A New Immune-Serum Scoring System:Role In Predicting The Prognosis Of Lung Adenocarcinoma

ObjectiveTo comprehensively investigate the effects of 18 immune cells in the immune microenvironment and related clinicopathological parameters on the prognosis of lung adenocarcinoma.Integrating the key impactors to construct a nomogram that can be used to predict the prognosis and stratify patients with the same TNM stage into different risk subgroups,which can further supplement the prognostic significance of TNM staging and guide individualized treatment.MethodsA total of 171 cases of postoperative pathologically confirmed lung adenocarcinoma were selected from September 2012 to March 2013 at the Tianjin Medical University Cancer Hospital.The clinical informations and corresponding formalin-fixed paraffin-embedded?FFPF?were collected.Immunohistochemistry?IHC?was used to detect the expression of 18 different types of immune cells in the core of tumor?CT?and the invasive margin?IM?of tumor in lung adenocarcinoma,including:T cells?CD3?,cytotoxic T cells?CD8?,B cells?CD20?,memory T cells?CD45RO?,naive T cells?CD45RA?,NK cells?CD57?,neutrophils?CD66b?,macrophages?CD68?,and regulatory T cells?FoxP3?.The expression of the above 18 of immune indicators was divided into low density group and high density group by"minimum P value method".Kaplan-Meier method were used to evaluate OS?overall survial?.171 cases were randomly divided into training cohort and validation cohort according to the ratio of 6.5:3.5.LASSO regression was used to further screen the key clinicopathologic parameters and immune markers related to OS.Integrating the above features to construct a nomogram.The predictive accuracy and consistency were evaluated by C-index and calibration curve.In addition,based on the point of each feature in the nomogram,a new immune-serum scoring model is constructed by summing the corresponding value.The ROC curve was used to compare the predictive accuracy of the scoring system and TNM staging.OS was calculated by Kaplan-Meier between different scoring groups with the same clinical stage.Results1.In the basic clinical pathological information,Kaplan-Meier analysis showed smoking history?P=0.013?,tumor size?P<0.001?,lymph node metastasis?P<0.001?,distant metastasis?P<0.001?,pathological type?P=0.017?,TNM stage?P<0.001?,preoperative serum NSE,CEA and Cyfra21-1 levels?P<0.001,P<0.001,P=0.020?significantly affected 5-year OS.2.The distribution of 18 immune cell expression heat maps showed the relative high expression levels of CD20tumor core?CT?,CD2Oinvasive margin?IM?,CD20CT,CD2OIM,CD3CT,CD3_IM,CD45RO_CT,CD45RO_IM,CD45RA_CT,CD8_IM.Inversely,the overall expression levels of CD8_CT,CD45RA_IM,CD68_CT,CD68_IM,CD66b_CT,CD66b_CT,CD57_CT,CD57_IM,FoxP3_CT,FoxP3_IM were relatively low.In addition,the correlation matrix showed an apparent cluster of CD45RA_CT and CD20_CT and a cluster of CD68_CT and CD68_IM.3.In Kaplan-Meier analysis,the group of patients with high CD8_CT?P=0.001?,CD8_IM?P<0.001?,CD45RO_CT?P=0.014?and CD45RO_IM?P=0.002?had a significantly better5-year OS.Meanwhile,the group of patients with high CD66b_CT?P=0.002?and FoxP3_CT?P<0.001?had a significantly worse 5-year OS.None of the other immune cells either CT or IM had significant prognostic value.4.The features obtained by LASSO screening including 3 clinicopathological characteristics?TNM stage,preoperative serum NSE,CEA?and 4 immune features(CD8_CT,CD8_IM,CD45RO_CT,FoxP3_CT)in the training cohort.5.Based on the variables screened by LASSO regression,a nomogram is established to predict the prognosis of lung ADC in the training cohort.The C-index for OS prediction was 0.89,indicating that the prediction performance of the model is very high.The calibration curve showed a prominent agreement and an acceptable agreement in the validation cohort between the actual observation and the nomogram prediction for 2-year and 5-year OS.6.Based on the point of each feature in the nomogram,a new scoring system?immune-serum scoring model?is constructed by summing the corresponding value.Sorted by total score,patients were grouped with cutoffs of 60 and 120?total score:0-60,61-120,>120?.5-year OS was significantly different between different subgroups patients?P<0.001?,and the accuracy of OS prediction was higher than TNM staging?AUC=0.861 vs.AUC=0.827?.7.To further investigate the significance of 4 immune indicators(CD8_CT,CD8_IM,CD45RO_CT and FOXP3_CT)and 2 serum indicators?NSE and CEA?in lung ADC,patients were divided into 4 group with cutoff values of 25,50,and 75 points?total score:0-25,26-50,51-75,>75?.The prognosis of patients in different groups was significantly different?P<0.001?.Compared with TNM staging,this group can better show the difference in prognosis between the 2 group and 3 group.Moreover,similar results were observed in patients with the same clinical stage when selecting 50 point as cutoff value to divide patients into low-scoring??50?and high-scoring?>50??P=0.007,P<0.001?.ConclusionsA new scoring system?immune-serum scoring system?established by the combined immune parameter and serum index,which OS predictive accuracy is higher than TNM staging.In addition,the scoring system can stratify patients with the same TNM stage into different risk subgroups,which can supplement the prognostic significance of TNM staging and further guide individualized treatment.