BACKGROUND:Soluble Fas (sFas) plays multiple roles in tumorigenesis and cancer dissemination by avoiding apoptosis via connection to Fas ligand. We assessed associations of serum sFas levels with the incidence of extrahepatic biliary tract cancer (eBTC) in a prospective case-control study nested in the Japan Collaborative Cohort study. METHODS:A baseline survey was conducted from 1988 using blood samples from 39,242 individuals (35% of 110,585). Patients diagnosed with eBTC, before the end of 1997, were regarded as cases. Odds ratios (ORs) for cancer incidence associated with sFas were estimated by conditional logistic regression. RESULTS:This study included 73 cases and 215 controls. Participants with high sFas showed an elevated risk of cancer (P-trend = 0.01) and the third tertile of sFas showed a higher risk compared to the first tertile (OR 3.27, 95% confidence interval [CI] 1.37-7.80). In both gallbladder and extrahepatic bile duct cancers (n = 33 and 40, respectively), high sFas was associated with elevated risk (P-trend = 0.04 and 0.02, respectively). In women and the non-elderly, subjects in the highest tertiles showed higher cancer risk. Limiting subjects to those followed for over 2 years, high sFas was related to cancer risk (P-trend = 0.02) and the third tertile showed a higher risk compared to the first (OR 3.03, 95%CI 1.14-8.07). CONCLUSIONS:High serum sFas may be related to future risk of eBTC. IMPACT:These findings highlight the need for additional studies of this biomarker in large prospective cohorts and through integrated analysis involving other cohorts.
BACKGROUND:Air pollution (AP) is a known cause of lung cancer (LC); growing evidence links AP with breast cancer (BC) risk. Many epidemiological studies analyze PM2.5, though AP is a complex mixture of many pollutants. New metrics are necessary to fully represent the carcinogenic constituents of AP. Therefore, we aimed to develop cancer-specific metrics to determine where and what types of AP drive incident cancers. METHODS:We combined databases of AP across the continental US in 2005-2020, including the Risk-Screening Environmental Indicators (RSEI) model and CACES land-use regression-estimated concentrations of criteria air pollutants (CAP). We joined AP concentrations at census tracts and employed meta-analytic cancer-specific risk estimates to weight the CAP and produce a set of combined metrics (AP for BC (APBC), AP for LC (APLC)) from numerous sources. RESULTS:APBC and APLC exhibited positive skew, with certain tracts having values approximately an order of magnitude above the mean, and were typically highest in large cities and along transportation corridors. Relative to RSEI, APBC and APLC were larger in certain regions including Los Angeles and New York City. Both varied with respect to race, ethnicity, and housing and exhibited significant spatial clustering. CONCLUSION:We developed novel AP metrics incorporating emissions from industrial facilities, along transportation corridors, and from power plants to characterize cancer-specific risks. IMPACT:Future research will use these metrics that represent multiple simultaneous sources of AP to identify specific constituents of the AP exposome relevant for BC and LC, informing strategies to reduce their incidence and disparities.
The COVID-19 pandemic exposed critical vulnerabilities in cancer screening systems while revealing strategies that enabled resilience. Analysis of data from the National Cancer Institute's Population-based Research to Optimize the Screening Process (PROSPR) consortium across 10 US health systems documented dramatic declines in cervical, colorectal, and lung cancer screening during April to May 2020, with some sites experiencing drops of up to 96%. Systems with established mailed fecal testing programs recovered more rapidly, underscoring the value of home-based screening infrastructure. Drawing on these findings, this commentary outlines actionable strategies for sustaining cancer screening during future public health emergencies: expanding home-based testing options, implementing risk-stratified prioritization frameworks, leveraging digital communication tools, investing in telehealth infrastructure, strengthening patient navigation and community partnerships, and reducing structural barriers through extended hours and mobile services. These strategies are not solely crisis measures. They represent best practices for equitable, patient-centered care delivery in ordinary circumstances. As future disruptions are inevitable, health systems must proactively embed these approaches, conduct regular preparedness planning, and invest in research to model long-term impacts on cancer outcomes. Doing so will protect continuity of care, particularly for the most vulnerable populations. See related article by Del Vecchio et al., p. 1527.
The advent of new multicancer early detection tests and publication of early diagnostic results have generated expectations of clinical benefit from multicancer screening. The clinical benefit of a cancer screening test depends critically on disease natural history, which is typically learned from prospective screening studies. Retrospective studies of stored blood specimens are important in learning about a test's preclinical diagnostic performance but have rarely been used to infer natural history. The extent to which these studies might be harnessed to also learn natural history is discussed in the context of an article in this issue that infers the combined natural history of a range of cancers targeted by a multicancer early detection test using a case-control subsample of specimens from a large cohort study. The critical question concerns the identifiability of key transition rates in multistate models of natural history alongside state-specific sensitivities. The article suggests that these parameters are estimable within a Bayesian framework that leverages prior information about test sensitivity from diagnostic studies. We offer a heuristic discussion of identifiability in this setting and encourage formal study to determine the extent to which models with varying degrees of complexity may be learned from stored-specimen studies. See related article by Dai et al., p. 1535.
Previously considered primarily only in the context of dental diseases, the oral microbiome is now recognized as a contributor to a variety of systemic diseases, including cancer and cardiovascular diseases. This commentary explores the evolving view of the oral cavity as a gateway to the body's broader physiologic networks, implicating oral microbiome dysbiosis in a spectrum of chronic conditions. Drawing on current evidence, we propose a re-envisioned healthcare model that integrates oral and systemic health and outline research gaps and clinical priorities to harness the oral microbiome for preventive and therapeutic gains, emphasizing cancer-relevant biomarkers and intervention opportunities.
Rare cancers account for almost one fourth of all cancers. Sarcomas belong to the group of cancerous diseases with an incidence of less than 6 cases per 100,000 inhabitants. Over the past decade, activities were launched worldwide to elucidate the peculiarities of many of the more than 100 sarcoma subtypes described in the World Health Organization handbook. The major contributor to exact diagnosis is molecular pathology. The subgroup of ultrarare sarcomas (URS) poses a significant problem as each URS type has its own morphology, biology, natural history, and prognosis. In 2020, 35 international sarcoma centers agreed to standards of evaluating URS. The threshold was set to an incidence of less than 1 case per 1,000,000 inhabitants, and 77 URS subtypes were defined. Also quality criteria for centers to be selected for retrieving data to registries were consented. This issue of Cancer Epidemiology, Biomarkers & Prevention contains the first article to validate these principles of URS using the data from a nationwide cancer database. The authors from Taiwan also pointed out limitations of the approach. Combination with the national death database allowed to calculate overall survival (OS) and identified age as a significant factor for OS per URS type. These new data might foster future research on diagnosis and treatment of URS. See related article by Lee et al., p. 1654.
BACKGROUND:Tobacco and opium consumption are prevalent in Iran and have been linked to increased risk of upper GI cancers. To investigate their role in lower GI cancers, the associations of tobacco and opium use with liver and colorectal cancer mortality were examined in this study. METHODS:From 2004-2008, participants aged 40 to 75 years were recruited through random systematic cluster sampling and primary healthcare worker contact in Golestan, Iran. Baseline assessment of tobacco and opium use, demographic information and cancer-relative risk factors were conducted through validated questionnaires. Cancer diagnoses and deaths were collected through regular follow-up between 2004 and March 6, 2023. Multivariate Cox proportional regression model evaluated the association between tobacco and opium use and liver and colorectal cancer mortality. RESULTS:During a median 15 years of follow up, 112 participants developed liver cancer and 201 developed colorectal cancer. Cigarette smoking (hazard ratio [HR] = 1.76, 95% confidence interval [CI] 1.05, 2.94) and water-pipe smoking (HR = 3.20, 95% CI 1.12, 9.13) were associated with liver cancer. Opium was not associated with liver cancer. We observed no associations between tobacco or opium use and risk of colorectal cancer. CONCLUSIONS:Cigarette and water-pipe smoking were associated with increased liver cancer mortality, whereas no significant associations were observed for tobacco and opium use and colorectal cancer mortality. IMPACT:Tobacco control could be a strategy for reducing liver cancer mortality in high-risk populations. Larger prospective studies are required to confirm the associations of tobacco and opium use with lower GI cancers.
BACKGROUND:Compared with non-Hispanic White (NHW) women, Black women with epithelial ovarian cancer (EOC) experience poorer survival, partly due to differences in treatment and access to care. However, these factors do not fully explain survival disparities, and outcomes among other groups remain understudied. Emerging evidence also suggests racial differences in molecular subtypes. The Kaiser Permanente Northern California (KPNC) Research on Ovarian Cancer Survival (KPROCS) Study was established to evaluate how multilevel factors contribute to differences in EOC treatment and survival. METHODS:KPROCS includes women diagnosed at KPNC between 2000 and 2022 with invasive epithelial ovarian, fallopian tube, or peritoneal cancers at age ≥18 years. Diagnostic slides underwent centralized pathology review to classify histotypes according to the 2020 WHO criteria. Addresses at diagnosis were linked to geospatial data to assess healthcare accessibility to chemotherapy infusion centers and neighborhood characteristics. In 540 women with high-grade serous EOC, gene expression molecular subtypes are being characterized. RESULTS:The cohort includes 6,067 women with EOC, including 359 Black, 798 Hispanic, 906 non-Hispanic Asian American and Pacific Islander (AAPI), and 3,971 NHW women. Descriptive analyses demonstrate subpopulation differences that may influence survival. For example, Black women were least likely to receive surgery (75%), while AAPI women were the most likely (91%). Overall, 19% of the cohort received neoadjuvant chemotherapy, with the highest proportion among Black women (25%). CONCLUSIONS:This integrative approach will provide new insights into persistent differences in EOC treatment and survival across population subgroups. IMPACT:KPROCS is a unique resource for EOC disparities research.
BACKGROUND:Social determinants of health (SDOH) influence cancer prevention and outcomes. Among SDOH, economic stability-including food insecurity, income, employment, and housing stability-has a major impact on health. However, data linking food security status with cancer screening completion remain limited. We examined whether lower food security status was associated with lower completion of colorectal, breast, cervical, and prostate cancer screening. METHODS:We performed a retrospective analysis of the 2018-2023 National Health Interview Survey. Analyses were limited to cancer-specific complete-case cohorts: colorectal (N = 59,849), breast (N = 30,867), cervical (N = 56,229), and prostate (N = 20,518). Food security was categorized as high, marginal, or low/very low. Screening completion was the dependent variable. Multivariable logistic regression estimated adjusted odds ratios (aORs) for screening by food security status, controlling for sociodemographic and clinical covariates. Sensitivity analyses used age-stratified eligibility definitions across survey years. RESULTS:Compared with high food security, marginal food security was associated with lower odds of screening across cancers (aOR range, 0.63-0.88), and low/very low food security showed similar or larger deficits (aOR range, 0.60-0.88). Associations were significant for all four cancers and were similar in sensitivity analyses. CONCLUSIONS:Lower food security status was associated with lower completion of cancer screening. Addressing food insecurity may improve screening equity and reduce cancer disparities. IMPACT:Lower food security status was associated with lower completion of guideline-recommended screening across four cancers, even after multivariable adjustment. Future work should test whether interventions addressing food insecurity can improve screening.
BACKGROUND:Neural regulation contributes to pancreatic ductal adenocarcinoma (PDAC) development but effects of neural-targeting medications on presentation or outcomes remain unclear. METHODS:We conducted a retrospective study using the Veterans Affairs (VA) Corporate Data Warehouse (CDW) to identify patients with pancreatic cancer (2000-2020). Exposure to beta blocker, cholinergics or statins was defined by active prescriptions within 6 months before or 1 month after diagnosis. Outcomes included histologic subtype, stage, and overall survival (OS). Propensity score matching was performed for each medication class, with survival assessed using Kaplan-Meier and Cox regression models. RESULTS:Among 7,578 Veterans with pancreatic cancer, 76% had adenocarcinoma and 60% presented with stage IV disease. Beta blocker use was associated with lower odds of advanced stage (OR 0.55, 95% CI 0.5-0.63, p<0.0001) and improved OS (HR 0.89, 95%CI 0.84-0.95, p<0.0001). Cholinergic agonist use was associated with reduced likelihood of adenocarcinoma histology (OR 0.64, 95% CI 0.41-1.01, p=0.051) and advanced stage (OR 0.53, 95%CI 0.34-0.85, p=<0.0001), but not OS. Statin use was associated with adenocarcinoma histology (OR 1.24, 95%CI 1.09-1.43, p=0.002) and lower odds of advanced stage (OR 0.77, 95%CI 0.68-0.88, p=<0.0001). CONCLUSIONS:Beta blockers were associated with earlier stage and improved survival. Cholinergic agonists and statins were associated with earlier stage. Cholinergic agonists were linked to lower likelihood of adenocarcinoma histology, while statins to higher likelihood. These findings suggest neural and metabolic pathways may shape early PDAC biology. IMPACT:Autonomic and metabolic pathways may influence PDAC biology. Beta blockers merit mechanistic and clinical evaluation as adjunctive therapies.
BACKGROUND:Recurrence is a key oncologic endpoint but is difficult to automatically capture from electronic health records (EHR). METHODS:We evaluated rule-based algorithms to detect recurrence and its timing using a publicly available clinico-genomic database of patients with breast, colorectal, non-small cell lung, or pancreatic cancer. Algorithms evaluated varying anchor dates, defined as the time at which patients were assumed eligible to recur for the purposes of the algorithm, including diagnosis and four-, six-, and twelve-months post-diagnosis, as well as varying criteria for subsequent evidence of cancer from radiology, pathology, medical oncology assessments, or cancer-directed regimen initiation. Algorithm-derived recurrence results were compared with manually curated institutional data. Performance was measured by sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and concordance of recurrence timing. RESULTS:The best algorithms, anchored at six months post-diagnosis and requiring two reports demonstrated high sensitivity (≥0.95), PPV (≥0.81), and NPV (≥0.84), reasonable specificity (≥0.75), and estimated recurrence within a median of one month of institutional reference data. CONCLUSIONS:This scalable method enables derivation of real-world endpoints from EHR-based data. IMPACT:The algorithm may be used to characterize outcomes for patients treated outside of prospective clinical trials and may inform the creation of synthetic control cohorts to support regulatory approvals for new drugs in rare tumor or biomarker defined populations.
BACKGROUND:Low absolute lymphocyte count (ALC) is associated with breast cancer-specific mortality (BCM) and overall mortality (OM) in non-Hispanic (NH) White women with triple-negative breast cancer (TNBC). However, little is known about the relationship of ALC with BCM or OM in other breast cancer subtypes, racial or ethnic groups. Here, we investigated the association of ALC with BCM and OM by breast cancer subtype and evaluated for modification by race or ethnicity. METHODS:We used the Oncoshare database, linking data from electronic medical records of academic and community institutions to data from the California Cancer Registry, to identify women diagnosed with stage I-III breast cancer in 2000-2017 who had a post-diagnosis ALC within five years. We used logistic regression to identify factors associated with low ALC(<1 K/µL) and Cox models to evaluate the association of minimum ALC (minALC) post-diagnosis with BCM and OM. RESULTS:In 11,669 women, low ALC was associated with neutropenia, chemotherapy receipt, advanced stage, more comorbidities, and aggressive tumor features (all p≤0.001). Lower minALC was associated with higher OM(HR=0.96 per 1 K/µL increase, 95% CI 0.93-0.99), most strongly in TNBC(HR=0.73, 0.59-0.91). Race and ethnicity modified associations of minALC with BCM and OM (interaction p<0.001), more strongly in NH Black and Hispanic than NH White women. CONCLUSIONS:Low peripheral ALC was associated with higher OM in women with estrogen and/or progesterone receptor-positive, HER2-negative breast cancer or TNBC. IMPACT:Low ALC is more strongly associated with OM and BCM in women from minoritized racial and ethnic groups.
BACKGROUND:We leveraged observational data to conduct a target trial emulation (TTE) framework evaluating the association of exercise on distant disease-free survival (DDFS) in a pan-cancer cohort. METHODS:To emulate a randomized controlled trial (RCT), a "historical" exercise assessment (Th) identified non-exercising (<10 metabolic equivalent of task-hours per week [MET-h/week]) patients. Repeat assessment 9-15 months later (T0) was used to classify patients into an exercise group (≥10 MET-h/week: increased exercise from Th) and a control group (up to 10 MET-h/week: minimal increase in exercise from Th). Kaplan-Meier methods and Cox proportional hazards models from the T0 assessment were used to estimate hazard ratios (HRs) for the exercise versus control group on DDFS and overall survival (OS). RESULTS:A total of 3,022 patients were classified into the exercise group and 1,183 into the control group. With 5.2 years (IQR 2.7, 7.0 years) median follow-up from T0, the 5-year DDFS was 95% (95% CI, 93-96) in the exercise group and 92% (95% CI, 91-94) in the control group. Compared with the control group, the adjusted HR for the exercise group was 0.80 (95% CI, 0.59-1.09, p=0.20) for DDFS and 0.78 (95% CI, 0.55-1.11, p=0.20) for OS. CONCLUSIONS:In this pan-cancer TTE, we estimated a potential 20-30% improvement in DDFS in the exercise group compared to control. IMPACT:These data add to a growing body of evidence supporting further development of exercise as a therapeutic-intent strategy in primary solid tumors.
BACKGROUND:Talc is a known risk factor for some lung diseases due to its small particle size, commonly aerosolized form, and possible contamination with asbestos, a known lung carcinogen. Personal talc-based body powder (talc) use is common among U.S. women, but the association of talc use and lung cancer has not been studied in non-occupational settings. METHODS:At enrollment (2003-2009), Sister Study participants reported talc use in the last year and ages 10-13. We used Cox proportional hazards models to estimate covariate-adjusted hazards ratios (HRs) and 95% confidence intervals (CIs) for the associations between talc use and incident lung cancer. RESULTS:Among 49,592 eligible women, 41% reported using talc in the last year and 641 reported a new lung cancer diagnosis. Self-reported talc use was not consistently associated with incident lung cancer: HR=0.91 (95% CI: 0.78-1.07) for talc use in the last year; HR=0.94 (0.80, 1.10) for talc use ages 10-13. CONCLUSIONS:No consistent associations were observed between talc use and lung cancer overall, though some suggestive positive associations observed among never-smokers and for lung adenocarcinomas may warrant further investigation. IMPACT:Population-based research on personal care product use is critical to improving our understanding of the impacts of potentially harmful substances, and their known or unknown contaminants, on human health.
BACKGROUND:Ovarian cancer is characterised by high mortality and lacks effective screening, making prevention critical. Polygenic risk scores (PRS), which aggregate the effects of multiple common alleles, may capture a proportion of currently unexplained genetic risk. While PRS have been evaluated for risk prediction, their association with treatment response and survival remains unclear. This study assessed the utility of a PRS for predicting high-grade serous ovarian cancer (HGSOC) risk in an Australian population and its association with chemotherapy response and outcomes. METHODS:PRS were calculated for 1,097 HGSOC, and 812 controls using data from Australian research programs. Associations between PRS, OC risk, chemotherapy response, and survival were analysed. RESULTS:Each standard-deviation increase in PRS was associated with a 40% increase in HGSOC risk (OR 1.40, p < 0.001). Women in the top 1% of the PRS distribution had a lifetime OC risk approaching 3%. Higher PRS values showed a trend toward poorer outcomes, however these associations were not consistent across analyses. CONCLUSIONS:PRS were not clearly associated with chemotherapy response or survival but represent a significant risk factor for the development of HGSOC. Incorporating PRS into clinical models may improve risk stratification and support targeted prevention. IMPACT:As the first study to evaluate how PRS relate to both HGSOC risk and chemotherapy response and outcomes, we show that PRS are unlikely to serve as therapeutic biomarkers but support their use for enhanced risk-stratified prevention.
Colorectal cancer is the leading cause of cancer death among adults younger than 50 years in the United States. Recent updates to colorectal cancer screening guidelines only accommodate adults 45 to 49 years, who comprise about half the estimated incident early-onset colorectal cancer cases. Adults 18 to 44 years without access to average risk screening are likely to present symptomatically and require a well-defined system for symptom identification and workup to avoid diagnostic delays and worse outcomes. The study conducted by Chubak and colleagues sought to understand the role of symptoms as diagnostic indicators by measuring their positive predictive values for colorectal cancer. They provided evidence supporting the effectiveness of the combination of symptomatic identification and fecal testing to identify high-risk individuals requiring faster early-onset colorectal cancer diagnostic evaluation. Given its use in the United States for colorectal cancer screening and as a symptomatic triage test in other countries like the United Kingdom, symptomatic fecal testing has the capacity to be readily implemented in practice. However, further research is needed to develop an effective strategy in the United States to ensure that those individuals at risk for early-onset colorectal cancer without access to average-risk colorectal cancer screening experience little to no diagnostic delay. See related article by Chubak et al., p. 1276.
Although research demonstrates that cancer survivors who are sexual and gender minorities (SGM), racial and ethnic minorities, or of lower socioeconomic status experience a higher prevalence of cancer-related pain, limited studies have examined how these intersecting marginalized identities and additional social experiences interconnect and affect pain outcomes. In their recent study, Boehmer and colleagues conducted an intersectional analysis of cancer-related pain among SGM adults and other marginalized backgrounds. Using data from the Behavioral Risk Factor Surveillance System, they found that SGM individuals reported a higher prevalence of cancer-related pain. SGM adults with a history of cancer who were younger than 65 years, Black or Hispanic, had low income, unemployed, never or formerly married, and reported poor access to care demonstrated higher likelihood of pain prevalence than their heterosexual cisgender counterparts. In this commentary, we explore minority stress, defined as the socially based and chronic stress associated with stigma, fear of discrimination, and prejudice, as a possible contributing factor to this higher prevalence of pain among those with intersecting marginalized identities, through its effect on biological function and symptoms. The findings of Boehmer and colleagues' study underscore the need to further explore intersectional approaches to pain management in cancer care and research. See related article by Boehmer et al., p. 1309.
BACKGROUND:Ambient air pollution may influence ovarian cancer risk but remains understudied. We examined long-term exposure to ambient nitrogen dioxide (NO2) in relation to the risk of invasive and borderline epithelial ovarian tumours. METHODS:In a 2011-2016 population-based case-control study in Greater Montreal (Island of Montreal, North and South Shores), 473 cases and 887 controls provided residential postal codes at study participation, which were linked to mean NO2 concentrations from a national land-use regression (LUR) model. Logistic regression estimated odds ratios (ORs) and 95% confidence intervals (CIs) for ovarian cancer overall. Polytomous logistic regression estimated associations for borderline and invasive tumours separately. Sensitivity analyses used NO2 concentrations from an alternative LUR model specific to the Island of Montreal. RESULTS:For each interquartile range (IQR) increment (9.6 ppb) in ambient NO2, the OR (95% CI) of ovarian cancer overall was 1.07 (0.89-1.30), adjusted for age, education and area-level income. For borderline tumours, the adjusted OR (95% CIs) per IQR of NO2 was 1.36 (0.99-1.87), while for invasive cancer was 0.98 (0.79-1.21). Results were similar using NO2 estimates specific to the Island of Montreal, and with adjustment for fine particulate matter. CONCLUSION:Our findings do not strongly support an association with invasive ovarian cancer, but suggest that air pollution, as indicated by NO₂ concentrations, may increase the risk of borderline tumours. IMPACT:This study addresses an important gap, given limited previous work on air pollution and ovarian cancer, and brings new findings as the first to analyze borderline and invasive tumours separately.
BACKGROUND:Survivors of childhood cancer face elevated risks for treatment‑related late effects, including subsequent malignant neoplasms (SMNs). In the general population, accelerated aging is associated with increased malignancy risk, however this has not been evaluated in survivors. METHODS:Survivors enrolled in the Childhood Cancer Survivor Study with measures of physiologic frailty, Cumulative Illness Burden for Geriatrics (CIRS-G), or Deficit Accumulation Index (DAI) data were included. Associations between these measures of accelerated aging with SMNs, benign meningiomas, and non-melanoma skin cancers (NMSC) were estimated with piecewise exponential models, adjusted for sociodemographic, treatment, and modifiable factors. RESULTS:Physiologic frailty and CIRS‑G were not associated with risk of SMNs, meningiomas, or NMSCs. Higher DAI was only associated with increased risk of benign meningioma, but not SMNs or NMSCs. CONCLUSIONS:Accelerated aging measures were not consistently associated with SMN risk in survivors of childhood cancer. IMPACT:Our results confirm that established risk factors, especially treatment exposures, remain critical in the development of SMN in survivors, reinforcing the importance of risk‑adapted therapies and long‑term surveillance in survivorship care.
BACKGROUND:Smoking-related metabolic perturbations may persist long after smoking cessation. We aimed to develop a metabolite signature capturing persistent smoking effect (MSPS) and prospectively assess its association with lung cancer (LC) risk among former smokers. METHODS:Global untargeted metabolomics data from 1732 participants of the Shanghai Men's Health Study (SMHS) and Southern Community Cohort Study (SCCS) were used to construct MSPS. Linear regression and Elastic net regression were applied to select persistent smoking-related metabolites and develop MSPS. We evaluated the association of MSPS with LC via conditional logistic regression among 229 case-control pairs nested in former smokers of the SCCS and SMHS. RESULTS:We identified 130 metabolites significantly associated with current smoking status (PFDR<0.05). After excluding tobacco-specific metabolites, we selected 16 metabolites associated with former smoking (6 lipids, 6 xenobiotics, 2 amino acids, and 2 carbohydrates) to derive the MSPS. The MSPS showed a significant association with increased LC risk [Odds ratio (OR), 95% confidence interval (CI):1.44 (1.15-1.81) per standard deviation (SD)], independent of smoking pack-years and quitting duration. The association was consistently seen in Whites (OR=1.91, 95%CI:1.19-3.07 per SD), Blacks (OR=1.66, 95%CI:1.07-2.56 per SD), and Asians (OR=1.15, 95%CI:0.81-1.63 per SD). CONCLUSIONS:We developed an MSPS that was robustly associated with LC risk independent of smoking history for former smokers. If validated, the MSPS can be applied to identify former smokers at high risk of developing LC. IMPACT:The MSPS provides a novel tool to stratify LC risk among former smokers, beyond current guidelines based on smoking history.
