Invited Commentary
Public Health
Lung Cancer Screening in Lower-Risk Individuals—Less Can Be More
Carl Martin Tammemägi
JAMA Netw Open 2025;8;(7):e2523050. doi:10.1001/jamanetworkopen.2025.23050
Although evidence demonstrating the efficacy of computed tomographic lung screening (CTLS) is established, global uptake of CTLS has been slow, and the optimal methods of implementing CTLS have not been determined and may vary in different populations. One of the pieces of the puzzle that has needed clarification is the ideal screening interval. de Nijs and colleagues1 conducted simulation studies to evaluate the cost-effectiveness of biennial vs annual CTLS in lower-risk individuals. They compared a variety of adaptive intervals according to the participants’ age, sex, and smoking exposures. They found that biennial screening, in particular at younger ages and lower smoking exposures, was cost-effective at a willingness-to-pay threshold of $100 000 per quality-adjusted life-year (QALY) gained. The biennial screening in lower-risk individuals led to 4.1% fewer deaths averted while performing 20.6% fewer scans. The findings of this study are not surprising, as many studies have already described encouraging findings regarding increasing CTLS intervals beyond annual screening,2-5 ongoing studies are actively evaluating biennial screening intervals,6 and biennial screening has already been incorporated into the provincial screening program in British Columbia, Canada.7
Many people believe that if CTLS works well in high-risk individuals, it must work somewhat or moderately well in low-risk individuals, when in fact the mortality reduction in low-risk individuals is minimal or nonexistent.8 The initial US Preventive Services Task Force (USPSTF) eligibility criteria already selected a sizeable proportion of lower-risk individuals. To improve sensitivity and reduce racial, ethnic, and sex disparities, the USPSTF broadened its eligibility criteria in its 2021 recommendations. This further increases the proportion of lower-risk individuals who are eligible for CTLS and increases the likelihood of potential harms, such as invasive clinical investigations in the absence of cancer. If an expanded screening interval approach were combined with the USPSTF eligibility criteria, the number of screenings of low-risk individuals and potential for harm would be reduced. An alternative approach to the USPSTF inclusion of low-risk individuals is to base screening selection on accurate, validated lung cancer prediction model–estimated risk. Lung cancer risk prediction model estimates can direct individuals to biennial screening and can also recommend no screening for low-risk individuals who are unlikely to benefit. The idea of modifying the screening interval has been advanced even further in the ENGAGE (Individualized Lung Cancer Screening Decisions) framework. ENGAGE uses simulation to incorporate model-based risk, life expectancy, and additional information to provide personalized screening intervals. These can be shorter or longer than the 1-year interval.9 In the future, individualized screening intervals with an ENGAGE-type approach may further improve CTLS mortality benefits and cost-effectiveness.
An important benefit from increasing screening intervals for lower-risk individuals is a reduction in harms done, including harms associated with false-positive results such as unnecessary invasive procedures for benign diseases. Although the rate of these harms is already low in high-quality CTLS programs, there are opportunities for further improvements. An additional potential benefit of introducing appropriate biennial screening may be improvement in uptake and adherence to screening. Currently in the US, CTLS uptake and adherence have been disappointingly low. Making CTLS simpler and less expensive is likely to improve uptake and adherence.
The most serious harm that can occur from adopting extended screening intervals is the failure to diagnose some early-stage lung cancers and thus failing to avert the deaths caused by them. For the adaptive program that starts with biennial screening from 50 to 59 years of age and annual screening from 60 to 80 years of age, there is a 4.1% reduction in deaths averted and a 20.6% reduction in the number of scans performed. Compared with annual scanning, 1 fewer death will be averted per 1000 individuals undergoing CTLS (Table 2 in de Nijs et al1). Given the expected sojourn time (time from cancer detectability to clinical symptoms) of 4.2 years for adenocarcinoma in women and 3.7 years in men, the estimated number of deaths not averted seems realistic.
Being able to extend the screening interval appears to be reassuring to some. However, it may be concerning to others at the same time. CTLS that finds abnormal results that do not lead directly to lung cancer diagnoses following clinical investigations may be associated with increased future lung cancer risk even more than 3 years later.10 Having the option to double the screening interval for those at lower risk should not lead to the impression that the seriousness of abnormal results can be discounted.
The study by de Nijs et al1 contributes supportive knowledge regarding optimizing the CTLS interval. Although the primary comparator in the report was the USPSTF lung screening recommendations, the findings are expected to generalize broadly to other categorical age-smoking CTLS eligibility criteria. Application of this information has the potential to make lung screening more efficient and cost-effective, can lead to reduced potential harms, and can promote CTLS uptake and adherence. Science, knowledge, and evidence and the policies that they drive advance in iterative, incremental steps. The approach to optimize the screening interval supported by the findings of the current study is expected to further improve clinical and public health CTLS practice.
