Unlocking Early Colorectal Cancer Detection With Artificial Intelligence

Findings from a new study presented at the ASCO Gastrointestinal Cancers Symposium highlights the potential of an artificial intelligence (AI)-prediction model, called C the Signs, in improving colorectal cancer (CRC) detection, offering an innovative approach to early diagnosis.¹

Seema Dadhania, consultant clinical oncologist, Imperial College London Department of Surgery & Cancer

“We were trying to understand if the C the Signs platform could pick up the signs or risk of colorectal cancer earlier than the patient was actually diagnosed,” explained study investigator Seema Dadhania, consultant clinical oncologist, Imperial College London Department of Surgery & Cancer, in an interview with The American Journal of Managed Care^® (AJMC^®). “So, [we’re] trying to understand if there is a way to pick up earlier signals of colorectal cancer by using an aggregation of symptoms or other pieces of data which were present within the [Mayo Data Platform].”

CRC is increasingly diagnosed, particularly among younger populations, posing a growing public health challenge. Despite advances in screening, many CRC cases are still identified only after symptoms appear, often at later stages when outcomes are poorer.

Early-stage CRC can mimic advanced disease, but individual symptoms alone have low predictive value, while clusters of symptoms are more indicative of malignancy. Unfortunately, these symptom clusters are often associated with benign conditions, contributing to diagnostic delays. C the Signs offers the potential to address these challenges by identifying high-risk individuals earlier, even before clinical suspicion arises.

In less than 30 seconds, C the Signs can identify which cancers a patient is at risk of and recommend the appropriate test or specialist to diagnose their cancer.²

The retrospective study utilized data from primary care settings to evaluate the performance of C the Signs. Using the Mayo Data Platform, researchers analyzed a comprehensive dataset of electronic medical records (EMRs) spanning 20 years, from January 1, 2002, to December 31, 2021. Patient records included demographic information, clinical symptoms, and tests results, which were analyzed using the AI model to assess its sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Additionally, the model’s performance was benchmarked against other screening methods, including colonoscopy and Fecal Immunochemical Testing (FIT).

The researchers explored the model's ability to identify high-risk patients with nonspecific symptoms or those outside typical screening criteria. Statistical analyses were conducted to compare the diagnostic efficiency of the AI model with traditional methods, focusing on its integration into routine primary care workflows.

Among the 894,275 patients analyzed, 7348 were diagnosed with CRC during the study period. The model achieved a sensitivity of 93.8% and a specificity of 19.7% in identifying patients at risk of CRC. Notably, the model identified 29.4% of these patients as high-risk up to 5 years earlier than primary care physicians, underscoring its potential for early detection. Despite its lower specificity, the model’s high sensitivity and early identification capability highlight its value in facilitating timely interventions and improving patient outcomes in CRC care.

“In order to get a patient to have a colonoscopy, they have to have a set of criteria or a group of symptoms that warrant the appropriate referral,” explained Dadhania. “I think what this tool is allowing is it’s capturing those patients who will already get through the system, but it’s also capturing a proportion of patients who won’t get through the current system as it stands, to get a colonoscopy.”

References

1. Dadhania S, Herrick B, Bakshi B, et al. Shifting the paradigm: Early identification of colorectal cancer with C the Signs clinical decision support. Poster presented at ASCO Gastrointestinal Cancers Symposium. January 23-25, 2025; San Francisco, CA.

2. Early cancer detection: C the signs. C the Signs. Accessed January 23, 2025. https://www.cthesigns.com/.