In the past decade, there have been substantial changes in the number and nature of drug launches in the United States. Oncology has been one of the drivers of these changes, accounting for 30% of the 46 US drug approvals in 2015. Specialty drugs are also increasingly common, displacing previously primary care-focused launches. We examined US drug launches to assess whether commercial success is now more difficult to achieve, given the increased competition, payer sophistication and price scrutiny.
To assess the success of a pharmaceutical launch, our study used three metrics: peak revenues, the shape of the revenue curve and the percentage of analysts' launch expectations achieved after 5 years. We examined US launches from 2006 to 2013 to evaluate 3- and 5-year data (not available for launches in 2012 and 2013) after launch, using revenue data from the EvaluatePharma database and multivariate regression analysis. Here, we present the key findings from our analysis (see Supplementary information S1 (box) for details of the data set and analysis).
Successful launch is harder to achieve
The average performance of launches over the period studied has declined for all three of the metrics we used as proxies for success. First, the percentage of drugs failing to reach US$200 million in peak annual sales rose from 29% in 2006 to 47% in 2011, while the percentage of blockbuster drugs (those exceeding $1 billion in annual sales) fell from 39% to 21% of launches (Fig. 1a).
Figure 1: Trends in drug launch performance.
Trends in drug launch performance.
a | Drugs with sales in the specified ranges in the first 5 years post-launch for each drug for the years 2006–2015 were summed. Data for 2006 (for 28 drugs) and 2011 (for 34 drugs) are shown, indicating that blockbusters have become less common; the trend remains the same for all other years. b | Drug launches over the past 25 years for which complete US sales data were available were assessed and grouped into four types of launch revenue growth trajectory: failed (non-monotonic growth); slow ramp-up (faster growth in later years); linear; and fast ramp-up (faster growth in early years). Data for drugs launched in two periods are shown, indicating that slow ramp-ups have become more common. c | EvaluatePharma's Archived Forecasts module was used to determine the percentage of expected peak sales achieved for drugs launched in 2006 (for 28 drugs), 2008 (for 28 drugs) and 2010 (for 25 drugs) in the 5 years after drug launch, indicating that fewer drugs are meeting analysts' expectations. See Supplementary information S1 (box) for details. Sources: EvaluatePharma; IMS; PharmaProjects; McKinsey analysis.
Second, sales growth after launch has slowed. We evaluated drug launches from 1990 to 2015 and divided launches into four categories based on their sales trajectory in the first 5 years: fast ramp-up; linear ramp-up; slow ramp-up; and failed launch. From 1990 to 1994, linear or fast ramp-ups were the norm, accounting for 73% of launches, but after years of steady decline, their share halved to 36% for launches in 2013–2015, while slow ramp-ups more than doubled in frequency, from 23% of launches in 1990–1994 to 58% in 2013–2015 (Fig. 1b).
Third, fewer drugs are achieving average and best-in-class performance relative to analysts' expectations; indeed, around half of the launches in 2010 did not even reach 25% of expected peak sales (Fig. 1c). Pharma is not the only industry in which analysts do badly at predicting launch revenues, but it seems that accurate predictions have become harder to make. Overall, these three sets of findings indicate that a commercially successful launch has become substantially harder to achieve in today's environment.