Investors are continuing to pour money into biotech companies. Dr Sarah Houlton looks at some of the developments in this marketplace that are encouraging such financial optimism
The biotech sector has changed dramatically over the past decade or so. Far from the early days of the biotech boom where investors rushed to put their cash into start-up companies in the hope of making a quick buck when the next big thing was floated on the stock market, the "exit route" investors now foresee is invariably a trade sale rather than an IPO - selling the company on to Big Pharma at a, hopefully, substantial profit.
"Most biotechs are set up for trade sale in the UK now," biotech investor Andy Richards told the recent FT Global Pharmaceutical & Biotechnology conference. "There's also a recycling of people - if they've done it before, they say they can do it again, and the venture capitalists are investing in them again. But there's no publicity - it's not necessary if there's going to be a trade sale!
"But those companies already floated on AIM and those set up to aim for it are suffering. Biologics are massively fashionable now, but that's OK as long as management recognise it."
And it's true - the pharma giants are, one way or another, strengthening their involvement in biologics. Much of this is through acquisition and in-licensing and, sure enough, many of the most interesting developments in biopharmaceuticals are coming from some of the start-up companies that have been swallowed up by Big Pharma.
AstraZeneca, for example, has stated that by 2010 it wants 25% of its pipeline to be biologics rather than small molecules, and with its recent acquisitions of MedImmune and Cambridge Antibody Technology, it's up to 20% already. Similarly, GlaxoSmith Kline (GSK) aims to have 20% of its pipeline filled with biologics, and is acquiring and in-licensing to achieve this. Pfizer has been making biologics acquisitions, such as Covex, Rinat and Coley, to expand its footprint in the field. These are just three of many.
Although much of the technology that has been acquired is as yet un-proven, at least some of these gambles are bound to pay off in the long run.
While biologics are more complex and expensive to manufacture, in a sense this is an advantage as it renders them less vulnerable to generic competition on patent expiry. There are now a handful of generic biologics on the market in Europe - Sandoz's version of human growth hormone was the first a couple of years ago, and has been followed by two versions of erythropoietin - but the US FDA has yet to approve any. The cost of production and the difficulty of proving it is biosimilar to the original both serve to dissuade generic versions.
Biologics are now big business. In 2006 for the first time, two of the top 10 global biggest selling medicines were biologics - the synthetic erythropoietin Aranesp (darbepoetin alfa) from Amgen used to treat anaemia, with global sales of $5bn, and Amgen and Wyeth's Enbrel (etanercept), a fusion protein for treating rheumatoid arthritis, with sales of $4.5bn. With the current stampede into biologics, the number of really big selling treatments will only increase in the future.
Many of the biologic medicines that have reached the market recently are monoclonal antibodies. More than 20 of these have been approved to date, some of which are listed in Table 1, and in 2006 US trade association PhRMA estimated that about 160 more were in the pipeline. Since the first monoclonal antibody was introduced in the mid-1980s to treat organ transplant patients, they have proved particularly useful in treating a variety of cancers, and also a selection of autoimmune diseases such as rheumatoid arthritis. If the antibody is designed to bind to antigens that are specific to cancer cells, then they should be able to induce an immune response at that specific cell type, leaving normal cells unaffected.
Monoclonal antibodies (MAbs) are either humanised or chimeric. Because the normal method for making the MAbs uses mouse cells, the antibodies that are produced will be mouse antibodies. Although these are very similar to human antibodies, they will still be identified as foreign by the human immune system and removed from the body.
Because of the ethical issues of producing human antibodies directly, antibodies are made using mouse DNA spliced with human DNA to make antibodies that are more human-like. Chimeric MAbs have regions of mouse antibodies grafted onto human regions, and are about two-thirds human.
Humanised ones involve amino acid domains of mouse antibodies being grafted into human antibodies, and as a result are 95% human, but these can have substantially lower binding ability; this can be improved by further genetic manipulation.
Because monoclonal antibodies are so effective in hitting precise targets on cells and elsewhere in the body, they can also be used to deliver small molecule drugs to the correct site of action. This is particularly useful in cancer treatment, where many of the side- effects of drugs result from the fact that they attack healthy cells as well as cancerous ones. One drug already on the market that works in this way is Wyeth's Mylotarg (gemtuzumab ozagamacin), where the monoclonal antibody gemtuzumab is linked to the cytotoxic agent calicheamicin to give an agent used to treat acute myelo-genous leukaemia. The humanised antibody targets the CD33 antigen on the surface of leukaemia cells, delivering the highly toxic calicheamicin more effectively to where it is needed. In trials, it was shown that 15% of patients experiencing their first relapse experienced complete remission when treated with Mylotarg.
Another type of antibody drug that is being developed is the BiTE antibody, based on German company Micromet's proprietary technology. It is developing several on its own, and has licensed products to Merck-Serono Astra-Zeneca's MedImmune unit. BiTE antibodies incorporate the small binding domains by which antibodies recognise their antigens linked together on one polypeptide chain. They can be made in mammalian cell culture systems.
The aim is to direct the bodies own T cells to attack tumour cells; the BiTE antibodies induce an immunological synapse between a T cell and a tumour cell, in much the same way as happens when T cells attack within the body naturally. The synapse mediates the delivery of cytotoxic proteins from the T cell to the tumour, ultimately inducing apoptosis. The T cells move on to the next tumour cell once they have killed one, so the antibodies are active at very low concentrations. The T cells also proliferate at the tumour site, and this may also have a positive effect on the patient's immune system.
Two are already in clinical trials: adecatumumab in collaboration with Merck Serono; and MT103 with MedImmune. Adecatumumab is a recombinant IgG1 human monoclonal antibody with a binding specificity for epithelial cell adhesion molecule, or CD326, which is commonly overexpressed on solid tumours such as breast, colon, gastric, lung, ovarian, pancreatic and prostate. This overexpression leads to proliferation and invasiveness of the tumour cells, and is often also associated with decreased survival rates.
MT103 has completed Phase I trials in patients with late stage non-Hodgkin lymphoma with MedImmune. Of the 15 evaluable patients, two experienced a complete response, two partial responses and a further two a minimal response. Six more achieved stable disease, while the remaining three had disease progression. The drug is being developed to treat several different types of B cell lymphoma, such as chronic lymphocytic leukaemia and mantle cell lymphoma.
One problem that dominates the whole class of monoclonal antibodies is that they are large and, as a result, expensive to manufacture. What if only a small fragment of the antibody also had a biological effect? Cambridge-based Domantis, which was taken over by GSK last year, specialises in what it calls domain antibodies. These are much smaller than whole antibodies, and can be designed to have the specificity and high affinity for a biological target one would expect from an antibody, but with more of the advantages of size and delivery methods of small molecules.
Domain antibodies are the smallest functional binding units of antibodies, and typically have a molecular weight of about 13 kDa - about a tenth of the size of a more traditional MAb. They are also easier to make than a MAb, because they are well expressed in bacterial, yeast and mammalian cell systems, and are also more amenable to normal manufacturing processes such as freeze-drying and heat.
Domantis claims another advantage is they can be used to interact with targets that conventional antibodies cannot. MAbs are not suitable for many cell surface receptors, for example, because they can bind dimerically, which could lead to cross-linking and activation of the receptors.
Domain antibodies can also be used against targets that are less accessible, such as enzyme active sites or receptor binding clefts, because their binding sites are more compact. Their size should also make them more useful for targeting tumours and penetrating tissues, and they also ought to be able to be delivered directly to the lungs to treat pulmonary conditions.
Because they are soluble and can be freeze dried, it means that oral formulations should be possible - a major advantage over conventional antibodies where parenteral delivery is essential. The company has also managed to create dual targeting antibody molecules. These fully human molecules bind two targets at the same time: potential applications include antibodies that act against cytokine targets for inflammatory conditions, or two different tumour antigens that are present on the same cell surface to make targeting of cancer treatment even more selective.
Wyeth's Enbrel (etanercept) is an extremely successful example of a fusion protein. These are created by fusing the genes that code for two different proteins, with the resulting protein made by this DNA having some of the properties of each of the two originals.
Typically, these genes are made by removing the stop codon from the sequence of DNA that codes for the first protein, and a second stretch of coding DNA, for the second protein, then attached. It can be engineered to include all of the proteins, or just parts of them. DNA codes to add a string of "spacer" amino acids between the two can also be included, which means that the two proteins are more likely to fold and behave in the same way as the two parent proteins.
Etanercept is the result of the fusion of two naturally occurring soluble human tumour necrosis factor receptors, linked to an Fc portion of an Ig-G1, and as a result it mimics the inhibitory effects of naturally occurring TNF receptors. Because it inhibits the binding of TNF to TNF receptors on the surface of the cells, it has been shown to improve the symptoms of a number of TNF-mediated conditions, including rheumatoid arthritis and psoriatic arthritis.
One area that has caused much excitement in recent years is RNA interference technology. RNA is the step between DNA and proteins, as it is the intermediate messenger that translates the code included in the DNA molecule into proteins. When DNA reproduces, it unwinds the double helix and splits into two single strands, giving a template for rebuilding the whole helix. This is simple as the four nucleic acid bases that make up DNA always pair up in the same way - cytosine with guanine, and adenine with thymine.
To create the messenger RNA that makes proteins, the DNA unwinds and splits again, but this time the nucleic acid bases attach to ribose rather than the 2-deoxyribose that is present in DNA, and build the second string up again, only this time the base uracil replaces the thymine. These two strands then separate again, leaving behind a single strand of mRNA to make the ribosomes that lead to protein synthesis.
By creating small pieces of RNA that have two strands instead of the usual one, these can be used to turn genes off, with the potential to prevent cancer cells from proliferating or to fight off viruses, for example. This process is known as RNA interference, and the pairs of nucleic acid bases in the double strand RNA, or dsRNA, match up with those of the mRNA, which blocks it and switches the gene off. This gene silencing process occurs in nature, but in the human body such sequences of dsRNA will be recognised as alien and destroyed. However, shorter fragments of maybe 20 base pairs do not trigger this response, so such short interfering RNA fragments, or siRNA, may have potential as drug treatments. They form RNA-induced silencing complex, or RISC, in the cells, which is able to seek out and destroy the mRNA they are matched to.
Several siRNA drugs have already reached the clinic. Sirna, a company set up to exploit the technology, was acquired by Merck & Co in 2006. Its pipeline of drugs includes Sirna-027, being developed in collaboration with Allergan under the codename AGN-745 to treat age-related macular degeneration. The siRNA is designed to target vascular endothelial growth factor receptor 1, which is involved in the growth of new blood vessels.
Age related macular degeneration results from the formation of unnecessary new blood vessels that leak blood and protein under the macula, causing damage to vision. By targeting VEGF-1 with a siRNA, the hope is that this angiogenesis will be prevented. AGN-745 is now in Phase II trials, and in Phase I studies all subjects experienced stabilisation of their visual acuity eight weeks after a single injection of the drug was given. A decrease in the thickness of the fovea was experienced by some patients, showing that it does indeed have biological activity.
Our knowledge about how human diseases work and how biological processes can be interfered with to provoke effects has increased dramatically in the years since the first reports of the decoding of the human genome. Because of their agility, small companies are far better placed to make the sort of exciting developments that will lead to the drug treatments of the future. And, as a result, the trend for Big Pharma to expand its horizons by acquiring these companies or in-licensing their technologies is bound to continue.