The most common problem with clinical trials is the lack of them - particularly good, randomized, double-blinded, controlled studies. There are often few data for dose refinement, or even basic recommendations, in children, for example. In too many instances, the efficacy of many current - and often very expensive - therapeutic options has simply not been demonstrated in controlled, randomized trials. Sadly, there are only a handful of clinical trials currently for GBS, MG, and CIDP. With a couple of hundred current trials, the treatment horizon is more hopeful for MS and SLE patients.
Other problems arise from design flaws - sometimes unavoidable ones. For ethical reasons, placebo control treatment arms are often lacking when some other treatment has become the mainstay of therapy (e.g., corticosteroids in MG). Another flaw is that although the therapeutic aim is to improve both short- and long-term prognoses, clinical trials often examine only short-term time points.
Further, scales used to measure disease severity or progression are imperfect. One needs a standardized scale that can be used to provide a quantitative measure of a patient's clinical status and course or stage of progression. Scales currently in use have limitations - the Kurtzke Expanded Disability Status Scale (EDSS) and the Multiple Sclerosis Functional Composite score used in MS, for example.
Several patient recruitment issues can negatively impact trials. Disease classification criteria are imperfect, and diagnoses are often flawed - leading to inappropriate inclusion in or exclusion from clinical trials. Racial, ethnic, and socioeconomic issues can influence outcome in clinical trials,1 yet are often overlooked at either the patient recruitment or analysis stages. Inclusion of patients with mild or stable disease does not allow for an effect size sufficient to show statistical differences in treatment arms. Often, only patient subsets are included in trials - for example, when, for ethical reasons, new therapies are tested on patients who have failed conventional therapies; this may not pose a problem for safety studies, but it can skew early results for efficacy trials. Together, these problems diminish the power of the conclusions that can be drawn.
Statistical power is also negatively impacted by low patient numbers, which are mandated in Phase I trials. Even when patient numbers are statistically sufficient at the outset, patient attrition inevitably occurs as a result of side effects, an inability to follow a particular treatment regimen, patient expectations not being met, or patients being allowed to switch therapies. Some measures can be taken to minimize dropout rates by, for example, permitting patients to take the highest tolerated dose instead of fixed doses, and making sure that patients are well informed at the outset and that their expectations are reasonable.
Interpretation of data is complicated in progressive and relapsing diseases by the fact that deteriorations can be due either to ongoing disease progression or the effect of the treatment. In diseases with a relapsing-remitting course, or with drugs that have a slow onset of action, conclusions are particularly difficult to draw from anecdotal observations. Finally, even in an optimal clinical trial, clinicians are faced with the question of the applicability of the overall results to a particular clinical subset or patient.
Was this article helpful?