# Epidemiological Classification Changes and Incidence of Early-Onset Colorectal Cancer

## TL;DR

Changes in cancer classification guidelines significantly affected early-onset colorectal cancer incidence estimates, especially for those aged 15 to 39.

## Contribution

The study uniquely compares colorectal adenocarcinoma and neuroendocrine neoplasms incidence patterns before and after a classification change.

## Key findings

- The 2013 NEN classification change significantly increased EOCRC incidence for those aged 15 to 39.
- NEN rates increased from 2004 to 2013 but declined afterward, with no change in ADC rates.
- EOCRC with distant metastasis increased in both sexes over the study period.

## Abstract

This cohort study examines age-specific incidence rates of colorectal adenocarcinoma and neuroendocrine neoplasms in individuals with invasive colorectal cancer.

How does changing cancer registration guidelines affect the estimation of colorectal cancer incidence?

In this cohort study of 63 780 patients with colorectal cancer, changes in the definition of neuroendocrine neoplasms (NENs) affected the incidence of colorectal cancer only for people aged 15 to 39 years, for whom the proportion of NENs was sufficient to change incidence. Time-series patterns for nonmetastatic NENs were similar across all age classes, while metastatic NENs remained rare.

Findings of this study suggest that public health decision-makers should consider the artifactual increase in NEN incidence amid increasing colorectal cancer rates, especially regarding the current debate about initiating colorectal cancer screening at an earlier age.

Previous studies showed an increasing incidence of early-onset colorectal cancer (EOCRC) during a period when cancer registration guidelines were changed. None of these studies simultaneously compared the patterns in colorectal adenocarcinoma (ADC) and neuroendocrine neoplasms (NENs) according to age and stage extension.

To describe patterns in colorectal cancer incidence according to histopathological type and tumor extension at diagnosis in individuals younger than 50 years compared with older patients.

This cohort study used data from the French Network of Cancer Registries (FRANCIM) on inhabitants of 7 French administrative areas. The cohort included individuals newly diagnosed with invasive colorectal cancer from 2004 to 2021. This period covered the change in NEN classification published in 2013. Statistical analysis was conducted from November 2024 to March 2025.

Age-specific incidence rates of ADC and NENs were primary outcomes. Incidence was calculated and modeled to estimate annual percent changes (APCs) using a flexible multivariable Poisson model. In a sensitivity analysis, Joinpoint regressions consolidated the results.

A total of 63 780 patients (35 266 males [55.3%]; mean [SD] age, 63.8 [12.8] years) were diagnosed with all colorectal cancer histologic types (CRC-All). The incidence of EOCRC increased in females (APC, 2.9%; 95% CI, 1.6%-4.3%) and males (APC, 2.6%; 95% CI, 1.2%-4.0%) aged 15 to 39 years but remained stable in those aged 40 to 49 years. In both sexes and all ages, the rates of ADC did not vary while NEN rates increased. In EOCRC, the increase in NENs was marked from 2004 to 2013 (between 10.1% [95% CI, 4.0%-16.5%] per year and 12.6% [95% CI, 7.2%-18.2%] per year) and disappeared thereafter (ranging from −1.1% [95% CI, −1.9% to −0.2%] per year to −1.7% [95% CI, −3.0% to −0.4%] per year). It affected the incidence of colorectal cancer only for patients aged 15 to 39 years, for whom the proportion of NENs was high (29.7% [278 of 935]) compared with 5.7% (132 of 2333) of patients in the 40 to 49 years age class and 1.4% (856 of 60 512) of patients in the 50 years or older age class). In EOCRC, ADC M1 (distant metastasis) incidence increased in both sexes over the entire study period (between 1.2% [95% CI, 0.3%-2.1%] per year and 2.3% [95% CI, 1.0%-3.6%] per year between 2013 and 2021).

In this cohort study, the 2013 change in NENs classification had a prominent role in EOCRC incidence patterns. Public health decision-makers should consider this artifactual increase.