Score Predicts Advanced Colorectal Neoplasia in Young Adults

Researchers have developed and internally validated a simple score using clinical factors that can help estimate the likelihood of advanced colorectal neoplasia in adults younger than age 45 years.

While colorectal cancer (CRC) incidence has declined overall due to screening, early-onset CRC is on the rise, particularly in individuals younger than 45 years — an age group not currently recommended for CRC screening.

Studies have shown that the risk for early-onset advanced neoplasia varies based on several factors, including sex, race, family history of CRC, smoking, alcohol consumption, diabetes, hyperlipidemia, obesity, and diet.

A score that incorporates some of these factors to identify which younger adults are at higher risk for advanced neoplasia, a precursor to CRC, could support earlier, more targeted screening interventions.

The simple clinical score can be easily calculated by primary care providers in the office, Carole Macaron, MD, lead author of the study and a gastroenterologist at Cleveland Clinic, Cleveland, told Medscape Medical News. “Patients with a high-risk score would be referred for colorectal cancer screening.”

The study was recently published online in Digestive Diseases and Sciences.

To develop and validate their risk score, Macaron and colleagues did a retrospective cross-sectional analysis of 9446 individuals aged 18-44 years (mean age, 36.8 years; 61% women) who underwent colonoscopy at their center.

Advanced neoplasia was defined as a tubular adenoma ≥ 10 mm or any adenoma with villous features or high-grade dysplasia, sessile serrated polyp ≥ 10 mm, sessile serrated polyp with dysplasia, traditional serrated adenoma, or invasive adenocarcinoma.

The 346 (3.7%) individuals found to have advanced neoplasia served as the case group, and the remainder with normal colonoscopy or non-advanced neoplasia served as controls.

A multivariate logistic regression model identified three independent risk factors significantly associated with advanced neoplasia: Higher body mass index (BMI; P = .0157), former and current tobacco use (P = .0009 and P = .0015, respectively), and having a first-degree relative with CRC < 60 years (P < .0001) or other family history of CRC (P = .0117).

The researchers used these risk factors to develop a risk prediction score to estimate the likelihood of detecting advanced neoplasia, which ranged from a risk of 1.8% for patients with a score of 1 to 22.2% for those with a score of 12. Individuals with a score of ≥ 9 had a 14% or higher risk for advanced neoplasia.

Based on the risk model, the likelihood of detecting advanced neoplasia in an asymptomatic 32-year-old overweight individual, with a history of previous tobacco use and a first-degree relative younger than age 60 with CRC would be 20.3%, Macaron and colleagues noted.

The model demonstrated “moderate” discriminatory power in the validation set (C-statistic: 0.645), indicating that it can effectively differentiate between individuals at a higher and lower risk for advanced neoplasia.

Additionally, the authors are exploring ways to improve the discriminatory power of the score, possibly by including additional risk factors.

Given the score is calculated using easily obtainable risk factors for individuals younger than 45 who are at risk for early-onset colorectal neoplasia, it could help guide individualized screening decisions for those in whom screening is not currently offered, Macaron said. It could also serve as a tool for risk communication and shared decision-making.

Integration into electronic health records or online calculators may enhance its accessibility and clinical utility.

The authors noted that this retrospective study was conducted at a single center caring mainly for White non-Hispanic adults, limiting generalizability to the general population and to other races and ethnicities.

Validation in Real-World Setting Needed

“There are no currently accepted advanced colorectal neoplasia risk scores that are used in general practice,” said Steven H. Itzkowitz, MD, professor of medicine, oncological sciences, and medical education, Icahn School of Medicine at Mount Sinai in New York City. “If these lesions can be predicted, it would enable these young individuals to undergo screening colonoscopy, which could detect and remove these lesions, thereby preventing CRC.”

Many of the known risk factors (family history, high BMI, or smoking) for CRC development at any age are incorporated within this tool, so it should be feasible to collect these data,” said Itzkowitz, who was not involved with the study.

But he cautioned that accurate and adequate family histories are not always performed. Clinicians also may not have considered combining these factors into an actionable risk score.

“If this score can be externally validated in a real-world setting, it could be a useful addition in our efforts to lower CRC rates among young individuals,” Itzkowitz said.

The study did not receive any funding. Macaron and Itzkowitz reported no competing interests.