Wednesday, January 15, 2020

When a cancer diagnosis predicts future good health

The American Cancer Society recently reported the largest recorded one-year decline in the U.S. cancer death rate (2.2 percent) from 2016 to 2017, driven by lower lung cancer and melanoma mortality. Although the relative contributions of screening / early detection, changes in risk factors (e.g., smoking, sunscreen use), and new cancer therapies can be debated, the continuing decline is positive for these patients, even in the context of troubling increases in overall death rates.

Despite this apparent progress, hardly anyone would call a cancer diagnosis "good news." But as my longtime readers know, all cancers are hardly created equal. Some are almost invariably lethal, some are "cancer" in name only, and most reside somewhere in between. A less heralded research letter in JAMA Internal Medicine in December presented relative and disease-specific survival statistics of 10 early-stage cancers in three illustrative graphs. (I wish that I could reproduce the figures here, but the American Medical Association takes copyright infringement quite seriously and requests prohibitive fees for reprinting its content, even if the data are derived from a public source such as the National Cancer Institute's Surveillance Epidemiology End Results (SEER) registry - which these are.)

Dr. Andrea Marcadis and colleagues found three patterns in the SEER data from 2004 to 2015. The first one you might expect: for early oral cavity/pharyngeal, bladder, and lung/bronchus cancers, relative survival (comparing survival of a cancer population to that of a matched population without cancer) was worse than disease-specific survival (the proportion of patients not dying of their cancer), likely related to a higher smoking prevalence in these patients. The second pattern, which held true for testicular and pancreatic cancer, showed relative survival that was similar to disease-specific survival, even though the prognosis of testicular cancer is generally excellent, while the prognosis of pancreatic cancer is usually terrible.

The third pattern was the most interesting: for early prostate, breast (including ductal carcinoma in situ, which was considered separately), thyroid cancer, and melanoma, relative survival was not only better than disease-specific survival, but greater than 100%. In other words, patients with these particular early cancer types were more likely to survive than similar individuals without cancer.

How can a cancer diagnosis possibly predict future good health? The answer emerges from what these cancers have in common that the others don't: we screen patients for them, even though the evidence of net screening benefit varies quite markedly. Patients who undergo regular screening are also more likely to follow their doctors' advice and practice other health-enhancing behaviors. The researchers noted:

These statistics can help physicians explain to certain patients that their actions prior to diagnosis may mean more for their survival than the diagnosis itself. This may be reassuring and mitigate the risk of depression and anxiety associated with cancer diagnosis. ... Relative survival may help patients reorient their expectations by providing a comparison to a similar person without cancer, revealing that not every cancer diagnosis necessarily portends an earlier death.

I don't mean to suggest that anyone welcome a cancer diagnosis. Even cancers with a great relative prognosis cause emotional stress, more medical visits and expenses, potentially uncomfortable or risk-associated diagnostic procedures, and possibly chemotherapy, radiation therapy, or surgery. But it's critical for patients and clinicians to understand that some cancer (over)diagnoses are more the result of doctors looking too hard for them (see South Korea's "thyroid cancer epidemic" caused by an ill-advised national screening program) than one's genetics or behavior.