Lung cancer remains a leading global health burden and the most common cancer type in recent estimates. Population screening with low-dose CT (LDCT) is cost-effective primarily in people at higher risk, notably current or recent cigarette smokers. Unlike many other cancer screening programmes that use age alone, LDCT screening eligibility is commonly determined by predicted absolute risk, often estimated with validated models such as PLCOm2012 and LLPv2.
Quantifying tobacco exposure is central to lung cancer risk prediction. Pack-years or cigarette consumption (duration and number per day) are widely used, but tobacco exposure can also come from cigars, cigarillos, pipes and smokeless products. Many risk models have historically focused on cigarette use because the development cohorts included few users of other products, particularly in US-derived models like PLCOm2012. Systematic reviews indicate that cigar and pipe smoking, with or without concurrent cigarette use, are associated with significantly increased lung cancer risk compared with never-smokers.
Because screening eligibility thresholds are sensitive to tobacco exposure, failure to account for multiple tobacco products could misclassify individuals with substantial non-cigarette use as having insufficient risk and therefore make them ineligible for LDCT. The SUMMIT study was used to examine the effect of cigars, cigarillos and pipe smoking on measured lung cancer incidence and on screening eligibility when these products are included in risk estimates.
SUMMIT is a prospective cohort study set within a LDCT screening service operating across four sites in North and East London. It recruited people aged 55–77 who were registered at participating GP practices and had a medical code indicating they had smoked within the past 20 years. The study aimed both to evaluate delivery of LDCT screening to a high-risk population and to validate a multi-cancer early detection blood test.
At the baseline Lung Health Check, individuals were eligible for LDCT and for the study if they met either the USPSTF 2013 criteria (minimum 30 pack-year history and, if former smokers, quit within the past 15 years) or had a PLCOm2012 risk estimate ≥1.3%. For these initial eligibility criteria, only cigarette consumption was considered.
In addition to standard cigarette history, participants were asked about cigar, cigarillo and pipe use. Frequency categories included occasionally (less than weekly) or regularly (at least weekly). Regular users reported current versus former status, age they started using the product, and the amount used per week. These data were used to evaluate how inclusion of non-cigarette tobacco products affected individual estimated risk and resultant screening eligibility under the two validated models used in the UK programme: PLCOm2012 and LLPv2.
Among approximately 13,000 participants who were current or former cigarette smokers, 511 participants—about 1 in 25—were regular users of cigars, cigarillos, or pipes in addition to cigarette smoking. The observed 5-year cumulative incidence of lung cancer among people who currently smoked both cigarettes and other tobacco products was 8.85%, compared with 5.74% among current cigarette-only smokers.
When assessing the subset of 511 dual-product users for changes in screening eligibility after including their other tobacco consumption, the proportion whose estimated risk moved from below to above the screening threshold differed by model. Using LLPv2, 0.6% of these participants crossed from below to above the eligibility threshold when non-cigarette use was included. Using PLCOm2012, 8.2% of participants moved from below to above the eligibility threshold after including other tobacco products in their tobacco exposure estimate.
These findings indicate that, within this screened cohort, concurrent use of cigars, cigarillos or pipes was associated with higher observed lung cancer incidence and that inclusion of these products in risk estimation increased the number of individuals meeting screening risk thresholds—more markedly for PLCOm2012 in this dataset.
SUMMIT data demonstrate that regular use of cigars, cigarillos or pipes alongside cigarettes is associated with a higher 5-year lung cancer incidence compared to cigarette-only current smokers. Importantly, ignoring non-cigarette tobacco consumption when estimating lung cancer risk can result in some individuals being incorrectly classified as below screening eligibility thresholds, particularly when using models that were not developed with significant non-cigarette product prevalence in their derivation cohorts.
The authors conclude that all tobacco consumption should be considered when determining LDCT screening eligibility to avoid missing individuals who may benefit from screening.
This study is registered at ClinicalTrials.gov, identifier NCT03934866.