Why oncology is such an attractive category for pharma companies

It’s obvious to anyone who takes a cursory look at the pharmaceutical industry that cancer is where it’s at. There are over 1300 cancer drugs in development as of 2019, compared to roughly 8000 drugs in total in development. Cancer drugs bring in $145 billion in revenue each year, the greatest of any drug category by far (the next category is diabetes, with about $50 billion).

Why is cancer so popular for pharmaceutical companies to target? A few reasons.

Funding wise, cancer is the only disease with its own $6 billion federal budget allocation, not to mention the funding from huge nonprofits like the American Cancer Society. Cancer drugs also have, due to complex reasons, very favorable economics, and can end up costing $100,000 per month per patient.

There are also political reasons. The US Patent Office offers a special expedited review for cancer immunotherapy. Similarly, the FDA offers a specific center for cancer “excellence”, which they don’t do for other diseases.

All of these are easy enough to find and pretty public. However, I want to talk about another reason, which is less known but very important. Cancer drugs are held to a lower standard than other drugs when it comes to approval. This is done by using bullshit measurements in cancer trials, measurements that aren’t allowed to be used in any other drug trials.

These bullshit measurements are called progression free survival (PFS) and objective response rate (ORR), respectively. PFS is based on the idea that cancer gets worse over time. So, PFS measures the time that a person’s cancer doesn’t “progress”, with the idea that a cancer drug works better the longer it can stave off progression. ORR, meanwhile, is a percentage calculation that measures the number of people who “respond” to the drug. The idea here is that a cancer drug works better if more people respond to the drug.

Both the “progression” in PFS and the “response” in ORR are based around the same idea, the idea that the progression that matters is the size of a tumor. If the tumor doesn’t grow, then that’s not a progression. If the cancer shrinks in response to a drug, that’s a response.

You might think that these aren’t necessarily meaningful things to measure. You’d be right. By this calculation, it doesn’t matter if everyone dies from taking the drug in 6 months. If their tumors shrank before they died, they responded to the medication.

Now, this might seem like just a technicality. Surely this isn’t how these measurements are supposed to be used. It probably seems unlikely that the FDA would approve a drug which would have no impact on overall survival, even if it halted or reversed tumor growth. Well, unfortunately, that’s actually better than a lot of drugs that the FDA approves.

Let’s take a look at one of the PFS offenders. Here’s the FDA’s announcement of their approval of tivozanib, a drug for renal cell carcinoma, a kidney cancer. I didn’t pick this drug specifically because it’s bad, I literally just picked it off the top of the list of most recent cancer drug approvals.

According to the FDA’s own press release, when tivozanib was compared against soranefib, the current best case drug, tivozanib extended “progression-free survival” by about a month and a half. Not amazing, but maybe ok, right? Well, tivozanib decreased survival by 3 months.

So, your tumors don’t grow for a month and a half, but you die 3 months sooner. If I do some bad math, I get to find out that you get to spend a month and a half in the grave with your tumors remaining the same size. That’s what PFS is like.

Now, let’s take a look at one of the ORR offenders. Again, I’m not trying particularly hard: this is 4th on the list of most recent drug approvals. Melphalan flufenamide is a drug for refractory multiple myeloma (i.e. multiple myeloma that hasn’t responded to other drugs). According to the press release, 23% of patients “responded” (i.e. had their tumors shrink) with the drug. The median “duration of response” (how long the tumors shrank for) was 4 months.

Notice what’s missing? Not only is this missing literally every single symptom that a patient with multiple myeloma would care about (survival, anemia, bodily pain), it doesn’t even define how much the tumors shrank. Literally, it’s just binary. Did it shrink at all? And, to add insult to injury, there’s not even a placebo. There’s no measuring of whether tumors would shrink without the drug or not.

No other drug category is measured like this. Every other drug category has quantifiable, clinically relevant endpoints. Even if the symptoms are difficult to quantify, like depression, doctors will at least ask the patients a battery of questions about how they’re feeling. Plus, every other drug category has some sort of comparison, whether it’s placebo or another treatment. But not cancer.

This isn’t to say there are no good cancer drugs. There are. But there are also a lot of bullshit ones, and the bullshit ones can still make boatloads of money, before and after approval. And that’s why pharmaceuticals are so cancer-focused.