Smarter Trial Design Saves Money and Produces Better Drugs

Get Permission

William Douglas Figg, Sr, PharmD

We have to have drugs that have proven benefit for patients, especially given their huge costs to our system. The more drugs that get to market simply because they show some activity, the more difficult it becomes to get meaningful endpoints, especially overall survival.

—William Douglas Figg, Sr, PharmD

The process of identifying a promising molecule and moving it from the laboratory through the highly complex series of clinical trials necessary to garner U.S. Food and Drug Administration (FDA) approval is a costly scientific gauntlet during which many new agents fail. New trial design, pharmacology principles, and emerging biomarkers will play a large role in ensuring the future of a robust pipeline of antineoplastic agents. The ASCO Post recently spoke with William Douglas Figg, Sr, PharmD, Deputy Chief, Genitourinary Malignancies Branch, National Cancer Institute (NCI). Dr. Figg elucidated the role of clinical trial design, addressed the need for clinically meaningful endpoints, and offered his expertise on other important issues in oncology.

Clinically Meaningful Endpoints

Some experts contend that too many cancer clinical trials are underpowered, and their endpoints, such as progression-free-survival and objective response, leave too much room for interpreting responses that do not necessarily benefit patients. What are your thoughts on this issue?

The approval of regular new drug applications is based on a demonstration of clinical benefit or an effect on an established surrogate marker. We definitely need to revisit the statistical design of trials and ask the question of whether the size and scope of the study allow us to further assess endpoints in specific patient subsets.

One of the most important aspects in designing a late-phase trial is to choose a clinically meaningful endpoint for the target population in the study. Demonstrable clinical benefits vary with cancer type and status of disease. We need to understand the target agent being investigated and how it works for that particular patient population in testing and then adapt the clinical trial methodology (design) to appropriately identify early markers of clinical benefit, which may correlate with survival and allow for patient selection.

Drugs With Proven Benefit

There is a move in the community to raise the bar for a drug’s approval, in an effort to limit the amount of drugs being brought to market that offer minimal survival advantage. However, some experts contend that we need to lower the bar, which would allow more drugs on the market, thus giving researchers the ability to test more indications and drug combinations. What are your thoughts?

I would argue against lowering the approval bar. We have to have drugs that have proven benefit for patients, especially given their huge costs to our health-care system. Moreover, the more drugs that get to market simply because they show some activity, the more difficult it becomes to get meaningful endpoints, especially overall survival.

One of my main concerns is the crossover trial design, in which the investigators allow those patients who fail to respond to the control arm, ultimately receive the investigational drug. The trial then gauges its success on patients who have received standard of care plus the investigational agent. Therefore, if that investigational drug is truly active, are you ever going to get a meaningful endpoint on the trial, such as overall survival?

Adaptive Trial Design

In recent years, the use of adaptive trial design methods in clinical research and development based on accrued data has become very popular due to its flexibility and efficiency. What are your thoughts on this direction?

I was slow to embrace the adaptive trial design methods. But I now see that this method more quickly identifies drugs that have a therapeutic effect, and it zeros in on patient populations for whom the drug is appropriate. Plus, the adaptive design both lowers costs and saves time by eliminating the number of study failures; however, the adaptive design works best in exploratory phases through phase II.

It’s been shown that the adaptive trial design saves sponsors hundreds of millions of dollars. Amending protocols and more quickly shutting down one arm of the study and adjusting statistics on the fly as you see activity not only save time and money but help us better inform which agents are destined for early failure.

Improving the Preclinical Environment

Pharmacology, pharmacokinetics, and biomarkers have shown merits in oncology research, especially in preclinical drug development. Where are we in this area, and what do you see on the horizon?

With the number of late-stage clinical trial failures and the cost of drug development increasing, there is a strong need to continue to improve the preclinical environment, such as through incorporating predictive biomarkers into preclinical models. We also need to enhance rigorous methodology with improved validation strategies and proper controls and adopt more robust preclinical cancer models (eg, translation and pharmacodynamic indicators and murine models that are appropriately designed to mimic specific disease subtypes).

The idea is to prioritize and move drugs with optimal preclinical validation, so the preclinical data can be translated to success at the bedside. Simply put, the goal is to improve the trial success rate and limit the testing of what would ultimately prove to be ineffective targeted therapy.

In addition, another limitation of preclinical evaluation is the inability to accurately predict the toxicity of a particular agent, especially in combination with other chemotherapeutic agents. Another area on the horizon that should be addressed is understanding how to apply preclinical pharmacokinetic and pharmacodynamic data and evaluation to guide the phase I trial setting in areas such as dose selection and minimize toxicity. The overall objective is to make drug development an iterative process to continue to improve the accuracy of preclinical models in predicting clinical benefit.

Biomarker Signature

The screening debate in cancer is ceaseless, but detection is also part of clinical trial design. As a prostate cancer expert, do you foresee a prostate cancer–specific biomarker down the pike?

Because cancer is heterogeneous and multiple tumor pathways are altered, identifying and validating a “biomarker signature” (gene or serum proteins) would be a more appropriate global platform for detection or even prognosis. For prostate cancer, a combination of markers will improve the predictive accuracy and decrease the number of unnecessary biopsies.

In fact, we are currently seeing this explosion in genomic data and can use this information to identify disease subtypes; differentiate indolent vs aggressive prostate cancer; and, further down the line, distinguish between responders and nonresponders to treatment.

Antiangiogenic Therapy

In early work, the promise of antiangiogenic therapies was the buzz of the oncology community. Are we stalled or rebooting this intriguing way to kill tumors?

The field has clearly taken a hit, but antiangiogenic therapy still has a role in the cancer treatment paradigm for specific tumor types. A lot of these drugs work very well in showing progression-free-survival, but when it comes to overall survival, they seem to fail. I’ve said for years that it’s not rational to think that an antiangiogenic agent is going to control a tumor without help from combinations of drugs. In fact, we may need two or more antiangiogenic agents that have different mechanisms plus chemotherapy and molecular targeted  therapies to have the most activity.

As with any cancer targeting strategy, we need to fine-tune the antiangiogenic process by understanding the mechanisms of evasive resistance—for example, ongoing effects to target known resistance mechanisms—identify predictive markers of response and resistance, and develop combination therapies that can synergize with angiogenesis inhibitors.

To that end, ongoing studies are looking at combining angiogenic inhibitors with immunotherapies. So, antiangiogenic therapy holds promise, we just need to keep studying this type of therapy until we hone into more specific ways to use it. 

Challenges Ahead

As a last thought, although our research efforts have delivered drugs that have greatly improved survival, are NCI’s research capabilities ready for the challenges ahead?

The researchers are ready, but unfortunately without proper funding, our efforts will be stalled. Over the past few years, when you factor inflation into the equation, the NCI’s research funding has flatlined. It’s important to note that every single cancer drug on the market has had some help getting there from the NCI. Although we (the NCI) may not have discovered the molecule, the target that the drug hits was most likely discovered by an NCI basic science grant, or the trials were most likely conducted in cancer centers that were funded by the NCI, or the investigators conducting the trials were most likely funded by the NCI. Each year, the top two cancer drugs drive more income for their respective companies than the entire NCI budget. The NCI needs more funding. What we do is valuable, because it directly improves the lives of cancer patients. It’s that simple. ■

Disclosure: Dr. Figg reported no potential conflicts of interest.