Could fresh human tissues be the missing link for stratified medicine? By Dr David Bunton
The drugs don’t work (sometimes)
Leukotriene antagonists are extremely successful drugs for the treatment of asthma. By successful, I mean that they work in most patients; in fact, a significant proportion of patients (perhaps as many as one-third) fail to respond sufficiently to gain a clear clinical benefit (such patients are often termed “non-responders”)1, 2.
It turns out the same is true for many other drugs used to treat asthma. It’s also true of drugs used to treat most conditions. For many patients, drug treatment is simply ineffective, and this only becomes apparent after a period of unsuccessful drug therapy. Wasn’t personalised medicine supposed to change all this?
The problem with pharmacogenomics: too much genomics, not enough pharmacology
It has been 12 years since the term “personalised medicine” was coined following the mapping of the human genome; however, the expected flow of personalised drugs has not yet materialised. For some drugs, where a single mutation predicts “response” or “no response”, then pharmacogenomics follows a clearly defined path. Benefits arise for both the pharmaceutical company (by providing evidence of a clinical benefit and potentially allowing the design of less expensive clinical trials that exclude non-responders) and patients (greater certainty that the drug will effectively treat their condition)- but what about the vast majority of conditions where a single gene fails to predict drug responses?
For this type of drug, multiple genes, or more likely, one or more “continuous” biomarkers (e.g. serum levels of a hormone rather than a yes/no genetic test), are required to predict who will turn out to be a responder to the drug. It is interesting to consider how the development process might be altered to gain an early understanding of efficacy in such a target patient population. What if the efficacy of a drug in different patients could be tested before clinical trials and could run in tandem with the development/validation of biomarkers?
Predicting clinical endpoints using fresh, functional human tissues
Functional demonstrations of human efficacy can be made using fresh, intact human tissues obtained from patients with the relevant clinical condition, in other words, a mini clinical trial can be conducted in a lab before the actual clinical trial. The predictive value of fresh tissues is maintained by 1) using the tissue as rapidly as possible 2) minimising any changes to the tissue (thereby retaining its predictive value of the actual patient responses) and 3) by measuring end-points that translate to the clinical goals of the drug.
For example, the end-points for a drug to treat asthma might be bronchodilatation (relaxation of isolated human airways) and anti-inflammatory activity (reduction in cytokine production from cultured sections of intact human airways).
The inter-individual responses of patients to the test drug can be further examined in relation to the production of biomarkers by the tissue or presence of biomarkers in blood from the same patient that the tissue was obtained from. By freezing the tissues and measuring biomarkers or gene expression in tissues from the same patients, the missing link between biomarkers, genomics and clinical efficacy is filled.
Why this matters: cost savings and added commercial value
Two key benefits emerge from this approach. Firstly, improved clinical trial designs (smaller patient numbers) and secondly, an increased chance of demonstrating efficacy during phase II trials.
In most clinical trials, it is the average safety and efficacy of a drug that is measured (an “all-comers” approach).Testing in a subpopulation of patients who have a greater likelihood of response to the drug changes the economics of the clinical development process. Even though the ultimate market size may be reduced (by only treating responders and not all-comers), the savings from a smaller clinical trial (fewer patients needed to demonstrate benefit above placebo), together with the concomitant reduction in risk of failure at phase II and III (greater certainty of clinical efficacy) means that the net present value prior to entering development may in fact be increased. Clearly, the decision to embrace a stratified approach is dependent on the level of confidence in predictive biomarkers. Fresh human tissues form part of a platform of evidence of the likely clinical effectiveness of the drug.
Paul et al.3 calculated that based on an average cost to market for a NCE of $1.78bn, a reduction in phase II attrition from 66% to 50% would reduce the cost to around $1.28bn. Three factors are particularly important in driving economic value for stratified medicines4:
- Relative therapeutic performance of the drug (measure in vitro efficacy in fresh human tissue)
- The prevalence of the biomarker (measure release of the biomarker from the tissue or in the matched blood)
- Clinical performance of the biomarker (compare the release of the biomarker to the in vitro measure of efficacy)
Assays in fresh, functional human tissues provide:
- Increased evidence of clinical efficacy at an early stage
- Input into stratified medicine approaches
- Increase net present value of compounds by increased ability to predict clinical success
- Wenzel SE. Antileukotriene drugs in the management of asthma. JAMA 1998; 280:2068–9.
- Drazen JM, Israel E, O’Byrne PM. Treatment of asthma with drugs modifying the leukotriene pathway. N Engl J Med 1999; 340:197–206.
- Paul, S.M. et al. How to improve R&D productivity: the pharmaceutical industry’s grand challenge. Nature Reviews Drug Discovery 2008, 8, 203-214.
- Trusheim, M.R. et al. Quantifying factors for the success of stratified medicine. Nature Reviews Drug Discovery 2011; 10, 817-833.