Society: AGA

Background and Aims: Although high body mass index (BMI) increased the risk of colorectal cancer (CRC), the impact of BMI change
on cancer risk has been rarely reported. We investigated the impact of BMI change on colorectal cancer risk.
Methods: Individuals, who underwent general health examination at 2009 and 2013 through the National Health Insurance Service System, were enrolled and followed up through 2017. Persons who had been diagnosed with any cancer until the second measurement of BMI were also excluded. BMI at baseline (2009) and follow-up (2013) were extracted and was classified according to
Asian Pacific guideline; underweight (<18.5 kg/m²), normal (18.5-22.9 kg/m²), overweight (23-24.9 kg/m²), obesity I (25-29.9 kg/m²),
and obesity II (≥30 kg/m²). To investigate cancer risk according to BMI change, proportional Cox regression analysis using hazard ratio (HR) and 95% confidence interval (CI) were conducted.
Results: Among 2.8 million persons, 12439 patients developed colorectal cancers during follow-up.
Persistent high BMI increased the risk of CRC. Adjusted HR is 1.11 (95% CI 1.06-1.18) for persistent obesity I and 1.24 (95% CI 1.08-1.41) for persistent obesity II comparing to persistent normal BMI group. Among baseline obese persons (obesity I), weight reduction decreased the risk of CRC comparing to persistent obesity I group (HR 0.79 in persons with change from obesity I to normal BMI and 0.92 in persons with change from obesity I to overweight group). Change from obesity I to obesity II increased the risk of CRC (HR 1.23; 95% CI 1.04-1.46) comparing to persistent obesity I group (Figure).
CONCLUSIONS: Persistent high BMI increased the risk of CRC. Furthermore, change of BMI dynamically reduced or increased the CRC
risk among baseline obese persons.

This work was supported by the National Research Foundation, republic of Korea (NRF-2022R1A2C2013044).

Background: Polygenic risk scores (PRS), which calculate genetic risk for colorectal cancer (CRC), may help prioritize individuals for more intensive versus less intensive CRC screening and surveillance. These tools have many potential benefits over current clinical risk stratification algorithms, but have yet to be widely validated. We investigated an existing PRS to determine associations with prevalent and incident advanced neoplasia (AN) in a population undergoing screening colonoscopy and follow up.

Methods: The CSP#380 screening colonoscopy cohort includes 10 years of clinical follow-up and a biorepository comprised of selected individuals with baseline AN (defined as CRC or adenoma ≥10mm or villous or high-grade dysplasia) and matched individuals without neoplasia. A PRS was constructed for each individual from 136 pre-specified CRC-risk single nucleotide polymorphisms. Multivariate logistic regression was used to evaluate the PRS for associations with AN prevalence at initial screening colonoscopy (n=594), or incident AN in participants with at least one follow-up colonoscopy (n=354). The PRS was analyzed categorically by quintiles (based on the distribution among controls) and as a continuous variable in stratified analyses.

Results: Participants with a PRS in the lowest quintile had more than a 70% decreased risk of AN at baseline screening colonoscopy (OR 0.29, 95% CI 0.14-0.58; p<0.001) compared to participants with a PRS in the middle quintile (Table 1). Further, only one CRC was identified in the lowest PRS quintile, compared to 19 cancers in the other PRS categories. Using the PRS positivity level of greater than 20% as high-risk, the sensitivity/specificity/NPV for AN at baseline colonoscopy is 91.8%/20.1%/85.6% and for CRC is 95.2%/17.1%/99.0%. In stratified analyses (Table 2), a one point increase in the PRS had the greatest impact on baseline AN risk in those with a family history of CRC (OR 12.6, 95% CI 1.20-15.50; p=0.04), Europeans (OR 7.80, 95% CI 2.30-27.52; p=0.001), and participants aged 65-75+ (OR 4.44, 95% CI 1.39-14.52; p=0.01). However, we did not observe a relationship between the PRS and incident AN during follow up (Table 1). Further, the PRS was not significantly associated with incident AN during follow-up in the stratified analyses by age, race, or baseline colonoscopy findings (Table 2).

Conclusions: Our results provide evidence which support the ability of a PRS to identify patients at risk for prevalent AN, particularly in those with a family history of CRC. Ongoing work will determine whether this PRS can identify a subset of individuals at sufficiently low risk across diverse populations who could safely delay or undergo less frequent or non-invasive screening. However, more research is needed to augment genetic tools to predict the likelihood of developing incident AN and CRC during long term follow up.

Thomas Redding

Appearances