An Analysis Of Correlated Factors About The Chemoradiotherapy Sensiitvity And Radiation-Induced Pneumonitis In Stage â…¢ Non-Small Cell Lung Cancer Patients

BACKGROUNDCurrently, chemoradiotherapy (CRT) is the standard of care for locally advancednon-small cell lung cancer. Patients have different responses to CRT due to individualdifferences, thus there is still lacking of sensitive markers to predict treatmentresponse in non-small cell lung cancer (NSCLC) patients treated with CRT. We havefiltered several serum protein in our previous study, apolipoprotein A-1(ApoA-1) andalpha-1-antitrypsin (α1-AT), which were closely associated with NSCLC. However, itis uncertain whether these two proteins can predict treatment sensitivity of NSCLCpatients treated with CRT.OBJECTIVETo study the differences in ApoA-1andα1-AT levels among sub-groups ofdifferent clinical characteristics, and to investigate the role of these two proteins and related parameters in predicting treatment sensitivity of NSCLC patients treated withCRT.MATERIALS AND METHODSA total of50patients with Stage III NSCLC were evaluated. All the patients weretreated with CRT. Blood samples were obtained by venipuncture before CRT.Enzyme-linked immunosorbent assay was used to measure the levels of ApoA-1andα1-AT. The treatment response was evaluated2months after the treatment accordingto Response Evaluation Criteria in Solid Tumors version1.0(RECIST v1.0).One-way ANOVA was used to study the differences in ApoA-1andα1-AT levelsamong sub-groups of different clinical characteristics. Independent t test wasconducted to find the difference in the levels of these two proteins between response(CR+PR) and non-response (SD+PD) groups.RESULTSThe levels of ApoA-1and α1-AT were significantly different between maleand female patients (p=0.036and0.006). The α1-AT level was correlated withTNM stage and histology (p=0.031,0.043). There was37patients in sensitive group(10patients got CR,27got PR),13patients in resistive group (9patients got SD,4got PD). α1-AT level before treatment had the ability to predict insensitivity ofCRT (p=0.014, AUC=0.731), the ratio of post/pre ApoA-1had the ability to predictsensitivity of CRT (p=0.011, AUC=0.759). And the specificity, sensitivity, accuracyvalue was61.5%,75.7%,72.0%and81.2%,63.6%,70.0%, respectively.CONCLUSIONSThe α1-AT level was associated with tumor stage and histology in NSCLCpatients. The level of α1-AT of sensitive group was significantly lower than that ofresistive group, which indicated thatα1-AT might be a potential predictor fortreatment response of NSCLC patients treated with CRT. BACKGROUNDDuring radiotherapy for thoracic carcinomas, some of the normal lung tissue isunavoidable to be irradiated. The risk of radiation-induced pneumonitis (RP) was thepredominantly dose-limiting factor of thoracic radiation, which can have aconsiderable impact on patient survival and quality of life. Because the developmentof RP is always hysteretic, and there is no effective therapeutic method for serous RP,to predict and prevent the incidence of serious RP seems urgently necessary. Singlepositron emission computed tomography (SPECT) lung perfusion scan can be directlylinked to local function of the lung tissue. SPECT images can be matched to thehighly geometrically reliable CT image and be used in the radiotherapy planningoptimization. Thus functional lung can be better protected, and lung toxicity can bedecreased at some extent. However, the threshold of defining functional lung inSPECT perfusion images is still controversial home and abroad.OBJECTIVETo investigate the correlation between RP occurred after treatment and clinicalfactors of the enrolled patients, such as age，sex，radiation dose, tumor stage, histology,KPS，smoking history, and chemotherapy regimen. Through analyzing the abilities ofdifferent functional dose-volume histogram parameters determined by SPECT lungperfusion images of different thresholds, SPECT-weighted mean lung dose (SWMD)and volume percentage of lung receiving more than20Gy (FV20), to establish anoptimal one for the definition of functional lung. MATERIALS AND METHODSA total of40patients of stage â…¢ NSCLC treated with CRT, who came from ourfirst part and were conducted SPECT lung perfusion scan were retrospectivelyrecruited in the study. RP development was evaluated after treatment throughfollow-up, and grade2or higher RP was considered to be of significance. All of themobtained SPECT lung perfusion scans and positron emission tomography/computedtomography (PET/CT) scans for treatment planning, and the images wereco-registered in the treatment planning system. Total lung and perfusion lung volumescorresponding to10,20,…,60%of the maximum SPECT count were segmentedautomatically. The SWMDx%and Fx%V20of different thresholds were investigated,using SPECT-weighted DVHs. RP was evaluated and graded according to theNational Cancer Institute Common Terminology Criteria for Adverse Events (NCICTCAE) version3.0. Univariate logistic regression models were fit to assess theassociation between grade2or higher RP and covariates of interest, including age, sex,histology, tumor stage, KPS, smoking history, and chemotherapy regimen. And thedifferences among SWMDx%and Fx%V20were compared respectively in Friedman test.Receiver operator characteristic (ROC) curves were used to identify SWMDx%andFx%V20of different thresholds in predicting the incidence of grade2or higher RP.RESULTSThe results from univariate logistic regression models showed that the risk of grade2or higher was significantly associated with KPS (p=0.029), but there was nosignificant correlation between grade2or higher RP and other patient characteristics(all p＞0.05). Total lung volume and functional lung volumes at the threshold of10%,20%,…,60%of the maximum SPECT count were calculated. As the thresholdincreased, the lung volume decreased irregularly. SWMDx%and Fx%V20fromfunctional DVHs at different thresholds were significantly different among groups (both of p value＜0.001). Through follow-up, eleven patients (27.5%) developedsignificant RP (grade2,3and4were12.5%,10.0%, and5.0%, respectively) aftertreatment. The largest area under the ROC curve was0.881for the ability of SWMDto predict RP with20%as threshold and0.928for the ability of FV20with20%asthreshold, with the specificity100%and90.9%, sensitivity79.3%and86.2%, acuracyvalue85.0%and87.5%. The cut-off value of SWMD20%and F20%V20were16.3Gyand29.8%, respectively. Combining SWMD20%and F20%V20obtained from SPECTlung perfusion images, RP could be better predicted.CONCLUSIONRP development was significantly correlated with patients’ performance status, thepatient with poor status tended to develop RP. Further study is urgently needed toverify the results. SPECT lung perfusion images showed heterogeneity of lungperfusion. Different areas of the lung may have different degrees of perfusion and thefunctional lung region differed as different threshold was used in SPECT lungperfusion imaging. A significant correlation was demonstrated between the risk of RPand DVH parameters, SWMD and FV20. In our study, SWMD20%å'ŒF20%V20have theability of predicting RP. Using20%of the maximum SPECT count as threshold, thedefined functional lung could be better protected through modifying beam to avoidfunctional region. This might reduce the incidence and severity of RP and potentiallyprovide the probability of dose escalation.