Changes to primary endpoints in immunotherapy trials: common and untransparent

In specific cases, adjustment of a primary endpoint in studies is warranted. However, the broad scientific consensus is to exercise caution and ensure transparency when doing so. The researchers found the opposite to be true in their analysis, with transparent reporting of modifications in only 8 out of 24 studies. Although the consequences of such modifications are often difficult to determine, there are instances where the modifications resulted in approval for broader patient populations for the investigated drug. The researchers urge colleagues to be vigilant about changes to primary endpoints in studies and to critically evaluate them.

Sixty-three percent of the studies included modifications

The researchers examined the frequency of changes to primary endpoints in 38 studies. They reviewed publications, protocols and public study registrations for all comparative studies involving immunotherapy in patients with bladder, kidney and lung cancer. In the analysis, the researchers assessed three aspects of each study; the outcome measure, the patient population and - for studies with more than two study groups - between which groups the comparison was made.

In 24 of these studies (63%), an adjustment to primary endpoints was made after the study was initiated. The most common adjustments were adding overall survival as an outcome measure and adding additional comparisons in biomarker-positive patients. Adjustments were made both early in the study and after completion of inclusion and randomization. 

Positive outcome leads to approval for broader patient group

Most of the randomized trials in this analysis aimed to determine the efficacy and safety of a drug for specific cancer patients. A positive outcome often results in approval by the EMA and/or the FDA. As such, these studies form the direct input for treatment guidelines in Europe and beyond and determine the clinical practice for hundreds of thousands of patients (and their families) worldwide.

It is difficult to ascertain in what way adjustments to primary endpoints influence the final conclusions of a study. For some studies however, researchers see clear implications. In the publication, they provide two examples where the adjustments led to an approval for a broader patient group that otherwise would not have been the case. In one of the studies, the group of renal cancer patients to be analyzed was expanded from patients with a poor to moderate prognosis to include all randomized patients, including those with a good prognosis, despite the subgroup analysis for the latter category showing no favorable treatment effect. This broad group was adopted in the American and European treatment guidelines, with the result that the recommendation regarding this treatment now focuses not specifically on patients with a moderate or poor prognosis, but also on patients with a good prognosis.

I was looking for information about the reporting of a secondary endpoint of a particular trial, when I discovered that the endpoints had been modified. Since I was under the impression that this is not customary, I decided to investigate this further.

Senior researcher Dr. Anke Richters, Netherlands Comprehensive Cancer Organisation (IKNL)

Signal changes to primary endpoints and be critical

The researchers believe it is important that the best possible information is available for both caregivers and patients, to make appropriate treatment choices. It is crucial that the scientific rationale and social relevance of the chosen primary endpoints can be transparently and completely assessed by the scientific community. The changes observed in the study and the lack of transparency in particular, undermine this process. Therefore, the researchers urge authors, sponsors, peer reviewers and editors of scientific research to be more attentive to potential changes to primary endpoints, ensure transparency, and critically evaluate any modifications made during the research process.

Source

JAMA Oncology: Changes to Primary End Points in Randomized Clinical Trials on Immune Checkpoint Inhibitors in Urothelial, Renal Cell, and Lung Cancer: A Systematic Review. Anke Richters, Hilin Yildirim, Christopher M. Booth, Francisco E. Vera Badillo, Lambert A.L.M. Kiemeney, Katja K.H. Aben.

Learn more
Do you have a question about this study or are you a journalist and would you like more information about this analysis? If so, please contact our spokesperson Jolanda Sinha (j.sinha@iknl.nl or +0031 6 34 62 02 29).

 For (high.res.) images or the embed code of the interactive summary of this study, please contact senior communications advisor Irma Koorn.