1.FOXC1 mediates the proliferation and apoptosis of colon cancer cells through the Rap1 signaling pathway
Fu XIAOXIA ; Li RUI ; Duan RUIMIN ; Hao LIYAO ; Jin YING
Chinese Journal of Clinical Oncology 2025;52(13):649-655
Objective:To investigate the expression characteristics and clinical significance of FOXC1 in colon cancer,and decipher its mo-lecular mechanism in regulating tumor cell proliferation and apoptosis.Methods:The GEPIA database was employed to analyze the expres-sion of FOXC1 and its correlation with prognosis in colon cancer.Differential expression of FOXC1 was detected by qRT-PCR and Western blot in colon cancer cells(HCT116 and SW620)and normal colon epithelial cells(NCM460),and stable FOXC1-knockdown(sh-FOXC1)cell lines were established.Western blot,flow cytometry,CCK-8,and plate colony formation assays were performed to analyze the effects of FOXC1 knockdown on cell proliferation,cell cycle,and apoptosis.Furthermore,the downstream signaling pathway was verified using Rap1 overexpression rescue experiments.Results:FOXC1 mRNA expression was significantly higher in colon cancer tissues than in normal tissues(P<0.001).FOXC1 overexpression was nearing significance in relation to tumor staging(P=0.053),and patients with high FOXC1 expression had a shorter overall survival(Log-rank P=0.013).After FOXC1 knockdown,the expression of CyclinD1 and Bcl-2 decreased,whereas the ex-pression of Bax increased(P<0.01).The proportion of cells in the G0/G1 phase increased,while the proportion of cells in the S phase de-creased(P<0.001),and the cell proliferation activity and number of colonies formed decreased(P<0.001).Mechanistic studies demonstrated that after FOXC1 knockdown,Rap1 expression was reduced,while the expression of Rap1GAP increased(P<0.05).After restoration of Rap1 expression in FOXC1-knockdown cells,the downregulation of CyclinD1 and Bcl-2 expression and the increase in Bax expression were re-versed(P<0.05),the S phase ratio was increased(P<0.05),and cell proliferation activity and colony formation abilities were also re-scued.Conclusion:FOXC1 promotes colon cancer progression by facilitating Rap1 expression and downregulating Rap1GAP.Targeted inter-vention of the FOXC1-Rap1 signaling axis may emerge as a potential therapeutic strategy.
2.Comparison of efficacy of different biological scaffolds for pulp regeneration therapy in immature permanent teeth:a Bayesian network meta-analysis
Kaigang WANG ; Dongsheng HAO ; Pei MA ; Shuo ZHOU ; Ruimin LI
Chinese Journal of Tissue Engineering Research 2025;29(34):7447-7460
OBJECTIVE:There are many kinds of biological scaffolds used for pulp revascularization in clinical practice,and the difference of efficacy between different scaffolds is controversial.The efficacy of nine kinds of biological scaffolds in endodontic revascularization was evaluated by Bayesian network meta-analysis.METHODS:The computer was used to search the literature in CNKI,VIP,WanFang,China Biomedical Literature Service System,PubMed,Cochrane Library,Web of Science,Embase,and Scopus databases.Randomized controlled trials of different biological scaffolds for the treatment of pulp revascularization in young permanent teeth meeting inclusion criteria were collected from each database up to April 1,2024.Two researchers sifted through the literature,data collection,sorting and extraction were completed independently,and the quality of the included literature was assessed for risk of bias.A network meta-analysis was performed using BUGSnet1.1.1 package of R4.2.0 software.RESULTS:A total of 22 studies with 926 affected teeth and 9 different interventions were included in this study.The results of network meta-analysis showed that:(1)Clinical success rate(primary goal):platelet-rich fibrin was superior to blood clot[OR=1.45,95%CI(0.32,2.69)],and the top three ranking results were:concentrated growth factor(82.77%)>platelet-rich fibrin(75.38%)>modified platelet-rich fibrin(62.39%).(2)Increased root length(secondary goal):There was no difference among the 7 biological scaffolds at 1-6 months of follow-up(P>0.05),the top 3 ranking results of rank probability were:concentrated growth factor(86.25%)>platelet-rich plasma(53.76%)>platelet-rich fibrin(51.11%).When followed up for>6 months and<12 months,concentrated growth factor was superior to blood clot[MD=9.59,95%CI(0.52,18.40)],the top 3 ranking results of rank probability were:concentrated growth factor(92.42%)>platelet-rich plasma(56.03%)>platelet-rich fibrin(55.76%).When followed up for more than 12 months,concentrated growth factor was superior to modified platelet-rich fibrin[MD=11.01,95%CI(0.02,22.72)],the top 3 ranking results of rank probability were:concentrated growth factor(86.95%)>platelet-rich fibrin(68.61%)>blood clot combined with collagen(52.5%).(3)Increased root wall thickness(secondary goal):at 1-6 months of follow-up,platelet-rich fibrin was superior to blood clot[MD=11.37,95%CI(4.74,17.71)],the top 3 ranking results of rank probability were:platelet-rich fibrin(93.66%)>concentrated growth factor(63.11%)>modified platelet-rich fibrin(50.48%).At>6 months and<12 months of follow-up,there was no difference among the 6 biological scaffolds(P>0.05),the top 3 ranking results of rank probability were:modified platelet-rich fibrin(73.63%)>platelet-rich fibrin(62.36%)>concentrated growth factor(56.25%).When followed up for more than 12 months,there was no difference among the 9 biological scaffolds(P>0.05),and the top 3 ranking results of rank probability were:blood clot combined with collagen(81.9%)>platelet-rich plasma(62.67%)>modified platelet-rich fibrin(59.49%).(4)Pulp vitality(third-level goal):there was no difference among the 6 biological scaffolds(P>0.05),and the top 3 ranking results of rank probability were:blood clot combined with collagen(84.22%)>concentrated growth factor(67.71%)>platelet-rich fibrin(48.79%).CONCLUSION:Existing evidence shows that the clinical success rate of different scaffolds is higher in pulp revascularization,among which platelet-rich fibrin is better than blood clots.In terms of comprehensive comparison of root length and root wall thickness increase,concentrated growth factor performs best in the follow-up period of 1-6 months and>6 months and<12 months,while blood clot combined with collagen performs best after follow-up of more than 12 months;concentrated growth factor performs outstandingly in all levels of goals,and may be more conducive to the continued development of the tooth root than other scaffolds,and has great potential in pulp regeneration treatment.Limited by the quality and quantity of literature,the above conclusions still need to be verified by more high-quality studies.
3.Protective effect of exenatide on oxidative stress in hypothalamus of diabetes mice and its mechanism
Lu ZHENG ; Haohao ZHANG ; Feifei WU ; Jiaqi GUO ; Youqin WANG ; Ruimin HAO ; Lihui FENG ; Yan LI
The Journal of Practical Medicine 2025;41(3):330-338
Objective To explore the effect of exenatide on oxidative stress in the hypothalamus of diabetes mice and its potential mechanism.Methods After one week of adaptive feeding,C57BL/6J mice were randomly divided into the CON group(normal chaw diet),the T2DM group(high-fat diet,HFD),and the T2DM+Exe group(HFD+exenatide).After 8 weeks of HFD,mice in the T2DM+Exe group were intraperitoneally injected with exenatide[24 nmol/(kg·d)]for 8 weeks.The weight and glucose and lipid metabolism levels of the mice were measured,and the levels of inflammatory and adipokine factors in mice were detected using the ELISA method.Western Blot was used to detect the expression of melanocortin receptor-4(MC4R)and proopiomelanocor-tin(POMC)in the hypothalamus.Hypothalamic mitochondria were extracted,and the content of mitochondrial reactive oxygen species(ROS)was measured using a flow cytometer.The content of malondialdehyde(MDA)and the activities of superoxide dismutase(SOD)in the mitochondria were detected using assay kits.Changes in the ultrastructure of mitochondria were observed using a transmission electron microscope.In vitro experiments,pal-mitic acid(PA)and exenatide were used to treat hypothalamic GT1-7 cells,and short hairpin RNA(shRNA)was used to silence the melanocortin 4 receptor(MC4R),and observe the cellular oxidative stress and lipid deposition.Results Compared with the CON group,the T2DM group mice showed a significant increase in glucose and lipid metabolism indicators,pro-inflammatory factors,and adipose factor levels(P<0.05),the expression of MC4R and POMC proteins in the hypothalamus were decreased(P<0.05),and the mitochondrial ROS and MDA content in the hypothalamus significantly were increased(P<0.05),while SOD and CAT activities were decreased(P<0.05).Mitochondrial morphology was abnormal.After intervention with exenatide,the above indicators were signifi-cantly improved.After inhibiting MC4R expression in vitro experiments,compared with the intervention group with exenatide,the ROS and MDA content was significantly increased(P<0.05),SOD activity was decreased(P<0.05),and lipid deposition occurred in the cells.Conclusions Exenatide exhibits a protective effect on hypotha-lamic oxidative stress injury in diabetic mice,and this mechanism may be associated with the upregulation of MC4R expression.
4.Comparison of efficacy of different biological scaffolds for pulp regeneration therapy in immature permanent teeth:a Bayesian network meta-analysis
Kaigang WANG ; Dongsheng HAO ; Pei MA ; Shuo ZHOU ; Ruimin LI
Chinese Journal of Tissue Engineering Research 2025;29(34):7447-7460
OBJECTIVE:There are many kinds of biological scaffolds used for pulp revascularization in clinical practice,and the difference of efficacy between different scaffolds is controversial.The efficacy of nine kinds of biological scaffolds in endodontic revascularization was evaluated by Bayesian network meta-analysis.METHODS:The computer was used to search the literature in CNKI,VIP,WanFang,China Biomedical Literature Service System,PubMed,Cochrane Library,Web of Science,Embase,and Scopus databases.Randomized controlled trials of different biological scaffolds for the treatment of pulp revascularization in young permanent teeth meeting inclusion criteria were collected from each database up to April 1,2024.Two researchers sifted through the literature,data collection,sorting and extraction were completed independently,and the quality of the included literature was assessed for risk of bias.A network meta-analysis was performed using BUGSnet1.1.1 package of R4.2.0 software.RESULTS:A total of 22 studies with 926 affected teeth and 9 different interventions were included in this study.The results of network meta-analysis showed that:(1)Clinical success rate(primary goal):platelet-rich fibrin was superior to blood clot[OR=1.45,95%CI(0.32,2.69)],and the top three ranking results were:concentrated growth factor(82.77%)>platelet-rich fibrin(75.38%)>modified platelet-rich fibrin(62.39%).(2)Increased root length(secondary goal):There was no difference among the 7 biological scaffolds at 1-6 months of follow-up(P>0.05),the top 3 ranking results of rank probability were:concentrated growth factor(86.25%)>platelet-rich plasma(53.76%)>platelet-rich fibrin(51.11%).When followed up for>6 months and<12 months,concentrated growth factor was superior to blood clot[MD=9.59,95%CI(0.52,18.40)],the top 3 ranking results of rank probability were:concentrated growth factor(92.42%)>platelet-rich plasma(56.03%)>platelet-rich fibrin(55.76%).When followed up for more than 12 months,concentrated growth factor was superior to modified platelet-rich fibrin[MD=11.01,95%CI(0.02,22.72)],the top 3 ranking results of rank probability were:concentrated growth factor(86.95%)>platelet-rich fibrin(68.61%)>blood clot combined with collagen(52.5%).(3)Increased root wall thickness(secondary goal):at 1-6 months of follow-up,platelet-rich fibrin was superior to blood clot[MD=11.37,95%CI(4.74,17.71)],the top 3 ranking results of rank probability were:platelet-rich fibrin(93.66%)>concentrated growth factor(63.11%)>modified platelet-rich fibrin(50.48%).At>6 months and<12 months of follow-up,there was no difference among the 6 biological scaffolds(P>0.05),the top 3 ranking results of rank probability were:modified platelet-rich fibrin(73.63%)>platelet-rich fibrin(62.36%)>concentrated growth factor(56.25%).When followed up for more than 12 months,there was no difference among the 9 biological scaffolds(P>0.05),and the top 3 ranking results of rank probability were:blood clot combined with collagen(81.9%)>platelet-rich plasma(62.67%)>modified platelet-rich fibrin(59.49%).(4)Pulp vitality(third-level goal):there was no difference among the 6 biological scaffolds(P>0.05),and the top 3 ranking results of rank probability were:blood clot combined with collagen(84.22%)>concentrated growth factor(67.71%)>platelet-rich fibrin(48.79%).CONCLUSION:Existing evidence shows that the clinical success rate of different scaffolds is higher in pulp revascularization,among which platelet-rich fibrin is better than blood clots.In terms of comprehensive comparison of root length and root wall thickness increase,concentrated growth factor performs best in the follow-up period of 1-6 months and>6 months and<12 months,while blood clot combined with collagen performs best after follow-up of more than 12 months;concentrated growth factor performs outstandingly in all levels of goals,and may be more conducive to the continued development of the tooth root than other scaffolds,and has great potential in pulp regeneration treatment.Limited by the quality and quantity of literature,the above conclusions still need to be verified by more high-quality studies.
5.Protective effect of exenatide on oxidative stress in hypothalamus of diabetes mice and its mechanism
Lu ZHENG ; Haohao ZHANG ; Feifei WU ; Jiaqi GUO ; Youqin WANG ; Ruimin HAO ; Lihui FENG ; Yan LI
The Journal of Practical Medicine 2025;41(3):330-338
Objective To explore the effect of exenatide on oxidative stress in the hypothalamus of diabetes mice and its potential mechanism.Methods After one week of adaptive feeding,C57BL/6J mice were randomly divided into the CON group(normal chaw diet),the T2DM group(high-fat diet,HFD),and the T2DM+Exe group(HFD+exenatide).After 8 weeks of HFD,mice in the T2DM+Exe group were intraperitoneally injected with exenatide[24 nmol/(kg·d)]for 8 weeks.The weight and glucose and lipid metabolism levels of the mice were measured,and the levels of inflammatory and adipokine factors in mice were detected using the ELISA method.Western Blot was used to detect the expression of melanocortin receptor-4(MC4R)and proopiomelanocor-tin(POMC)in the hypothalamus.Hypothalamic mitochondria were extracted,and the content of mitochondrial reactive oxygen species(ROS)was measured using a flow cytometer.The content of malondialdehyde(MDA)and the activities of superoxide dismutase(SOD)in the mitochondria were detected using assay kits.Changes in the ultrastructure of mitochondria were observed using a transmission electron microscope.In vitro experiments,pal-mitic acid(PA)and exenatide were used to treat hypothalamic GT1-7 cells,and short hairpin RNA(shRNA)was used to silence the melanocortin 4 receptor(MC4R),and observe the cellular oxidative stress and lipid deposition.Results Compared with the CON group,the T2DM group mice showed a significant increase in glucose and lipid metabolism indicators,pro-inflammatory factors,and adipose factor levels(P<0.05),the expression of MC4R and POMC proteins in the hypothalamus were decreased(P<0.05),and the mitochondrial ROS and MDA content in the hypothalamus significantly were increased(P<0.05),while SOD and CAT activities were decreased(P<0.05).Mitochondrial morphology was abnormal.After intervention with exenatide,the above indicators were signifi-cantly improved.After inhibiting MC4R expression in vitro experiments,compared with the intervention group with exenatide,the ROS and MDA content was significantly increased(P<0.05),SOD activity was decreased(P<0.05),and lipid deposition occurred in the cells.Conclusions Exenatide exhibits a protective effect on hypotha-lamic oxidative stress injury in diabetic mice,and this mechanism may be associated with the upregulation of MC4R expression.
6.FOXC1 mediates the proliferation and apoptosis of colon cancer cells through the Rap1 signaling pathway
Fu XIAOXIA ; Li RUI ; Duan RUIMIN ; Hao LIYAO ; Jin YING
Chinese Journal of Clinical Oncology 2025;52(13):649-655
Objective:To investigate the expression characteristics and clinical significance of FOXC1 in colon cancer,and decipher its mo-lecular mechanism in regulating tumor cell proliferation and apoptosis.Methods:The GEPIA database was employed to analyze the expres-sion of FOXC1 and its correlation with prognosis in colon cancer.Differential expression of FOXC1 was detected by qRT-PCR and Western blot in colon cancer cells(HCT116 and SW620)and normal colon epithelial cells(NCM460),and stable FOXC1-knockdown(sh-FOXC1)cell lines were established.Western blot,flow cytometry,CCK-8,and plate colony formation assays were performed to analyze the effects of FOXC1 knockdown on cell proliferation,cell cycle,and apoptosis.Furthermore,the downstream signaling pathway was verified using Rap1 overexpression rescue experiments.Results:FOXC1 mRNA expression was significantly higher in colon cancer tissues than in normal tissues(P<0.001).FOXC1 overexpression was nearing significance in relation to tumor staging(P=0.053),and patients with high FOXC1 expression had a shorter overall survival(Log-rank P=0.013).After FOXC1 knockdown,the expression of CyclinD1 and Bcl-2 decreased,whereas the ex-pression of Bax increased(P<0.01).The proportion of cells in the G0/G1 phase increased,while the proportion of cells in the S phase de-creased(P<0.001),and the cell proliferation activity and number of colonies formed decreased(P<0.001).Mechanistic studies demonstrated that after FOXC1 knockdown,Rap1 expression was reduced,while the expression of Rap1GAP increased(P<0.05).After restoration of Rap1 expression in FOXC1-knockdown cells,the downregulation of CyclinD1 and Bcl-2 expression and the increase in Bax expression were re-versed(P<0.05),the S phase ratio was increased(P<0.05),and cell proliferation activity and colony formation abilities were also re-scued.Conclusion:FOXC1 promotes colon cancer progression by facilitating Rap1 expression and downregulating Rap1GAP.Targeted inter-vention of the FOXC1-Rap1 signaling axis may emerge as a potential therapeutic strategy.
7.Effects of Rap1 GTP-activating protein on proliferation,invasion and migration of colon cancer cells
Ying JIN ; Xiaoxia FU ; Ruimin DUAN ; Liyao HAO ; Feng LI
Chinese Journal of Clinical and Experimental Pathology 2024;40(2):144-149,157
Purpose To investigate the corr-elation between Rap1 GAP expression in colon cancer tissues and clinicopatho-logical features and prognosis.Methods Immunohistochemistry was used to detect Rap1 GAP protein expression in 125 cases of colon cancer,and its correlation with clinicopathological features and prognosis was analyzed.Rap1 GAP protein expression in co-lon cancer LOVO,HCT116,SW480 cells and normal colon epi-thelial HCoEPiC cells was detected by Western blot.The expres-sion of Rap1 GAP was down-regulated and up-regulated in LO-VO,HCT116 and SW480 cells by lentivirus transfection,and di-vided into no-load group(sh-NON,LV-NON),sh-Rap1 GAP group(low expression Rap1 GAP)and LV-Rap1 GAP group(overexpression Rap1 GAP)according to different treatments.The transfection efficiency was verified by Western blotting.MTT assay and Transwell assay were used to detect cell proliferation,invasion and migration in each group.Results In 125 colon cancer samples,83 cases(66.4%)had the loss of Rap1 GAP expression,which was higher than that in paracancer control(7.2%,P<0.001).The rate of loss of Rap1 GAP expression was correlated with the degree of tumor differentiation(x2=6.152,P=0.011)and the presence of mucinous adenocarcino-ma(x2=4.908,P=0.028),but not with gender,age,tumor location,tumor stage,or lymph node metastasis(P>0.05).Western blotting results showed that compared with HCoEPiC(0.189±0.081)cells,Rap1 GAP protein expression was in-creased in colon cancer LOVO(0.238±0.008)cells.Rap1 GAP protein expression was decreased in HCT116(0.064± 0.002)and SW480(0.152±0.026)cells(F=159.6,P<0.05).After LOVO cells were transfected with Rap1 GAP low expression lentivirus,the expression level of Rap1 GAP in sh-Rap1 GAP-1 group(0.733±0.071)and sh-Rap1 GAP-2 group(0.559±0.136)and sh-Rap1 GAP-3 group(0.606±0.037)was significantly lower than that in LOVO group(1.880± 0.129)(F=49.57,P<0.05).Compared with sh-NON(1.260±0.109)group,the proliferation ability of sh-Rap1 GAP-2(1.569±0.059)and sh-Rap1 GAP-3(1.548±0.087)cells was significantly increased at 72 h(F=28.36,P<0.05).Its invasion and migration ability were significantly increased(P<0.05).After HCT116 cells transfected with overexpression lentivirus,the expression of Rap1 GAP protein in LV-Rap1 GAP group(1.395±0.137)was relatively higher than that in LV-NON group(0.485±0.097)(P<0.05).The results of MTT assay showed that compared with LV-NON(0.652±0.047)group,the proliferation ability of cells in LV-Rap1 GAP(1.212 ±0.038)group was decreased,and the invasion and migration ability were significantly decreased(P<0.05).The transfection results,proliferation,invasion and migration of SW480 cells were consistent with those of HCT116 cells.Conclusion The loss rate of Rap1 GAP expression is related to the differentiation degree of colon cancer and whether it is accompanied by mucin-ous adenocarcinoma.The up-regulation of Rap1 GAP expression can inhibit the proliferation,invasion and migration of colon cancer cells,providing a theoretical basis for exploring the occur-rence and development of colon cancer.
8.Construction and validation of a predictive model for ventilator-associated pneumonia in premature infants undergoing mechanical ventilation in the neonatal intensive care unit
Yongqin GUO ; Yingying DOU ; Jianli LI ; Ruimin CHANG ; Yanan HAO
Chinese Journal of Practical Nursing 2024;40(33):2578-2587
Objective:To construct a prediction model for ventilator-associated pneumonia (VAP) in mechanically ventilated preterm infants in the neonatal intensive care unit (NICU) and to test its clinical effect.Methods:This was a cross-sectional study. A total of 740 preterm infants admitted to the NICU for mechanical ventilation from July 2018 to June 2023 were retrospectively selected as the study subjects, and were divided into the modeling set (518 cases) and the validation set (222 cases) according to the ratio of 7∶3 using the computer-generated random number method. The modeling set was divided into the VAP group (181 cases) and the non-VAP group (337 cases) according to whether VAP occured, and 21 clinical characteristics were analyzed, using single factor difference analysis to screen predictive factors, the independent risk factors of VAP in mechanically ventilated preterm infants were determined by multivariate Logistic regression analysis, and the nomogram model was made by R software. Then, the nomogram model was tested by validating the data of the validation set. The receiver operating characteristic (ROC) curve, Hosmer-Lemeshow goodness-of-fit test, calibration curve and clinical decision curve were used to evaluate the efficacy and practical value of the model.Results:There were 88 males and 93 females in the VAP group, with 156 cases of gestational age<34 weeks and 25 cases of gestational age≥34 weeks. There were 155 males and 182 females in the non-VAP group, with 196 cases of gestational age<34 weeks and 141 cases of gestational age≥34 weeks. Birth weight ( OR=0.114, 95% CI 0.044-0.268, P<0.05) and oral care of breast milk ( OR=0.124, 95% CI 0.0.057-0.249, P<0.05) were protective factors for VAP in mechanically ventilated preterm infants, and Apgar score at 5 min after birth ( OR=2.895, 95% CI 1.318-6.419, P<0.05), serum prealbumin at 72 h of mechanical ventilation ( OR=4.837, 95% CI 2.643-9.063, P<0.05), gastric contents reflux ( OR=6.754, 95% CI 3.156-15.240, P<0.05), and time of mechanical ventilation ( OR=7.784, 95% CI 3.491-18.160, P<0.05) were independent risk factors for VAP in mechanically ventilated preterm infants. The area under the curve (AUC) of the ROC curve of the modeling set was 0.929 (95% CI 0.907-0.950, P<0.01), and the validation set (AUC) was 0.917 (95% CI 0.882-0.952, P<0.01), the model has good discrimination. The C indices of the modeling set and the validation set were 0.93 and 0.92 respectively by sampling 500 times by the Bootstrap method, indicating that the model had good consistency, and the decision curve suggested that the prediction model was far from the extreme curve and the net return value was high, indicating that the nomogram prediction model constructed this time had high prediction value. Conclusions:Birth weight, Apgar score at 5 min after birth, time of mechanical ventilation, oral care of breast milk, serum prealbumin at 72 h of mechanical ventilation, and gastric contents reflux are independent influencing factors for VAP in mechanically ventilated preterm infants. The nomogram prediction model constructed can provide a visual and simple evaluation tool for early identification of high-risk children and reducing the occurrence of VAP.
9.Prediction of EGFR mutant subtypes in patients with non-small cell lung cancer by pre-treatment CT radiomics and machine learning
Jiang HU ; Ruimin HE ; Pinjing CHENG ; Xiaomin LIU ; Haibiao WU ; Linfei LIU ; Baiqi WANG ; Hao CHENG ; Junhui YANG
Chinese Journal of Radiological Medicine and Protection 2023;43(5):386-392
Objective:To evaluate the feasibility and clinical value of pre-treatment non-enhanced chest CT radiomics features and machine learning algorithm to predict the mutation status and subtype (19Del/21L858R) of epidermal growth factor receptor (EGFR) for patients with non-small cell lung cancer (NSCLC).Methods:This retrospective study enrolled 280 NSCLC patients from first and second affiliated hospital of University of South China who were confirmed by biopsy pathology, gene examination, and have pre-treatment non-enhanced CT scans. There are 136 patients were confirmed EGFR mutation. Primary lung gross tumor volume was contoured by two experienced radiologists and oncologists, and 851 radiomics features were subsequently extracted. Then, spearman correlation analysis and RELIEFF algorithm were used to screen predictive features. The two hospitals were training and validation cohort, respectively. Clinical-radiomics model was constructed using selected radiomics and clinical features, and compared with models built by radiomics features or clinical features respectively. In this study, machine learning models were established using support vector machine (SVM) and a sequential modeling procedure to predict the mutation status and subtype of EGFR. The area under receiver operating curve (AUC-ROC) was employed to evaluate the performances of established models.Results:After feature selection, 21 radiomics features were found to be efffective in predicting EGFR mutation status and subtype and were used to establish radiomics models. Three types models were established, including clinical model, radiomics model, and clinical-radiomics model. The clinical-radiomics model showed the best predictive efficacy, AUCs of predicting EGFR mutation status for training dataset and validation dataset were 0.956 (95% CI: 0.952-1.000) and 0.961 (95% CI: 0.924-0.998), respectively. The AUCs of predicting 19Del/L858R mutation subtype for training dataset and validation dataset were 0.926 (95% CI: 0.893-0.959), 0.938 (95% CI: 0.876-1.000), respectively. Conclusions:The constructed sequential models based on integration of CT radiomics, clinical features and machine learning can accurately predict the mutation status and subtype of EGFR.

Result Analysis
Print
Save
E-mail