Experts urge caution regarding study of angio-linked contrast nephropathy

Experts are urging physicians to remain skeptical about controversial findings that show patients with compromised kidney function face a more than one in 10 chance of death, stroke, or myocardial infarction after experiencing contrast-induced nephropathy from coronary angiography.

The study, reported June 25 in the online version of the Journal of the American Society of Nephrology, found that 13% of patients with moderate to severe kidney disease and CIN from a coronary angiogram will experience a major adverse cardiac event in the 12 months following the procedure.

The second-year results drawn from Cardiac Angiography in Renally Impaired Patients (CARE), a multicenter, prospective, double-blinded, randomized trial, indicate that the adjusted rate of adverse events was twice as high in the year after cardiac angiography for patients who experienced CIN as for those who did not incur contrast-related kidney injury.

Negative health effects were reported in the 12-month follow-up period for 31% of the patients. They included end-stage kidney disease, coronary revascularization, and the development of congestive heart failure.

Among the criticisms of the study were that the lead researcher redefined contrast induced nephropathy to improve sensitivity to study questions and that it lacked the statistical power necessary to establish a solid link between CIN and the adverse events.

Overall, the findings suggest to chief investigator Dr. Richard J. Solomon, director of the hypertension and nephrology division at the University of Vermont, that seemingly minor and reversible kidney damage from an exam enhanced with an iodinated contrast agent can lead to serious health problems.

"That supports the hypothesis that acute kidney injury [from CIN] is the determinant for long-term adverse events," Solomon said in an interview.

Secondarily, the study identified fewer adverse reactions when patients were administered iopamidol (Isovue) made by Bracco Diagnostics, the study's sponsor, than with iodixinol (Visipaque), a competing product from GE Medical Diagnostics.

The two vendors have battled through their advertising and in court over the relative safety of the two agents for years. In March, a New Jersey court ruled that GE had falsely advertised that Visipaque is safer than Isovue. It was ordered to pay Bracco $11.4 million in damages and required to publicly retract misleading statements.

CARE was originally conceived as a way to compare the safety profiles of iopamidol and iodixinol. Initial findings, published in Circulation in 2007, did not establish a statistically significant difference for CIN incidence for the two agents, according to Solomon. Bracco Diagnostics sponsored the trial and contracted with Solomon and several of his collaborators as paid consultants.

Solomon and colleagues modified the way they defined CIN to produce a more sensitive measure for the current study. CIN, defined by the conventional measure of elevated serum creatinine, arose for 11.1% of 414 patients in the original study.

A year later researchers were unable to follow up with 120 subjects for various reasons, leaving 294 for the current trial. The CIN rate for 242 of these patients was calculated using cystatin C, an alternative marker for glomerular filtration rate. Based on this measure, Solomon established an overall CIN rate of 17.4%. The higher incidence rate raised the likelihood of achieving statistical significance for the central questions of the trial. But it also led to criticism from Dr. Eric Cantor, a nephrologist and head of medical and professional affairs for GE Medical Diagnostics, that Solomon changed his methods to generate desirable results.

"In the end, it is bad science," he said.

Dr. Richard Cohan, chair of the ACR committee on drug and contrast media, said that the study was not statistically powerful enough to rule out explanations other than CIN for the adverse cardiovascular events that occurred up to a year after the contrast-enhanced procedure.

"We need to be cautious before jumping to that conclusion," he said. "There may be something about these fragile patients that we don't yet understand."

In general, from 17% to 25% of cardiac patients are known to also have abnormal renal function, according to Dr. Joshua A. Becker, a professor of radiology at New York University, but a study involving thousands of patients would be required to determine if CIN or other factors triggered future adverse events.

"This has been talked about for at least 20 years, and we still don't have an answer," he said.

The findings may have direct implications for cardiologists, but it is harder to determine how they apply to the contrast-enhanced CT and x-ray procedures that radiologists usually perform, Cohan said. Radiologists should be aware of the relationship between CIN and adverse events for patients who have kidney disease and have undergone cardiac catheterization, he said.

Becker's advice to radiologists for handling at-risk patients boils down to a single sentence:

"Hydrate them well and hope that nothing happens," he said.

For Solomon, the prudent response turns back to secondary findings about the contrast media safety. The overall adverse event rates for iopamidol and iodixinol were 27% and 36% respectively. Iopamidol had an 11% rate for major events, while iodixinol had a 13% rate. Both differences were statistically significant in the follow-up trial, he said.

"It really doesn't matter [for the primary findings] which contrast agent came out on top," he said. "But it turned out that one did a better job with regard to kidney safety, and that translated into a reduction in long-term adverse outcomes."