For as long as clinical trials have existed, traditional procedures have been the to-go way to conduct a study. That is, the whole process that normally lasts 10+ years and costs about $2.5 billion dollars is pre-determined by a protocol – basically, a guideline stating in details what do to and how to do in order to get the desired results.
However, such conventional approach is causing problems due to its rigidity and at times realistic inapplicability. To untangle this dilemma, adaptive clinical trials have been brought into view.
The phrase is self-explanatory. Rather than strictly following one single outline, researchers can modify some aspects of the study based on the interim analysis of data previously generated through such study. These changes can be in terms of trial procedures such as treatment procedure, study dose, study endpoints, diagnosis procedures, or statistical procedures including sample size, randomization, study design, statistical analysis plan and/or methods of analysis.
Of course, these alterations are not made out of the blue. Permission for such modifications must be granted prior to study execution by sponsors and contract research organizations (CROs) alike, and thus must be clearly pre-specified in the protocol. For example, study dose and sample size are open for changes if highly recommended by analyzed data, yet study design should always stay the same.
By allowing more flexibility in the process, clinical trials with adaptive design are proving to be more economical and efficient. In the traditional approach, data are collected step by step before being analyzed after a specific amount of study duration, normally one to two years. However, with the adaptive clinical trial model, a great amount of time and money are saved through constant effectiveness evaluation and immediate modifications towards more beneficial directions.
Patients also receive increased benefits. For instance, if a type of medicine shows apparent perks towards a disease, researchers can decide to increase the sample size so that the studied drug can get to the hand of more patients in need. Especially in case of cancer, this adaptation serves undeniably well.
However, the reason companies are still somewhat hesitant to adaptive clinical trials is its inherent cons. Cutting treatment arms that respond poorly to the regimen can eliminate the extreme cases, or badly influence the generalizability of the study. In a broader sense, altering the study can create data bias.
Next comes the concern for patient safety and equality. According to Donald Berry, a biostatistician at the University of Texas MD Anderson Cancer Center in Houston, “faster drug access, at its extreme, can jeopardize patient safety.” This argument not without foundation, because a focus on a group of patients may mean adverse effect on other unstudied groups.
In a survey carried out in 2015 by University of Michigan, biostatisticians expressed their concern about inequality towards patients in an adaptive clinical trial. Specifically, patients who participate in the earlier stage receive less benefit than those who come later due to modifications with regards to dosage. This ethical matter attracts public attention, since if there is a conflict of interest among those beneficiaries the trial will no longer be purposeful.
Yet in the face of disadvantages, clinical trials are witnessing a shift to this model in CROs. Highly aware of the aforementioned problems, related entities are taking careful consideration of each modification they make so as to minimize risks and maximize study efficiency. Meanwhile, the U.S Food and Drug Administration (FDA) encourages the use of adaptive clinical trials supported by customizable and configurable electronic data capture (EDC) system. In his speech, while running for the position of FDA’s head, Dr. Scott Gottlieb – the current Commissioner of FDA, showed his enthusiasm to promote the wider use of this approach.
With all the support, future opportunities are open for adaptive clinical trials. As a result, technical solutions such as eClinical tools should be flexible and customizable in order to better serve clinical studies.