Development Of Prognostic Models For Lung Cancer Patients With Malignant Pleural Effusion

Background Malignant pleural effusion(MPE)can be caused by lung cancer with pleural metastasis.Patients with MPE usually have a poor outcome with a median survival of 3 to 12 months.Some studies demonstrated that non-small cell lung cancer(NSCLC)patients usually survived approximately 8.5 months.LENT score and PROMISE score are prognostic models to predict the survival of patients with MPE,which include clinicopathological characteristics.Whereas,LENT and PROMISE models are used to assess the survival of all cancer patients.They may not apply to lung cancer patients with MPE as the result of different tumor origins and treatment.We aimed to develop models to predict the survival of lung cancer patients and lung adenocarcinoma patients.Methods Lung cancer patients with MPE were included in Jinling Hospital from January 01,2008 to June 30,2018.Patients were followed and the clinicopathological features like gender,age,smoking,ECOG PS,relapse,WBC,NLR,Hb,PLT,CRP,liver and kidney functions,albumin-Globulin ratio,pleural LDH,pathological types,TNM stage,mutations were collected.We selected risk factors using multivariable Cox proportional-hazards analysis in the development cohort.The risk models were created according to the risk ratio(RR)value.Results 367 lung cancer patients were included in the development cohort,of which,54 patients were lost to follow-up.Lung adenocarcinoma accounted for the majority of all the pathological types with the proportion of 77.4%.The prediction efficiencies of LENT and PROMISE models were low for lung cancer patients with MPE.Based on the results of the multivariate Cox proportional-hazards analysis,RECLS(Relapse or not,ECOG PS,CRP,pleural LDH and TNM Stage)scoring system was created for lung cancer patients with MPE.Lung cancer patients were categorized into low-risk group(0-4 points),moderate-risk group(5-10 points),high-risk group(11-15 points)with the median survival of 716 days,224 days and 38 days,respectively.Low-risk patients had a median survival of 716 days and the survival rates of this population at1 month,6 months,12 months were 99.2%,92.3%,75.7%,respectively.Patients in high risk group had a significantly shorter median survival of 38 days,and 51.9%,7.4%,3.7% of the subjects survived to 1 month,6 months and 12 months.The moderate risk group showed intermediate prognosis of 90.1%,54.8% and 37.9% at 1month,6 months,1 year,with a median survival of 224 days.Area under the curve(AUC)values for the RECLS model were 0.911,0.845 and 0.754,respectively at 1month,6 months and 12 months,which indicated good performance of RECLS model.We also developed a RECLSAM score for lung adenocarcinoma patients: Relapse or not,ECOG PS,CRP,pleural LDH,TNM Stage,Albumin-Globulin ratio and activating gene Mutation.The median survival of low-risk population,moderate-risk population and high-risk population were 783 days,227 days,38 days,respectively.Conclusion We developed the prognostic models for lung cancer patients with MPE based on retrospective single center analysis.RECLS score and RECLSAM score are practical models with good prediction efficiency for lung cancer patients and lung adenocarcinoma patients with MPE,which can help guide the the selection of optimal treatment strategies for the population in clinical work.