RHEUMATOID ARTHRITIS GUIDANCE REMISSION CLAIM CRITERIA TOO STRINGENT, WORKSHOP PANELISTS SAY; ACTIVE CONTROL, COMBINATION THERAPY TRIALS RECOMMENDED

FDA's draft guidance criteria for a rheumatoid arthritis remission claim is too rigorous and would bar many agents from receiving the indication, David Felson, MD, Boston University School of Medicine, suggested at a March 27 FDA workshop on RA.

"The definition of remission" in FDA's working draft for rheumatoid arthritis studies "is just too tough," Felson asserted. The remission criteria would "put the bar so high that no one can get there," he maintained. Felson is a member of FDA's Arthritis Advisory Committee.

The draft guidance, "Designing Clinical Programs for Developing Human Drugs, Medical Devices, or Biological Products Intended for the Treatment of Rheumatoid Arthritis," proposes that a claim for a remitting agent be based on statistically significant improvement in remission rate using the 1981 American College of Rheumatology criteria of remission sustained over six months. The remission rate must be at least 20% higher than that of standard therapies in a study of at least one year duration.

Center for Drug Evaluation & Research Director Janet Woodcock, MD, acknowledged that it is "very hard to achieve the kind of remission laid out in the draft." Woodcock said the agency rheumatoid arthritis working group, which developed the guidance, "understood that" but felt that a remission claim would require a "very active agent [that is] somewhat curative."

Panel member Michael Weinblatt, MD, Brigham & Women's Hospital, maintained that the problem with the remission claim as described in the guidance is not the FDA-suggested six-month duration of remission but the use of the ACR definition. The 1981 ACR definition of remission requires five or more of the following criteria: morning stiffness, absent or not exceeding 15 minutes; no fatigue; no joint pain; no joint tenderness; no joint or tendon sheath swelling; and no elevation of erythrocyte sedimentation rate.

Weinblatt asserted that a majority of patients in methotrexate trials do not achieve remission because they "fail to obliterate all the swelling because of synovial thickening. The ACR remission criteria says no swelling." He noted that "only 3% of our patients achieved ACR criteria." Panel member Vibeke Strand, MD, Stanford University, remarked that the FDA guidance remission criteria should remain as "a hurdle to aspire to" but that duration of remission should not be specified.

Remission claim presenter and panel member Harold Paulus, MD, University of California at Los Angeles School of Medicine, suggested that "all we need is 90% ACR improvement criteria or 90% Disease Activity Score and call that the drug-induced remission, which might be attainable. It is probably about as good as we are going to be able to do without a real cure."

The rheumatoid arthritis public workshop was organized by FDA to discuss the claims, criteria and clinical study design included in the agency's early draft guidance released March 6 ("The Pink Sheet" March 11, T&G-14). FDA will consider suggestions from the workshop and comments submitted to the agency in its formulation of a final guidance, which will be presented to FDA's Arthritis Advisory Committee. The deadline for comments on the workshop is May 30.

FDA's proposed criteria for a claim of reduction in signs and symptoms of rheumatoid arthritis based on a composite set of indices like those in the ACR preliminary definition of improvement, published in 1995, would set an appropriate common standard for rheumatoid arthritis therapies, Felson said. He also agreed with a 12-week minimum duration for trials to show symptom improvement. Other panelists commented that 12 weeks would only be appropriate for a nonsteroidal anti-inflammatory drug but that longer trials should be required for other kinds of agents.

Under the ACR definition, "patients in order to improve will have to have experienced a 20% improvement in their tender joint count and a 20% [improvement] in swollen joint count," Felson explained. In addition, patients must have a 20% improvement in three of the following five indices: patient-assessed pain, patient global assessment, patient-reported disability, physician global assessment, and either Westgren erythrocyte sedimentation rate (ESR) or C-reactive protein (CRP) level.

Pfizer Director of Experimental Medicine Bruce Littman, MD, who spoke from the audience, questioned the requirement for improvement in composite indices, noting that with certain drugs "you may never achieve significant effects in overall indices."

Littman maintained that the criteria are based on drugs that treat overall disease and might not be useful for targeted drug therapy. "Nowadays, you pick drugs on the basis of a molecular target [and] it is similar in the biological world...You might end up with an agent that only affects a certain spectrum of these signs and symptoms," Littman said.

Woodcock responded: "I don't think there is an intent in the document to limit this, strictly indices, but to include indices to treat the signs and symptoms overall." Agents that are more targeted to one particular aspect of the disease will receive "more specific claims," Woodcock suggested. Felson contended that preliminary data from studies of biological agents "suggest a spectrum of response that would work fine with indices."

David Scott, MD, Kings Healthcare, Dulwich Hospital, London, discussed the guidance criteria for prevention of structural damage or radiographic progression claims. Scott concurred with the guidance recommendation for measurement of retardation of X-ray progression using the Larsen or modified Sharp score. Concurrent clinical endpoints that correlate with Larsen score include tender and swollen joints, while patient physical and global assessments and acute-phase reactant values do not seem to correlate highly, Scott noted.

Scott said that highly effective drugs should be able to show significant difference from another agent over 12 months by showing less of an increase in Larsen or Sharp score, or fewer new erosions, or more non-progressive cases. However, weakly effective drugs may not be able to show significance due to high dropout rates and patients who stay on therapy might have less progressive rheumatoid arthritis than those who withdraw. He recommended comparisons of six-month and 12-month X-rays for changes in progression.

Scott further recommended that patient retention rate for X-ray progression studies be 75%-80% instead of 85% proposed by the agency in the guidance, since a 15% to 20% dropout rate is common. BRI International's Frank Hurley, PhD, maintained that 85% is "unobtainable" for a one-year study and recommended a completers analysis. Woodcock responded that "when we refer to 85%, we are talking about number of patients that would get X-rays, not the number of patients remaining on treatment."

Prevention of disability claims may be conducted in postmarketing studies, suggested Strand. "Observational long-term cohort may be the way we get to this claim. This claim may be a postmarketing claim after activity has already been established." Panel member Theodore Pincus, MD, Vanderbilt University, said: "If you can maintain functional status over five years in a patient...I think that would be the basis of a claim." Pincus pointed out that any new drug for prevention of disability would have to be tested in combination with methotrexate, the current standard of therapy.

Panel member Strand made several recommendations for patient enrollment and design of RA studies. She suggested that Phase I trials enroll patients with active disease who have failed current therapies and that Phase II look at patients with earlier disease to maximize treatment effect of the test product. Strand strongly recommended placebo control and assessment of the safety of combination therapy in Phase II. Phase III trials should compare the drug alone or in combination to current therapies, Strand said.