Risk factors for poor short-term prognosis after minimally invasive resection of primary liver cancer and construction of individualized warning model
Objective To analyze the risk factors for poor short-term prognosis after minimally invasive resection of primary liver cancer,and construct an individualized warning model.Methods The clinical data of 357 patients with primary liver cancer undergoing minimally invasive resection admitted to the First Affiliated Hospital of Zhengzhou University from January 2018 to May 2022 were retrospectively analyzed,and they were divided into modeling set(n=238)and validation set(n=119)according to the ratio of 2∶1.The modeling set patients were divided into poor prognosis group(n=57)and good prog-nosis group(n=181)based on the occurrence of poor prognosis within 1 year after surgery.The risk factors for poor short-term prognosis after minimally invasive resection of primary liver cancer was screened through logistic regression analysis,and an individualized warning model was constructed and validated.Results Logistic regression analysis results of the modeling set data showed that tumor diameter＞5cm,poorly differentiated,vascular invasion,portal vein infiltration,liver cirrhosis and alpha fetoprotein(AFP)≥400 μg/L were the risk factors for poor short-term prognosis after minimally invasive resec-tion of primary liver cancer(P＜0.05).Based on the above 6 indicators,the individualized warning model for poor short-term prognosis after minimally invasive resection of primary liver cancer was established.The calibration curves of the modeling set and validation set showed that the calibration curve of this model had good consistency with the ideal curve,and the consisten-cy index of them were 0.816(95％CI:0.763～0.862)and 0.805(95％CI:0.724～0.870)respectively.The receiver operat-ing characteristic showed that the area under the curve predicted by this model for poor short-term prognosis after surgery in the modeling set and validation set were 0.823(95％CI:0.768～0.869)and 0.810(95％CI:0.727～0.876)respectively.The decision curve showed that the modeling set could obtain net benefits when the threshold probability was 0-1.0,while the validation set could obtain net benefits when the threshold probability was between 0-0.81.Conclusion Tumor diameter＞5 cm,poorly differentiated,vascular invasion,portal vein infiltration,liver cirrhosis and alpha fetoprotein(AFP)≥ 400μg/L are risk factors for poor short-term prognosis after minimally invasive resection of primary liver cancer,and the indi-vidualized warning model constructed based on them has good predictive efficiency and applicability,and it can be used to predict the prognosis of patients with primary liver cancer.
Primary liver cancerMinimally invasive resectionPoor prognosisRisk factorIndividualized warning model
NETL
NSTL
万方数据
