Monday, 20 January 2014

When Moore’s law stopped working

Nearly fifty years ago, Gordon Moore, a director of R&D at Fairchild semiconductors, wrote an internal paper about the progress of integrated circuits. In his paper, Moore predicted, based on an extrapolation from six previous years following the invention of the integrated circuit in 1959, that every year a computer chip will double the number of components whilst halving in price.

Maybe it was a self-fulfilling prediction because Intel employed Gordon Moore shortly after and used his prediction in their business model. Nevertheless, the same rule applied to all types of technologies and was branded the ‘Moore’s Law’. Technological improvement leads to an increase in production which leads to a subsequent price reduction. 

Progress in healthcare also follows the same law. Since the 1970s, production has sped up and in a way got cheaper.

Working in the pharmaceutical industry for the last decade, I have noticed that even with the economic downturn of the last three years, there is still huge progress in medical research and the speed of change is exponential. Technologies are developing and reaching larger audiences, however, when one hears about the cost of treatments, one may start to wonder if Moore’s Law could ever be applied to medicine.

Why should Moore’s Law apply to medicine?

Moore’s law should apply to medicine because scientific research is dependent on computing power. For example: efficiency of equipment used for diagnostics has always been limited by computer memory and processor speeds. The introduction of LED lasers instead of the cumbersome expensive noble gas lasers, MRI, 3D ultrasound, computerised tomography (CT),  DNA Sequencing, microscopy, flow cytometry, computer associated drug design, robotics, high-throughput screening and automated bioreactors in manufacturing, all advanced due to computing power in the last 10 years.

Therefore, scientific research that took 3 years to perform in 2000 may take just 3 hours today.

Moreover, structural changes within the pharmaceutical industry moved research and manufacturing to specialised contract partners (CROs and CMOs), reducing R&D costs, and improving productivity. Very few companies today are autonomously developing their pipelines from conception to manufacturing. Many pharmaceutical brands have become virtual, macro managing entities that employ specialist contractors to perform the different aspects of clinical research.

Why Moore’s Law does not apply medicine?

The pharmaceutical industry is cash dependent. Return on investment takes over 10 years to achieve and in many cases does not happen at all. In the meanwhile, biotech and pharmaceuticals survive on constant cash flows, sometimes only from investors to pay for overheads and R&D. When a drug finally makes it to the market it may have to pay for the total investment of all conceived but undeveloped posterity.

In-order to attract investors’ interest, large pharmaceuticals have to bring to the market at least one blockbuster drug a year (a blockbuster achieves over $1B sales per annum).

Regulation has improved and therefore has become more expensive. Today’s applications to the FDA, EMA and Japan are costlier than ever. For example an average phase-I trial cost $7m-$10m, and a phase-III trial may cost 10x that. New drug application consists of about 100,000 pages and takes up to 2 years to approve, costing an additional $1M-$2M.

As a consequence healthcare costs are rising much faster than GDP.

Figure 1 Healthcare costs vs. GDP

Greed and money

If you have followed my trail of thought this far, you might think that I am heading towards blaming the escalating costs of healthcare on pharmaceutical companies’ ‘greed’.  Well, I am not!

When I got my first job in a small biotech, based in Welwyn Garden City, my friends warned me “Be prepared for disappointment. Don’t get attached! Because projects that you live and breathe and become part of you like a beloved pet, tend to die (or be sold off)”.

Unlike many other businesses (mining, banking and insurance), the pharmaceutical industry is predominantly based on good intentions – finding cures for diseases, helping humanity, prolonging life and curing cancer. Nevertheless, good intentions come at expense, and the only way to fund them is to run a business, attract investment and make a profit.

Moore’s law cannot apply to medicine because although drug discovery has come a long way from the invention of penicillin, we are still only at the beginning. When we achieve cures for many of our ailments, only then, will we see healthcare costs come down.

 Figure 2 Moores’ extrapolation (from Wikipedia)

 Our January blog entry was written by Shai Senderovich, a Research Executive at Branding Science.

No comments: