What data can tell us about drug type and launch performance

The FDA approved 50 drugs in 2021, continuing a trend of increasing drug approvals over the decade. The European Medicines Agency approved 54. These approvals range from first-in-disease drugs and those targeted to specific patient subtypes to those with new mechanisms of action for treating disease and so-called “me too” drugs—next-in-class entrants into competitive markets.

Many observers have been quick to criticize this last type of drug for being less risky to manufacturers and of lower value to patients. This New York Times opinion piece is typical of the genre. In it, the authors argue that “less innovative drugs offer low financial risk to manufacturers because they are based on existing products or work via similar biochemical pathways.”

So, which type of drug is most likely to find commercial success? The more novel the drug, the more likely it will be to succeed, right? Or do next-in-class drugs succeed more often because they launch into established markets with incumbent patients who may be open to seeking alternative treatment options? Research into launch success by drug type is sparse, so the ZS Research Center set out to quantify the launch success of new drugs and investigate which drivers correlate with success and failure. We wanted to test if drug type really matters when launching a drug and, in particular, if the assertion above has merit, namely: that pharmaceutical companies launch next-in-class drugs to "stack the deck" for market success rather than patient benefit.

Before we could investigate how different drug types fared in their launch years, we first had to define our drug types and the share of launches each represented. We grouped them into four mutually exclusive categories. Despite some minor year-to-year noise, the distribution of launches in each of the four drug types has been largely stable.

• First in disease. New molecular entities (NMEs) that address a historically untreated disease with significant unmet needs. These represented 16% of all drug launches we studied. (In this article, we use "drug" and "new molecular entity" interchangeably.)
• Specific patient. NMEs that provide a viable treatment option for subpopulations within, or a specific variant of, a broader disease. Patients may already have access to more general treatments, such as chemotherapy not tied to a specific biomarker. They represented 20% of all launches.
  
• New mechanism of action (MOA). NMEs with a novel MOA to treat diseases with competing options. They represented 27% of all launches.
• Next in class. NMEs that follow other drugs that treat diseases with established markets. They represented 38% of all launches. 

We also assigned each NME to one of our four mutually exclusive drug categories: acute, cancer, chronic and rare disease. 

We observed notable differences when breaking down drug type by targeted disease category (Figure 1). First-in-disease drugs represented more than 40% of drugs launched in rare diseases, making this by far the most popular category for this type of drug. Next-in-class and new MOA drugs, meanwhile, made up more than 80% of launches targeting chronic diseases, such as Alzheimer’s disease, diabetes and HIV.

Figure 2 provides an overview of how each drug type fared in its launch year compared with analyst forecasts. Next-in-class drugs were significantly more likely than other types to miss their launch-year forecasts, with the median next-in-class drug attaining just 82% of forecast and missing nearly 60% of the time. At the other end of the spectrum, the median first-in-disease drug attained 147% of analyst forecasts in its launch year and missed only 26% of the time. The median specific-patient drug fared almost as well, attaining 120% of forecast and missing just 31% of the time. The median new MOA drug, meanwhile, attained 101% of forecast but still missed 44% of the time. These results suggest two things: that novelty is a strong predictor of success and that more novel drugs tend to overperform expectations to a greater degree than less novel ones.

Other than novelty, what might lead an NME to miss its launch-year forecast? We investigated two potential factors: clinical efficacy and disease category. 

For new MOA and next-in-class drugs, we measured clinical superiority using a variety of methods, including comparing clinical efficacy scores of NMEs against competitors, head-to-head clinical trials with competitor drugs and cross-trial comparisons. For first-in-disease drugs and those targeting specific patient subpopulations, we evaluated efficacy based on drug performance against a comparator arm in regulatory trials.

We then placed each drug into one of three categories:

• Drugs offering efficacy superior to benchmark
• Drugs offering efficacy comparable to benchmark
• Drugs offering efficacy inferior to benchmark

To our surprise, we found that NMEs with comparable or inferior efficacy were about equally as likely to miss their launch-year forecasts. Each missed roughly half the time (Figure 3). While more efficacious drugs missed only about a third of the time, the more important finding here might be that even superior efficacy is not enough to guarantee launch success.

If efficacy doesn’t guarantee launch success, perhaps launching in certain disease categories is more predictive of success? We found that launches in rare diseases met or exceeded launch-year expectations more often (67%) than other launches—and the more novel the drug, the more likely it was to meet or exceed forecasts. Cancer drugs were successful slightly less often (63%), but they followed the same pattern of novelty begetting success.

Chronic disease markets, though, tend to have large patient populations and are often more competitive—either from generic entry or other branded drugs. Launching a drug for chronic illness ends up being a riskier endeavor. Overall, 58% of these launches missed their forecasts—even when they were the first drug to market in years. Next-in-class drugs made up just less than half of all launches in chronic conditions but nearly 60% of launch failures in this category. If a company plans to launch a drug to treat a chronic disease, it will need to tailor its strategy to the drug type it is launching. Figure 4 offers a list of considerations that can help.

What does this all mean for launch team decision-making? Here are four insights leaders can use to reexamine their pipelines and launch plans in light of our research.

1. Novelty is undervalued. For the period we analyzed, the median first-in-disease drug dramatically outperformed its forecasted performance in its launch year. Some of these drugs, like Dupixent and Shringix, exceeded even lofty initial forecasts. Manufacturers clearly demonstrated these drugs’ efficacy and built robust systems to find the right patients. When first-in-disease drugs fail, it is often because of questionable efficacy. The fact is efficacy still matters for novel drugs.

2. Succeeding with next-in-class drugs isn’t easier, it’s harder. Based on our findings, launching a next-in-class NME is a risky endeavor. These launches have missed external expectations more than twice as often as first-in-disease drugs. Why? It may be that external analysts overestimate the ability of next-in-class entrants to drive revenue in the initial period after launch, setting up these drugs to miss their targets. But there is no easy solution here: Even superior efficacy is no guarantee for success for these drugs. Regardless of efficacy, next-in-class NME launches missed forecasts in their launch year more than half the time.

3. Pursuing next-in-class drugs isn’t always a bad idea. Byetta was the first GLP-1 receptor agonist approved to treat Type 2 diabetes. But next-in-class options such as Trulicity and Ozempic offered providers and patients superior options to also improve blood sugar control, reduce the risk of major adverse cardiovascular events and promote weight loss. Similarly, Spikevax is no less valuable a tool for controlling the COVID-19 pandemic because Comirnaty arrived first.

4. Drug type should dictate launch strategy. Pharma leaders may consider adjusting early performance expectations. But more importantly, they should adapt launch playbooks to address the most salient challenges of the type of drug they’re launching.

By bearing in mind the drivers of success for each drug type and planning tailored launch strategies accordingly, pharma leaders can assure themselves of fewer surprises and more commercial wins.

The authors would like to acknowledge Surbhi V Gupta, Deeksha Parashar and Yashika Jain for their assistance with this research. Howard Deutsch served as editor for this publication to ensure it met ZS Research Center quality standards and research integrity.

For the purposes of this study, we evaluated the initial launches of new FDA-approved NMEs from 2017 to 2021. We omitted follow-on launches of already-launched molecules' route of administration or formulation. We also did not measure the effect of new indications on a molecule's success.

To evaluate launch success, we grouped drugs by both NME category and broad disease type—acute, cancer, chronic and rare disease—and then compared launch-year revenue to external analysts’ consensus forecasts to determine whether they beat forecast (attained >110% of forecasted sales), met forecast (attained 90%-110%) or missed forecast (attained <90%). Launches that either beat or met their forecast were deemed successful.

One limitation of our study is its reliance on forecasts from external analysts. Another is its focus on launch-year results only. Some may challenge our assessments, given that an NME might have met or exceeded internal expectations or driven significant revenue beyond the launch year. External analyst forecasts, however, often set early market expectations and the resulting consensus for whether a drug is poised for success or failure. And there is another, simpler reason we used external forecasts: They are publicly available, and they offer a reliable data source against which to quantify launch success.

This research used forecast data from Evaluate Pharma and efficacy data from Datamonitor, as well as open secondary sources including accessdata.fda.org (/label), clinicaltrials.gov, fda.gov, ncbi.nlm.nih.gov, pubmed.ncbi.nlm.nih.gov, drugs.com, fiercepharma.com and globenewswire.com.