Changing the definition of “cancer” isn’t going to be a cure-all

Scanning our language for the right way to talk about cancer.
Scanning our language for the right way to talk about cancer.
Image: Reuters/Jim Bourg
We may earn a commission from links on this page.

The field of linguistics has a subset called “medical discourse,” a term we can at least trace back to the publication of Peter Middleton’s eponymous work in 1769. It’s through this lens that we can examine the recommendation made in The Journal of the American Medical Association that we redefine cancer—not as a malignant neoplasm, not as kαρκινος—but by such names as “indolent lesions of epithelial origin.” While there is a medical reality to consider, I can’t help but greet this idea with caution and reserve. Language is capable of too many things (subtle and otherwise) to have the idea of changing a term brought to the fore “just like that.”

The reason for the proposed change—as The New York Times puts it, is that:

Some premalignant conditions, like one that affects the breast called ductal carcinoma in situ, which many doctors agree is not cancer, should be renamed to exclude the word carcinoma so that patients are less frightened and less likely to seek what may be unneeded and potentially harmful treatments that can include the surgical removal of the breast.

Lauren Brown previously wrote in Quartz about Angelina Jolie’s mastectomy, and how—for some women—“studies have suggested that about a quarter of screening-detected cancers might have gone away on their own,” which raises the question as to whether redefining some aspects of cancer will stem unnecessary treatment. The article in the Times partially speaks to that, but over-diagnosis doesn’t just occur in cancer—only 38% of adults in the US meet the 12-month criteria for depression, for instance. It begs the questions of whether patients are being sufficiently informed about statistics like this and the idea that their “cancer” could simply go away. (This stands in fascinating contrast with a 1989 court case in Japan that once ruled that doctors didn’t need to tell their patients they had cancer at all.)

I asked Dr. James Coyne, the director of the Behavioral Oncology Program at the University of Pennsylvania, whether he thought renaming types of the disease was sufficient at preventing unnecessary surgery. He told me this:

Some of our worst off are young women with a diagnosis of “stage 0” DCIS [as in, a DCIS has been detected—EF.] They don’t know what to make of the uncertainty and often receive aggressive interventions when it is not really known whether their cancer will progress or even if it is appropriately labeled cancer.

I asked him whether he thought it is sufficiently communicated to patients that a quarter of screening-detected cancers may go away on their own and he told me that the word “cancer” and “the indoctrination [associated with it] has been far too powerful.”

Dr.  Kenny Lin, deputy editor of the American Family Physician Journal, agreed with Coyne.

“To me,” he wrote in an email, “‘cancer’ should be defined not only by the appearance of cells under a microscope, but also its behavior” —whether or not it spreads, how quickly, how often it causes certain symptoms and whether or not it kills.  “I think that we can go even farther along these lines,” he added. “If a pseudo-cancer (or IDLE, or whatever you call it) isn’t going to bother a patient during his or her lifetime, then we shouldn’t look so hard for it in the first place. Less aggressive approaches to screening are an integral part of this strategy.”

I find it difficult to agree wholeheartedly. Neither really answers the question when it comes to the nature of the word itself. Coyne’s assertion is troubling: the term “cancer” is so loaded that it rebuffs attempts at communicating various degrees and permutations. There is no way to communicate; therefore, the word itself must change. What is needed is actually more communication, and more clear terminology, not less.

The discussion around the word cancer belies the fact that even basic communication between doctors and patients is not only fraught but altogether absent. According to the New England Journal of Medicine from 2006:

According to one study, no interpreter was used in 46 percent of emergency department cases involving patients with limited English proficiency. Few clinicians receive training in working with interpreters; only 23 percent of U.S. teaching hospitals provide any such training, and most of these make it optional. Data collection on patients’ primary language and English proficiency is frequently inadequate or nonexistent. Although no federal statutes require the collection of such information, no statute prohibits it, either.

In 32 of 35 malpractice cases studied by the National Health Law Program (pdf) at UC Berkeley in 2008, no competent interpreter was used despite the fact that language barriers existed. The study found that the inability to communicate results in poor health outcomes and even death and billions of dollars in malpractice cases.

To tackle this problem, Washington, Iowa, Indiana, and Oregon have developed, or are developing “interpreter certification programs.” California frequently provides written summaries of various treatments in a language understood by the patient. Georgia only has a law on the books to have a proper interpreter on hand to explain “the risk of carrying an unborn child to term in a language understood by the patient … when she is considering an abortion,” says another paper (pdf) put out by the NHLP at UC Berkeley.

But as this debate reveals, even speaking the same language does not ensure clear communication.

Shifting one set of words in favor of another could also create a “Will Rogers Phenomenon”—what happens when one element of a data set is recategorized to another data set, and the averages of both increase. Dr. Arjen Noordzij, who practices urology in the Netherlands, tweeted out this very paradox:

Lin seconds Noordzij’s thinking:

Will reclassifying some cancers as non-cancers make the prognoses of the remaining true cancers appear worse than before? That’s the likely outcome, and the transition may cause havoc with cancer statistics. But in a sense it’s returning to a more honest assessment of cancers in the years before screening tests became widespread. Prostate cancer 5-year survival rates, for example, skyrocketed after PSA screening started—mostly because we were inflating the pool of prostate cancers with things that didn’t behave like cancer, not because screening and early treatment were really that effective.

There is the medical reality to consider. Anything that encourages greater accuracy in understanding and treating disease is worthy of our attention. But simply changing the words without the necessary dialogue will likely leave patients sick as well as confused.