Search

Background: To evaluate the value of six tumor markers including squamous cell carcinoma antigen (SCCA), cytokeratin 19 fragment (CYFRA21-1), carcinoembryonic antigen (CEA), glycoconjugate antigen 125 (CA125), neuron-specific enolase (NSE), along with gastrin-releasing peptide precursor (ProGRP) in the diagnosis and pathological staging of lung cancer (LC) alone or in combination. Materials and Methods: A retrospective analysis was performed on 300 patients of which 62 cases were diagnosed as small cell lung cancer (SCLC), 105 adenocarcinoma (LUAD), 41 squamous cell carcinoma (LUSC) and 92 large cell lung cancer (LCLC) by histopathology. Another 364 patients with benign lung disease in the same period were chosen for the benign lung disease group. Results: The detection results of 6 tumor markers in serum of LC, SCCA, NSE and ProGP patients with different pathological types were higher compared with lung benign lesions group (P<0.01). LC can be separated into SCLC along with non-Small cell lung cancer (NSCLC) by NSE+ProGRP combination. The detection rate of lung squamous cell carcinoma by SCCA+ cytokeratin 19 fragment (CYFRA21-1) was higher than that by CEA+CA125 combination for adenocarcinoma. The combined determination of NSE and ProGRP was highly sensitive to SCLC. In NSCLC, the combined detection of SCC-Ag and CYFRA21-1 showed high sensitivity to LUSC. CEA+CA125 combined detection was highly sensitive to LUAD. Conclusion: The combined determination of NSE and ProGRP was highly sensitive to SCLC. CEA+CA125 combined detection was highly sensitive to LUAD, and SCCA+CYFR21-1+CEA+CA125 was the best combined detection analysis of multiple indexes. [ABSTRACT FROM AUTHOR]

Colorectal cancer is the most common malignant tumor of digestive tract, and the incidence of colorectal cancer in China is especially characterized by middle and low rectal cancer. In recent years, with the progress of computer science, artificial intelligence technology has developed rapidly, and has achieved a lot of application results in the medical field. At present, artificial intelligence technology has covered various stages of colorectal cancer, including screening, individualized assessment, auxiliary diagnosis and treatment decision-making, refined surgery and prognosis judgment, providing help for the accurate and individualized treatment of rectal cancer. However, the lack of standardized, systematic, and scalable AI models remains a major pain point for the field. Therefore, it is necessary to carry out large-scale prospective clinical studies on artificial intelligence model to further confirm its application value in the clinical diagnosis and treatment of rectal cancer.

Objective: To reassess the prognostic value of minimal residual disease (MRD) and IKZF1 gene deletions in adults with B-cell acute lymphoblastic leukemia (B-ALL) who received pediatric-specific chemotherapy regimens during the Nanfang Hospital PDT-ALL-2016 trial. Methods: We retrospectively analyzed the prognosis of 149 adult patients with B-ALL who were admitted to Nanfang Hospital from January 2016 to September 2020. Prognostic factors were identified using Cox regression models. Results: The complete remission rate was 93.2% in 149 patients, with a 5-year overall survival (OS) rate of (54.3±5.0) % and a cumulative incidence of relapse (CIR) of (47.5±5.2) %. The Cox regression analysis revealed that MRD positivity at day 45 (MRD(3)) after induction therapy was independently associated with relapse risk ( HR =2.535, 95% CI 1.122-5.728, P =0.025). Deletion of IKZF1 gene was independently associated with mortality risk ( HR =1.869, 95% CI 1.034-3.379, P =0.039). Based on MRD(3) and IKZF1 gene status, we categorized adult patients with B-ALL into the low-risk (MRD(3)-negative and IKZF1 gene deletion-negative) and high-risk (MRD(3)-positive and/or IKZF1 gene wild type) groups. The 5-year OS and CIR rates were (45.5±6.0) % vs (69.4±8.6) % ( P <0.001) and (61.6±8.3) % vs (25.5±6.5) % ( P <0.001), respectively, in the high-risk and low-risk groups, respectively. The multivariate analysis showed that the high-risk group was an independent risk factor for OS ( HR =3.937, 95% CI 1.975-7.850, P <0.001) and CIR ( HR =4.037, 95% CI 2.095-7.778, P <0.001) . Conclusion: The combined use of MRD and IKZF1 gene in prognostic stratification can improve clinical outcome prediction in adult patients with B-ALL, helping to guide their treatment.