A High Percentage of Early-Age Onset Colorectal Cancer is Potentially Preventable

Introduction

Although overall colorectal cancer (CRC) rates have been decreasing, the incidence of early-age onset CRC (EAOCRC; CRC diagnosis <50y) is increasing. This alarming trend has led to a focus on optimized screening recommendations.

Family history of CRC in a first-degree relative (FDR; children, siblings, parents) is well-known to increase risk and most guidelines recommend that individuals with a FDR with CRC at any age begin colonoscopy at 40y or 10 years younger than the affected relative.^{1, 2} Of note, a registry-based study estimated that 25% of CRC cases diagnosed between 40–49y could have been identified early or prevented if these guidelines had been followed.³ Similarly, certain inherited cancer syndromes have guidelines recommending more aggressive screening.⁴

The American Cancer Society (ACS) recommends average-risk screening at 45y, but this has not been widely applied and other guidelines continue to recommend starting at 50y.^{1, 2, 5}

Our aim was to determine what proportion of EAOCRC from a prospectively-accrued, population-based study could have been potentially prevented if existing high-risk screening guidelines were followed and average-risk screening was initiated at 45y.

Methods

Ohio Colorectal Cancer Prevention Initiative

The Ohio Colorectal Cancer Prevention Initiative (OCCPI) was a statewide study that prospectively recruited adults newly-diagnosed with invasive CRC (ClinicalTrials.gov identifier: NCT01850654). The OCCPI is described in detail elsewhere; all EAOCRC patients received germline multi-gene panels.⁶ Institutional Review Board approval was obtained by individual programs or by ceding to the Ohio State University Institutional Review Board.

Study Analysis

Inclusion criteria for this analysis was EAOCRC enrolled into OCCPI with completed genetic evaluation. Baseline characteristics including reported family history were assessed. Patients included in the family history cohort had a FDR with CRC (independent of subsequent genetic evaluation). Patients included in the hereditary cancer syndrome cohort were found to have a pathogenic or likely pathogenic variant in a gene that would have led to changes in CRC screening based on National Comprehensive Cancer Network (NCCN) guidelines.⁴ Patients included in the sporadic cohort did not meet either of these criteria.

Guidelines from NCCN, United States Multi-Society Task Force, American College of Gastroenterology and ACS were utilized for analysis.^{1, 2, 4, 5, 7} Based on estimates that the effectiveness of polypectomy to prevent malignancy is >80% within a five-year interval, we assumed colonoscopy carried out five years or more prior to the patient’s diagnosis age would have potentially prevented CRC.⁸ We considered screening that would have started one year or more prior to CRC diagnosis as likely to have led to at least earlier diagnosis (inclusive of those that may have been potentially prevented). Descriptive statistics were utilized.

Results

The OCCPI included 713 patients with EAOCRC that met criteria. This included 566 (79.4%) without family history or identified hereditary cancer syndrome, 64 (9.0%) with only a family history, 50 (7.0%) with only a hereditary syndrome and 33 (4.6%) with both. When family history and hereditary cancer guidelines were applied and otherwise screening at age 45 was considered, it is estimated that 368 (51.6%) would have at least been diagnosed earlier with 117 (16.4%) possibly prevented.

There were 97 (13.6%) patients with a family history. In this cohort, the median diagnosis age was 45y (20 – 49). The median recommended age to initiate colonoscopy was 7y prior to diagnosis (Figure 1). It is estimated that 80 (82.5%) patients would have been diagnosed earlier and 65 (67.0%) had potentially preventable CRC if guidelines were followed. If early initiation of screening was based on FDR ≤60y or ≥2 FDRs, it is estimated that 52 (53.6%) patients would have at least been diagnosed earlier with 44 (45.4%) possibly prevented.

Figure 1.

Waterfall plot presenting the difference between age of colorectal cancer onset and recommended age to initiate colorectal cancer screening based on family history of a first-degree relative with colorectal cancer. Each bar represents a single patient, with green representing those who would have started screening prior to their diagnosis and red those who would have started screening after diagnosis.

There were 83 (11.6%) patients with at least one hereditary cancer syndrome with established CRC screening guidelines (10 APC, 6 biallelic MUTYH, 23 MSH2, 17 MLH1, 12 MSH6, 11 PMS2, 3 CHEK2, 1 SMAD4, 1 BMPR1A). The median recommended age to initiate colonoscopy was 18y prior to diagnosis, with 81 (97.6%) at least having an earlier diagnosis and 74 (89.2%) with potentially preventable CRC.

There were 566 (79.4%) patients with sporadic EAOCRC. The median diagnosis age was 44y (17 – 49). It is estimated that 234 (41.3%) patients would have been diagnosed earlier with initiation of screening at 45y as recommended by the new ACS guidelines.⁵

Discussion

EAOCRC is increasing in frequency and is a growing concern. Our data suggests 51% of EAOCRC could be at least diagnosed earlier with 16% potentially prevented if current screening guidelines were followed.

As part of efforts to optimize screening for CRC, early initiation of colonoscopy is recommended for those with family history. Our data supports early initiation of screening independent of age of FDR with CRC as roughly 50% more patients would have been diagnosed earlier or potentially prevented than if previous age restrictions were in place. Encouraging stricter adherence to early CRC screening in this population is fertile territory for reduction in EAOCRC. This needs to be impressed on providers and patients to ensure these recommendations are delivered to family members.

Many patients may have met criteria for genetic counseling and testing. Importantly, current NCCN guidelines for high-risk conditions recommended starting CRC surveillance before the age of diagnosis in >97% of these patients, including the recent update to start MSH6 and PMS2 carriers at 30y. If genetic diagnoses occurred in early adulthood rather than after CRC diagnosis, this would be another opportunity to decrease EAOCRC. The identification of hereditary cancer syndromes in unaffected individuals is complicated, however, as many would not meet current criteria for genetic testing and would require population-level screening to be identified. At a minimum, all EAOCRC patients should be referred for genetic assessment to allow for high-risk surveillance and predictive genetic testing for relatives.⁴

The decision to change the starting age of CRC screening in the average-risk population is complex and has led to discrepant recommendations. In our population, the widespread adoption of starting at 45y would have likely led to the earlier diagnosis in 41.3% of the sporadic EAOCRC population. This is especially important in the EAOCRC population as they often present with more advanced disease.

This study is subject to ascertainment bias as all participants were recruited after CRC diagnosis. However, this was limited by the prospective recruitment strategy that included CRC across the state of Ohio. In addition, the effectiveness of early colonoscopy would also require standard surveillance intervals be followed.

In summary, EAOCRC patients with family history of CRC are commonly diagnosed after recommended age to initiate screening and effort is needed to increase the uptake of early colonoscopy. Earlier identification of hereditary cancer syndromes and widespread adoption of an earlier average-risk screening age would also have a significant impact on decreasing EAOCRC.