By Steven J. Mento, PhD., president and CEO, Conatus Pharmaceuticals
There are several general principles that can help accelerate clinical development, beginning and ending with a focus on getting the drug to the patients that need it.
CHOOSING THE RIGHT PATIENT POPULATION
One major obstacle in clinical development is that the requirements of the drug approval process can only approximate the actual practice of medicine. Researchers usually are trying to determine if results observed in a specific model — often a preclinical laboratory model with a genetically defined mouse — can be duplicated in human patients. Regulators typically want drug developers to focus on one particular category of patient. To prove efficacy, well-controlled clinical trials often are conducted in a specific subgroup of the total potentially treatable patient population. Although clearly defined inclusion criteria provide an appropriate scientific context in which to measure effectiveness of the drug, those criteria define a subgroup that may represent only a small portion of the actual patient population that a physician would typically treat. Determining how to match the controlled population of a clinical trial to the broad patient population faced by physicians is challenging.
For example, there can be various causes — alcohol, viral infections, or obesity — that lead to liver damage and cirrhosis. Although these causes, or insults, lead to similar outcomes, disease progression, prognosis, and intervention or treatment opportunities may vary with the insult. Patients who have advanced liver cirrhosis as the result of nonalcoholic steatohepatitis (NASH) represent a relatively pure patient population in which to conduct a clinical trial, but they represent only about one quarter of all the cirrhosis patients a hepatologist would see in practice. A drug developer might conduct well-controlled clinical trials in NASH cirrhosis patients leading to approval only to create a situation in which three quarters of the patients a physician would hope to treat are not included.
In oncology, as another example, not all patients with a particular cancer are diagnosed at the same stage of disease. They also may have attempted other therapies and failed, or may not have been previously treated at all. In addition, a clinical trial may require selecting a particular subset of patients based not only on stage of disease, but also on prior treatment status. In any therapeutic area, decisions have to be made to limit the scope of patients in a clinical trial from across a spectrum of disease. One useful general principle is to conduct trials in patients with more advanced stages of disease, as higher unmet medical need translates directly into shorter clinical development pathways. Patients with more advanced disease will exhibit clinically relevant progression much sooner than patients with early-stage disease. Patients treated with a drug that can delay, prevent, or reverse progression may show benefit more quickly compared with patients on placebo or standard of care.
Choosing a specific population for clinical development is an iterative process, as knowledge gained in the process will feed back into decision making. Collecting data from trials across diverse patient categories helps during discussions with regulatory authorities regarding appropriate populations for future trials and also the breadth of the drug label. Conducting initial signal-seeking studies may help identify the right development path. By conducting trials on a group that is broadly representative of what a physician would typically treat, a developer can gain information across a range of patient categories and identify subgroups for further development. For example, with liver disease, studies conducted in patients with a mix of disease etiologies — patients with alcoholic liver disease, NASH, hepatitis C, autoimmune disease affecting the liver — could identify characteristics of patient subgroups in which the effects of a drug are particularly evident. Following these relatively short-term trials, the appropriate disease stage or etiology for the next step in clinical development can be more clearly defined.
CHOOSING THE RIGHT ENDPOINTS
Another approach that can streamline the pathway to approval is taking advantage of surrogate endpoints, which are indirect measurements likely to predict clinical outcomes. For example, in the early days of HIV drug development, researchers tried to prevent the disease state of AIDS. Over time, with the improvement of assays, drugs were developed to decrease viral load and improve CD4 levels. An increased CD4 level or a reduction in viral load is not in itself a measure of clinical benefit, but higher CD4 levels and lower viral loads have been established as predictive of clinical benefit, and are more convenient to measure in a clinical trial.
In recent years, authorities have opened the door to additional applications of surrogate endpoints, especially in diseases with high unmet need and limited treatment options. For example, portal hypertension (high blood pressure in the vein delivering blood into the liver) can lead to bleeding, fluid accumulation, and decline in brain function. Most physicians agree that persistent portal hypertension will lead to negative clinical consequences for the patient. Portal hypertension is a surrogate endpoint that may be suitable to support approval.
Demonstrating an effect on a surrogate endpoint rather than clinical consequences may be one way to shorten the pathway to approval. Surrogate endpoints that are not fully established may require an extension study that takes place after marketing approval to show that the changes in the surrogate endpoint translate into bonafide positive clinical outcomes. It is also prudent to reach prior agreement with regulators on appropriate surrogate endpoints for a specific indication and how those endpoints should be measured.
DISTRIBUTING RISK BY CONDUCTING MULTIPLE TRIALS
Another way to increase the likelihood of approval is to distribute risk by generating the information needed for approval across multiple small trials rather than in a single large trial. In general, demonstrating safety requires a trial with a substantially larger number of patients than a trial of sufficient size to demonstrate efficacy. However, safety data can be cumulative across multiple trials. Distribution of risk may be less important in an indication like high blood pressure, with multiple drugs available and an established development pathway. But in a higher risk area where there are no approved treatments, it may be most efficient to conduct multiple trials in parallel. Since there can be significant diversity in stage of disease and cause of a disease, it can be challenging to select the single best patient population subgroup for demonstrating efficacy. Instead, it can be advantageous to conduct multiple trials in different subgroups using different endpoints. Only one successful efficacy outcome may be needed to move forward. By conducting a series of trials, we can gather a broad, diverse patient population very relevant for safety while focusing on tightly defined patient populations and tailored efficacy endpoints in smaller trials.
A series of trials not only distributes risk but also allows exploration of potential benefits. Multiple trials may provide an understanding of a drug’s activity across a broader spectrum of patients outside of the disease category listed on the initial label. Depending on the strength of the trial data, smaller parallel trials can show similar trends and open possibilities for broadening the label.
INCLUDING THE PATIENT’S PERSPECTIVE
Developers are increasingly recognizing the importance of incorporating the patient voice in the design of clinical trials, which also can help speed development. While there may be a standard perception of the amount of risk that is appropriate in a disease area, in certain cases, patients may be willing to accept a different risk standard in order to treat their conditions. This is an important consideration in a development plan for any area. Patient quality-of-life measurements are important and should be considered as potentially relevant for any trial. There has been a recent shift among regulatory authorities to include quality-of-life measurements as a key element in the approval process, as opposed to just an aside.
Careful consideration of patient selection and patient needs, along with general strategies like distribution of risk and identification of surrogate endpoints, are all likely to help unearth the most efficient development pathway in any particular disease area.