The Clinical Study On The Therapeutic Evaluation And Prognosis Of PET/CT In Lung Cancer

Objectives To investigate the clinical value of ¹⁸F-FDG PET/CT in the evaluation of curative effect and prognosis of lung cancer.PERCIST1.0 proposed using SUL_peak to replace the traditional SUV_max.In this study,we discussed the stability of SUL,and the consistency of SUL_peak automated measurement methods and manual measurement methods,to provide theoretical and operational basis for the application of the index firstly.Then the influence of PET metabolic parameters on the prognosis of lung cancer was analyzed.And focused on the comparison of PERCIST1.0 and RECIST1.1 in the evaluation of curative effect of lung cancer,in order to find and solve the problems in the evaluation of PERCIST in the treatment of solid tumors.To provide the preclinical basis for PERCIST better service in clinical,as well as further prospective,multicenter clinical studies.Methods 1)Retrospective analysis fifty patients with primary lung cancer confirmed by ¹⁸F-FDG PET/CT.We measured SUV_mean,SUV_max,SUL_mean,SUL_max and SUL_peak of lung cancer.SUV_mean,SUV_max,SUL_mean and SUL_max of liver?center region of right lobe?were also measured.Then T/N ratios(lesion SUV_max/ liver SUV_mean,lesion SUL_max/ liver SUL_mean and lesion SUL_peak/liver SUL_mean)of lung cancer lesions were calculated.The correlation between ¹⁸F-FDG metabolic parameters and BMI,gender of patients were analyzed.2)We measured SUL_peak–auto and SUL_peak-manual,which defined as the average SUL_max of five random pixels in fixed sized?about 1cm3volume?spherical VOI drawn over lesions calculated by two doctors and the same doctor at different time.Bland-Altman analysis and paired T test were used to evaluate the repeatability and consistency of SUL_peak–auto and SUL_peak-manual.3)Retrospective analysis patients with lung cancer who underwent PET/CT examination in our hospital.Baseline scan was performed before operation and at least 5 years follow-up.Evaluated of the relationship between PET metabolic markers and overall survival by multivariate COX proportional hazards regression model.Three points to treat each metabolic index,using K-M method to draw survival curves,and the survival of each group was compared by Log rank test.4)Two cases of non surgical lung cancer patients before and after the treatment of PET/CT in our hospital were retrospectively analyzed,and interval less than 6 months.Compared the efficacy of PERCIST1.0 with RECIST1.1 and visual evaluation of lung cancer by Kappa test.Compared the relationship among PFS,PERCIST1.0 and RECIST1.1 in evaluating the therapeutic effect in patients with lung cancer by Log rank test.Multivariate COX was used to evaluate the correlation among clinical indexes,metabolic indexes and their changes,PERCIST1.0,RECIST1.1 and PFS.The results of PET and CT joint evaluation were similar to the gold standard,the ROC curve was used to find the best node to judge the therapeutic response.Univariate and multivariate COX analysis was performed to evaluate the results of PFS data before and after two PET/CT intervals of less than 3 months.P < 0.05 was statistically significant.Results 1)The SUV_mean and SUV_max of liver were positively correlated with BMI?? =0.38,0.36,P < 0.05?,but there was no significant correlation between SUL_mean and BMI(SUL_max)?P > 0.05?.SUV_mean,SUV_max,SUL_mean,SUL_max and SUL_peak had no significant correlation with BMI?P > 0.05?in lung cancer.However,the T/N ratio(SUV_max)of lung cancer was only negatively correlated with BMI?? =-0.28,P < 0.05?in the T/N ratio.2)The bias of SUL_peak-auto was 0 g/ml* and repeatability was100% for liver and lung cancer lesions.While liver SUL_peak-manual of two doctors and the same doctor at different time were lower than SUL_peak-auto?P < 0.05?.Lung cancer lesions SUL_peak-manual of two doctors and the same doctor at different time were higher than SUL_peak-auto?P < 0.05?.3)There were statistically significant associations of OS with MTVpetvcar,MTV2.5,MTV25%,MTV42%,TLGpetvcar,but there were no statistically significant associations of OS with MTV50%,MTV75%,TLG2.5,SDS,all SUV and SUL.The subgroup analysis according to the pathology showed that there were statistically significant associations of OS with stage,MTV50%,MTV42% in adenocarcinoma group,with SUL_mean,MTV25%,MTV42% in squamous cell carcinoma group,but with none in others group.There were statistically significant associations of OS with MTV42% and MTV50% in early stage,but none in late stage.4)The agreement between PERCIST1.0 and RECIST1.1in the evaluation of lung cancer was moderately good?Kappa = 0.524,P = 0.00?,with a statistically significant correlation with PFS?P < 0.05?.SUV_mean,?MTV and ?TLG were associated with PFS,which mean the three were helpful to predict PFS.In addition,the trichotomy showed ?TLG was the most influential factor to PFS.There was general consistency between single lesion and multiple lesions ?SUL_peak in evaluation of the therapeutic effect of lung cancer.Joint evaluation was better than single evaluation.ROC curve showed that ?SUL_peak selected-26.05% as PR,0.45%for PD;?TLG selected-53.1% as PR,20.35% for PD;?MTV selected-39.5% as PR,12% for PD;?SUL_mean selected-22.9% as PR,17.65% for PD;?SUV_max selected-28.9% as PR,22% for PD;?SUV_mean selected-23.1% as PR,21.2% for PD.The prognosis of PERCIST1.0 was better than RECIST1.1.Conclusion SUL and T/N ratio?SUL?affected by patients BMI and gender were not obvious,which could be suitable to quantitative analyze and assess treatment response.The SUL_peak-manual mainly depended on manual measurement,whose repeatability and consistence were inferior to SUL_peak-auto,while the SUL_peak-auto had good reliability.The accuracy and reliability of SUL_peak-auto from PETVCAR could be suitable to assess treatment response,and helpful to PERCIST clinical application and promotion,which evaluated response mainly by SUL_peak.¹⁸F-FDG PET/CT MTVpetvcar,MTV2.5,MTV25%,MTV42%,TLGpetvcar had the prognostic value with OS.Select the appropriate metabolic parameters,may predict prognosis of non-small cell lung cancer.PERCIST1.0 and RECIST1.1 had better consistency,better relationship between PERCIST1.0 and prognosis,and earlier diagnosis and prognosis of lung cancer than RECIST1.1,and the best combined assessment.PERCIST1.0used multiple lesions to judge prognosis better than single lesion.The changes of each metabolic index can be used to judge the prognosis.TLG may be the best prognostic parameter in metabolic parameters.TLG taked-53.1% as the PR boundary point,selected 20.35% for the PD boundary point to determine the prognosis may be better.The results of this study need to be further validated by prospective,multicenter clinical studies.