| Background:Lung acinar cell carcinoma(LACC)is a rare histologic subtype of lung cancer,with few studies on its prognosis because of its low incidence.The aim of this study was to identify independent prognostic factors in patients with LACC and to construct two nomograms for the prediction of overall survival(OS)and cancer-specific survival(CSS)in LACC patients,based on which individualized survival prediction and prognostic assessment could be achieved,which would help to guide treatment strategies and improve patient prognosis.Methods:This study retrospectively analyzed clinicopathological data from patients diagnosed with LACC from the Surveillance Epidemiology and End Results(SEER)database(SEER*Stat 8.3.6 software)from 2010-2015.According to inclusion and exclusion criteria,patients were randomized into two study cohorts:the training cohort(70%)and the validation cohort(30%).The training cohort was used to identify independent OS-related and CSS-related factors and to construct nomogram,while the data of LACC patients in the validation cohort was used for external validation of the constructed nomogram.Potential prognostic factors were first identified in the training cohort using univariate Cox regression analysis,and statistically significant(P<0.05)variables were further included in the multivariate Cox regression analysis to determine the independent OS-related and CSS-related factors for LACC patients,with hazard ratio and 95%confidence intervals(CIs)calculated for each variable.Based on the prognostic variables identified in the study,the"rms" and "regplot" packages of R software were used to construct nomogram for predicting the OS and CSS rates of patients with LACC at 1-,3-and 5-year,respectively.The newly developed nomogram was evaluated and validated using the Harrell concordance index(C-index),receiver operating characteristic(ROC)and area under curve(AUC),calibration curves and decision curve analysis(DCA)in the training and validation cohorts to evaluate the discriminative power,predictive accuracy and clinical application of the nomogram.Results:A total of 1599 patients were included in this study,of which 1120 patients were classified into the training cohort and another 479 patients were classified into the validation cohort.Age,tumor size,gender,tumor pathological grade,AJCC stage and surgery were identified as independent prognostic factors for OS in LACC patients based on Cox regression analysis(P<0.05).In addition,age,tumor size,AJCC stage and surgery were determined as the independent prognostic factors for CSS in LACC patients.Two nomogram models were constructed to predict the OS and CSS of patients with LACC at 1-,3-and 5-year respectively,and both prediction models performed well with excellent discriminatory power and good fit.The ROC curves indicated that the model had good predictive performance,in the training cohort,the AUC values for the OS nomogram at 1-,3-and 5-year were 0.819,0.768 and 0.756,for the CSS nomogram at 1-,3-and 5-year were 0.843,0.790 and 0.770.In the validation cohort,the AUC values for the OS nomogram at 1-,3-and 5-year were 0.788,0.745 and 0.713,for the CSS nomogram at 1-,3-and 5-year were 0.871,0.731 and 0.712.The calibration curves indicated that both nomograms performed well in both the training and validation cohorts,with high agreement between the actual observed and predicted values.Moreover,the DCA results suggest that the newly constructed prediction model has good clinical application.In addition,the ROC curve results showed that the AUC of the OS and CSS nomogram was consistently greater than the AUC of each single independent prognostic factor in both the training and validation cohorts.Conclusion:This study demonstrates that age,tumor size,gender,tumor pathological grade,AJCC stage and surgery are independent prognostic factors for OS in patients with LACC,and that age,tumor size,AJCC stage and surgery are also independent prognostic factors for CSS.The two nomograms were constructed to predict 1-,3-and 5-year OS and CSS. |
PDF Full Text Request