Using Adaptive Approaches in Oncology Studies

The setting

Clinical trial sponsors often find it difficult to enroll adequate numbers of subjects, required to fulfill sample size requirements in confirmatory studies, especially in the most serious disease areas. Many studies involving rarer cancer sub-groups pose the greatest challenges with regard to patient recruitment.

Adaptive trial is a new breed of clinical study that has been shown to be especially suitable for oncology studies. For sponsors, such adaptive methods are providing a much more favorable setting to continue with their studies that otherwise would not have been possible with conventional approaches.

Why adapt?

Adaptive trials are different from conventional, fixed studies, and enable predetermined alterations to be made to an ongoing study on the basis of interim analysis. There are many forms of design adaptations including phase 3 designs that increase the possibility of study success by re-estimating the sample size, i.e., the patient population.

Adaptive sample size re-estima­tion approaches are approved by FDA and EMEA and they improve the size of the samples without affecting the statistical validity and integrity of the study. While it can be a costly prospect to treat large patient populations with an acute disease, re-evaluating patient needs on the basis of interim analysis can make it viable for sponsors to run even late-stage studies confidently.

Why interim analysis?

An interim analysis, with lack of overwhelming efficacy, can accomplish a Promising Zone that can be defined as conditional power from 30% to 80%.

  • If there are no results in the Promising Zone, the study should be continued without an increase in sample size
  • If within the Promising Zone, the sample size should be increased to obtain a conditional power of 80%

Defining the “Promising Zone”

Regulatory considerations

  • Trials based on this concept have been accepted by the FDA and EMEA. Regardless of the potential increase in sample size, suitable methods are used to maintain trial statistical validity.
  • A separate data monitoring committee (DMC) is given a comprehensive charter that also includes predetermined rules for increase in sample size, as well as the flexibility to cancel the same in case of extraordinary situations such as unexpected safety problems.
  • The DMC alone has access to interim results: the sponsor continues to be blinded; thus the disclosure of premature study data does not affect the trial’s integrity.

The Cytel advantage

Cytel was chosen by sponsors to both design and assist with the implementation of new trial methods that raise the chances of success, regardless of difficult situations. Cytel was also selected for qualified independent statisticians to serve on DMCs and to appear at trial design regulatory review meetings with the sponsor representatives.

Ethical advantages

By nature, adaptive trials also provide unique benefits to patients when compared to conventional, fixed clinical trials. For example, in any adaptive dose finding trial, there is a greater probability that the patient will receive a meaningful dose of medicine as opposed to conventional studies that do not provide any probability of moving from an ineffective to a more effective dose group. The possible adaptive advantage is especially meaningful in therapeutic studies focused on severe diseases, including cancer, and could even be life-saving.


If interim analysis results enter the so-called Promising Zone, then engaging the sample size increase can significantly improve the chances of success. If the interim result does not enter the Promising Zone, then the sponsor neither benefits nor loses than with a traditional study design (without a reassessment option of sample size). So, instead of committing to the larger sample size up front, the sponsor can wait until the study shows enough proof that the cost of enrolling more patients is advisable, based on a likely successful result.

Last updated: Dec 14, 2017 at 11:41 AM

Other White Papers by this Supplier