Precedents play a central role in the drug development industry. Clinical trials are based on earlier studies with ostensibly similar drug candidates. Business models are judged against the current landscape. Assets are valued by looking for comparables. Indeed, big pharma consider the institutional accumulation of such experience to be a significant competitive advantage.
But if you stop and think about it, such an approach makes no sense if you seek true innovation. The premise of all innovation is that it breaks the mould, changes the goalposts, disrupts the marketplace. Aiming to change the world, and then judging your progress against historical norms isn’t just meaningless – some precedents can lead you down blind alleys or drive altogether the wrong decisions.
DrugBaron examines a number of recent examples of misapplied precedents to illustrate why, when searching for the breakthrough that changes everything, forwards is the only way to look.
One of the most common, and sensible, uses of precedent is in clinical study design. The need to follow well-trodden paths to win regulatory approval encourages repetition and stifles productive change. If someone else persuaded the regulator to approve their drug with a given trial design, why consider anything else?
That approach works just fine if your molecule is broadly similar – in terms of general mode of action, degree of expected efficacy and side-effect profile. In other words, a “me too” or “me better” drug candidate. The same trial design will be ideally placed to show your equivalence or superiority.
But what if your candidate works in a different way? A wholly different selection of patients may respond; a different clinical end-point may be much more relevant. More importantly, if you drug is an order of magnitude more effective than previous candidates, a much smaller or shorter (or both) study may be ample.
A classic case relates to hard end-point studies in cardiovascular disease. Demonstrating reduced rates of heart attack in a phase 3 population is considered an enormous challenge, requiring thousands of patients. The Sanofi Phase 3 program for their anti-PCSK9 candidate alirocumab will enroll more than 20,000 patients. Such huge numbers are needed when drugs are only marginally effective, or (as with the anti-PCSK9s) potentially only marginally more effective than the current, established gold standard.
Demonstrating small differences requires long, large and hence costly studies
This has become the norm in cardiovascular disease: any advantage of Brillanta™ ticagrelor from AZ over generic clopidogrel required tens of thousands of patients to quantify; showing that Factor Xa inhibitors such as Xarelto™ rivoroxaban (J&J/Bayer) and Eliquis™ apixaban (BMS / Pfizer) are superior to warfarin required many huge studies. Demonstrating a hard-end point benefit with fish oils likely requires a study so large that its unclear whether Amarin can even afford to finish it.
The result? The entire industry tends to shy away from developing drugs for these cardiovascular indications because precedent says the trials need to be huge, and the failure rate is very high. Despite the massive prevalence, and unmet medical need, of coronary heart disease, therefore, it remains a disfavoured area for investment and innovation – it is the victim of ‘asset favouritism’.
But what happens if you turn up a drug candidate that’s a big step change from everything that’s gone before? Although still in preclinical testing, the anti-thrombin antibody ichorcumab, being developed by the Index portfolio company XO1, is an example of a product candidate that has the potential to re-write the rulebook. If this agent prevents clotting without causing bleeding in the clinic, as it does in preclinical models, should the potential size of the Phase 3 program preclude development? Of course not. In the limit, where it was impossible to have a heart attack while treated with this antibody, trials with a few hundred patients would be more than adequate to unequivocally define its efficacy.
Another Index portfolio company, Epsilon-3 Bio has had a similar experience. Its lead candidate, also in preclinical testing, is a highly specific stimulator of cell debris clearance through the autophagy pathway. This compound has shown remarkable efficacy in the MRL lpr/lpr mouse model of lupus – but, thanks to the power of precedents, the company now faces a challenging decision.
History (and hence experts in clinical development in the field) tell us that clinical proof of concept has been difficult to achieve in lupus. The only recent approval in this indication, Benlysta™ belimumab (an anti-BLyS/BAFF marketed by GSK), required large Phase 3 studies, themselves founded on somewhat equivocal findings in Phase 2. Most severe cases of lupus (those with kidney involvement), where the treatment need is highest, were excluded. Judging by this precedent (and those of failed studies with other anti-inflammatory agents), clinic development in lupus is indeed a serious challenge.
But should the company be beginning development on the assumption that the new candidate is only as effective as Benlysta™ (whose efficacy was considered “marginal” by the FDA reviewers prior to approval)? If they have reason to be that unexcited about the prospect for the Epsilon-3 Bio drug candidate in lupus then why proceed into development at all?
Once again, if a strong case can be built for game-changing efficacy then demonstrating that in the clinic will be, if not a walk in the park, nevertheless readily achievable. Three month treatment in a hundred patients with severe lupus would be sufficient to demonstrate efficacy for an agent with genuine disease-modifying properties.
What these examples tell us is not that indications such as coronary heart disease, lupus or even sepsis are impossible arenas for clinical development, but that too many drugs have been progressed in the past whose profile suggested (at best) a minor increment in clinical outcomes could be achieved.
Some areas lend themselves to incremental innovation (usually where there is a validated biomarker accepted by regulators, such as LDL cholesterol or blood pressure, or other reasons why clinical development is unusually straightforward). For the rest, the difficulties that have been encountered in the past should not preclude new attempts in the future: instead they should teach us only that candidates must have game-changing potential.
And the simple corollary: if you have game-changing potential, past development challenges are a poor guide to your future development pathway
If precedents can mislead innovators when it comes to clinical trials, they are arguably even more dangerous when it comes to disruptive business models.
Consider a new venture DrugBaron is working on in the field of neutraceuticals and functional foods. This is a massive market: over $140billion globally. Yet it is sustained by a relatively small number of ingredients whose health benefits are now widely accepted. The number of new (and, critically, proprietary) ingredients with marketable health claims has slowed to a trickle as regulators have rightly tightened the rules. With the bar for health claims associated with food products approaching the height of the bar for drugs, growth in this market is slowing as the supply of novel product candidates dries up completely.
Enter a new company with a platform designed to deliver proprietary new food supplements, accompanied by marketable health claims acceptable to regulators. How should you assess such an investment opportunity?
The usual approach – benchmarking – fails completely. The nearest comparable are either companies attempting to launch variations on existing products (new formulations, combinations or simply new brands) or else those developing a single new candidate food supplement identified by serendipity. Neither of these comparator companies look particularly attractive as investments: re-branding (in all its forms) is a marketing play, rather than a technology play, with accompanying wafer-thin margins; and the evidence supporting novel food supplements rarely stands up to proper scientific rigour, and in today’s regulatory environment, stands little chance of supporting any kind of health claim useful in the market place.
Attempting to shoehorn a validated neutraceutical discovery platform into the business models of these existing players squeezes out the magic that makes it special. You might even conclude that such a venture had little to recommend it – none of the comparators are thriving, and the new company has on top of those issues, the technical risk associated with its platform.
But to take that view misses the point altogether. You cannot judge the potential for a disruptive business through any kind of detailed assessment of the current landscape. Who attempted to value Google or Amazon in the late 1990s by reference to the profitability of bricks and mortar libraries or shops? Certainly not the backers who saw them grow to be among the largest and successful global businesses of today by disrupting the landscape, not evolving it.
This is not so much a lesson that benchmarks and precedents are not useful, as a reminder not to stretch those comparisons too far. If the business you are assessing is incremental – evolution not revolution – benchmarks will serve you well enough. But, as with clinical trial design, when the newcomer encompasses real innovation an over-reliance on precedents may lead you to the wrong conclusion altogether. On the way to changing the world, there are no signposts.