There's been talk lately in the gastroenterology crowd: Instead of screening most everyone for colorectal cancer when they turn 50, maybe we could save more lives if we screen patients at the age when their own risk factors (such as smoking or obesity) indicate they would benefit.
It sounds reasonable, and it's in line with current thinking about personalized medicine. Uri Ladabaum, MD, head of the Gastrointestinal Cancer Prevention Program at Stanford, shares the enthusiasm but strikes a cautious note:
Personalized screening would be ideal, but we have to make sure we can implement it in practice. If we get fancy and start thinking about who's a smoker, who's overweight or has diabetes, put this and that into the recommendations, will doctors embrace this? Will patients embrace this? Or do we risk creating so much confusion that people end up paralyzed from too much information and we have lower screening participation rates?
Sometimes keeping it simple is better.
Most physicians in the United States recommend a colonoscopy or alternative screening tests when their patients reach 50, earlier if there's a family history of the disease. Colonoscopy, in which a physician runs a scope through the colon, is highly successful at finding early cancer and in removing pre-cancerous polyps. Death rates from colorectal cancer have dropped by half in the last 40 years, a change that Ladabaum said is largely due to screening.
Risks for developing the disease vary: Patients who are smokers, who are overweight, who drink alcohol or who smoke are more likely to develop colorectal cancer. Conversely, taking aspirin for five or more years reduces the risk of developing the disease.
Personalizing screening has great promise, but it also carries its own risks, Ladabaum writes in a Gastroenterology article co-authored with Douglas Robertson, MD, a professor at Dartmouth's medical school. Ladabaum outlines a few of the promises and problems:
Successful risk-stratification and tailored screening would optimize clinical outcomes and resource use. But we are not sure that current risk prediction models are accurate enough yet to avoid missing an unacceptable number of cancers in people identified by the models as low risk.
Also, we need an easy way to issue recommendations for the age to start screening, and when to repeat it, because overworked primary care physicians are unlikely to employ complex models to come up with different age recommendations.
Ladabaum added that research he's conducted suggests that when patients are told they have a lower risk of colorectal cancer, they may be less likely to undergo screening, even when it's recommended. "We need to be careful not to discourage the lower-risk group, because their level of risk is still clinically relevant," he said.
He pointed out that the current practice of screening everyone by age 50 has worked well overall. "What we're doing has caused dramatic decreases in deaths," he said. "Could we do better? Yes. But if we're going to change things, we should proceed with caution."