| Background and purposeThe morbidity and mortality of lung cancer in China is the first place,and the early diagnosis of lung cancer has a significant impact on the treatment and prognosis.At present,the diagnosis of lung cancer by molecular epidemiological analysis,diagnosis,histological and molecular biology make a preliminary judgment,the final pathologic diagnosis was obtained.CT is currently the diagnosis of chest common lung cancer,but only in the CT level to explain anatomical lesions.Especially for the silence onset malignant solitary pulmonary nodules,CT was difficult to make a qualitative diagnosis on its cell metabolism function.PET/CT are the main methods to monitor the development of lung cancer which can be combined with anatomic and metabolic function effectively.However,the single examination cannot provide evidence for the early diagnosis of lung cancer.In this study,the PET/CT and serum tumor markers combined application was designed to investigate the value of differential the benign from the malignant lesions by using combination group.SUVmax has always been a semi quantitative and important parameter to judge the benign and malignant lung cancer,but there is still no deterministic value and the study of the optimal threshold value of SUV is good for the differential diagnosis of lung cancer.And finally,the correlation between maximum standard uptake value(SUVmax)and clinical features such as gender,age,tumor size,tumor location,tumor pathological types,tumor stage and tumor differentiation degree were analyzed,etc.Material and MethodsWe retrospectively studied the clinical data of 239 cases with lung neoplasm in Daping Hospital from January 2013 to June 2015.These cases underwent PET/CT and serum tumor markers(CEA,CA125,CY21-1,NSE)inspections within ten days prior to the operations.The sensitivity,specificity,accuracy,positive predictive value(PPV)and negative predictive value(NPV)and the correlations between clinicopathologic features(age,sex,tumor size,smoking history,tumor size,pathological type,pleural reaction,lymph node metastasis and anatomical location)with standard uptake value(SUVmax)were evaluated(t-Test and One-Way ANOVA),respectively.The diagnostic value of tumor markers,PET and the combined group in the different maximum diameter(≤2,2-3,3-5 and >5cm)of the tumor was compared with the accuracy.Furthermore,applying the logistic regression and ROC curve to compare the diagnostic efficacy of the three groups under different maximum diameter of lung cancer.The cutoff value of SUVmax was calculated without considering the effect of Computer tomography image according to the principle of maximum Youden index.Applying the Pearson correlation and Spearman rank correlation to analysis the relations between the SUVmax with different sex,age,tumor size,pathological stage,pleural invasion and histological type.Results1.193 patients and 46 patients were diagnosed as lung cancer and benign lung diseases,respectively.With regard to lung cancer,the sensitivity,specificity,accuracy,PPV,NPV of the serum tumor markers and PET/CT were 62.2% vs.91.7%,73.9% vs.47.8% and 64.4% vs.83.2%,89.6% vs.88.1%,31.8% vs.57.9%,respectively.Interestingly,the sensitivity,specificity,accuracy,PPV and NPV by using tumor markers plus PET/CT were 97.9%,45.76%,87.9%,88.3%,84.0%.Sensitivity and NPV of the combination group was significantly higher than sole use of serum tumor markers or PET/CT(p<0.05).2.Area under the ROC curve(AUCROC)of serum tumor markers,PET/CT and their combination were 0.751(CI95%,0.684-0.819),0.790(CI95%,0.718-0.862),0.832(CI95%,0.772-0.892).The diagnostic efficacy of combination group was higher than the serum tumor markers or PET/CT.3.Applying ROC curve and the maximum principle of Youden’s index to calculate the SUVmax cutoff which was 0.864(when the sensitivity was 76.8% and specificity was 75.5%).4.According to the lung cancer with different maximum diameter(≤2,2-3,3-5,> 5cm),serum tumor marker,PET / CT,combined group lung cancer diagnosis accurate rate was 51.6%,63.7%,59.0%,77.0%;70.0%,75.4% and 84.2%,96.7%.80.6%,85.5%,85.8% and 96.7%,respectively.When the maximum diameter of the tumor was ≤2 and 2-3cm,the area under the ROC curve of the tumor marker were 0.555 and 0.738,respectively.When the maximum diameter of the tumor was 3-5 and >5cm,the area under the ROC curve of the tumor marker were 0.749,0.867.When the maximum diameter of lung cancer was ≤2,2-3,3-5 and more than 5cm,the diagnostic efficacy of PET/CT in the diagnosis of lung cancer were 0.700,0.761,0.819 and 0.898,respectively.Similarly,The area under the curve of the combined group of lung cancer with different maximum diameter(≤2,2-3,3-5,> 5cm)were 0.732,0.811,0.891 and 0.949(95%CI,0.871-1.00),respectively.5.SUVmax expression in lung cancer was significantly different in tumor size,smoking history,anatomical location,pathological type,differentiation degree,and staging group.The SUVmax in male group was higher than female group(p< 0.05).Tumor size ≥3cm was higher than tumor size <3cm(p < 0.05).SUVmax in different anatomical location(peripheral type 12.52±6.90,central type 17.20±6.28,p<0.05)was significantly different.SUVmax in different smoking history(Yes,No)was significantly differently(P <0.05).SUVmax in different pathological type(adenocarcinomas12.89±6.74,squamous15.41±7.52,Small cell lung cancer12.76±8.21,Other types13.02±7.69,P <0.05)was significantly different.The SUVmax in different Staging group(Ⅰ8.15±7.27,Ⅱ 15.06±6.38,Ⅲ15.38±6.62,14.25±5.55,P<0.05)were significantly different.According to whether lung cancer have pleural reaction divided into pleural reaction group and no pleural reaction,the pleural reaction group SUVmax is 13.92 ± 6.56,no pleural reaction group,SUVmax is 13.26 ± 7.38,the SUVmax difference between pleural reaction and no pleural reaction was not statistically significant(P>0.05).According to the lymph node metastasis of lung cancer were divided into the group.Transfer group SUVmax was(14.89 + 6.43)and the average value of non lymph node metastasis was(12.09 + 7.52).Ultimately,the transfer group SUVmax was higher than that in the non metastasis group(P < 0.05).6.SUVmax was positively correlated with tumor size(r=0.498,p<0.01)and tumor location(r=-0.615,P<0.01).In the lung beign disease.7.SUVmax and lung cancer lesion size is moderate positive correlation(r=0.498,P < 0.01),in the lung benign disease,SUVmax and the primary tumor size is weak correlation(r=0.327,P < 0.05).8.SUVmax was weakly correlated with lymph node metastasis of lung cancer(r=0.209,P < 0.05).9.The relationship between smoking history and SUVmax was weak(r=0.207,P < 0.05).10.There was no correlation between the pathological type of lung cancer and the age of the patients with SUVmax(P > 0.05).Also,there was no correlation between lung cancer pleural invasion and SUVmax(r=0.072,P > 0.05).11.Among many factors,the largest diameter of primary tumor of lung cancer and the Lymph node metastasis is the independent imaging factor of SUVmax.Conclusion1.The combined use of PET/CT and serum tumor markers is contributive to the diagnosis of lung cancer.By the way,the combined diagnosis can provide a relatively reliable diagnostic method for patients without chinically pathological examination or pathological failure in the case of favorable avoidance of false positive and false negative interference.2.The greater the diameter of lung cancer was,the higher the diagnostic efficiency of the three groups was.In clinical practice,we should take the means of joint detection in order to avoid missed diagnosis and misdiagnosis when evaluating the largest diameter of the tumor less than 3cm pulmonary nodules.3.Our study apply ROC curve and the maximum principle of Youden’s index to calculate the SUVmax cutoff according with the actual situation of our hospital which may provide the clincians in our hospital a reliable clues distinguishing between benign and lung cancer.4.The size of the primary lesion,smoking history and lymph node metastasis was the independent influencing factor of SUVmax. |
PDF Full Text Request