True extent of NSAID problems now becoming clearer

An August 1999 warning about corneal problems and topical ophthalmic NSAIDs kicked off a chain reaction of statements, rebuttals and clarifications by physicians, manufacturers and the American Society of Cataract and Refractive Surgery (ASCRS), and within 90 days a multimillion-dollar-a-year drug was withdrawn from the market.

The true number and nature of the clinical cases that triggered the ASCRS statement and all that followed will remain unclear until Oliver Schein, MD, of the Wilmer Eye Institute, issues his detailed report on the survey data. That will have to wait until patient records and detailed questionnaires are obtained. In the meantime, many ophthalmologists are wondering how many patients actually were affected and what drugs were involved. Here is what Ocular Surgery News could find out.

Doing the numbers

When reports of postoperative corneal melts started to circulate early in 1999, it was no consolation to patients or their ophthalmologists that it was an exceedingly rare phenomenon. According to data from the health-care market research firm Scott Levin Associates that were provided to Ocular Surgery News by Allergan, there were 1,083,000 prescriptions written for Acular (ketorolac tromethamine, Allergan), 557,000 for Voltaren (diclofenac sodium, CIBA Vision) and 61,000 for DSOS (diclofenac sodium ophthalmic solution, Falcon) in 1998. (DSOS was not available on the U.S. market until August 1998.) ASCRS said that according to the manufacturers, there were approximately 1 million units of Acular, 500,000 units of Voltaren and 500,000 units of DSOS sold in the 12 months preceding the NSAID survey. The survey found only about 200 cases with adverse events, and there were few details on those.

It is hard to know what to say about the survey results today because a number of conflicting statements about the data have been released.

The initial Aug. 3 ASCRS statement did not number or describe in any detail the cases of postoperative corneal complications, but it did say that “all currently used topical NSAIDs were implicated.” None should be used postoperatively, the statement said flatly, until the nature of the problem could be determined.

Eight days later, after a review of the survey responses and discussions with doctors and manufacturers, ASCRS revised its position and released some figures.

The Aug. 11 “NSAID Update” was ambiguous, however. The warning against using any postoperative NSAIDs was gone, but not specifically revoked. On the one hand, ASCRS wanted to “reiterate our significant concerns.” On the other, “we are also aware of the otherwise excellent and long track record of NSAIDs.” Sometimes there were three hands: “the majority of cases have been associated with generic diclofenac” (that is, the Falcon brand DSOS), but “Voltaren has been mentioned” – but then again, maybe those patients didn’t get Voltaren at all. There had not yet been any attempt to verify which drug was dispensed.

At least the statement on Acular seemed clear enough: “eight cases of varying degrees of keratitis have been reported” and “No cases of corneal melts.”

But then 4 weeks later, Kenneth Rosenthal, MD, told the annual meeting of the European Society of Cataract and Refractive Surgery that the ASCRS survey found eight corneal melts associated with Acular (two of which required major intervention or resulted in permanent and severe loss of vision) and 16 cases of keratitis. Those figures, based on the initial tabulation of the ASCRS survey, were repeated in a news report in Ocular Surgery News and in an article Dr. Rosenthal wrote for Ophthalmology Times.

So what did the ASCRS survey find out about Acular?

Acular record clarified

In a letter to Ocular Surgery News, Peter C. Bartlett of Allergan’s department of scientific information and medical compliance wrote that “upon investigation of these reports, they [ASCRS] found that there were no corneal melts associated with use of Acular after cataract or refractive surgery.”

According to David Garbe, director of the department, when he and his colleagues at Allergan investigated the eight reported cases of “corneal melt” in the ASCRS survey, they found that some cases were not a corneal melt as the survey had defined it

“Some doctors saw corneal thinning,” Mr. Garbe told Ocular Surgery News, “but not what most people would consider a corneal melt.” One case of ‘corneal melt’ was in fact a neurotrophic ulcer in a very ill cancer patient.

In other cases, Acular had not actually been dispensed, or there was some factor other than routine NSAID use that was a more likely cause of the melt. The two eyes reported in the ASCRS survey to have had corneal melts requiring intervention were in one patient with Sjögren’s syndrome who had been using topical NSAIDs up to 10 times a day for months. Allergan has not been able to substantiate that case, however.

“Allergan has not received, to date, any post-marketing reports of corneal melt which are associated with use of Acular after cataract or refractive surgery,” Mr. Bartlett told us in his letter.

Don’t feel bad if you’re not sure what the ASCRS survey said about diclofenac. The story has been very confusing.

Let’s start by getting the products straight. There were two brands of topical ophthalmic diclofenac available in the U.S. at the time of the ASCRS survey. One was CIBA Vision’s Voltaren, introduced to the market in 1991 and still available today. The other was DSOS, introduced in August 1998 by Alcon subsidiary Falcon and withdrawn from the market in September 1999.

DSOS was not a generic drug. Diclofenac was the active ingredient in both Voltaren and DSOS, but the formulations were different. DSOS came to market only after the FDA reviewed a New Drug Application and Alcon and CIBA Vision agreed to settle a patent dispute. An approvable letter for DSOS was issued based on a clinical study demonstrating safety and efficacy equivalent to Voltaren.

Now, what did the ASCRS survey say about these two diclofenac products?

The Aug. 11 “NSAID Update” from ASCRS said that “at least 200 cases of corneal toxicity” ranging in severity from superficial punctate keratitis to “full corneal melts” had been reported, that “the majority of cases have been associated with generic diclofenac,” but that “we have yet to determine what agent was actually dispensed.”

In his ESCRS presentation on the ASCRS survey, Dr. Rosenthal said that there were 21 cases of corneal melt associated with Voltaren, none of which required intervention. In his later article, however, he specified that “patients in this group with severe reactions who had been treated with Voltaren or Acular all had underlying rheumatological disease, which may have contributed to the clinical picture.”

Dr. Rosenthal also wrote that severe corneal problems associated with NSAID use began appearing only after the introduction of DSOS and seemed to occur exclusively in the U.S. – the only market in which DSOS was available.

But ASCRS placed a different emphasis in a Sept. 27 “NSAIDs Update.”

Clarifications, qualifications

“Following discussions with Alcon (Falcon) and further investigations … members of both the ASCRS Executive Committee and the Cataract Special Interest Group unanimously share the view that corneal complications can result from the use of any of the topical NSAIDs now on the market in the United States.” Here again there was a qualifier: “The majority of the most severe corneal melts occurred following the use of generic diclofenac.”

Soon thereafter, DSOS was voluntarily taken off the U.S. market by Alcon.

“The FDA classified this as a product withdrawal, not a recall,” said Fred Pettinato, Alcon’s vice president and general manager for pharmaceuticals, in an interview with Ocular Surgery News. “There will be further study by the ASCRS, and we’re participating in that. I think we have developed a closer relationship with the ASCRS as a result of this.

“We still feel strongly that this was a class issue,” Mr. Pettinato said.

“The incidence of corneal melt [with NSAIDs] is very low,” he emphasized. “It appears to be closely correlated to use other than approved dosage and indications. When used properly, NSAIDs are safe and efficacious.”

So where does that leave Voltaren users? According to Dan Myers, the president of North American ophthalmics for CIBA Vision, it leaves them most of all with a better understanding of how to use NSAIDs properly.

“Maybe this has raised awareness of problems that might not have been reported otherwise,” Mr. Meyers said. “Of all the cases that ASCRS sent to us, we interviewed every one [of the doctors].” Many of the reports turned out to be questionable even at the most basic level of what was being reported.

“The wording matters,” Mr. Meyers said. “What some call corneal melt versus thinning versus keratitis is somewhat subjective. Intervention has to be the boundary between serious problems and other corneal adverse events.”

Of the reports in the ASCRS survey that identified Voltaren as the drug being used, CIBA Vision found only three eyes in two patients that had corneal melt requiring some type of intervention. One case was a patient with rheumatoid arthritis.

“I think we all know now that that is probably a population that is predisposed to this,” Mr. Meyers said, “and you should be cautious using NSAIDs if you use them at all” in those patients.

The other patient from the ASCRS reports with a confirmed corneal melt was a woman who had been given a sample of Voltaren, then somehow managed to obtain refills that she continued using for 6 months before returning to her ophthalmologist.

The ASCRS survey is not the only source of information about problems with topical ophthalmic NSAIDs. Even before Alcon, Allergan and CIBA Vision started investigating the ASCRS case reports, the FDA had already received some reports of serious corneal conditions that physicians thought were caused by topical ophthalmic diclofenac.

MedWatch protocols

The only source of adverse event reports on medical devices and drugs that is guaranteed to be available upon demand to the public is the database of MedWatch — the Adverse Event Reporting System (AERS) maintained by the Office of Postmarketing Drug Risk Assessment (OPDRA) of the FDA.

This database includes all adverse events reported to the FDA concerning drugs on the U.S. market since the system was activated on Nov. 1, 1997. Information on reports made before that date is also on file, but it is in an electronic format that is much more difficult to search.

The ongoing MedWatch program has been set up to make it easy to report adverse events associated with drugs or medical devices. Health professionals, for example, can call (800) FDA-1088, fax (800) FDA-0178 or go to the FDA Web page http://www.fda.gov/medwatch/report/hcp.htm to submit a report.

According to OPDRA, 90% of the reports received by the FDA still come from drug manufacturers, which are required to notify the FDA of any adverse events reported to them. The other 10% are from health care professionals and consumers, who may report directly to the FDA but are under no obligation to do so.

Most of the MedWatch data is public information available through a Freedom of Information request, but the program says it will not disclose anything that could be used to identify the patient or health care professional who made the report. MedWatch does note, however, that “to allow for timely follow-up in serious cases, the reporter’s identity may be shared with the manufacturer unless specifically requested otherwise” on the adverse event reporting form.

MedWatch on diclofenac

Ocular Surgery News requested and obtained the AERS data on adverse events reported for diclofenac or ketorolac. No adverse events involving ophthalmic ketorolac were found.

By searching for key words in an electronic document containing more than 1,100 AERS Individual Safety Reports on diclofenac, we identified 52 in which the ophthalmic form was listed as the primary suspected agent in adverse events reported between Nov. 1, 1997, and Oct. 18, 1999 (the latest date for which reports were available at the time we made our request). Those reports were all dated after March 1999. In 41 of those cases, an outcome of “Required Intervention to Prevent Permanent Impairment/Damage” was noted. Those cases are listed in the table at right.

Although AERS emphasizes that “the information contained in the reports has not been scientifically or otherwise verified” and that “accumulated case reports cannot be used to calculate incidence or estimates of drug risk,” the data do document the recent history of ophthalmic diclofenac.

The story in brief

For 16 months, from the time the database begins in November 1997 until March 1999, there were no adverse event reports mentioning topical ophthalmic diclofenac. For most of that time, the only brand of ophthalmic diclofenac was Voltaren.

According to several independent sources interviewed by Ocular Surgery News and published news reports, surgeons started noticing unexplained corneal melts in some of their postoperative cataract and refractive surgery patients in January 1999. Alcon had introduced DSOS to the U.S. market in August 1998, and several doctors became convinced that the corneal problems were related to that drug.

Some surgeons notified the drug manufacturers, who then investigated each case and filed the required FDA reports.

The first one to reach the FDA is dated March 31, 1999. It lists corneal erosion as an adverse event in a male patient receiving diclofenac, with Alcon named as the manufacturer. No outcome was specified in that case, but on April 2 there was another report of corneal erosion, this time accompanied by postoperative wound breakdown. This report classified the problem as one requiring intervention to prevent damage. In May, there were two cases requiring intervention.

While MedWatch logged another 10 serious cornea cases in June and eight in July, the ASCRS was gathering information in its member survey. Three more case reports came in to the FDA in August, and ASCRS issued its first warnings. Nine more reports came in September and eight more in October. By then, DSOS was withdrawn from the market. By the end of the year, however, Allergan and CIBA Vision had launched educational and promotional efforts encouraging ophthalmologists to continue using Acular and Voltaren at the right dosage in the right patients.

Throughout the debate on the issue, Alcon maintained that improper dosing of NSAIDs and poor patient selection were responsible for the corneal complications. There were clearly cases where diclofenac or ketorolac use had gone far beyond the four-times-daily maximum in the approved product labeling. MedWatch reports cases where diclofenac was prescribed “Qhs.”

Ophthalmologists who have studied NSAIDs agree that overdosing will cause corneal complications and that certain classes of patient are at significantly greater risk for problems. They emphasize that the boundaries of safe practice are clearly marked (see Related Article). These surgeons have continued to use Voltaren as well as Acular in their postoperative patients, and they encourage their colleagues to use the NSAID class without fear.

For Your Information:

• MedWatch reports can be made by telephone to (800) FDA-1088, by fax to (800) FDA-0178 or at the FDA Web page http://www.fda.gov/medwatch/report/hcp.htm. If you have a question about submitting a report about an adverse event associated with a drug, call the FDA Reports Evaluation Branch at (301) 827-3170. Adverse event reporting forms (FDA Form 3500) can be obtained from Consolidated Forms and Publication Distribution Center, Washington Commerce Center, 3222 Hubbard Road, Landover, MD 20785.
• Ocular Surgery News coverage of the ophthalmic NSAID controversy is available on the “Back Issues” page of our Web site.
• "ASCRS alert: beware of NSAIDs postoperatively," September 1, 1999. "Alcon halts NSAID distribution," September 15, 1999. "ASCRS clarifies its NSAID alert,” October 1, 1999. "Alcon recalls its generic diclofenac," November 1, 1999.