How can we fix the clinical trial system?
In recent months, there have been quite a few articles on the need to fix the clinical trial system. Among the most recent articles is the one by Boston-based Nature writer Heidi Ledford, Ph.D. published as a News Feature in the 29 September issue of Nature. In my humble opinion, this is the best article on the subject among those that have been published recently.
The pharmaceutical/biotechnology industry is frustrated with the increasing expense and the low output of the clinical trial system. This low productivity is economically unsustainable. The current clinical trial paradigm is over 50 years old. Back in the 1960s, the norm was to conduct single trials at single sites, each designed to answer a single question.
Nowadays, the norm is the large, multicenter clinical trial, especially for Phase 3 trials. “Multicenter” means that a trial is conducted at multiple sites, often in different countries, and could involve thousands of investigators and staff members. As pointed out in Dr.Ledford’s article, the large trials are mandated by the need in our more risk-adverse world to detect safety issues that occur in only a small percentage of patients, and to obtain good statistics for drugs that confer only a small benefit over the standard of care. However, certain major diseases require large trials of long duration even for drugs that may confer large benefits. For example, because the percentage of patients per year in cardiovascular disease (CVD) trials who experience cardiovascular events is small, these trials must be large and multiyear, in order to see any benefit even for a breakthrough drug.
The advent of personalized medicine–developing drugs and combinations of drugs that are specific for the molecular mechanism behind a patient’s disease–has put additional burdens on the clinical trial system. A disease may be found to be a collection of rare diseases in terms of mechanistic subtypes, each of which affects only a small number of people. This makes patient recruitment difficult.
As stated by Dr.Ledford, “Solving the problem may require fundamental changes to the clinical-trial system to make it faster, cheaper, more adaptable and more in tune with modern molecular medicine.”
Don’t use an “e-commerce” approach to determining drug efficacy!
Other commentators have recently noted the need to make clinical trials “faster, cheaper, and more adaptable.” Several of them have suggested bringing in strategies from other industries, especially e-commerce and social media.
For example, in an editorial published in the 23 September issue of Science, Andrew Grove, the former Chief Executive Officer of Intel, proposes moving towards an “e-trial” system, based on such large-scale e-commerce platforms as that of Amazon.com. Under the proposed e-trial system, the FDA would ensure safety only, not efficacy, and would continue to regulate Phase 1 trials. Once Phase 1 trials have been successfully completed, patients would be able to obtain a new drug through qualified physicians.
Patients’ responses to a drug would be stored in a database, along with their medical histories. There would be measures to protect a patient’s identity, and the database would be accessible to qualified medical researchers as a “commons.” The response of any patient or group of patients to a drug or treatment could then be tracked and compared to those of others in the database who were treated in a different manner or were untreated. These comparisons would provide insights into a drug’s efficacy, and how individuals or subgroups (perhaps defined in part via biomarkers) respond to the drug. This would liberate clinical trials from the “tyranny of the average” that characterize most trials today. As the database grows over time, analysis of the data would also provide information needed for postmarketing studies and comparative effectiveness studies.
Dr. Grove’s proposal is one of several in which the mandate of the FDA (and regulatory agencies in Europe, Japan, etc.) is to regulate safety only (via Phase 1 clinical trials) not efficacy. Efficacy is then determined via some sort of open system, with information gathered and provided to patients and physicians electronically, via systems reminiscent of e-commerce or social media.
We are opposed to removing efficacy from the oversight of the FDA and other regulatory agencies. There are two reasons for this, both of which are illustrated graphically in Box 1 of Dr. Ledford’s article, entitled “the clinical trial cliff”. Approximately half of Phase 2 clinical trials between 2008 and 2010 failed due to inability to demonstrate efficacy. (Around one-third of Phase 2 failures were due to safety, and the remaining failures were mainly due to strategic decisions to terminate a drug.) Among Phase 3 failures between 2007 and 2010, around two-thirds were due to efficacy, and around one-quarter were due to safety. These results indicate that the majority of drugs entered into clinical trials lack efficacy.
The second reason is that many safety problems–especially the rarer safety issues that occur in only a small percentage of patients–are typically not detected in Phase 1, but in Phase 3 and even the postmarking period.
Reduce clinical attrition with new trial designs and improved animal models
Dr. Ledford’s proposals for fixing clinical trials leave regulatory agencies in charge of overseeing both safety and efficacy. They mainly focus on improving clinical trials by reducing “attrition”–i.e., failure of drugs in the clinic, especially in Phase 2 and Phase 3, and on improving patient recruitment. Haberman Associates has produced publications–as well as articles on this blog–during the 2009-2011 period that provide a more in-depth discussion of strategies for reducing attrition than is possible in a 3-page article such as Dr. Ledford’s.
Two of Dr. Ledford’s strategies involve modifications of clinical trial design. Both of these are discussed in Chapter 6 of our book-length Cambridge Healthtech Institute (CHI) Insight Pharma Report, Approaches to Reducing Phase II Attrition. The first is the “Phase 0” trial. This is a type of pre-Phase 1 clinical trial, which uses microdoses of a drug to assess such parameters as pharmacokinetics and target occupancy. As Dr. Ledford suggests, in some cases Phase 0 trials can reduce or eliminate pharmacological testing in animals, and allow researchers to get human data more quickly.
The other trial design strategy mentioned in Dr, Ledford’s article is the use of adaptive clinical trials. This type of trial allows researchers to change the course of a trial in response to trial results. For example, this may mean assigning new patients to specific doses, changing the numbers of patients assigned to each arm of a trial, and changes in hypotheses or endpoints. Monitoring and changing the trial is typically done by an independent data monitoring committee [DMC] so that ideally, double-blind conditions are maintained.
As Dr. Ledford states, adaptive clinical trials may result in shortening the time and cost of the clinical trial process. But, as with Phase 0 microdosing trials, there are many controversies surrounding adaptive clinical trials. Both of these strategies are works in progress.
The other strategy for reducing attrition discussed in Dr. Ledford’s article is to use improved animal models (i.e., animal models designed to more faithfully model human disease) in preclinical studies. We discussed this strategy in Approaches to Reducing Phase II Attrition, and in greater detail in another book-length report, Animal Models for Therapeutic Strategies. I also recently led the workshop “Developing Improved Animal Models in Oncology and CNS Diseases to Increase Drug Discovery and Development Capabilities” at Hanson Wade’s 2011 World Drug Targets Summit.
Several articles on our Biopharmconsortium Blog also focus on improved animal models for predicting efficacy of drug candidates in discovery research and in preclinical studies. Our April 15, 2010 blog post, based on an article in The Scientist, focused on “co-clinical mouse/human trials”. This type of clinical trial was developed by Pier Paolo Pandolfi, MD, PhD (Director, Cancer and Genetics Program, Beth Israel-Deaconess Medical Center Cancer Center and the Dana-Farber/Harvard Cancer Center) and his colleagues.
These trials utilize genetically engineered transgenic mouse strains that have genetic changes that mimic those found in specific human cancers. These mouse models spontaneous develop cancers that resemble the corresponding human cancers. In the co-clinical mouse/human trials, researchers simultaneous treat a genetically engineered mouse model and patients with tumors that exhibit the same set of genetic changes with the same experimental targeted drugs. The goal is to determine to what extent the mouse models are predictive of patient response to therapeutic agents, and of tumor progression and survival. The studies may thus result in validated mouse models that are more predictive of drug efficacy than the currently standard xenograft models.
The new Ledford Nature article discusses co-clinical trials as a means to develop more predictive animal model studies–not only using improved, potentially more predictive animal models, but also treating these animals in similar way (in terms of doses, formulations, schedules of medication, etc.) to the humans in the parallel human clinical trial.
The Ledford article mentions the animal-model portion of a co-clinical trial, which was published in January 2011. This trial utilized two genetically-engineered PDGF (platelet-derived growth factor)-driven mouse models of the brain tumor glioblastoma multiforme (GBM), one of which has an intact PTEN gene and the other of which is PTEN deficient.
Unlike the “standard” mouse xenograft models, these models more closely mimicked the human disease, including growth of tumors within the brain, not subcutaneously. Thus any drug administered to these mice systemically (e.g., intraperitoneally, as was done in this study) had to cross the blood-brain barrier (BBB), as in the case of human clinical trials. This would not be the case with a standard xenograft model, which is one deficiency of these models for brain tumors such as GBM.
GBM is both the most common and the most malignant primary brain tumor in adults. It has a poor prognosis. PDGF-driven GBMs, which results from deregulation of the PDGF receptor (PDGFR) or overexpression of PDGF, account for about 25-30% of human GBMs. These mutations result in the activation of the phosphatidylinositol 3-kinase (PI3K)/Akt/mammalian target of rapamycin (mTOR) pathway. These tumors may also exhibit mutation or loss of heterozygosity of the tumor suppressor PTEN, which also upregulates the PI3K/Akt/mTOR pathway.
The researchers tested the Akt inhibitor perifosine (Keryx Biopharmaceuticals, an alkylphospholipid) and the mTOR inhibitor CCI-779 (temsirolimus; Pfizer’s Torisel; originally developed by Wyeth prior to the Pfizer merger and approved for treatment of renal cell carcinoma), both alone and in combination, in vitro and in vivo. Specifically, the drugs and drug combinations were tested in cultured primary glioma cell cultures derived from the PTEN-null and PTEN-intact mouse PDGF-driven GBM models, and in the animal models themselves.
The studies showed that both in vitro and in vivo, the most effective inhibition of Akt and mTOR activity in both PTEN-intact and PTEN-null cells or animals was achieved by using both inhibitors in combination. In vivo, the decreased Akt and mTOR signaling seen in mice treated with the combination therapy correlated with decreased tumor cell proliferation and increased cell death; these changes were independent of PTEN status. The co-clinical animal study also suggested new ways of screening GBM patients for inclusion in clinical trials of treatment with perifosine and/or CCI-779.
According to Dr. Ledford’s Nature article, the National Cancer Institute (NCI) invested $4.2 million in Dr. Pandolfi’s co-clinical trials in prostate and lung cancer in 2009. In addition to the co-clinical trials with genetically-engineered mouse models run by Dr. Pandolfi and others, researchers at the Jackson Laboratory are conducting co-clinical trials with mouse xenograft models that receive tumor cells from patients to be treated in human clinical trials.
Use patient registries in recruitment of patients for clinical trials
In Dr, Ledford’s article, she discusses a crucial factor other than clinical attrition that hinders progress in conducting clinical trials–patient recruitment. According to the article, at least 90% of trials are extended by at least six weeks because of failure to enroll patients on schedule. Only about one-third of the sites involved in a typical multicenter trial manage to enroll the expected number of patients. As a result, clinical trials are longer and more expensive, and some of them are never completed.
Personalized medicine, in which researchers use biomarkers or other criteria to determine what fraction of patients with a particular disease are eligible for a trial (e.g., cancer patients with an activating mutation in a kinase that is the target of the drug to be tested), makes recruitment harder. That is because researchers must screen large numbers of patients to identify the fraction of patients that would be eligible for the trial. So they need to recruit (and screen) a much larger number of patients than in conventional clinical trials with no patient stratification.
Therefore, researchers, “disease organizations”, and patient advocates are devising new strategies to facilitate recruitment of eligible volunteers. Dr. Ledford cites the example of the Alpha-1 Foundation (Miami, Florida), a “disease organization” that focuses on the familial disease alpha-1 antitrypsin deficiency. (This disease renders patients susceptible to lung and liver diseases.) This foundation has created a registry of patients with alpha-1 antitrypsin deficiency who are willing to be contacted about and to participate in clinical trials.
There are also cancer registries. Dr. Ledford mentions the Total Cancer Care program run by the Moffitt Cancer Center (Tampa, Florida). This program, which involves 18 hospitals, compiles medical history, tissue samples (stored for future analysis) and genetic information about each patient’s tumor. Patients can consent to doctors contacting them about trials. There are other similar programs being developed in the Netherlands and elsewhere. Dr.Ledford mentions the difficulty in negotiating agreements between institutions, and the need for adequate, ultra-secure networks to support registries that connect multiple hospitals and research centers.
Patient registries that are designed to proactively support recruitment for clinical trials have some resemblance to a “social media” approach to recruitment. However, there is a big difference–the need to secure the privacy of patient records. The current trend in social media (and in some e-commerce platforms) is anti-privacy. This is yet another important reason why a social media or e-commerce approach to clinical trials or other aspects of biotech/pharma R&D is not a suitable model. (To his credit, Dr. Grove mentions the need to maintain patient privacy and confidentiality. But this is not the norm with e-commerce and social media.)
Cutting red tape for faster and cheaper clinical trials
Dr Ledford also mentions ways to deal with more bureaucratic issues that can slow down or block the progress of clinical trials. The NCI is now initiating a data-management system that will standardize data entry across all 2,000 sites that conduct NCI-sponsored trials. This should help reduce costs and cut down on record-keeping errors and omissions.The FDA is also looking into ways to reduce reporting requirements and paperwork. so that investigators can submit summaries of case reports rather than each individual document.
To adapt to the multicenter nature of clinical trials, the US Office for Human Research Protections (Rockville, Maryland), which oversees NIH-funded human studies, has proposed changes to its guidelines that would require designation of a single review board for each project. This may greatly improve the current situation, in which multicenter trials must get approval from each center’s institutional review board. This can take months or even years. Despite the definite advantages of more centralized review, individual research centers may be reluctant to give up their direct oversight of clinical trials.
Something important was not in Dr. Ledford’s article
The space limitations for Dr. Ledford’s “News Feature” article, plus its strict focus on clinical trials per se, did not permit her to include something of crucial importance to reduce clinical attrition. That is utilizing such strategies as biology-driven drug discovery in the research phase of drug development. These strategies are designed to select the best targets and to discover drugs that are more likely to be efficacious in treating a particular group of patients. These research strategies are then coupled with early development strategies that emphasize designing clinical trials aimed at obtaining rapid proof of concept in humans. Such trials typically involve the use (and often the discovery) of biomarkers.
We discussed these issues extensively in our report, Approaches to Reducing Phase II Attrition, as well as in an article published in Genetic Engineering and Biotechnology News (and available on our website) “Overcoming Phase II Attrition Problem“. We also discussed a specific case of the use of this strategy in our October 25, 2010 article on this blog.
Given the low productivity of pharmaceutical R&D, it is tempting to take an envious look at the success of e-commerce and social media, and to attempt to devise strategies that apply methodologies from these industry sectors to the biotech/pharmaceutical industry. We should remember, however, that not so long ago some pharmaceutical executives attempted to apply methodologies from such industries as aerospace, computer hardware, and the auto industry to pharma R&D. Not only did that not work too well for the pharmaceutical industry, but as we all know, the industries that served as a model for these approaches haven’t done very well in recent years either.
In contrast, pharmaceutical and biotechnology companies that have formulated strategies that embrace the uniqueness of biology, such as Novartis and Genentech (the latter now merged with Roche), have done a lot better.
There are other strategies for making clinical trials faster, cheaper, and better that are now under discussion in the biotech/pharma industry and the FDA. These strategies are based on clinical experience, not e-commerce. We shall discuss them in further blog posts.
As the producers of this blog, and as consultants to the biotechnology and pharmaceutical industry, Haberman Associates would like to hear from you. If you are in a biotech or pharmaceutical company, and would like a 15-20-minute, no-obligation telephone discussion of issues raised by this or other blog articles, or of other issues that are important to your company, please click here. We also welcome your comments on this or any other article on this blog.