The sale of Avila to Celgene may mark the pinnacle of the market for kinase inhibitors in cancer. Where might early stage drug discoverers in cancer start looking now if they want to create successes like Avila in five to eight years time?
Excitement for different kinds of therapeutic target in cancer goes in waves – and with cancer by far and away the leading indication for biotechs (judged by the number of molecules in development), predicting the “next big thing” is an important and valuable skill for early stage investors. What comes next after kinase inhibitors and epigenetic modulators?
With anti-infectives, all therapeutic strategies rely, in some way or another, on exploiting the (often subtle) differences between pathogen and host. So too with cancer – only the differences tend to be an order of magnitude smaller. It has been known for a long time that different metabolic pathways dominate in cancer cells, but recent research has focused on a couple of key metabolic “switches” that might be different enough from normal cells to represent viable therapeutic targets.
Will these metabolic switches prove to be effective new targets, particularly in combination with other strategies that deprive tumours of their blood supply and so make them more reliant on anaerobic metabolism? Its too early to tell, but the science has moved far enough forward now to make it worth renting a few square metres of this virgin new territory just in case metabolic inhibitors become the focus of the next big “land grab” in cancer.
The discovery of oncogenes in the 1970s was the origin of the first “cancer bubble”, overflowing with optimism that these abnormal variants of genes involved in the regulation of cell proliferation that were expressed by tumour cells would provide targets that could “cure” cancer.
While oncogenes quickly revolutionized our understanding of cancer biology, it has taken much longer to translate that knowledge into useful therapeutics. And while there clearly have been some big successes, such as the impact of Genentech’s Heceptin™ trastuzumab in HER2+ breast cancer or more recently the impact of Gleevec™ imatinib from Novartis that targets the several tyrosine kinases, there is also a growing recognition of the problem with kinase inhibitors that target oncogenes: the sheer diversity of pathways controlling cell proliferation, and the pragmatic and time-consuming process of matching a particular indication to the (often small) population in which it will be effective.
It seems likely that this wave is now cresting. Impressive exits in the field are happening now (such as Celgene’s acquisition of Avila, and their BTK inhibitor for $350M up-front just this week). But if the exits are happening now, surely its too late to join that particular party. There are hundreds of kinase inhibitors targeting different oncogenic kinases in development, and for many it is becoming clear that the benefit on survival, while statistically significant (and, no doubt to the individual sufferers of meaningful magnitude) is at best only on the cusp of being commercially viable for payors. Many of these drugs are destined to be “busters” (drugs that achieve regulatory approval but not market acceptance).
As one wave breaks, the next is already clearly formed behind it: epigenetic modulators. These targets control chromatin condensation, either by modulating histones or even the methylation pattern of the DNA itself. The basic science is well advanced, and suggests that HDAC inhibitors, for example, may have impressive therapeutic benefit. The first hints of clinical efficacy are now being reported by the vanguard, stoking the enthusiasm still further and few, if any, of the serious players in the field lack some kind of stake in the ground in epigenetics.
But for the surfer who likes to invest at the earliest stages of the drug discovery pipeline, it may be too late to catch even the epigenetics wave. If your programme is not already underway, it may be hard to play catch up. For these wild-eyed optimists, which of the ripples on the horizon is going to turn into the next wave?
“When you hear about something once it hardly registers; the second time makes an impression. But three times from different sources and you know you are on to something. Thats the principle of triangulation”
Using the triangulation principle DrugBaron first learnt from Kevin Kinsella, a highly successful west-coast VC with Avalon Ventures, the crosshairs have firmly settled on tumour cell metabolism. Twice in just a week, two exciting papers coming at the problem from different angles have caught our attention (as those who follow @sciencescanner on Twitter will already know).
The first, published in the Journal of Experimental Medicine (JEM), looked at the effect of inhibiting expression of the M2 isoform of pyruvate kinase (PK M2). On the surface, pyruvate kinase, a boring metabolic enzyme in the ubiquitous glycolysis pathway that releases energy from glucose, does not look like a particularly promising target. Glycolysis is an important battery for all cells, so inhibiting it would seem to be a pretty non-selective strategy.
But as is often the case, the detail reveals a more elegant story. Not only are their two isoforms of pyruvate kinase, one (PK L/R) specific for liver (where it is also important for gluconeogenesis) and red blood cells, and the other (PK M) in all other tissues, but this M isoform also exists in two variants (PK M1 and PK M2) generated by alternative splicing.
It is the M2 splice variant of PK that dominates in proliferating cells. But the diversity doesn’t stop there. PK M2 exists in both tetrameric and dimeric forms. The tetrameric form is a typical pyruvate kinase, rapidly channeling glycolytic intermediates into energy production. But the dimeric form has almost no conventional activity, and as a result the glycolytic intermediates are diverted into synthetic pathways making amino acids, nucleotides and other building blocks for uncontrolled cell growth.
Tumour cells, it turns out, rely heavily on this dimeric PK M2 to generate biosynthetic intermediates. So much so, that blocking PK M2 expression not only halts growth but induces apoptosis. The balance between dimeric and tetrameric forms is in effect a switch between biosynthesis and energy production.
The paper in JEM used siRNA to prove that PK M2 was essential for tumour growth, and it may be a good target for RNA interference. But is also a pliable target for small molecule modulators. Metabolites such as fructose-1,6-bisphosphate and alanine are already known to be allosteric modulators of the enzyme, and to affect the balance between dimeric and tetrameric forms, suggesting this may be an accessible target for modern drug discovery.
Which of the ripples on the horizon is going to turn into the next wave? Metabolic Inhibitors
The second metabolic enzyme target in cancer supported by new data this week is IDH1 (type 1 isocitrate dehydrogenase). Two papers in the same issue of Nature demonstrate that tumour cells use this enzyme for biosynthesis of lipids essential for growth.
Tumour cells, unlike normal cells, operate at much lower oxygen tensions and as a result convert pyruvate almost exclusively into lactate, shutting off the provision of Acetyl-CoA build-blocks for lipid synthesis. Tumour cells, it seems, get round this problem by using glutamine to inject carbon skeletons into the TCA cycle, a reaction mediated by IDH1. Indeed, tumour cells carrying inactivating mutations in the Von Hippel Lindau tumour suppressor protein elect to rely on reductive glutamine metabolism even at normal oxygenation levels.
The simple take home message from these complex studies of intermediary metabolism is that tumour cells rely heavily on metabolic pathways that are entirely dispensable to normal cells under almost all conditions. That is surely the hallmark of an optimal drug target for cancer.
But it gets better: these metabolic pathways are sufficiently important that tumour cells cant do without them. Slowing tumour growth is clinically valuable, but inducing apoptosis and regression is even better.
Furthermore, unlike the oncogenic kinases, these metabolic pathways important for cell proliferation at low oxygen tensions seem to be conserved across many (possibly all) solid tumour types. The biggest challenge for kinase inhibitors is matching the right inhibitor to the right tumour type (and the resulting slicing and dicing of the market that ensues). First indications, then, are that similar problems will not bedevil drugs targeting cancer metabolism.
And metabolic enzymes are good prospects for drug developers (in the sense that finding small molecule inhibitors of enzymic activity is easier than finding small molecule blockers of receptor:ligand interactions), although the intracellular location of these enzymes presents perhaps the biggest challenge to exploiting these targets.
Best of all, there is every prospect of synergy. PK M2 and IDH1 are both involved in switching carbon skeletons from energy generation into biosynthesis. Blocking both at once should have better than additive benefit. This synergy may extend outside the class of metabolic inhibitors altogether: the reliance on PK M2 and IDH1 stems from the hypoxic conditions that prevail in the tumour micro-environment. And existing therapeutics, such as Avastin™ bevacizumab from Genentech, that target VEGF to disrupt tumour angiogenesis will only exacerbate the hypoxia – and the dependence of the tumour on PK M2 and IDH1. Metabolic inhibitors coupled with Avastin™ may show a degree of benefit hitherto only dreamt of.
It must now be worth renting a few square metres of this virgin territory before metabolic inhibitors become the focus of the next big “land grab” in cancer.
If you had to pick an indication to invest your biotech dollars in, cancer remains as good as any and better than most. If you had to pick a nascent field that might grow to provide the kind of worldwide excitement that drove a $350M up-front valuation for Avila, then metabolic inhibitors have as good a chance as any of fulfilling that promise. If exciting new data comes as thick and fast as it has done in January 2012, it may not even five years before we get there.