.

Although the patient in this case was not elderly, he did appear clinically malnourished, with a height and weight of 63 inches and 52 kg, respectively. His measured serum albumin concentration was 2.1 mg/dL (normal range, 3.5–4.8 mg/dL). This supports the theory that the combination of fluoroquinolone-induced insulin release and a decreased ability of the body to employ compensatory mechanisms to correct hypoglycemia may lead to persistent hypoglycemia.

Of note, both glipizide and glimepiride are highly protein bound. Because of this patient's malnutrition, the inhospital use of glipizide could have been associated with hypoglycemia attributable to a decrease in protein binding of the drug. Since glimepiride is more highly protein bound (99%) than glipizide (97%) and the patient did not complain of hypoglycemia episodes at home, the possibility that the decreased protein binding of glipizide was the sole cause of the patient's hypoglycemia is decreased.

It is difficult to determine the most likely mechanism for the hypoglycemia observed in this patient; however, his hypoglycemia was most likely due to the concurrent use of levofloxacin and glipizide, his diabetes, and malnutrition.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.