National Cancer Institute NCI Cancer Bulletin: A Trusted Source for Cancer Research News
June 16, 2009 • Volume 6 / Number 12

The information and links on this page are no longer being updated and are provided for reference purposes only.

A Closer Look

A Sense of Urgency: Rethinking the Clinical Trial Development Process

Low enrollment of adult cancer patients in clinical trials is an ongoing challenge in cancer research. Only about 3 to 5 percent of adults diagnosed with cancer enroll in a clinical trial. A new analysis of federally funded cancer clinical trials sheds light on a barrier to clinical trial enrollment that has until now been invisible: delays caused by the sheer complexity of getting a trial up and running.

"The development time of a clinical trial can be used as an early indicator of accrual success," said Dr. Steven Cheng at the Knight Cancer Institute of Oregon Health & Science University (OHSU), who presented the study findings May 31 at the 2009 American Society of Clinical Oncology (ASCO) Annual Meeting in Orlando. Dr. Cheng and his coauthors reviewed all cancer treatment trials for adult patients sponsored and developed by NCI's Cancer Therapy Evaluation Program (CTEP) between 2000 and 2007.

Trials that went from concept to study activation between 9 to 12 months were significantly more likely to reach their accrual goals than trials with a longer activation time. Trials with the longest activation time (more than 27 months) were the least likely to achieve their accrual goals. Strikingly, 40 percent of more than 550 trials examined fell short of their accrual goals.

Trials that don't meet their accrual goals may close without answering the scientific questions they were designed to address. Failed trials may delay getting potentially beneficial new cancer treatments to patients.

"Resources spent on clinical trials that do not enroll sufficient patients are misappropriated because the underlying scientific objectives are not met," said Dr. Cheng. "One must also consider the ethical implications of enrolling patients onto trials that have a diminished likelihood of achieving success."

Previous work by principal investigator Dr. David Dilts, formerly of Vanderbilt University and now with OHSU's Knight Cancer Institute, has shown that the activation time for a phase III trial sponsored by an NCI-supported cooperative group can range from 10 months to more than 4 years.

"Opening a clinical trial involves hundreds of steps, dozens of decision points, and frequent 'process loops'—points where a change occurs, sending a study protocol back to a previous step," said Dr. Dilts. "Every aspect of the trial is reviewed multiple times and approved by multiple entities.

"People want to do the best science," he added. "So they tweak the protocol to make it better—but every time they do that, it adds months to the process. Those months significantly decrease the likelihood that the study will accrue patients."

This graph shows the relative odds that a clinical trial with the indicated development time will meet its accrual goals. The dotted line indicates the median development time. Points (triangles) above the dotted line indicate greater success in meeting accrual goals; points below the line indicate less success. Trials with a development time of 9 to 12 months were significantly more likely to achieve their accrual goals, while those exceeding 27 months were significantly less likely to achieve their accrual goals. (Source: Cheng et al., 2009) This graph shows the relative odds that a clinical trial with the indicated development time will meet its accrual goals. The dotted line indicates the median development time. Points (triangles) above the dotted line indicate greater success in meeting accrual goals; points below the line indicate less success. Trials with a development time of 9 to 12 months were significantly more likely to achieve their accrual goals, while those exceeding 27 months were significantly less likely to achieve their accrual goals. (Source: Cheng et al., 2009)

A Question of Complexity

Scientific advances have increased the complexity of cancer clinical trials, said Dr. James Doroshow, director of NCI's Division of Cancer Treatment and Diagnosis. For example, many trials now include imaging studies or studies of genetic biomarkers, which provide valuable information about why a treatment works, doesn't work, or works only in some patients. "The most informative trials are often the ones that incorporate these studies, yet doing so adds to the complexity of trial development," he said.

Improving operational effectiveness was a major focus of NCI's Clinical Trials Working Group. Dr. Doroshow is a co-chair of the Operational Efficiency Working Group (OEWG), a task force that NCI Director Dr. John E. Niederhuber charged with devising strategies to reduce by half the activation time for clinical trials conducted by the cooperative groups and NCI-supported cancer centers.

One of the issues being considered by the OEWG is how to restructure financial incentives so that clinicians are able to make clinical trials a higher priority. At the moment clinical work is considered revenue-generating but clinical trials are not.

NCI's network of clinical trial activities developed over many years to meet the evolving needs of patients and clinical investigators, said Dr. Doroshow. "No one ever said, 'Let's make this a harmonized process.' That's our goal now—to create a clinical trials system that is efficient, streamlined, and coordinated."

NCI's Dr. Jeffrey Abrams, associate director of CTEP, agreed and said, "Dr. Dilts' data have driven consensus around the need for major changes in the clinical trials approval process." He also noted that "another challenge confronting NCI-supported Cooperative Groups in this tight funding environment is that increased efficiency may require significant downsizing in the number of trials conducted, which in turn could negatively impact the overall clinical research effort.

Streamlining the Process

Dr. Dilts, an OEWG member whose background is in operations management research, advocates applying process improvement strategies used by companies such as Toyota Motor Corporation and Southwest Airlines to clinical trial development.

"To improve a system's efficiency you first have to study it," he said. "You keep asking 'why?' until you identify the fundamental problem slowing the system down. How many activities are not adding value? Do you really need to have a pre-meeting meeting, a post-meeting meeting, and a post-meeting phone call? Why not get everyone in a room, decide what you're going to do, and be done in 2 hours?"

Finding ways to speed up decision-making is another strategy. Perhaps there should be firm deadlines for submitting clinical trial protocols as there are for R01s and other grant applications, suggested Dr. Dilts. "What if [NCI] said, 'We will fund your trial provided that you submit the final protocol within 90 days?' People generally perform when they are given a deadline.

In his new job as director of clinical research at the Knight Cancer Institute, Dr. Dilts has the opportunity to put his approach to the test. "My role is to create a model for conducting clinical trials efficiently, safely, and effectively," he said.

The Institute's director, Dr. Brian Druker, says his goal is to reduce within 2 years the activation time for a clinical trial to less than 90 days. "We need to create a sense of urgency," he said. "As a clinician I'm tired of talking to patients about a promising new agent, and 6 months later I still can't offer them a trial of that agent because it isn't open yet. Patients need new therapies today.

"It's absolutely clear there are numerous steps [in the development of a trial] that don't add to either its ethical conduct or its scientific validity," he continued. "If we can eliminate or streamline those activities, we can get trials open more quickly and accelerate progress against cancer."

Eleanor Mayfield