Several trends reported in evolution of randomized controlled trials in oncology

Issue: November 25, 2008

During three decades, randomized controlled trials have dramatically shifted toward industry sponsorship.

Randomized controlled trials in oncology today are larger and more likely to be sponsored by industry than similar trials conducted in the previous decades.

Researchers from the Princess Margaret Hospital examined changes in design, outcomes and sponsorship of randomized controlled trials in breast (48%), colorectal (24%) and non-smell cell lung cancer (28%) published between 1975 and 2004. All trials were published in one of six major journals.

“We observed several important trends in this review of 321 randomized controlled trials involving more than 170,000 patients with cancer,” the researchers wrote in Journal of Clinical Oncology.

During three decades, the researchers reported increases in the number and size of randomized controlled oncology trials. Today’s trials are typically larger and more likely to be multicenter and international, they wrote.

In addition, randomized controlled trials are more likely to be sponsored by industry. For-profit and/or mixed sponsorship increased from 4% to 57% during the study period (P<.001). Further, the authors of contemporary randomized controlled trials are more likely to endorse therapies in the experimental arm (P=.017).

“There has been a dramatic shift in sponsorship of randomized controlled trials in oncology from government to for-profit organizations. Although sponsorship status is not associated with increased effect size, industry-funded randomized controlled trials are more likely to strongly endorse novel treatments,” the researchers wrote.

An independent association was reported between a significant P value for the primary endpoint and authors’ endorsement of therapies used in the experimental arm (OR=19.6; 95% CI, 8.9-43.1). This association was also found between industry sponsorship and authors’ endorsements of the experimental therapies (OR=3.5; 95% CI, 1.6-7.5).

Use of a time-to-event as a primary endpoint increased from 39% to 78%, while use of a response rate endpoint decreased from 54% to 14% (P<.001). Effect size remained stable during the study period, according to the researchers.

“Our data provide information that may be of use in the reporting and interpretation of contemporary cancer clinical trials. The conclusions drawn from a randomized controlled trial can be influenced substantially by the quality of methodology and reporting,” they wrote. – by Katie Kalvaitis

J Clin Oncol. 2008;doi:10.1200/JCO.2008.16.5456.

There are some findings reported in this study that definitely are not surprising. The authors report that over the last three decades more randomized controlled trials — of breast, colorectal and non-small cell lung cancers — are funded by industry rather than non-profit organizations. With the expense of these large trials, the ability to obtain sponsorship in the non-profit or governmental sector is becoming much more difficult. Therefore, industry sponsorship is to be expected. The authors of this study wonder whether interpretation of the benefit of these trials is somehow biased by industry sponsorship. What they found over the last three decades is that the predictor of having a strong endorsement about the experimental arm is independently associated with a significant P value or whether the study was sponsored by industry. Whether that is because these are increasingly larger studies with increased power to show a positive effect or if there is some other reason, we just don’t know. The bottom line is that even though we hold randomized controlled trials as the gold standard, I think we still need to understand how individual studies are done, how well they are performed, and why they are being done and not just go by P values. The devil is definitely in the details of the study, not just simply looking at a bottom line P value.

– Samuel Silver, MD, PhD

HemOnc Today Editorial Board member