| Objectives We aimed to investigate predictive factors of occult lymph node metastasis(OLM)and to explore the diagnostic value of various standardized uptake value(SUV)parameters using 18F-FDG PET/CT in predicting OLM of non-small cell lung cancer(NSCLC)patients.Methods We retrospectively analyzed 124 patients who underwent 18F-FDG PET/CT at the Third Affiliated Hospital of Soochow University from January 1,2013 to December 31,2017 and preoperatively diagnosed as clinical NO stage(cN0).All patients underwent anatomical pulmonary resection with syste1atic lymph node dissections within three weeks after 18F-FDG PET/CT examinations.The patients were defined as occult lymph node metastasis group(OLM)if lymph node metastasis were detected and as non-occult lymph node metastasis group(NOLM)if didn't.Patients' clinical,laboratory characteristics and 18F-FDG PET/CT parameters of tumor,such as the maximum diameter(Dmax),tumor site,morphological features,maximum standardized uptake value(T-SUVmax),mean standardized uptake value(T-SUVmean),metabolic total volume(T-MTV)and total lesion glycolysis(T-TLG)were obtained.Standardized uptake ratio(SUR)were derived from tumor SUV to blood SUVmean standardized uptake ratio(B-SUR)and tumor SUV to liver SUVmean standardized uptake ratio(L-SUR).We used the independent sample T test,Mann-Whitney U test and Pearson's chi-square test to compare the parameters between groups appropriately.Using receiver operating characteristic(ROC)curve to evaluate the diagnostic value of different parameters.The univariate and multivariable logistic.regression were used to analyze the independent risk factors for OLM.Results Nineteen(15%,19/124)patients were diagnosed as OLM among 124 cN0 NSCLC patients.In OLM group,36 lymph nodes were pathologically positive for metastasis,including 25 N1 disease and 11 N2 disease.The carcinoembryonic antigen(CEA),cytokeratin 19 fragment(CYFRA21-1),lobulation and PET metabolic parameters of primary lesions(T-SUVmax,T-TLG,L-SURmax,L-TLGsur,B-SURmax and B-TLGsur)in OLM group were significantly higher than those in NOLM group(z=-3.853?-1.680,?2=7.521,all P<0.05).L-SURmax of primary lesion yielded the highest area under the ROC curve(AUC=0.778),which was higher than those of CEA,CYFRA21-1,lobulation,T-SUVmax,T-TLG,L-TLGsur,B-SURmax and B-TLGsur(AUC=0.625,0.684,0.671,0.767,0.707,0.724,0.774 and 0.718,respectively).However,the difference were not statistically significant.When L-SURmax>2.68,the sensitivity,specificity,positive predictive value and negative predictive value of L-SURmax for predicting OLM were 94.7%(18/19),57.1%(60/105),28.6%(18/63)and 98.4%(60/61),respectively.The multivariable logistic regression analysis showed that L-SURmax(OR=35.390,95%CI=3.957-316.489,P=0.001),CEA(OR=4.149,95%CI=1.069?16.110,P=0.040)and CYFRA21-1(OR=8.309,95%CI=2.072?33.311,P=0.003)were independent risk factors for OLM.Compared to L-SURmax alone and the combination of CEA and CYFRA21-1(Model 2),the Model 1 consisting of three independent risk factors achieved the highest AUC(0.901 vs.0.778 vs.0.780,P=0.021 and 0.0141).Conclusions L-SURmax of primary lesion were independent risk factor for OLM in NSCLC patients and it could sensitively predict OLM preoperatively.The Model consisting of three independent risk factors(CEA,CYFRA21-1 and L-SURmax)can effectively predict OLM in NSCLC patients. |
PDF Full Text Request