RESUMO
Data curation for a hospital-based cancer registry heavily relies on the labor-intensive manual abstraction process by cancer registrars to identify cancer-related information from free-text electronic health records. To streamline this process, a natural language processing system incorporating a hybrid of deep learning-based and rule-based approaches for identifying lung cancer registry-related concepts, along with a symbolic expert system that generates registry coding based on weighted rules, was developed. The system is integrated with the hospital information system at a medical center to provide cancer registrars with a patient journey visualization platform. The embedded system offers a comprehensive view of patient reports annotated with significant registry concepts to facilitate the manual coding process and elevate overall quality. Extensive evaluations, including comparisons with state-of-the-art methods, were conducted using a lung cancer dataset comprising 1428 patients from the medical center. The experimental results illustrate the effectiveness of the developed system, consistently achieving F1-scores of 0.85 and 1.00 across 30 coding items. Registrar feedback highlights the system's reliability as a tool for assisting and auditing the abstraction. By presenting key registry items along the timeline of a patient's reports with accurate code predictions, the system improves the quality of registrar outcomes and reduces the labor resources and time required for data abstraction. Our study highlights advancements in cancer registry coding practices, demonstrating that the proposed hybrid weighted neural-symbolic cancer registry system is reliable and efficient for assisting cancer registrars in the coding workflow and contributing to clinical outcomes.
RESUMO
BACKGROUND: Shisa3 is a novel tumor suppressor identified in lung cancer. However, its antitumor activity in other human cancers and the mechanism of gene inactivation remain unknown. METHODS: SHISA3 expression was measured by reverse transcription-PCR (RT-PCR) and quantitative RT-PCR (RT-qPCR). DNA methylation was determined by bisulfite sequencing and pyrosequencing. RESULTS: Down-regulation of SHISA3 expression was observed in all of 11 colorectal cancer (CRC) cell lines and was further confirmed in 34 (65.4 %) of 52 colorectal carcinomas by RT-qPCR. Four of six CRC cell lines could restore SHISA3 expression after treatment with 5-aza-2'-deoxycytidine. Tumor-specific methylation of five CpG sites in the first intron of SHISA3 was identified by bisulfite sequencing, and their methylation levels were quantified in 127 pairs of primary CRC tissues by bisulfite pyrosequencing. The methylation levels of SHISA3 in tumors were noticeably higher than that in their matched normal mucosae. In addition, SHISA3 hypermethylation was significantly associated with an increased risk of disease recurrence in patients with stage II and III disease (P = 0.007) and was an independent predictor of poor overall survival [hazard ratio (HR) 2.9, 95 % confidence interval (CI) 1.5-5.8; P = 0.002] and disease-free survival (HR 4.0, 95 % CI 1.6-10.2; P = 0.003) of CRC patients. CONCLUSIONS: SHISA3 gene is epigenetically inactivated in a substantial fraction of CRC, and its hypermethylation is of prognostic significance in predicting clinical outcome. The quantitative bisulfite pyrosequencing assay established could be a cost-effective tool for providing a potential biomarker of adverse prognosis in CRC.