Traditional drug development is a slow and inefficient process. By focussing on side effects, we can find unexpected new uses for existing drugs.
In 1989, Pfizer created the drug sildenafil as a treatment for cardiovascular problems. Although the drug lacked efficacy during clinical trials, it was having rather noticeable effects further down the male anatomy.
Instead of giving up on sildenafil altogether, Pfizer recognised that this side effect filled a gap in the market: up until then, there was no oral agent available for treating male impotence. Men had to use fine needles to inject drugs into their penis ‒ not very sexy, nor pleasant. Taking an oral drug for male impotence would be revolutionary.
In 1998, nine years after its creation, sildenafil gained Food and Drug Administration approval and was registered as Viagra, the first oral drug for erectile dysfunction.
Usually, we think that side effects of drugs are bad because they cause us discomfort or have unwanted effects on our body. But in drug development, as was the case with Viagra, side effects are inherently valuable and contribute to an efficient and more economical path of drug discovery.
Historically, medicines were discovered in flora and fauna by accident or through folklore and tradition. Yondelis, which is used to treat soft tissue carcinomas, exemplifies the difficulty and time taken to develop a drug from nature. It was the first anti-cancer marine drug to be approved by the European Union.
Dr Garth Maker, a biochemistry lecturer at Murdoch University's School of Veterinary and Life Sciences, has a special interest in drugs derived from sea organisms. He describes how trabectedin, the active ingredient in Yondelis, was discovered and developed.
“In 1969, the US National Cancer Institute found that the sea squirt, Ecteinascidia turbinata, possessed an anti-tumour activity, but they couldn’t isolate the active molecule," said Maker. "Fifteen years later, researchers at the University of Illinois finally purified it.”
Attempts made to farm the sea squirt using aquaculture failed because not enough of the active compound could be produced. In 1996, trabectedin was made synthetically at Harvard University. It entered the European market as an anti-cancer agent in 2007 (24 years after it was first isolated) and the US market in 2015.
Yondelis took 24 years and Viagra took nine ‒ is there are faster way to get a drug onto the market? Yes, there is.
Recently, drug discovery has started to shift from biotechnology companies synthesising novel molecules as putative active pharmaceutical ingredients, towards the repositioning of extant drugs. The pharmacopoeias (the “bibles” of drug specification and identification) offer a veritable feast of potential candidates, especially when explored alongside data repositories such as SIDER (Side Effect Resource) and pharmacovigilence reports.
Pharmacovigilance, or drug surveillance, occurs after a drug has entered the market. Drug regulators collect reports of adverse drug reactions from many sources: patients, the medical fraternity, carers, research and clinical studies, the pharmaceutical industry’s post-marketing observations, and articles in magazines, medical journals and the internet. In addition, drug sponsors in Europe, Japan and the USA (ICH countries) must also submit a Periodic Benefit-Risk Evaluation Report describing drug efficacy, patterns of use risk, clinical trial results and more. All of these reports are collated and stored for future reference.
The SIDER database offers an alternative resource to record and research side effects. It was created in 2010 by a team of German researchers who were looking at phenotypic side effects as a way to identify the target sites of different drugs. The researchers suggested that side effects would be useful to investigate new uses for marketed drugs.
But how useful is the information from the SIDER and pharmacovigilance databases? Data-mining these resources reveals a wealth of information, including causality and the ability to assess risk-benefit ratios. In other words, existing data offer a cost-effective and efficient option for drug development, unlike the traditional trial-and-error method of seeking and trialling novel molecules.
Since registered drugs have already undergone safety and toxicity tests, and their side effects have been documented, they offer major savings in time and money. Researchers can trawl the data for potential new drug applications. There are several ways to approach this.
Some researchers use in silico studies to assess a drug’s potential for use in other applications. Using computer simulation, virtual screening rapidly assesses thousands of compounds without a test-tube in sight. Interactions between a drug and biological targets, such as receptors, enzymes or nucleic acids, are predicted and scored by a computer, based on the drug’s structure and its ligands (functional sites).
A more targeted method involves uploading as many pre-existing data as possible, then letting a computer sort it out. This approach is used by Biovista in the US, who reposition drugs for specific diseases. Their technology platforms correlate masses of biomedical data to identify potential new drug applications in a fraction of the usual time. They estimate that “drug repositioning accelerates development by 15 to 20%”.
An alternative systematic approach is to explore phenotypic side effects for potential new therapeutic applications. For example, if a drug’s side effect is a fall in blood pressure, then the drug could be repositioned as an anti-hypertensive agent. This method is gaining favour and offers tangible potential — in fact, it has been patented.
As for Viagra, one of its side effects is manifest, unsurprisingly, on the cardiovascular system as hypotension. So we come full circle. Today, sildenafil is also marketed as Revatio for its original intended purpose: a cardiovascular treatment for pulmonary arterial hypertension.
Old-style drug research and development undertaken by pharmaceutical companies is long, expensive and gambles with failure. The future of drug discovery offers faster, more efficient and more economical outcomes. Bioinformatics and mining pre-existing data have the undeniable advantages of cost and time, putting a vast amount of drug, disease and side effect information at our disposal. Computer programs make rapid predictions that would in the past have taken years.
Even with intellectual property and legal issues present as obstacles, this path of drug discovery is good news. But perhaps even more exciting is the feasibility of treating rare diseases and conditions with repositioned drugs.
Edited by Andrew Katsis and Ellie Michaelides, and supported by James Garth.