Detection Of Early Response To Chemotherapy By Using Diffusion Weighted Image For EGFR Mutated And Wild-type Lung Adenocarcinoma

BackgroundLung cancer is the main cause of tumor-related death in whether China or developed countries (United State). As US official cancer statistics showed, in 2014, about 224,210 new patients were diagnosed with lung cancer in US, which with death cases of 159,260. According to data from Chinese Cancer Registry Annual Report in 2012 published by the National Central Cancer Registry, annual new cases of cancer in China were 3.12 million, death cases were 2.7 million. Compared to the 10 million of annual new cases of cancers and 5 million death cases in worldwide with a population of 6.6 billion, the proportions of Chinese were about 31.2% and 54% respectively. Lung cancer is also one of the tumors with the highest incidence in China (ranked first in males with a proportion of 23% and ranked second in females with proportion of 14.85%). Both active smoking and passive smoking are primary risk factors for causing lung cancers. Due to the occult of the incidence of lung cancer, no obvious clinical symptom can be observed in the early stage; when the neoplastic lesions can be observed, it is already too late for timing surgical treatment.In recent decades, a big progress has been made in terms of lung cancer research from screening to diagnosis, minimally invasive treatment, as well as targeted treatment and so on. EGFR gene is one of the meaningful targeted gene therapies for treating non-small lung cancer.Because of the low specificity and strong adverse effects of standard chemotherapies, in recent years, EGFR targeted drugs such as erlotinib, gefitinib and afatinib, as well as the developing of new tyrosine kinase inhibitor have always been research hotspots of non-small lung cancer treatmentEpidermal growth factor receptor (EGFR) is a member of erbB gene family, and the other three closely related receptor tyrosine kinases are Her2 (Neu, ErbB2), Her3 (ErbB3), and Her4 (ErbB4). EGFR is a type of glycoprotein, and it is a receptor tyrosine kinase with the structures of an extracellular ligand binding domain, transmembrane region and an intracellular tyrosine kinase domain. It forms a dimer structure after binding with the ligand, and forms the second messenger by auto-phosphorylation, which can then participate in the signaling of cells, and finally participate in the regulations of growth, division and apoptosis of cells. EGFR illustrates overexpression/mutation status in breast cancer, lung cancer and many other tumor tissues in either human or animal models. Many experiments also suggest that the overexpression/mutation in non-small lung cancer are associated with the growth, invasion and metastasis of tumor vessels. It is located on the cell membrane. Thus it is a good gene target for targeted treatment.Diffusion weighted imaging (DWI) is the only noninvasive approach to detect movement of water molecules in cells, and it is widely applied especially for brains . The combination of DWI and spin-echo echo planar imaging (SE-EPI) dramatically reduce the imaging time of DWI, and significantly reduce the effect of breathing and heart movements on imaging quality.There are increasing functions of DWI in tumor diagnoses; breast cancer, prostate cancer and brain glioma can be measured by the apparent diffusion coefficient (ADC), and benign or malignant tumor be diagnosed with semi-quantitative or even quantitative methods [18,19]. This has great significance for both patients and clinicians.Research StatusCurrently, PET-FDG test with F-18 marker is recommended by National Comprehensive Cancer Network (NCCN) for staging of non-small lung cancer and observations of treatment prognosis. However, this method is expensive, and it has a high false positive rate for tumor lesions with the diameter of less than 10mm. Moreover, ionizing radiation can be produced during the testing process, which limits the widely application of this test project. In recent years, DWI had great significance for detection, diagnosis, staging, prognosis and treatment efficacy evaluation of brain tumors. ADC value can provide semi-quantitative or even quantitative research results for these experimental results. Some studies suggested that DWI is as good as PET/CT in terms of non-small lung cancer, especially for staging of N and M. It can also play a guiding role in chemotherapy regimens development for patients who cannot receive surgical treatment. There has been an article that investigates the changing rate of the apparent diffusion coefficient (ADC) illustrated in the early stage of tumor treatment by using DWI after non-small lung cancer chemotherapy, and the prediction capacity had similar performance as PET/CT. Other articles reported that dynamic contrast-enhanced MRI was applied to predict the early efficacy of first line lung cancer chemotherapies:bevacizumab, gemcitabine and cisplatin, and good experimental results were achieved. However, those research described above did not perform unified study for small cell lung cancer or non-small cell lung cancer pathological type, and did not carry out category observation about whether EGFR had been mutated in lung cancer.Therefore, our research focused on the non-small cell lung cancer of lung adenocarcinoma and used whether EGFR was mutated as grouping criteria to investigate clinical application value of DWI for efficacy prediction of early chemotherapy with EGFR mutations.Aim:To investigate the clinical application value of early chemotherapy efficacy prediction capacity of DWI for lung adenocarcinoma patients with EGFR mutations Materials and methods:Materials and Methods:1. PatientsThis experiment had been approved by Ethics Committee of our hospital, and the subjects were 39 patients with primary lung adenocarcinoma who were firstly diagnosed from jan 1st,2015 to Nov lth,2015; those patients did not receive chemotherapy and surgical treatment before, wherein 29 patients were males and 10 patients were females, and the age was form 30 to 80. Tumor tissues from all patients were obtained by CT-guided percutaneous biopsy, and all obtained tissues were confirmed as adenocarcinoma, EGFR and ALK gene mutation test was subsequently performed. All the patients has no ALK mutated. EGFR test sequence include:exon 18 G719X (G719A、G719C、G719S); exon 19:E19-DEL; exon 20:EXIB20-ins. T790M. S768I; exon 21 L858R。2. CT-guided percutaneous biopsy:All experimental manipulations were finished on Philips 16-slice spiral CT, scanning conditions were 120kv 80mAs, layer thickness was 5mm, and layer spacing was 5mm. Cutting needle biopsy or needle aspiration biopsy was selected appropriately according to preoperative enhanced CT scan, PET/CT and other imaging data from our hospital or other hospitals, as well as whether there is a necrotic area in interior lesions and the size of a necrotic area. Before the biopsy, operation doctor and chief doctor had talked together with patients and let the patient sign informed consent. Routine examinations of clotting time, platelet count and prothrombin were performed before the operations. If patients had history of long-term use of anticoagulant, the medicine need to be ceased for over 2 weeks, and clotting time, platelet count and prothrombin need to be re-examed, and biopsy can be performed only when patients’ coagulation capacity returned back to normal.According to patients’ age, their situation, and the size and the actual location in the lung, perating doctors and assistant help patient to put their on a more comfortable position which convenient for operate the biopsy. The patient no need to hold breath. Take a metal stick in the focal region about surface markers, in after a scan location to determine the puncture point and angel.operate biopsy with "three-step" biopsy:1, anesthesia with 2% lidocaine 5 ml.inject the needle into the muscle, then confirm the direction and Angle 2. After the needle into the lung, run CT scan again to ensure the needle is on right path;3. Pull the needle into the lesion, according to the lesion size, cutting or suction suitable diameter for organization. Then pull out the needle, stick in the wound with sterile apply paste. After the biopsy, Row chest CT scan again to ensure there is no complication occurring, especially the presence of pneumothorax, serious bleeding and the air embolism.All biopsy operators have to be doctors with over high professional title and with biopsy experiences of over 20 years.3. MRI examination methodThe operations of Philips Achieva 3.0T superconductive MR scanners (made in the Netherlands) were examined. XL-torso coil was performed to scan the whole body. Patients laid down on the middle of the magnet center and were covered with the induction coil. Conventional scanning sequence:axial T1 WI Thrive (TR 3 ms, TE 1.4 ms), scanning range (Field of View FOV)350 X 350, layer thickness 4mm, layer spacing -2mm, T2WI sSSH-SPAIR(TR1100 ms, TE110ms), FOV 350 X 350, layer thickness 4mm, layer spacing 2mm. DWI scanning:EPI-DWI(echo planar imaging-diffusion weighted imaging) (TR 2000 ms, TE 60ms), b. value was 500,1000, FOV 350 X 350, layer thickness 3mm, layer spacing 2mm. Respiration sensing devices were placed on the abdomen of patients, respiration-triggered scanning was applied in both sSSH-SPAIR and DWI imaging, the triggering time was depended on the respiratory rate per minute of patients,0-300 ms triggered scanning was delayed according to the actual situations. Breath-hold scan was applied for thriving sequence, and the breath-hold scan time was 8-11s. Second order shim was performed before the scanning. The whole scanning time was around 10 min.4. DWI imaging processing and data measurementDWI raw data were imported into Philips Extended MR Workspace 2.6.3.1 software to obtain ADC images.The obtained images were analyzed by two imaging diagnosing physicians with high qualifications (higher than deputy director or with 15 years of MRI diagnosis experiences). Combined with conventional MRI images (T1WI, T2WI), plain CT scanning and enhanced scanning images, regions of interest (ROIs) were delimited at biggest area in tumor part of ADC images; calcifications, hemorrhages, hollows and other regions which may affect ADC values were avoided. Measurements were repeated for 3 times, and took the average as the final result. The formula of calculating The changing rate of ADC values (formula 1-1):△ADC=ADCafter-ADCbefore/ADCbefore X 100%(formula 1-1)After 4-6 weeks of chemotherapy, patients were requested to go back to hospital for re-examination. The first CT diagnosis and CT review were applied to measure the size of cancer foci, and to compare and observe whether there was any metastasis. CT scanning was performed at Philips 64 slice CT or i256 slice CT, and RECIST (response evaluation criteria in solid tumors) vl.1 standard was utilized as reference of chemotherapy efficacies. Patients were divided into effective group and non-effective group according to the changes in tumors. The effective group includes: Complete response (CR):foci was completely disappeared; partial response (PR):the diameter of foci was decreased by 30%; Non-effective group includes:Progressive Disease (PD):the diameter of foci was increased by 30%, or foci was not changed, but new foci was emerged; Stable disease (SD):the foci was increased by less than 30% or was decreased by less than 20%.5. Statistical analysisAll experimental data were analyzed by using statistical software SPSS 19.0, and normality tests were performed on measurement data. EGFR gene expression result was obtained by pathology genetic test, and ADC values of EGFR mutant group, wild type before and after treatment were calculated individually, and change rate of ADC values were calculated. (1)The ADC value comparisons of EGFR mutant group and wild type group before treatment were performed by using two independent-samples t-test; (2) the ADC value comparisons of effective group and non-effective group before treatment were performed by using two independent-samples t-test; (3)ADC paired T test in group before and after treatment were performed, (4)the ADC value comparisons of effective group and non-effective group were performed by using two independent-samples t-test. (5)The predicted treating efficacy valid ADC value changing rate threshold was obtained by using ROC before and after treatment. If the area under the ROC curve was over 50%, it was considered as significant, variation with p<0.05 was considered as statistically significant. When the area under the ROC curve was over 50% and P<0.05, diagnosis cut-off point of ADC value changing rate was calculated.ResultThe study included 39 patients with lung adenocarcinoma were treated with first line chemoradiotherapy.1. There is no statistical significant difference of mean ADC under b500 and 1000 value between mutated group and wild-type group before chemotherapy.2. There is no statistical significant difference of mean ADC under b500 and 1000 value between responding group and non-responding group before chemotherapy.3. Responding and non-responding group in b500 and b1000, the ADC values change rate difference was statistically significant (b500t= 2.791,p=0.011;b1000 t =2.286, p=0.028)4. The results of our analysis indicated that the slopes of the fitted regression lines, which indicated symmetry, justified use of the log odds ratio as a summary statistical measure for comparison of test performance. For assessment of all the cases without catalog EGFR mutated or wild-type, the area under the curve (AUC,0.768; 95% confidence interval:6.20%,9.17%, p<0.05) with b500, the area under the curve (AUC, 0.711; 95% confidence interval:5.44%,8.77%,p<0.05) with b1000.By using a cutoff value of 11.21% changing rate of b500 predict the responding, ADC had a sensitivity of 60%, specificity of 89.5%; a cutoff value of 8.57% changing rate of b1000, ADC had a sensitivity of 75%, specificity of 73.7%;5. For EGFR mutated group, the area under the curve (AUC,0.645; 95% confidence interval:0.405,0.884, p=0.250) with b500, the area under the curve (AUC, 0.661; 95% confidence interval:0.428,0.894, p=0.200) with b1000. there was no statistical significance for the AUC.For the EGFR wild-type group, the area under the curve (AUC,0.958; 95% confidence interval:0.873,1.00, p<0.05) with b500, the area under the curve (AUC, 0.806; 95% confidence interval:0.582,1.000,p<0.05) with b1000.By using a cutoff value of 9.71% changing rate of b500 predict the responding, ADC had a sensitivity of 88.9%, specificity of 87.5%; a cutoff value of 8.96% changing rate of b1000, ADC had a sensitivity of 77.8%, specificity of 87.5%。Conclusion:ADC is a promising biomarker for predicting chemotherapy responses for the egfr wild-type lung adenocarcinoma, but there still more proof and data for the egfr mutated adenocarcinoma.