A Foggy Forecast?

I was very interested to read Stephanie Sutton’s editorial comment, “Bringing Down the House” in the June issue of The Medicine Maker (1). The editorial points out the challenges that pharma and biotech companies of all sizes face when it comes to forecasting demand for products. Some inaccurate forecasts have resulted in manufacturing plants being built, then standing idle, and eventually being demolished without ever producing a single dose, such as Sanofi’s Montpellier facility in France. What an incredible waste of time, talent and resources that could be directed to new drug development.

The pharma industry is facing unprecedented challenges. Payers are demanding lower cost drugs, and yet the cost of getting a drug from discovery through to market approval is increasing – now estimated at $2.6 billion (2). In addition, a delay in launch is estimated to cost an average of $15 million per drug, per day. And research shows that a blockbuster drug will lose $1 billion in revenue annually until capacity is developed to meet demand (3). So whether you over-estimate or under-estimate demand, there will always be costs involved.

Unfortunately, the problem is not easily solved. Forecasting demand, particularly for a product launch, is mired in uncertainty. Each new drug’s success is susceptible to variations in the external environment, the uncertainty of drug development, and the unpredictable actions of competitors. It is not unusual for forecast and actual dosage to vary by a factor of three – and getting it wrong could mean foregoing profit or tying up capital unproductively, and can be catastrophic for a small company without a financial safety net. The impact of inaccurate forecasting is perhaps greatest for large molecules because of the cost of goods and the time it takes to access capacity; a biologics plant usually takes longer to construct than a small molecule plant – and build time pushes the biologics forecast windows out so far that you sometimes need to make a call on whether to build (or not) based on very early clinical data.

Instead of trying to make better predictions (which by their very nature will never be wholly accurate nor able to remove risk), I believe that biopharma companies would do well to consider alternative strategies that minimize risk. Historic approaches – building plants, for example – won’t always work given the dynamic nature of the modern pharmaceutical landscape. And perhaps there is no reason for everyone to build proprietary plants when there are so many outsourcing specialists available, particularly for companies that don’t have the resources or those that are concerned about forecasting. The outsourcing market is very mature – and often quick to adopt the newest technologies.

In my view, every decision should take into account the fact that patients are waiting – there is huge demand for new medicines at the right price. If the old ways of doing drug development aren’t compatible with today’s dynamic pharma landscape, we need to look for a fresh approach. And wasting money on facilities built for the wrong demand is not an option.