 |
|
|
 |
Is the STAR Trial Really a Safety Trial?
To the Editor: The results of the STAR trial (Safety Trial of Adalimumab in Rheumatoid Arthritis) by Furst and colleagues1 are of much interest to the rheumatology community. However, I think they merit further discussion. It is surprising the authors designate this well designed efficacy study a safety study. In the 1999 study of Moreland, et al2, a tumor necrosis factor-a (TNF-a) antagonist, etanercept, was shown to be superior to placebo among patients with rheumatoid arthritis (RA) who had poor response to traditional disease modifying antirheumatic drugs (DMARD) in a trial in which the failing DMARD had been discontinued at entry. The STAR trial now shows, very usefully, that the addition of a TNF-a antagonist, this time adalimumab, to traditional DMARD is superior to the addition of placebo, again among a group of patients who had poor response to traditional DMARD therapy. There are also some important issues related to the interpretation of the safety data. (1) The authors say that "A sample size of 300 patients per group was determined to demonstrate a specific adverse effect of 1%, or less, with a 95% confidence." This statistical reasoning is valid only if no adverse effects are observed3. It only puts an upper confidence limit to a zero observation and is not of much use to search for statistically significant differences in the frequencies of rare events between the 2 arms of a drug study. (2) The frequencies of the most feared side effects of the TNF-a antagonists, i.e., lymphoma, serious infections, and demyelinating disease, appear, fortunately, to be relatively small. Thus, thousands of patients need to be studied for properly powered clinical trials to probe the real safety of a new agent like adalimumab, and it is probably best to depend on postmarketing surveillance data to appreciate the real importance of these side effects in clinical practice. It is disappointing not to find any discussion of this phenomenon in the STAR report. (3) The authors use the annual incidence rates of serious infections in the general population of patients with RA as a comparator for the infections they observed in their study group. As suggested for lymphomas4, this appears unjustified. As with lymphomas, the onset of tuberculosis5 — and for that matter of demyelinating disease, as well6 — occurs, in many patients, only within months of use of TNF-a antagonist. To use the presumably random annual incidence among a group not using TNF-a antagonists as a comparator unjustifiably dilutes the frequency, and thus the real importance of the drug side effects. It is timely and healthy to question the real value of the controlled clinical trial as the ultimate guide for evidence based practice of medicine7,8. Sometimes, however, it is not the "controlled clinical trial," but its interpretation that needs to be put on trial. HASAN YAZICI, MD, Division of Rheumatology, Department of Medicine, Cerrahpasa Medical Faculty, University of Istanbul, Aksaray, 34303 Istanbul, Turkey. E-mail: hyazici@attglobal.net REFERENCES 1. Furst DE, Schiff MH, Fleischmann RM, et al. Adalimumab, a fully human anti-tumor necrosis factor-a monoclonal antibody, and concomitant standard antirheumatic therapy for the treatment of rheumatoid arthritis: results of STAR (Safety Trial of Adalimumab in Rheumatoid Arthritis). J Rheumatol 2003;30:2563-71. 2. Moreland LW, Schiff MH, Baumgartner SW, et al. Etanercept therapy in rheumatoid arthritis. A randomized, controlled trial. Ann Intern Med 1999;130:478-86. 3. Yazici H, Biyikli M, van der Linden S, Schouten HJ. The 'zero patient' design to compare the prevalences of rare diseases. Rheumatology Oxford 2001;40:121-2. 4. Yazici H. Lymphomas after tumor necrosis factor antagonist therapy: comment on the article by Brown, et al. [letter]. Arthritis Rheum 2003;48:2389. 5. Gomez-Reino JJ, Carmona L, Valverde VR, et al. Treatment of rheumatoid arthritis with tumor necrosis factor inhibitors may predispose to significant increase in tuberculosis risk: a multicenter active-surveillance report. Arthritis Rheum 2003;48:2122-7. 6. Mohan N, Edwards ET, Cupps TR, et al. Demyelination occurring during anti-tumor necrosis factor alpha therapy for inflammatory arthritides. Arthritis Rheum 2001;44:2862-9. 7. Dieppe P, Szebenyi B. Evidence based rheumatology [editorial]. J Rheumatol 2000;27:4-7. 8. Pincus T. Guidelines for monitoring of methotrexate: "evidence-based medicine" outside of clinical trials. Arthritis Rheum 2003;48:2706-9.
Dr. Furst replies To the Editor: I very much appreciate the thoughtful comments of Dr. Yazici. The study was conducted to obtain "real-world" safety data (similar to typical clinical practice). This data would include adalimumab with 0–3 or more disease modifying antirheumatic drugs (DMARD), nonsteroidal antiinflammatory drugs, or corticosteroids. The lenient entry criteria allowed for a more heterogeneous patient population and did not allow for true efficacy to be determined because of the small size of many groups. In that context, the sample size of 300 patients per group did allow us to determine at least one adverse event in a particular organ system of interest, with a 1% incidence at 95% confidence limit, and this would serve as a signal for further testing (i.e., precisely as Dr. Yazici commented—a confidence limit for zero observations). And, in agreement with Dr. Yazici, the study was not powered or intended to obtain an incidence of rare events. It was simply to see if relatively common adverse events occurred in a population of adalimumab treated patients taking various DMARD. The number of adverse events expected relates only to relatively common events, not rare events such as serious infections, lymphomas, or demyelinating disease. It is also for this reason that very few statistical conclusions were made regarding adverse events. The efficacy will be further analyzed in an exploratory analysis in the future. Thus, we believe our interpretation and discussion of this trial are appropriate to its stated purposes and it was designed appropriately for those purposes. DANIEL E. FURST, MD, Division of Rheumatology, Department of Medicine, UCLA, Rehabilitation Center, Room 32–59, 1000 Veteran Avenue, Los Angeles, California 90095-1670, USA
|