| Theme |
Colorectal cancer screening and surveillance -- Prospects towards standardization from new findings |
| Title |
Surveillance after endoscopic resection of T1 colorectal cancer |
| Author |
Katsuro Ichimasa |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
| Author |
Shin-ei Kudo |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
| Author |
Yuichi Mori |
Digestive Disease Center, Showa University Northern Yokohama Hospital / Clinical Effectiveness Research Group, University of Oslo |
| Author |
Masashi Misawa |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
| Author |
Hideyuki Miyachi |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
| Author |
AMED T1 Study Group |
|
| [ Summary ] |
In accordance with the current guidelines, most patients with T1 colorectal cancers (CRC) undergo surgical resection with lymph node dissection, despite the low incidence (-10 %) of lymph node metastasis. To provide these patients with a more minimally invasive treatment option and reduce their medical costs, it is necessary to build a model that predicts the presence of lymph node metastasis more accurately. The authors collected data of 4,073 patients with T1 CRC from seven hospitals in Japan from 1997 to 2017. Artificial intelligence (AI) was used to build a validated model to identify the T1 CRC cases at a risk of lymph node metastasis. The artificial neural network model used data regarding the patients' age and sex, as well as the size, location, morphology, lymphatic and vascular invasion, and histologic grade of the tumor. It outperformed the US and Japanese guidelines in identifying patients presenting T1 CRC with lymph node metastasis (area under curve: AI= 0.83, US guidelines= 0.73, and Japanese guidelines= 0.57). Thus, this model could be used to determine which patients require additional surgery after the endoscopic resection of T1 CRC. |