Is it time to replace ACR20 as a standard for treatment efficacy?

[Guest guest]

Rheumawire

June 3, 2004

Is it time to replace ACR20 as a standard for treatment efficacy?

Philadelphia, PA and Salt Lake City, UT - The American College of

Rheumatology 20 (ACR20) can identify effective agents for treating

rheumatoid arthritis (RA) but is not very good for discriminating among

active agents. Researchers have attempted to solve this problem by using

the more stringent 50% or 70% (ACR50 or ACR70) levels of improvement,

but Dr A Albert (University of Pennsylvania School of Medicine,

Philadelphia) and colleagues argue in the May 2004 issue of the Journal

of Rheumatology that neither is adequate. Albert reports that the

average number of ACR criteria met at a 20% or 50% level discriminates

more accurately among active treatments, as does the area under the

curve (AUC) of the number of ACR criteria met [1].

" We believe that all trials should report their results in a similar

standardized fashion. Our recommendation is that the conventional ACR20,

50, and 70 be supplemented with 1 of the 2 continuous measures we

propose. . . . This approach would generate a more comprehensive picture

of the efficacy of any particular therapeutic modality, " Albert writes.

Albert and colleagues at the University of Utah (Salt Lake City)

calculated ACR20, 50, and 70 for each of 6 arms in 2 clinical trials

from the ating Clinics for the Systematic Study of Rheumatic

Disease (CSRD) database. They also calculated the average number of

criteria met for the ACR20, 50, and 70 for each arm and their respective

areas under the curve over time.

A primary concern is that the ACR20 might be neither useful nor

appropriate for trials of newer disease-modifying antirheumatic drugs

(DMARDs), such as etanercept, leflunomide, and infliximab. " The ACR20

may have limited utility in discriminating among efficacious agents, "

the researchers write. They note that the ideal outcome measures would

be specific for detection of novel agents with the potential for

therapeutic utility and would be able to discriminate between effective

agents to generate " at least a rank order of effectiveness. "

" The ACR20 failed to discriminate among active therapies; however, the

ACR50 was too stringent, and only 1 patient in these trials satisfied

the ACR70, " Albert writes.

Within the mean number of criteria fulfilled (ACR-C), they recommend

that all 3 levels (20%, 50%, and 70%) be reported. " The 20%-level

indicates the number of criteria per person showing a minimally

clinically significant improvement, 50% indicates how many criteria show

a 'good' response, and 70% shows how many criteria achieve a near

remission. Using all 3 thresholds would result in a clearer picture of a

drug's utility, " Albert says. " Identifying promising therapies is most

appropriately done with ACR20, whereas deciding between therapies may be

more appropriately done with ACR-C or area under the curve. "

" We were surprised that therapies thought to be ineffective such as

auranofin and penicillamine were clearly efficacious if a more sensitive

outcome measuring tool was used to assess them. The moral is that your

results are highly dependent on the quality of the tools used to measure

them, " Albert tells rheumawire.

In an accompanying editorial, Dr T Felson (Boston University

School of Medicine, MA) notes that the American College of Rheumatology

has established a committee to reevaluate the ACR20 and outcome measures

in RA clinical trials [2]. However, with regard to Albert's proposal,

Felson notes that it treats tender and swollen joint counts just as

other core set measures, without requiring that they improve for a

patient to be characterized as improved. " If the approach recommended by

Albert et al were to be used, a therapy could be characterized as

efficacious if it produced no improvement in swollen and tender joint

count, a scenario that most rheumatologists would reject as not having

clinical validity. "

Janis

Sources

1. Albert DA, Huang G, Dubrow G, et al. Criteria for improvement in

rheumatoid arthritis: alternatives to the American College of

Rheumatology 20. J Rheumatol 2004 May; 31(5):856-66.

2. Felson DT. Whither the ACR20? J Rheumatol 2004 May; 31(5):835-7.

I'll tell you where to go!

Mayo Clinic in Rochester

http://www.mayoclinic.org/rochester

s Hopkins Medicine

http://www.hopkinsmedicine.org