Elderly patients taking the widely used antibiotic gatifloxacin were almost 17 times as likely to be hospitalized for very high blood sugar levels and four times as likely to be hospitalized for unusually low levels, a finding that is leading some physicians to call for the drug’s withdrawal from the market.
Overall, one of every 100 patients who took the drug was hospitalized, according to a study released online Wednesday by the New England Journal of Medicine. The study was released before its March 30 publication because of its health implications.
Several previous studies have shown an increased risk of glucose abnormalities and several deaths in diabetic patients who received the antibiotic, trade-named Tequin. The drug’s label was changed last month to say it should not be given to diabetics.
The latest study, which is larger and more definitive, showed that all patients were at risk, even those who were not diabetic.
The study’s authors said that because other antibiotics were as effective, there was no reason to continue prescribing gatifloxacin.
“Speaking as a clinician, I would never prescribe this drug,” said Dr. David N. Juurlink of the Institute for Clinical and Evaluative Sciences in Toronto, who led the study.
Dr. Sidney Wolfe of the Public Citizen Health Research Group in Washington, D.C., said: “This represents a unique danger in the absence of a unique benefit.... This is more than enough reason to think about petitioning the Food and Drug Administration to ban the drug, and we probably will.”
Eric Miller, a spokesman for Bristol-Myers Squibb Co., which makes Tequin, said that the findings “were consistent with the post-marketing experience we have had to date.” He said the labeling changes in February took the findings into account.
Miller said the company’s annual sales of the drug were about $100 million in the United States and $150 million worldwide -- a relatively minor portion of Bristol-Myers’ $19.4 billion in total revenue.
Gatifloxacin is a member of the family of broad-spectrum antibiotics known as fluoroquinolones. It is typically used to treat gonorrhea and lung, sinus and urinary tract infections. Physicians often use it when the nature of an infection is unknown, Juurlink said, because it kills a wide variety of bacteria.
But the quinoline family has proved problematic. Four other fluoroquinolones have been withdrawn from the market or had their use severely restricted: temafloxacin, for causing red blood cell damage, kidney failure and hypoglycemia; grepafloxacin and sparfloxacin because of heart problems; and trovafloxacin because of liver damage.
“That is an alarming proportion of drugs in that class that have been taken off the market,” said Wolfe of Public Citizen.
Gatifloxacin was introduced in 1999. By 2001, 3.3 million prescriptions per year were being written.
Researchers began noticing health problems that year, particularly alterations in glucose metabolism. Changes in blood sugar levels can induce coma and other serious problems, including death. Symptoms usually began five to 10 days after patients took the drug.
In most cases, the symptoms could be reversed when drug use was halted.
By 2003, 17 deaths had been linked to the drug and prescriptions were down to about 1.7 million per year.
One major contributor to that number is the Department of Veterans Affairs, which added the drug to its formulary and designated it an antibiotic of first choice, in part because Bristol-Myers offered the government a price of $1.35 per pill, compared with the $8 to $10 per pill charged for it and other fluoroquinolones at pharmacies, said Dr. Richard Frothingham of Duke University.
The government also chose it because the risk of glucose abnormalities caused by gatifloxacin did not seem to be any higher than that caused by other antibiotics in the class, Juurlink said.
In light of the new findings, he said, “the VA needs to very promptly revisit their policy.”
The Canadian researchers studied health records for 1.4 million Ontario residents over age 65. Among them were about 17,000 gatifloxacin patients.
Juurlink and his colleagues then looked at all patients who had been hospitalized after taking an antibiotic. They found 788 were hospitalized for excessively low blood sugar within 30 days of taking an antibiotic and 470 patients hospitalized for excessively high blood sugar.
Within that group, there were 61 gatifloxacin recipients with hypoglycemia and 86 with hyperglycemia. Compared with other patients in the group, the gatifloxacin recipients had four times the risk of hypoglycemia and 17 times the risk of hyperglycemia.
“This is the most compelling evidence to date showing a connection between gatifloxacin and glucose problems,” Frothingham said.
No other fluoroquinolone showed an unusual risk.
Juurlink said the study probably underestimated the risks of the drug. “If they died at home or in a pre-hospital setting, they would not have made it into the study,” he said.