Practice parameters and focusing research: Plasma exchange for myasthenia gravis

The American Academy of Neurology (AAN) recently published an evidence-based assessment of plasma exchange for treatment of myasthenic crisis or for preparation for thymectomy and concluded, ‘‘Because of the lack of randomized controlled studies with masked outcomes, there is insufficient evidence to support or refute the efficacy of plasmapheresis....’’ Many experts in the care of patients with myasthenia gravis (MG) expressed shock at the assessment, and the MG Foundation of America, the largest patient advocacy group in the USA, received comments of concern and confusion from patients. We appreciate the AAN’s efforts in providing such reviews and agree that rigorous assessments are important and useful in stimulating research. In fact, the 2000 practice parameter that evaluated the role of thymectomy for treatment of MG motivated clinical investigators throughout the world, including two of the authors (G.C. and H.J.K.), to help develop one of the few randomized comparative effectiveness trials in history of a surgical procedure to evaluate whether thymectomy added to corticosteroid treatment can benefit patients with generalized MG. However, we believe the process of formulating practice parameters for rare disorders retrospectively should be considered carefully. Although we specifically consider MG, our discussion can be generalized to other diseases. Our focus is on three points: (1) We disagree with the contention that, in the absence of controlled trials (Level U), an intervention cannot be assessed as to whether it does or does not work. (2) Further, practice parameters need to take into account the reality of performing randomized trials during a time of increasingly limited resources. (3) Despite the hope that the AAN parameters will not be interpreted as ‘‘negative recommendations,’’ they can be. In the case of plasma exchange, there is worldwide consensus that the therapy is rapidly effective for myasthenic exacerbation, and websites of the National Institutes of Health, Muscular Dystrophy Association, and numerous academic medical centers list plasma exchange as accepted therapy. The preponderance of clinical experience and published observation rises to a level that argues against the practical possibility of performing a randomized trial, which includes a placebo (shamexchange group). Many clinicians and patients would view such a trial design to be unethical for patients in myasthenic crisis, because treatment failure could mean death. We believe this point should have been made in the choice of parameters used to judge treatments. Although we recognize that further investigation is needed to improve treatment of MG, a randomized, shamcontrolled trial of plasma exchange, while ideal from an academic perspective to satisfy Class 1 evidence criteria, would: (a) not be in the best interests of patients; (b) be virtually impossible to conduct; and (c) likely place other important therapy questions in the field of MG on hold. We do not argue that such a trial would lack value. The practice parameter concluded that evidence for the application of plasma exchange for myasthenic crisis and treatment prior to thymectomy merits a Level U recommendation. We agree that there is a lack of placebo-controlled trials of plasma exchange in MG. For MG, and many orphan diseases, randomized, placebo-controlled studies are difficult to design and perform for numerous reasons such as: ethical concerns based on standards of care; lack of federal funding for the long-term nature of many of the clinically relevant outcomes; absence of economic incentives for industry sponsorship; poor infrastructure for performance of trials; and the challenge of coordinating clinical trials Correspondence to: H. J. Kaminski; e-mail: hkaminsk@slu.edu