The rise of innovative clinical trial designs: what’s in it for amyotrophic lateral sclerosis?

Abstract

The randomized, placebo-controlled trial (RCT) has long been recognized as the gold standard for assessing therapeutic benefit. The randomization ensures that both the observed and unobserved prognostic variables are balanced across treatment arms. Many regulatory agencies require, therefore, RCTs to obtain an unbiased estimate of a drug’s risk–benefit ratio. The conduct of a RCT demands, however, a considerable amount of resources and time. The median costs of phase II and III clinical trials are $8.6 million and $21.4 million, respectively (1); costs that are likely to further increase in the near future (2). These excessive expenditures are especially problematic for rare diseases such as amyotrophic lateral sclerosis (ALS), where funding, awareness and resources could be limited. Therefore, there is a need to improve the efficiency of clinical trial design while maintaining scientific rigor. In this issue of the Journal, Babu et al. (3) deviate from the classical two-armed RCT by applying a ranking and selection paradigm in order to determine which of three treatment strategies (i.e., creatine, tamoxifen 40mg or tamoxifen 80mg) is the most promising for a future pivotal trial. Notwithstanding the recent advancements in preclinical and early clinical models, phase II clinical trials are often the first opportunity for investigators to obtain a hint of therapeutic benefit. Improving the phase II selection process of promising ALS drugs is, therefore, essential, especially when one considers the expanding pipeline of ALS compounds (4). The selection paradigm as proposed by Babu et al. may give preliminary insights in the efficacy and safety of new drugs. As patients can only be allocated to active arms, enrollment rates could be positively affected. The lack of a concurrent placebo arm is, however, also its major drawback. This diminishes the interpretation of the safety data and may provide insufficient insight in the therapeutic benefit. In the end, this could lead to the erroneous selection of a treatment. Further refinement of the proposed methodology may help to ameliorate these limitations. In the last decades, there has been a steady increase in the use of innovative or adaptive design elements in RCTs (5), deviating from the fixed sample size, fixed eligibility criteria and fixed allocation ratios as seen in classical trial design. Adaptive designs use the accruing information during a RCT to make interim decisions regarding, for example, the continuation of subgroups, the re-estimation of sample sizes or to stop a trial early. These design adaptation may help to improve the efficiency of clinical trials when there is sufficient evidence for either superiority or futility. Especially the futility adaptations could prove value for ALS clinical trials (6). The probability that a compound will successfully pass the different phases of drug development, including regulatory approval, is less than 10% for rare diseases (7). In case of ALS, the success probability is even less than 5%: two FDA approvals (i.e., riluzole and edaravone) out of more than 60 tested compounds. Thus, in order to optimize the use and allocation of resources in ALS, the first selection question may not be “Which compound holds the most promise?”, but should rather be “Which compound holds no promise?”. The latter question is the primary aim of adaptive multi-arm, multi-stage (MAMS) clinical trials.