The resources required to bring a new drug to market are as well-known as they are staggering: upwards of $1.3 billion in costs, 10 to 15 years from target selection to approval, and an overall failure rate approaching 95%. No wonder pharmaceutical companies remain intrigued by the prospect of finding new uses for established small-molecule drugs—both those already on the market and those that have been abandoned after early-stage clinical trials demonstrated adequate safety and bioavailability profiles, but insufficient efficacy or other limitations.

Whether it’s called drug repositioning, repurposing, reprofiling, or rescuing, the process is based on the fact that drugs often interact with multiple targets, producing effects that can be bad or good. If a side effect is desirable, the drug might be repurposed for an additional indication. A classic example is thalidomide, which was withdrawn from the market in 1961 after causing thousands of severe birth defects. Later, scientists demonstrated its antiangiogenic and immunomodulatory properties, and the Food and Drug Administration (FDA) in 1998 approved Celgene Corp.’s Thalomid to treat leprosy and in 2006 for multiple myeloma.

In addition, drug targets are often associated with other diseases or indications. Merck & Co.’s Propecia (finasteride) was originally marketed to treat prostate enlargement by blocking the enzyme that converts testosterone to dihydrotestosterone. This method of action was later found to also prevent male pattern baldness, a secondary indication that has far surpassed the original.

Saving time and money
Repurposing an existing drug can save developers years of time and almost 40% of the cost of bringing a drug to market by eliminating the need for additional toxicological and pharmacokinetic assessments.1 The question is how to uncover promising new indications systematically. “Repositioning is not really a new idea,” says Andrew Reaume, PhD, MBA, president and chief executive of Melior Discovery Inc. (Exton, Pa.). “The concept was floated in the early 1990s, but rapidly gained momentum in the post-genomic era.”

Some of the companies involved in those early efforts included Sosei Pharmaceuticals (Tokyo), Gene Logic Inc. (now Ore Pharmaceuticals Inc., Cambridge, Mass.), and CombinatoRx Inc. (now Zalicus Inc., Cambridge, Mass.). Initial approaches involved gene expression databases and drug annotation technologies but have since evolved to encompass a range of often-proprietary computational and other approaches. These include bioinformatics, high-throughput screening using cell-based disease models, gene activity mapping using shared molecular profile databases, ligand-based chemoinformational approaches, and in vivo experimentation using animal models of diseases, among others.

Big Pharma’s perspective
Most large drug companies pursue some sort of drug repurposing activity, either formal or ad hoc. Those that have dedicated resources to repurposing include the New Indications Discovery Unit at the Novartis Institute for Biomedical Sciences, Bayer Healthcare’s Common Mechanism Research, and Pfizer’s Indications Discovery Unit (IDU). “Our mission is to find new uses, new indications for Pfizer’s development pipeline—compounds that either failed in the past or compounds that are active now,” says Steven L. Roberds, PhD, an IDU research project leader. “We’re trying to be opportunistic. We can’t change the compounds that we have; they’re the ones that Pfizer has already produced that have good properties, high affinity against their targets, and have been into human beings” through clinical trials, he told a seminar at the University of Pennsylvania Medical School last year.

Established in 2007 and headquartered in St. Louis, Pfizer’s IDU is a “virtual” unit that involves staff scientists in collaborations with outside academic investigators. Pfizer last year signed a $22.5-million agreement with Washington University School of Medicine in St. Louis, giving academic scientists access to proprietary data on drug candidates that are or have been in clinical testing. The goal is to leverage the university’s scientific expertise to identify new indications in such areas as Alzheimer’s, cancer, diabetes, and other metabolic disorders.

Pfizer is no stranger to repurposing. The company’s blockbuster Viagra (sildenafil citrate), a phosphodiesterase-5 (PDE-5) inhibitor, was developed to treat pulmonary arterial hypertension. In 1998, the company won approval for and began marketing the drug for treating erectile dysfunction, with PDE-5 being the same molecular target responsible for its famous side effect.

In search of similar discoveries, Pfizer and other drug companies are working with about two dozen smaller biotech and biopharmaceutical companies and contract research organizations hoping to hit upon promising new targets for old molecules. “Usually, informatics groups will be involved that will analyze targets and pathways for off-target effects that have been measured, and then will map these against potential indications that may be of interest to the company,” says Aris Persidis, PhD, president and cofounder of Biovista Inc. (Charlottesville, Va.) of these biotech and CRO collaborations.

Leaving no stone unturned
Biovista’s platform is called Clinical Outcome Search Space (COSS). The system searches the mechanism-of-action data of about 95,000 drugs and pharmacologically active compounds in the public domain against similar data for about 23,000 indications, 6,000 adverse events, and 20,000 human targets in the biomedical literature. “We leave no stone unturned. We miss practically nothing,” says Andreas Persidis, PhD, Biovista’s chief executive and cofounder of the company with his brother, Aris.

Biovista has its own repositioning pipeline, including drugs that have shown biological plausibility for a number of diseases and for which there is no prior art in the patent or scientific literature linking them to the new indications. These include drugs for eye disorders, diabetes and obesity, oncology, and other diseases.

The COSS platform has attracted the attention of Pfizer and Novartis, both of which have contracted with Biovista to identify new indications for their drugs. Biovista has also licensed the adverse event analysis portion of the platform to the FDA’s Office of Clinical Pharmacology. The agency is using the technology to study serious side effects of drugs and drug classes in order to help promote safer use of those medications.

Another company that employs systematic search methodologies is NuMedii (Menlo Park, Calif.). Founded in 2008, NuMedii relies on translational bioinformatics technology developed in the laboratory of Atul Butte, MD, PhD, assistant professor of medicine at Stanford University School of Medicine. NuMedii’s New Indications Discovery platform includes a database of genome-wide molecular profiles for more than 300 diseases along with comprehensive drug efficacy information. The system compares molecular profiles of different diseases, looking for similarities.

“In a lot of instances, a disease with no therapies would be markedly similar to a disease with a lot of therapies,” says Atul, NuMedii’s cofounder and chair of its scientific advisory board. “This encouraged us to think about how to share therapies across diseases that were newly discovered to be similar to each other from their expression profiles.” This has led to the discovery of two drugs that show promising results in animal models, he says. NuMedii is in the process of establishing partnerships with pharmaceutical companies, adds Gini Deshpande, PhD, the company’s other cofounder.

Systematic serendipity
Melior Discovery takes an entirely different approach, one that cofounder Andrew Reaume calls “systematic serendipity.” It eschews the so-called hypotheses-driven method of searching knowledgebases of molecular targets against known biochemical pathways and associations with disease states. Rather, Melior tests promising compounds in up to 52 in-vivo animal models of diseases to detect potential efficacy. “The collective knowledge base is very incomplete,” Reaume says. “We are not trying to follow a trail of breadcrumbs to get us somewhere when there may not even be any breadcrumbs.” The company has a 30% success rate in identifying clinical candidates from screened drugs, Reaume says. In addition to Pfizer and Merck, the company has also done work for Johnson & Johnson and Cephalon Inc. Melior’s own lead candidate is MLR 1023, a potential Type 2 diabetes treatment acquired from Pfizer now entering Phase 2 trials.

Companies involved in repurposing can be found worldwide, and range from very large database vendors such as GeneGo/Thomson Reuters (San Diego), to smaller biotechs, such as Odyssey Thera Inc. (San Ramon, Calif.), which concentrates on cell signaling; Horizon Discovery Ltd. (Cambridge, U.K.), a maker of translational genomics drug and diagnostic assays; In-Silico Biosciences Inc. (Lexington, Mass.), which develops computational models of neurological diseases; and Verva Pharmaceuticals Ltd. (Melbourne, Australia), which focuses on metabolic diseases.

Problems with IP 
Some large drug companies are not completely convinced of the value of drug repurposing. This is particularly true when drugs have gone off-patent and generics are available. While the FDA will give marketing exclusivity for a new indication, there is little to prevent physicians from prescribing a generic version in its place. “Highlighting new uses for off-patent drugs may be exciting, but it is a challenge for companies to get enough value out of it to fund clinical trials,” says Butte. “It’s an unanswered question right now.”

The National Institutes of Health (NIH) wants to help find answers. Earlier this year the NIH’s Chemical Genomics Center (NCGC) opened its Pharmaceutical Collection database for public screening of nearly 27,000 active pharmaceutical ingredients, including 2,750 approved small-molecule drugs and all compounds registered for human clinical trials.2 In April, NIH invited representatives from industry, academia, and government to discuss ways to make repurposing more practical and less burdensome. NIH hopes that drug companies will make their small-molecule libraries available to the agency and to academic collaborators for further discovery. Possible mechanisms discussed at the April roundtable included voluntary pooling of off-patent compounds and donating abandoned and stalled molecules to the common pool.

As a result of this initiative, NIH will be launching “a comprehensive effort to identify appropriate abandoned compounds, establish master agreements, match partners, make data resources available, and provide a central access point to relevant resources and expertise,” said NIH Director Francis S. Collins, MD, PhD in a written commentary after the meeting.3 The pharmaceutical industry’s willingness to participate remains to be seen.

About the Author
Contributing editor Ted Agres, MBA, is a veteran science writer in Washington, DC. He writes frequently about the policy, politics, and business aspects of life sciences.

References
1. DiMasi JA, Hansen RW, Grabowski HG. The price of innovation; new estimates of drug development costs. J Health Econ. 2003; 22(2):151-85.
2. Huang R, et al. The NCGC pharmaceutical collection: a comprehensive resource of clinically approved drugs enabling repurposing and chemical genomics. Sci Transl Med. 2011; 3(80):80ps16.
3. Collins FS. Mining for therapeutic gold. Nature Rev Drug Disc. 2011; 10(6):395.