Too little for too few: The average new cancer therapy in 2024 added 3 months of life for 1,600 patients
Devin Scannell is CEO of Decade, an oncology therapeutics company
Summary
Cancer death rates in the US have decreased by 34% since 1991 (~1% p.a.) but cancer deaths have risen from 521K to 618K due to population growth and aging
New cancer therapies approved in 2024 added 3 months incremental survival on average (vs 10-15 years lost) and benefit a median of just ~1600 patients
At this rate it will take >100 years to eliminate cancer deaths in the United States
Much bolder innovation is needed if we are to have an “Ozempic for Oncology”
Peak cancer is always next year
On Jan 17th this year Jean-Charles Soria, Head of Oncology for Amgen, reflected on the American Cancer Society (ACS) 2025 Cancer Statistics report: “Cancer death rates in the U.S. have dropped 34% since 1991, averting over 4.5M deaths”. The reduction in the death rate from 215 per 100,000 Americans in 1991 to 142 per 100,000 is a significant achievement that has been enabled by reduced smoking, earlier detection and novel therapeutics. On average, cancer patients lose 10-15 years of life.
Despite the falling death rate, the number of cancer deaths has actually risen from 521K in 1991 to 618K forecasted for 2025. A 1% per annum (34% in 34 years) reduction in the death rate has not been enough to keep pace with population growth and aging. “Peak cancer” is always next year.
A second issue relates to future improvements. The main driver of the steady gains in death rate from 1991 to today (red line) is reduced smoking which is not easily repeatable. Earlier detection has also been impactful for certain cancers (e.g., colorectal, prostate, breast) but faces challenges including technical, commercial and behavioral. Indeed, the ACS 2025 report highlights that death rate improvements are slowing for several cancers. Improvements don’t just happen.
Novel therapeutics have also contributed but despite meaningful impact on certain cancers - most notably hematologic and breast - the impact on overall cancer burden has been modest. Even breakthroughs such as anti-PD-1 drugs (e.g., Keytruda in 2014) are not visible on the chart above though their impact may be revealed over time.
The class of 2024
To better understand how novel therapies are impacting cancer burden, we looked at all 16 new cancer drugs approved by the FDA in 2024:
6 small molecule “precision medicines” that target specific mutations
6 antibodies (3 traditional and 3 bispecifics)
3 cell therapies including the first “tumor infiltrating lymphocyte”
1 protein product used in combination with a vaccine (Anktiva)
Overall, these therapies are thematically similar to the 13 approved in 2023. How much of an impact will they have on cancer deaths? We tackle this in three parts:
What is the incremental survival a cancer patient can expect from one of these therapies?
How many patients can potentially benefit from each newly approved drug?
Putting these together, how much of a dent does this make in overall cancer burden?
A typical new therapy adds 3 months of incremental survival
Before diving in, we need to consider how new therapies are approved: Randomized controlled trials (RCTs) compare patients getting a new therapy to those getting existing “standard of care” (SoC) therapy alone. They assess improved survival. When RCTs are infeasible or unethical, therapies may be evaluated based on “response rate”, the proportion of patients who appear to respond to the drug and “duration of response.” Such trials typically lack a control arm so they cannot assess the benefit over existing SoC but are typically faster and cheaper. In 2024, RCTs accounted for 6 of 16 approvals.
Here’s how RCT-approved drugs stack up in 2024 based on data from FDA approval notices:
The average overall survival improvement over SoC for an innovative cancer drug approved in 2024 is 3 months. At least one drug had no detectable overall survival improvement over SoC and no drug averaged more than 7.1 months. To be clear, individual patients could have more than 3.0 (or 7.1) months incremental benefit over SoC but the average improvement is modest relative to the 14 years a typical patient loses to cancer.
Here’s how the other novel therapies stacked up:
On average, fewer than half of patients (44%) appear to respond to a new therapy. No therapy had more than a 62% response rate and one had just a 21% response rate. For the patients judged to have benefited, the typical duration of response is 11.3 months but, without a control arm, we cannot say how these patients would have fared on SoC alone. Even assuming twice the benefit of RCT-approved drugs, the average across all 2024 approvals would be just 4.9 months of incremental survival.
The median new drug benefits 1,600 patients
Unlike survival and response rates, the number of patients who may benefit from a new therapy is not available from the FDA. For this information we turned to the pharma and biotech companies themselves (“sponsors”) as they frequently communicate the estimated number of addressable patients to investors.
The average number of patients who benefit from a new therapy is ~3,900 and the median is ~1,600 when applying the response rate (as discussed above) to the number of patients estimated by sponsors. Remarkably, only one drug is expected to benefit more than 20K patients and 3 benefit fewer than 500. The total across all therapies approved in 2024 is 61,000. Even assuming a doubling or tripling of addressable patients due to additional approved use cases in subsequent years, the average number of patients who benefit from a therapy is likely on the order of 10,000. Not quite Ozempic numbers.
These numbers partly reflect the success of “precision medicine” in cancer. Creating therapies directed to specific cancer mutations (e.g., IDH1 or NRG1 mutations in 2024) can result in increased efficacy, reduced toxicity and more cost-effective clinical trials. Unfortunately, mutations shared across patients are the exception not the norm so precision medicine has resulted in therapies that are increasingly targeted to very small populations.
Time we don’t have
What does all of this mean for addressing the overall cancer burden? For simplicity, we focus on the 618K Americans expected to die from cancer this year and assume an average of 14 years of life lost. This equates to an overall cancer burden of 8.6M years.
Assuming each new therapy adds 4.9 months of survival and benefits 10,000 patients, each new therapy adds ~4,000 years of life. This implies >2,000 (8.6M / 4000) new therapies are needed to offset current cancer burden. At 16 new therapies per year, it will take >130 years to eliminate cancer deaths.
Ozempic for oncology
What if we wanted to eliminate US cancer deaths in 20 years?
At the extremes there are two strategies. Increase the number of cancer approvals from 16 to >100 per year while holding therapeutic impact (incremental survival and number of patients) constant. The alternative is to innovate more impactful medicines: Therapies that reliably add three years (instead of three months) of incremental survival for 30,000 patients would rapidly eliminate cancer burden in the US.
Despite the evident difficulty of developing efficacious cancer therapies, we believe the focus must be on more impactful therapies for two reasons. First, the practical one: Patients do not want to take a different medicine every 3 months for 14 years. Clinicians - even with AI support - do not want the complexity of hundreds of different medicines. Evidence generation (i.e., clinical trials) for so many agents would be an enormous challenge for the system. The logistics are hard.
The second reason is that we are very far from having 100 therapies per year that can benefit patients. This can be seen from the FDA Accelerated Approvals which allow for drug launch on the basis of promising early data (e.g., Phase 2 or surrogate markers). One follow up study of cancer Accelerated Approvals indicated that only 36% had “high therapeutic value” while another concluded “41% (19/46) did not improve overall survival or quality of life.” If the class of 2024 has modest benefits (i.e., 3 months incremental survival), they are nevertheless the best we have.
If making effective therapies is so hard, why should we believe that highly effective therapies are possible? A glib answer is Ozempic. Obesity was a wasteland for a long time and now we have a class of drugs that could benefit 100M Americans. Big breakthroughs happen. A second reason is that we have learned a lot about why cancer therapies fail – and conversely what could make them succeed. These are the topics of two follow on posts.