Study Calls for Basing Lung Cancer Screening Guidelines on Individual Risk

Hormuzd A. Katki, PhD

An individualized, risk-based approach to screening current and former smokers for the early detection of lung cancer would dramatically expand the population that could benefit from the preventive strategy, while saving more lives than the current US Preventive Services Task Force (USPSTF) guidelines for lung cancer screening, according to a new study.¹

In 2013, the USPSTF recommended computed tomography (CT) lung cancer screening for asymptomatic adults aged 55 to 80 years with a 30 pack-year smoking history (average of 1 pack of cigarettes per day for 1 year) who are currently smoking or have quit within the past 15 years.² Under these guidelines, about 9 million people in the United States would currently be eligible for lung cancer screening.

However, screening guidelines could instead be based on choosing those at highest individual risk of cancer. With an individual risk-based selection model, the report contends, there might be a 20% relative increase in estimated deaths prevented through low-dose CT screening above the number forecast through the USPSTF’s recommendations, even if the same number of people were screened.

As a result, the researchers suggested in the Journal of the American Medical Association article that the criteria be broadened to include more people at a high risk of developing lung cancer.

Lead author Hormuzd A. Katki, PhD, a biostatistician with the National Cancer Institute, and colleagues developed and validated risk models that suggest that screening half of all ever-smokers, defined as someone who has smoked at least 100 cigarettes in their lifetime, might result in preventing 90% of screen-preventable deaths.

Among this classification are two subgroups ignored by USPSTF guidelines: current smokers who have smoked less than the 30 pack-year limit over many years, and those who smoked a great deal more but quit more than 15 years ago. Some members in these groups still have a high risk of having lung cancer, especially when combined with additional risk factors.

Additional risk factors include characteristics such as age, sex, race, education, smoking frequency and duration, quit years, body mass index (BMI), family history of lung cancer, and self-reported emphysema. These factors are fed into a formula that produces an individual’s risk of developing lung cancer that can be used in determining if a patient should undergo CT screening.

Katki and colleagues with expertise in risk communication are currently evaluating their online risk tool for lung screening. The user inputs the risk factors, and the tool calculates the risks of lung cancer and of lung-cancer death, both for declining or entering into a screening program, thereby simplifying the challenge of determining and communication the rate of risk. The tool also provides the risk of having a false-positive CT screen.

“It’s a lot of numbers and it’s well known that communicating risks is very difficult, that patients have trouble understanding them and doctors aren’t that much better,” Katki said in an interview with OncLive. “A lot of research is being done now to try to develop a reliable shared decision-making process for eliciting risk factors, for communicating risk estimates, and then using the risks to make decisions about screening in light of the patient’s personal values.”

Katki and his coauthors developed their models using data from the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO). They then validated their models using data from PLCO, the National Lung Screening Trial (NLST), and the 1997-2001 National Health Interview Survey (NHIS).

For illustration, they discussed two potential risk-based screening programs. First, the fixed-population validation program would screen the same population size (9 million) as the USPSTF guidelines, but those at highest 5-year lung cancer risk (≥1.9%) based on their risk models. Second, the fixed-effectiveness program would use the same number needed to screen (NNS) to prevent 1 lung cancer death as the USPSTF guidelines (NNS = 194). This results in a lower risk threshold (≥1.7%), allowing 3 million more ever-smokers to be eligible for screening.

Under USPSTF guidelines, an NLST-like CT screening program with 3 annual rounds of CT screening would prevent an estimated 46,488 (95% CI, 43,924-49,053) lung cancer deaths over 5 years (57% of estimated CT-preventable deaths).

In comparison, the risk-based fixed-population program would screen the same number of people while preventing an estimated 55,717 (95% CI, 53,033-58,400) lung cancer deaths (68% of estimated CT-preventable deaths). This is a 20% increase in prevented deaths versus USPSTF guidelines, yet screening the same number of ever-smokers. The estimated NNS was 162 (95% CI, 157-166), a 17% improvement in screening effectiveness compared with the USPSTF guidelines estimated NNS of 194 (95% CI, 187-201). This program was estimated to have 13% fewer false-positive screens per prevented death than USPSTF guidelines: 116 (95% CI, 113-119) compared with 133 (95% CI, 128-137).

The fixed-effectiveness program showed a 34% relative increase in modeled CT-preventable deaths over the USPSTF guidelines with 62,382 (95% CI, 59,567-65,196) preventable lung cancer deaths (76% of estimated CT-preventable deaths) over 5 years. Under this program, 12.1 million ever-smokers would be screened.

The risked-based approaches focused on those who were at highest risk, replacing many lower-risk USPSTF-eligible smokers with those who were ineligible but had higher risk.

“The next step is to conduct a cost-effective analysis to pick a risk threshold to guide screening eligibility,” Katki said. An analysis would determine if it makes sense financially to change the guidelines, and ultimately decide who should be screened that would benefit the most from such a program.

Implementing the risk-based approach described by Katki and his coauthors would also require outreach in communities that have more limited access to screening programs. Many people who have a higher risk for lung cancer are unaware of their options for screening, as well as of the risks, yet they must not be ignored, Katki said. “It’s one of the lessons of risk-based screening that we need to reach out to underserved populations, for example, African-Americans, people who have less education, and people who may be older,” he said.

Katki and colleagues note that “although CT screening can reduce lung-cancer mortality by approximately 20%, the majority of lung-cancer deaths are not screen-preventable at this time. The best way for smokers to avoid lung cancer, and all smoking-related illness, remains to quit smoking as early as possible.”

References

1. Katki HA, Kovalchik SA, Berg C, et. al. Development and validation of risk models to select ever-smokers for CT lung cancer screening [published online May 15, 2016]. JAMA. doi: 10.1001/jama.2016.6255.
2. Moyer VA, US Preventive Services Task Force. Screening for lung cancer: US Preventive Services Task Force recommendation statement. Ann Intern Med. 2014:160(5);330-338.