Background
Alcohol has been recognised as a Group 1 carcinogen by the International Agency for Research on Cancer. Alcohol consumption increases the risk of several cancers, but evidence for bladder cancer is inconclusive.

Methods
We analysed data from 30 cohorts in the Pooling Project of Prospective Studies of Diet and Cancer, including 2,366,448 participants, 20,705 with first primary bladder cancer (15,526 male; 5,179 female). Multivariable Cox regression estimated hazard ratios (HR) and 95% confidence intervals (CI) for bladder cancer associated with baseline alcohol intake (g/day), stratifying by age and year of questionnaire return and centre (EPIC only), and adjusting for race and ethnicity, education, smoking status and duration, physical activity, and intake of fruit, vegetables, and processed meat. Models were run separately by study and sex, and results combined by random-effects meta-analysis.

Results
For females, alcohol consumption of 15-29.9 g/day and ≥30.0 g/day was associated with HRs of 1.13 (95% CI: 1.02-1.26) and 1.18 (95% CI: 1.03-1.34), respectively, compared with 0.1-4.9 g/day (light drinking). For males, HRs were 0.98 (95% CI: 0.93-1.04) for 15-29.9 g/day and 1.04 (95% CI: 0.98-1.10) for ≥30.0 g/day, with evidence of heterogeneity by sex (P≤0.02). Non-drinking was associated with decreased risk for males (HR 0.94; 95% CI: 0.89-0.99), but not females (HR 0.98; 95% CI: 0.91-1.06), but the former result was largely driven by past-drinkers (not abstainers). Results were generally consistent among never-smokers, but fewer cases led to less precise HR estimates (e.g. consumption of ≥30.0 g/day relative to 0.1-4.9 g/day gave HR of 1.34 [95% CI: 0.97-1.85] for females and 0.98 [95% CI: 0.83-1.15] for males).

Conclusion
Results from a large consortium of prospective studies indicate that alcohol consumption is associated with increased risk of bladder cancer, particularly for females.

Background: Physical activity reduces the risk of several cancers; however, evidence is less established for liver cancer, and the competing risk of death must be carefully considered. We estimated the effect of physical activity on risk of primary liver cancer using a formal causal inference framework for competing risks.
Methods: This analysis included data for 263,184 participants (870 liver cancers) in the NIH-AARP Diet and Health Study. The total effect of physical activity on risk of primary liver cancer was defined by conceptualising competing death as a mediator on the causal pathway with a deterministic relationship with the outcome, such that if competing death = 1, primary liver cancer = 0. Risk differences (RDs) for the total effect were estimated as the differences in standardised, cause-specific cumulative incidence functions of primary liver cancer under exposure (≥ 4 hours/week of recreational moderate- to vigorous-intensity physical activity [MVPA]) versus no exposure (< 4 hours/week of MVPA). 95% confidence intervals (CIs) were obtained via the delta method. A sensitivity analysis was conducted for hepatocellular carcinoma (HCC), the most common histological subtype of primary liver cancer.
Results: RDs for the total effect of physical activity on risk of primary liver cancer at the average age of diagnosis (67 years) and average life expectancy (78 years) in the United States were estimated to be -18.1 (95% CI: -29.9, -6.2) and -58.4 (95% CI: -95.8, -21.1) cases per 100,000 persons, respectively. The magnitude of the RD increased from age 55 (RD per 100,000 persons = -2.0, 95% CI: -4.5, 0.6) to 85 years (RD per 100,000 persons = -79.0, 95% CI: -133.4, -24.7). Results for HCC were attenuated.
Conclusion: Physical activity may have a protective effect on primary liver cancer that increases with age.

Background:
Breast cancer (BC) patients are at increased risk of cardiovascular disease (CVD), largely due to cardiotoxic cancer treatments and pre-existing risk factors, such as hypertension, diabetes, tobacco smoking and low physical activity. Several measures/tools have been suggested or used in research to determine baseline risk and appropriate CVD management during and beyond cancer treatment. This review aims to scope the literature to identify baseline (pre-treatment) CVD risk assessment tools for BC patients proposed, developed, validated, or used in research.
Methods:
PubMed, Embase and Google Scholar were searched for articles published January 2013 – March 2024, using relevant terms. Eligibility was assessed and key data extracted from included articles independently by two reviewers. Publications included research articles (observational and experimental studies), position/policy, commentary and review papers that addressed the scoping review aim.
Results: 144 articles were included; 55% research articles. Thirteen tools assessed risk of CVD broadly (n=3), specifically cardiotoxicity or heart failure (n=8), venous thromboembolism (n=1) or CVD death (n=1) in people with BC. Four tools went through validation and performed poor to moderate in stratifying BC patients into correct risk categories. All tools included age as a risk factor; other common risk factors were BC treatment types and pre-existing hypertension or CVD. While clinical guidance and recommendations were identified, these were either for cancer patients broadly or for specific treatment types, rather than specifically for people diagnosed with BC.
Conclusion: This scoping review identified a number of existing tools with common risk factors but which performed poorly in correctly stratifying BC patients into risk categories. Further work is needed to optimise the effectiveness of baseline CVD risk assessments for monitoring BC patients during and beyond treatment to improve CVD health and outcomes.

Background: Prostate cancer is a major global health burden. Prostate-specific antigen testing is widely used for opportunistic screening but can lead to overdiagnosis and overtreatment. Risk prediction models could enable risk-based approaches that better balance the benefits and harms associated with screening. Polygenic risk scores (PGS) capture a substantial component of prostate cancer risk and hold promise for risk stratification. However, no validated risk tools currently combine PGS with non-genomic factors. We aimed to develop and validate a risk prediction model for prostate cancer that combines non-genomic factors with PGS, and to estimate the added predictive value of PGS.
Methods: Using data from UK Biobank – a large prospective cohort of over 500,000 participants with extensive questionnaire-based and linked health record data – we developed a risk model for absolute 5-year prostate cancer risk. Cox proportional hazards regression was used, incorporating non-genomic risk factors identified via systematic reviews, including age, family history, PSA testing history, medical/surgical history, and medication use. The model was then extended by integrating published PGS. Two-thirds (~125,000) of eligible participants were used for model development, while the remaining one-third (~62,500) was used for internal validation. Model performance was evaluated using discrimination and calibration metrics.
Results: Adding PGS significantly improved predictive performance. In internal validation, the non-genomic model had a C-index of 0.735 (95% CI: 0.723–0.746), while models incorporating PGS showed higher C-index values, with the most recently developed PGS achieving 0.804 (95% CI: 0.794–0.816) for PGS269 and 0.815 (95% CI: 0.804–0.825) for PGS451. The models were also well-calibrated.
Conclusion: Integrating PGS with non-genomic factors enhances prostate cancer risk prediction. External validation in the Australian 45 and Up Study, which also includes comprehensive questionnaire and linked health record data, is underway and will provide locally relevant evidence for potential risk-based screening strategies.

Abstract
Background: Colorectal cancer (CRC) survivors face an increased risk of multiple primary cancers (MPCs), but evidence on MPC-related mortality is limited. This study aimed to estimate the risk of dying from cancer-specific MPCs and the contribution of MPC to all-cause mortality in individuals first diagnosed with CRC.

Methods: Using data from the South Australian Cancer Registry (1982–2017), this retrospective study analysed CRC survivors diagnosed with MPCs, defined as distinct primary cancers arising ≥2 months after CRC diagnosis. Causes of death were categorized as index CRC, MPC, or non-cancer related. Poisson regression estimated cancer-specific mortality risk compared to the general population. Propensity score weighting was applied to balance covariate distribution between CRC survivors with and without MPC groups. A hazard ratio (HR) for all-cause mortality was estimated using a weighted dataset to assess the impact of MPC on overall survival.

Results: Among 26,093 CRC survivors (181,877 person-years follow-up), the age-standardized MPC-related mortality rate was 240 per 100,000 population. Gastrointestinal, lung, haematological, and urinary tract cancers were the most common MPC-related causes of death. CRC survivors had a 45% higher risk of dying from MPCs compared to the risk of cancer death in the general population (standardised mortality ratio = 1.45, 95%CI: 1.38–1.52). Adjusted analyses showed a 58% increase in all-cause mortality among CRC survivors with MPCs (HR = 1.58, 95%CI: 1.51–1.65).

Conclusions: CRC survivors with MPC face significantly worse survival compared to those with a single primary CRC. Early detection and management of MPCs are essential for improving long-term survival in individuals diagnosed with CRC.

Keywords: Colorectal cancer, Hazard ratio, Mortality, Multiple primary cancer, Standardised mortality ratio

Background: Ovarian cancer (OC) is often diagnosed at an advanced stage when prognosis is poor, highlighting the urgent need to establish approaches to allow for an earlier detection. We used a hypothesis-free proteome-wide Mendelian randomisation (MR) approach, with an aim to identify blood plasma proteins that could serve as predictors or potential drug targets of OC.
Methods: We conducted two-sample MR analyses using summary-level protein quantitative trait locus data covering 2,337 plasma proteins from the UK Biobank Pharma Proteomics Project (up to 54,219 participants), and genome-wide association study summary data on OC and its subtypes (up to 25,509 cases) from the Ovarian Cancer Association Consortium. Wald ratio or inverse-variance weighted MR analysis was used as the primary method, depending on the number of available instruments. We evaluated pleiotropy using MR-Egger intercept test and leave-one-out analysis.
Results: Of 2,337 plasma proteins, we found 19 associations (p <7.4 x 10-5) with OC or its subtypes, involving 12 unique proteins. Robust evidence linked follitropin subunit beta (FSHB) with endometrioid OC, where each standard deviation higher in plasma FSHB level corresponded to 2.41 times higher odds (95% CI 1.56, 3.71). Associations for the other 11 proteins could be explained by pleiotropy from ABO or MAPT-AS1 loci. We identified 12 additional suggestive associations (involving 11 plasma proteins) with OC or its subtypes at a nominal association threshold (p <0.05). Eight of these proteins, along with the receptor of follitropin, were identified as potential drug targets for approved or investigational drugs, highlighting a possibility of drug repurposing for the management OC or its subtypes.
Conclusion: Our study suggests FSHB and 11 additional plasma proteins as of potential interest in OC (or subtypes) prognosis, mostly representing potentially druggable targets. These findings demonstrate the value of proteome-wide MR in drug repurposing, including applications for OC.

Background
Processed meat is classified as a carcinogen and unprocessed red meat a probable carcinogen, but their associations along with those of poultry, seafood and egg intakes with risk of oesophageal squamous cell carcinoma (OSCC) and oesophageal adenocarcinoma (OAC) are presently unknown.
Methods
Individual-level data for nearly 2 million participants were pooled within the Pooling Project of Prospective Studies of Diet and Cancer for analyses of these food groups with risk of OSCC (17 cohorts) and OAC (14 cohorts) (mean follow-up 8.1-23.4 years across cohorts). Diet was assessed at baseline using food frequency questionnaires or diet histories generally covering the past year. Standardised definitions of the food groups evaluated were applied across studies (e.g., processed meat: all processed meat including sausages, hot dogs, bacon, ham and luncheon meats). We used Cox regression to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for associations with food groups. The models were adjusted for race and ethnicity; education; smoking status, duration and amount smoked; alcohol intake (OSCC); body mass index (OAC); physical activity; fruit intake (OSCC); vegetable intake; and energy intake.
Results
Analyses included 1,698 incident OSCC and 1,992 incident OAC cases. Processed meat (HR=1.23 for ≥30 vs <3 g/day; 95% CI: 1.01-1.50) and egg (HR=1.17 for ≥25 vs <5 g/day; 95% CI: 1.01-1.36) intakes were positively associated with OSCC risk. Poultry intake of ≥45 vs <5 g/day, was related to lower OSCC risk (HR=0.79; 95% CI: 0.67-0.95) while there was little evidence of associations for unprocessed red meat and seafood intakes. For OAC, unprocessed red meat intake was positively associated with increased risk (HR=1.24 for ≥80 vs <15 g/day; 95% CI: 1.03-1.48); other food groups were not related to OAC risk.
Conclusion
Our analyses suggest that processed meat and unprocessed red meat may have possible roles in OSCC and OAC aetiology.

Background
Colorectal cancer (CRC) is the third leading cause of cancer deaths worldwide. Early identification of individuals at highest risk is a public health need.
Methods
We used clinical and genetic data from large biobanks from Australia, Europe, America, and Asia to develop a well-powered genome-wide polygenic risk score (PRS) for CRC risk (103,401 cases and 1,344,953 controls). We evaluated its performance in 5 major ancestral populations from Australia, UK, Canada and USA.
Results
The PRS was strongly associated with CRC risk across different ancestries; European (EUR) (OR per SD =2.03, 95%CI=1.89-2.18, African (OR=1.32, 95%CI=1.09-1.60), Hispanics (OR=1.89, 95%CI=1.43-2.50), East Asia (OR=2.24, 95%CI=1.53-3.27), and South Asian (OR=1.78, 95% CI=1.39-2.28). In both European and non-European ancestries, it identified high risk individuals e.g 10% of EUR with 12-fold increased risk of CRC (OR=11.71, 95%CI=7.65-17.92), including ultra-high risk (top 1%) individuals with 17-fold increased risk (OR=16.86, 95%CI=9.52-29.86) and likely to develop CRC 15 years earlier than individuals with an average risk. It also improved the prediction of high risk individuals by up to 6% over the clinical risk factor approach (AUC = 77% vs 71%). It showed high predictive accuracy in clinically high risk individuals (e.g with family history), as well as to identify individuals at the risk of early onset of CRC (aged <50 years), showing its potential to guide clinicians when prioritising these individuals for diagnostic screening amid long waiting lists.
Conclusion
We provide comprehensive evidence to support potential integration of genetic screening for CRC into clinical practice guidelines for screening, early identification and prevention.

BACKGROUND
Obesity is associated with a higher risk of gastric cardia cancer; however, there is less evidence for associations with non-cardia cancer or other stomach cancer subtypes. It is also unclear whether associations are modified by other factors (e.g. smoking, race).
METHODS
Analysis was conducted using data from 2,050,714 participants across 21 cohort studies within the Pooling Project of Prospective Studies of Diet and Cancer; 8,451 incident cases of gastric adenocarcinoma (1,827 cardia, 4,499 non-cardia, 2,125 overlapping/unspecified/unknown) were documented. At baseline, body weight, height, and, in a subset of studies, waist and hip circumference were assessed. We used Cox regression to assess associations between stomach cancer and anthropometric measures including body mass index (BMI). We assessed whether associations varied by smoking status, race or tumour subtype.
RESULTS
Overall, we found positive associations between multiple body fatness measures and stomach cancer risk, e.g. multivariable-adjusted hazard ratio, HR: 1.13 [95% confidence interval, CI, 1.01-1.27] for BMI≥30 vs BMI18.5-<25 kg/m2. Using continuous measures, BMI (HR 1.08 per 5kg/m2 [95% CI, 1.03-1.12]), waist circumference (HR 1.08 per 10 cm [95% CI, 1.04-1.13]) and waist-to-hip ratio (HR 1.05 per 0.05-unit increment [95% CI, 1.03-1.08]) were positively associated with stomach cancer risk, while ‘a body shape index’ and hip circumference were not. Associations with BMI and waist circumference were stronger in never-smokers and White populations (all Pinteraction<0.05). In addition to cardia cancer (HR 1.22 per 5kg/m2 [95% CI, 1.16-1.30]), higher BMI was associated with increased risk of non-cardia cancer (HR 1.06 [95% CI, 1.01-1.10]), particularly distal non-cardia cancers of the lesser curvature (HR 1.16 [95% CI, 1.04-1.29]), pyloric antrum (HR 1.08 [95% CI, 1.01-1.15]) and pylorus (HR 1.11 [95% CI, 0.93-1.33]).
CONCLUSION
Global stomach cancer prevention policies should consider targeting obesity control measures given our novel findings of associations with far more common non-cardia cancer.