Cancer Medicine Approvals in the US

The medical journal JAMA Oncology published a research letter a couple of days ago that I found really interesting. A "research letter" is a kind of informal discussion of research results, very different from the long articles that report of clinical trial results. But it provides some historical content for cancer research as a whole that should be very encouraging for all of us.

The article is called "Cancer Medicine Approvals in the US," and it looks at the circumstances surrounding the approvals of cancer treatments by the FDA for the last 20 years (from January 1, 2006 to December 31, 2025). The research focused on treatments for solid tumors, so there were no blood cancer treatments involved. There are also no pediatric treatments or generic drugs or biosimilars.

But even without the blood cancer treatments that would be most relevant to us, it still gives us a picture of the kinds of trends that have happened in the last 20 years. I think those same trends are probably true for those other treatments that aren't included in the data.

The researchers found that from 2006 to 2025, the FDA approved 385 different cancer treatments for solid tumors. Of those, 154 (40.0%) were for brand new treatments, and 231 (60.0%) were for new indications for approved treatments. For example, a treatment might have been approved for a first-line breast cancer treatment that was already approved for relapsed breast cancer. 

Another breakdown: of the 385 approvals, 105 (27.3%) were approved through the Accelerated Approval process. This means they were approved using data from a smaller phase 2 clinical trial, rather than a larger phase 3 trial. Accelerated Approvals are granted when a treatment meets a particular need that isn't being met by any other available treatment. That number is lower than I would have expected. It seems like there are lots of Accelerated Approvals these days for blood cancers, and they don't always work out well. So I feel a little better knowing that about three quarters of approvals are from the regular process. Accelerated Approvals often work out great, but they concern me sometimes. 

Another interesting bit of data: Traditional chemotherapy accounted for 40% of all approvals from 2006 to 2010. However, for the last 5 years, they have accounted for only 3.9% of approvals. This is not at all surprising. We can see it in blood cancers -- R-squared, CAR-T, bispecifics, monoclonal antibodies. All of them are proving to be good alternatives to traditional chemotherapy. They are more targeted, so they do less damage to non-cancer cells and have a different set of side effects. And even though the word "cure" was recently applied to R-CHOP, an immunochemotherapy combination, the excitement lately comes from those non-chemo options. 

Another interesting bit of information from the research is about endpoint markers. These are the data that ultimately proves whether a trial is a success or not. The best endpoint should be Overall Survival -- how many patients were still alive after 5 years or 10 years or whatever length of time they need to compare to. If a treatment keeps more people alive, that should be the measure of success, right?

It's more complicated than that. Only 102 of the 385 trials (26.5%) had Overall Survival as an endpoint. And that number is declining -- it went from 40% of trials from 2006 to 2010 to 18.7% from 2021 to 2025. Instead of Overall Survival, most trials used a "surrogate endpoint" -- a different statistic that shows "success." In blood cancers, especially in Follicular Lymphoma, the surrogate endpoint is often PFS -- Progression Free Survival, or the number of patients whose disease did not get worse over a period of time. This makes sense for FL -- if the Median Overall Survival is now 18-20 years for FL patients, it could take much too long for a treatment to be approved. No drug company is going to wait 20 years for the data they need to get approval. So while PFS isn't a perfect surrogate, it's good enough for FL.

For this research on solid tumor treatment approvals, PFS was the most common surrogate from 2006 to 2015 (60% of all trials). But from 2016 to 2025, the most common was Response Rate (46.7%). This strikes me as more problematic. A response to a treatment is great. But some responses last only for months. At least PFS typically goes on for a couple of years. Of course, I understand that not all PFS measures are long, and not all Response measures are short, and some for some cancers with few treatments available, any response is a big success. It's all more complex than these numbers suggest. But in my observation, it seems like PFS, and not Response Rate, is the primary endpoint for most new treatments, which is good.

One final bit of data comes from the way trials are constructed. Many consider the "gold standard" for clinical trials to be a randomized, two-arm controlled trial. This type of trial involves randomly splitting patients into two groups. One group gets the "standard of care" -- the treatment that is accepted as the best available at the time. This is the "control" group. The other group gets whatever new treatment is being tested. In this way, the researchers can guarantee that the two groups are the same and that a fair comparison is being made between the old and the new. The alternative is a "single arm" study. There is no direct comparison because there is only one group. The data from the trial is compared to another trial that took place in the past. There is no guarantee that the old and the new are the same. It's much easier to run a single-arm study like this, and they are more common in phase 2 trials than phase 3 trials.

This research found an increase in approvals based on single-arm trials -- 12.9% from 2006 to 2010, up to 40% from 2021 to 2025. 

The researchers conclusion says "Continued decline in the proportion of approvals based on overall survival, increase in response rate–based approvals, and underutilization of the accelerated approval pathway all reveal concerning trends." Interesting that they want to see more Accelerated Approvals. 

As I said, this research doesn't include blood cancers, but the trends that they see are pretty close to what I have observed over 18 years of paying attention to the research on FL. I'll admit that I have become kind of a Clinical Trial Nerd over that time, and I have some definite opinions about some things. But I will also remind you that I'm not a doctor or a cancer researcher, so keep that in mind.

The big lesson for me in all of this is that FL has a number of treatment options available to us. The whole reason I've become such a nerd is that there is so much research for me to read. So even when a treatment doesn't get approved, or gets pulled after approval, we still have other options. Don't ever lose sight of that. 

Stay well. More soon.