Celeste Castillo Lee, a hospital administrator from North Carolina who suffers from nephrogenic systemic fibrosis, appears before an FDA advisory committee Tuesday. (Video copyright FDALive.com.)

Most members of a federal advisory panel recommended Tuesday that the Food and Drug Administration effectively ban, for patients with severe kidney disease, the use of two drugs used to create high-contrast images on magnetic resonance imaging (MRI) scans.

The drugs are Omniscan, owned by General Electric Co., and Optimark, made by the Mallinckrodt unit of Covidien.

The FDA didn’t ask the panel to take a formal vote, but it often adopts the advice of outside committees. After the daylong session, which included a last minute shift of position by one key member, FDA officials said they understood the group’s majority opinion to be that the two drugs should be "contraindicated" for severe kidney patients. They said they would take that and other input from the panel into consideration as they discuss what to do next in the coming weeks.

The FDA had asked the panel whether and, if so, how to tighten restrictions on use of several MRI agents, all of which contain the metal gadolinium and which are injected during magnetic resonance scans.

Read our original investigation into GE's Omniscan: Specter of MRI Disease Haunts GE

For the last three years, since the agents were first linked to a rare but often crippling or deadly disease, nephrogenic systemic fibrosis (NSF), the FDA has treated them as a class, requiring all drug makers to issue the same "black box" warnings for severe kidney patients. The term refers to a black outline calling extra attention to the warning. There were five agents on the market at the time; now there are seven.

But in recent weeks, in advance of the panel’s meeting, the agency had done further analysis and concluded that different, somewhat sterner warnings were warranted for three of the agents, which research indicates carry higher risks. The three are Omniscan, Optimark and Magnevist, the market leader, made by Bayer HealthCare.

Most of the panel went a step further than that, believing that Omniscan and Optimark warranted the even sterner step of contraindication. Since Optimark, in advance of the meeting, had voluntarily decided to change its label to indicate contraindication for severe kidney patients, the potential impact of the panel’s recommendation falls squarely on Omniscan, the second-most widely used brand.

There wasn’t a clear consensus among panelists about how to treat Magnevist; some said it too should be contraindicated, others agreed it was riskier than other agents but were unsure if a ban was warranted, still others were not sure if Magnevist was as risky as Omniscan and Optimark, and still others said more evidence was needed to make any such determination.

The lengthy deliberations were, in some ways, a dry debate among two dozen doctors and safety experts about how to manage the business and regulatory risks of drugs that have been safely used tens of millions of times to help doctors diagnose many diseases and injuries.

Some argued that more proof was needed before differentiating or banning drugs, even for the small population of patients with severe kidney disease. Others said the safety of that population couldn’t wait for better data, in part because doctors were already ahead of regulators in differentiating the relative risks of the agents. If all of the drugs are viewed as carrying similar risk, then the marketing prowess of a GE, for example, can be a determiner of market share. But if doctors conclude that the risks are higher for two or three of the drugs, and particularly if such conclusions are ratified by differential regulation by the FDA, then the share of the drugs viewed as less risky will rise. That is precisely what some panel members said is happening.

The market share of GE’s Omniscan has fallen by about half in the past few years, according to FDA data presented at the meeting. "We can’t wait until there is causative, definitive proof," said Dr. Judith M. Kramer, an associate professor of medicine at Duke University Medical Center, suggesting that the labeling needed to catch up with the realities of the market place.

Should the FDA adapt some of the restrictions recently recommended in Europe, such as mandatory screening of all patients for kidney disease before use of riskier drugs, the trends in market share, experts say, could accelerate: Omniscan, now at 17 percent, could drop further while one of Bracco Diagnostics' MRI drugs, MutliHance, which has gone from 3 percent in 2006 to 13 percent in 2009, could continue to gain more customers.

The meeting, at a suburban Maryland hotel, was open to the public and broadcast via the Internet. The audience included scores of doctors, scientists and executives of the various drug makers.

The morning session was dominated by presentations from the various manufacturers outlining the scientific and marketing histories of their drugs. GE argued that the FDA’s decision in 2007 to treat the drugs uniformly as a class "has been effective" and that "differential risk remains unproven." Bayer told the panel that there was insufficient data to draw a clear picture, but, if the panel chose to differentiate, the available "evidence supports a distinction between Omniscan/Optimark and all other" gadolinium agents.

Optimark’s representative said its decision to contraindicate "should be applied class-wide."

Bracco Diagnostics, the manufacturer of MultiHance and another agent, ProHance, argued that the very low incidence of NSF cases for its two drugs meant they should not be subject to more restrictive measures.

The FDA then gave its analysis of why the three drugs were riskier than the others. It focused on chemistry – Omniscan and Optimark were believed to be less stable – and various statistical analyses showing relatively more reported cases of NSF for the two drugs and Magnevist.

In the afternoon the panel, consisting of two committees, one on drug safety and another on cardiovascular and kidney issues, heard from the public.

In a presentation that seemed to rivet many members of the panel, an NSF patient from North Carolina described, in clinical detail, the "torture" she has experienced since contracting NSF after undergoing an MRI in 2006.

The patient, Celeste Castillo Lee, told the panel that "75 prcent of us are in wheelchairs" but that doesn’t adequately describe the actual disability. "It’s actually a torture," she said. She told the panel how an initial swelling in her ankles migrated up through her organs and body. Her bones, she said, felt like they were "in a vise."

Another presenter was Emanuel Kanal, a University of Pittsburgh professor of radiology who heads the American College of Radiology’s expert panel on magnetic resonance safety.

Kanal’s committee in 2007 had recommended more restrictive measures against Omniscan, as did European regulators and two separate medical reviews within the FDA, but the agency decided to adopt a more conservative, class-wide approach owing to a lack of data.

A few of the panel’s members expressed unhappiness with the lack of data and said that the FDA needed more "proof" before it could take further action.

The chairman, Dr. Robert A. Harrington, a professor of medicine at Duke University Medical School, steered the members to address the FDA’s primary question of whether the agency has done enough to address the risk of NSF and how it should proceed. As Harrington polled members two camps emerged. One, tied to Dr. Sidney M. Wolfe, a consumer representative on the panel and the head of health research at the Ralph-Nader-spawned consumerist group Public Citizen, favored restrictions on Omniscan and Optimark and perhaps on Magnevist. The other was aligned with Dr. Lawrence G. Hunsicker, a professor of internal medicine at the University of Iowa Hospitals and Clinics. Hunsicker at first described himself as being "at the other extreme" of Wolfe, favoring a "leave as is" approach or requiring more data before saying other agents present a lower risk.

As Harrington surveyed the members, a significant majority said they were aligned with Wolfe. As Harrington summarized the discussion, he drilled down further to see if the majority felt that Omniscan and Optimark should be contraindicated.

Two camps emerged again, though less clearly; those for and those against contraindication. Then, after Wolfe said "contraindication is appropriate," Hunsicker said "I’m going to switch sides" and endorsed contraindication for the two drugs. He explained that one of the advocates of the status quo had offered the possibility that the labels of the riskier agents say that "prudence would suggest you avoid the use" of the agent in severe kidney patients.

A prudency standard, Hunsicker said, was, in effect, "the same thing as contraindication." The meeting ended shortly thereafter.