For a decade investors in the development of therapeutic products have been coming to terms with the shrinking market sizes of approved products. The focus today is on avoiding “busters” (drugs that are registered, but never achieve any meaningful sales) rather than relying on “blockbusters”.
But for investors in diagnostics, the only bright light on a depressing landscape has been the vast numbers of potential users for such products. Margins may be much tighter on diagnostics than therapeutics, but at least the population of users was likely to be larger.
Not any more.
The dream for any diagnostic product was to be adopted for population screening: tests like mammography for the early detection of breast cancer, PSA measurements to catch prostate cancer and fecal occult blood (FOB) tests for colon cancer are applied in their millions each year. Even on thin margins, these kind of sales volumes make such products huge commercial successes.
Better still, the diagnostic performance of all three of these cancer screening tests are pretty awful. Even the best of them yields more ‘false positives’ than ‘true positives’ when used to screen a healthy population, and the worst of them may generate a hundred false positives to follow-up for every true positive. At first glance that sounds like a recipe for a good investment: an incumbent product that’s easy to beat and a vast market for the successful innovator.
Public support maintains demand for existing screening tests, even though their performance is unsatisfactory. New innovations will not be so lucky.
But a closer look reveals the problem. These tests are so far short of the performance needed to be useful as a population screening tool that even a dramatic improvement in that performance still leaves an innovative new product falling under the bar for adoption.
Just in the last month (October 2011), there have been two damning reports on the value of cancer screening, reviewed in BiomarkerBlog from Total Scientific. The conclusion, it seems, is that at best such screening programmes are a poor use of healthcare resources, and at worst they do more harm (through aggressive follow-up and unnecessary treatment of false positive cases) than good through early detection of true cases.
DrugBaron has noted before just how demanding the performance criteria are for a test to be useful in population screening. So these reports don’t amount to a step change in the objective scientific assessment of screening programmes. But that doesn’t mean they are not important. What they do signal is the hardening attitude of healthcare providers towards paying for population screening. After all, the commcerial success of a product (of any kind) depends not on how objectively useful it is, but on how many people are prepared to pay for it – a decision which depends at least as much on subjective factors as on cold, hard analysis.
Damning as these objective assessments seem to be (they cant be definitive, after all, because comparing harms and benefits necessarily has a degree of subjectivity to it) they are unlikely to lead to a cancellation of global screening programmes any time soon. Public pressure would resist such a change, not least because non-scientists struggle to understand how a non-invasive screening programme could possibly do more harm than good.
This public support for cancer screening will sustain existing programmes in the face of almost overwhelming objective evidence against them (at least for a while). But that same public support will not influence the decision to adopt new (or improved) screening tests: for these, the cold logic of clinical benefit and sensible economics will be the gatekeeper.
And these latest reports are creating a framework for the scrutiny of these new or improved screening products that will be immensely difficult to live up to. Comparing favorably to the current gold standard will not do – you can perform much better, you can be comparable on price, but you still will not pass these objective assessments of utility and the product will not be adopted, certainly not on a population-wide scale.
It is easy, as an investor, therefore to be seduced by a diagnostic product candidate that is demonstrably better (much better) than those currently used all over the developed world – but unless you really understand just how good the performance of such a test has to be in order to be useful, this is the kind of golden egg that will likely hatch to reveal a stillborn goose.