We’re naming cancers all wrong, oncology leader says

Naming cancers solely by the organs they originate in is getting a bit old, according to Fabrice André, a medical oncologist at Gustave Roussy in France and the president-elect of the European Society of Medical Oncology. Instead, André hopes to push for a new naming system that emphasizes the molecular characteristics of a cancer, regardless of its tissue of origin.

That’s because, in the last several decades, science has uncovered the ways genetic alterations can drive the growth and development of cancers — and how those alterations can be targeted with medicines to melt tumors away. In many cases, these mutations aren’t limited to cancers of a single organ, nor do all cancers from an organ share the same mutations. Two patients may both have breast cancer, but if one is a triple-negative cancer and another is packed with HER2 proteins, the treatment will look very different, André said. That can cause confusion for patients.

“Sometimes we see patients, and they say I have breast cancer. My friend also has breast cancer, and our treatments are totally different. It means one of our two doctors is wrong,” André said. “And it’s very difficult to reassure the patient. There are many examples like this just because we don’t represent cancer based on its real identity.”

That real identity being, of course, the genetic underpinnings of the tumor. If this sounds all very academic, André would argue the naming conventions for cancer can have real-world consequences for patients. One example is drugs that target the immune checkpoint PD-1, a protein that can tamp down the activity of immune cells when activated. Some tumors take advantage of this by carrying a protein called PD-L1 that can activate PD-1 on immune cells, essentially cloaking themselves from the immune system. Antibodies that block one of these proteins interrupt that interaction, essentially revealing the tumor to the immune system and allowing immune cells to attack.

Melanoma was the first testing ground for drugs targeting the immune checkpoint PD-1, and the FDA approved the landmark anti-PD1 therapies Keytruda and Opdivo for patients with advanced melanoma in 2014. But many other tumor types can also benefit from PD-1 therapy, and patients with those cancers had to wait many years for trials to be completed on their tumor type.

“Millions of people with tumors expressing high levels of PD-L1 were not able to access relevant drugs because trials had not yet been conducted for their type of cancer,” André wrote in a commentary published in Nature on Wednesday. “Those with certain breast or gynecological cancers expressing PD-L1 had to wait 7-10 years to access PD1 inhibitors.”

STAT spoke with André about how scientists might reconsider the popular naming of cancers, and how that could benefit patients. This interview has been edited for length and clarity.

So, do you think we should do away with organs of origin when it comes to naming or describing cancers?

It’s not going to be that we cancel organ-based classification. That’s not what the paper is saying, and nobody wants that. We are just saying, look, cancer is a biological disease, so please name it by its biology. The current classification is kind of the heritage of local therapy. For surgery, if the cancer is in the breast, then it’s different than if it’s in the lung. But nowadays, many therapies actually target specific proteins. These proteins can be common to many cancers.

Some may say, well, some targets are specific to some organs. Yes, of course. But let’s say you are treating a metastatic prostate cancer. Ultimately, you are treating a cancer driven by androgen receptor. You don’t need to know that it’s prostate cancer. It’s an androgen receptor-driven cancer, and you give your treatment based on this understanding of the biology of the disease.

We’re doing more and more genetic profiles for treatment decisions. We see that more and more patients present with genomic alterations that would make them eligible to a therapy if the origin of cancer was different. It’s a little bit strange. So, that’s the rationale to develop a more biological-based classification. We’re not the first or the only one to say that.

Don’t scientists and physicians already refer to many cancer subtypes by their molecular characteristics? Breast cancer is often referred to as HER2+ or triple-negative breast cancer, for example.

Yes, that’s true. But there’s also an impact on patients. We think, and there’s no evidence for this, but we think if we change the way we name cancer by its biology, we think it will be easier for patients to understand — therefore be more engaged with the treatment. If you understand how your treatment works, it will be easier to be engaged in your care. And it’s something very important for the representation of the disease.

The overall message is that nowadays we should represent the cancer disease by its biology rather than just its anatomy.

So, you’d envision a world in the future where one person may describe their cancer to another patient by saying like, “Oh, I have an EGFR mutant cancer that’s in my lung,” instead of just, “I have lung cancer?”

Yes, yes. I think we underestimate the importance of this for patients.

How would such a classification system accelerate access to new medicines for patients?

We’re not saying it was a mistake to classify cancers based on anatomy or that people should have done things differently in the past. But, look, the framework is no longer adapted to the types of treatments that are available. Making 20 to 30 trials sequentially, testing a drug in different patient populations, delays access to therapy. New drugs should be developed based on molecular alterations and, for example, you include patients based on the presence of biomarkers and include multiple cancers. You can see if all those cancers present the same sensitivity, then you develop the drug.

This speeds up development, and the benefit is some patients will get access to drugs quicker. Also, it may benefit some patients with rare cancers that are not included in drug development. If we focus on biomarkers or molecular characteristics for inclusion, then it will also include some rare cancers where drugs will never be considered for approval because we changed the definition.

It’s so difficult when the patient dies, and they could’ve had access to effective medicine. It’s difficult. It’s about putting the patient first.

Correction: An earlier version of this article misspelled the name of the drug Opdivo.