Abstract
Pulmonary hypertension has enjoyed a great deal of interest from the pharmaceutical industry in recent years. Unfortunately, this is now waning, because the challenge of addressing the disease is perceived to be large, and the size of the market is perceived to be small. For new drugs, there is still the attraction of asking a premium price after licensing. But with a high attrition rate, the risks of bringing a new chemical entity to the market are still great. 1 So how can academia help?
One approach is to explore drug repurposing, in which a biochemical pathway newly associated with the disease can be targeted pharmacologically by a drug licensed for another indication. The risks are lower for repurposed drugs, because there will be data on safety over the dose range used for its original indication. There remains uncertainty, because the drug may not demonstrate efficacy, and lack of efficacy is still the main cause of failure in clinical trials. 1
Identifying and pursuing drug targets in which there is a high degree of confidence is a key challenge. An already validated drug target (e.g., endothelin receptors and phosphodiesterase type 5) is “low-hanging fruit.” Although there may be benefits in terms of pharmacokinetics, developing different chemical entities that address the same receptor or enzyme does not advance treatment in major steps. An alternative powerful indicator of a valid target is genetics. A gene with a causative link to the disease defines a molecule or pathway that merits investigation for discovering new therapies. It also acts as a biomarker, defining the subpopulation that might benefit most from the therapy.
Whatever the target and therapy, we need to be smarter about the early steps in exploring the value of a new therapeutic agent. The current model for drug development does not work well, particularly when testing a number of treatments for an orphan disease. Much has been said and published about the limitations of animal models in selecting new therapies for pulmonary hypertension. Animal studies have their place, but their imperfect representation of human pathology mandates that those first studies in patients should be careful experiments designed to capture signals that enable early go/no-go decisions on whether to pursue the drug or move on to another.
This is not the philosophy of current practice, in which drugs proceed through a sequence of phases of development, gradually increasing patient numbers and culminating in a study based on entry criteria chosen to facilitate the rapid recruitment of a sizeable patient population, and in which success or failure depends on a single end point. When the drug fails, it is often unclear why, and this is usually because no attempt was made to collect the relevant data during the trial.
We have a duty to make every effort to capture as much information as we can about each patient drug exposure to better understand the benefits and risks on the basis of mechanistic data and, at the very least, to use the pharmacological response to better understand the disease; pharmacology is a very powerful tool for dissecting disease mechanisms. This is the value of a good experimental medicine study. Experimental medicine occupies a space at the interface of bench-to-bedside and is a two-way process. In drug development, it seeks to gather information about whether a molecule has potential as a medicine but also to gather data as the pharmacological agent probes the disease. A positive signal suggests that the biochemical pathway under interrogation participates in the pathology, as hypothesized, whereas a negative signal might point to the contrary and suggest that we return to the laboratory and try again.
What does a good experimental medicine study look like? From the point of view of interpreting drug response, it has to answer the following 3 primary questions. 2
Does the drug reach the tissue of interest? More specifically, is the tissue exposed to the drug at an appropriate concentration for a time period suitable for the drug to have an effect?
How does the drug interact with the target? For example, what is the percentage receptor occupancy, or does the drug alter activity?
What does the dose-response relationship look like? For example, can a safe dose be identified that has a significant effect on a biomarker linked to the mechanism of action?
From the viewpoint of understanding the disease, it is important to gather as much data as possible on patients recruited to studies, particularly data at the molecular level (e.g., genomics, proteomics, and metabolomics). This is a plea for deep phenotyping. Matching drug response to the molecular characteristics of the patients offers deeper insight into the pathology of the disease and lays the basis for stratified or more personalized medicine.
An experimental study design will be more complex and demanding than a traditional phase I/phase IIa study, but it is data rich, requires fewer patients for a confident signal to continue development or stop and look elsewhere, and will inform that next move, whether through the design of the next study (e.g., dose selection, patient selection, and biomarker) or through indicating why a drug has failed (e.g., because of pharmacokinetics or because it did not significantly affecting target activity). This should reduce the unnecessary exposure of patients to drugs that will not work for them and allow those patients to enroll in other studies.
Experimental medicine as so described requires close collaboration between academia and the pharmaceutical industry. It also requires engaging patients in the discussion. This is now the direction of travel for many diseases, and pulmonary hypertension should not miss out.