Diffusion-weighted MR Imaging: Evaluation Of Lung Cancer

【Background】Primary lung carcinoma is the leading cause of death among cancers worldwide.Early detection and accurate tumor staging is essential for choosing the appropriate surgical treatment and strategy for lung cancer, which decides the survival rate in the great degree. Although computed tomography (CT) is considered as "golden standard", and is widely used in the preoperative evaluation of lung cancer by morphology and haemodynamics characteristics, the results of many studies have shown that CT has some limitation in the display of inner ingredient and evaluation of histopathology characteristic.As the highly thought of radiation dose of CT in the international society, a new non-radiative and non-invasive method is demanded to evaluate and monitor lung cancer.With the improvement and development of magnetic resonance imaging (MRI) techniques, several disadvantages of MR imaging of pulmonary parenchyma have been conquered, which has proceeded the potential for investigation of pulmonary disease. Recently, more and more studies performed have been providing proof to investigate the role of functional MR imaging (fMRI) in the evaluation of different orgens. Experience in using diffusion-weighted imaging (DWI) to abdomen speeds up the procedure to lung.Diffusion MR imaging plays important roles in the detection, diagnosis, staging, and follow-up evaluation of tumors.It is well known that diffusion is caused by random translational molecular motion, also known as brownian water motion. DWI is the only imaging method that can be used to evaluate the diffusion process in vivo. In our study, multi-b value diffusion-weighted MRI is used in lung cancer imaging and has its unique advantages. Diffusion-weighted imaging is a noninvasive technique, which is used for the characterization, diagnosis, and the grading and staging of lung cancer, as well as prediction of the treatment response of patients with lung cancer to chemotherapy.It can provide a valuable tool in the evaluation of tumor qualitatively and quantitatively. Meanwhile,the ADCs has the potential to be associated with tumor prognosis factors and reflect tumor molecular biology features. Part One Multi-b Value Diffusion-weighted MR Imaging:Differentiation Diagnosis of Pulmonary Disease【Objective】To prospectively investigate the feasibility of multi-b value diffusion-weighted imaging (DWI) for pulmonary diaease. To evaluate the application value of ADCs and signal intensity in diagnosis and differentiation diagnosis of lung carcinoma.【Materials and Methods】Conventional MRI and DWI (b=0,300,600,900,1200s/mm2) were performed in 95 patients with pulmonary lesions all conformed by pathology and clinical diagnosis. Signal intensity of the lesions, thoracic spinal cord and thoracic muscle at the same image were measured and compared in DWI. ADCs were calculated by formula to compare the statistic difference between malignant and benign lesions, histopathologic types, different differentiation degree and diameters of lung cancer. With a 5-point rank scale,a further comparation of sensitivity,specificity and accuracy for differention diagnosis of pulmonary lesions was evaluated.【Results】1. ADCs of all pulmonary lesions decreased by the increasing of b value.①ADCs of the malignant and benign:b=300 s/mm2, there was significant difference between them (P=0.0153<0.05), there were no significant differences at other b values (P>0.05).②ADCs of different histopathologic types:b=300 s/mm2, there were significant differences between adenocarcinoma and inflammation, squamous cell carcinoma and SCLC. There were no significant differences as b value increases between them and at other groups.③analysis of different differentiation degree:There were notable statistical significance between well and poorly differentiated lung cancer, well and poorly differentiated adenocarcinoma.④analysis of different diameters:There were no statistical significance between groups according to different sizes of lung cancer and adenocarcinoma. 2.The AUC (area under the ROC curve) value of muscle (0.738)was higher than that of thoracic spinal cord(0.701). With the thoracic muscle as a reference,the mean score of malignant pulmonary nodules (4.34±0.97) was significantly higher than that of benign nodules (3.25±1.37) (P=0.0036). As a score of 3 was considered as a threshold, the sensitivity, specificity and accuracy were 81.58%,45.00%, and 75.86%, respectively. [Conclusion] Preliminary results revealed both ADCs and signal intensity for DWI are helpful to differentiate pulmonary malignant lesions from benigns. DWI could be a potential method for differential diagnosis of lung carcinoma.Part Two Multi-b Value Diffusion-weighted MR Imaging:to Evaluate Diagnosis Potency of Lung Cancer and Optimize b Value【Objective】To investigate the diagnosis potency of multi-b value diffusion-weighted MRI in discrimination of pulmonary malignant lesions from benigns. To optimize b value of multi-b value DWI.【Materials and Methods】Conventional MRI and DWI (b=0,300,600,900,1200s/mm) were performed in 95 patients with pulmonary lesions all conformed by pathology and clinical diagnosis. Signal intensity of the lesions were measured in DWI. ADCs was calculated by formula. ROC curves were analyzed to compare the diagnosis ability at differentiation of malignant and benign lesions. AUC (Area Under ROC) value, sensitivity, specificity and accuracy for differential diagnosis of pulmonary lesions were evaluated.【Results】1. ROC analysis for the malignant and benign:There was significant difference between quadruplet overall (P=0.0278 <0.05).AUCs were aquired to 0.699,0.642,0.615,0.559 (b=300,600,900,1200 s/mm2). As 1.25×10-3 mm2/s was considered as a threshold (b=300s/mm2),the sensitivity, specificity and accuracy were 51.4%, 87.0% and 62.0%, respectively.2. ROC analysis for the adenocarcinoma and benign:There was significant difference between quadruplet overall (P=0.0394<0.05). AUCs were aquired to 0.717,0.644,0.628,0.570 (b=300,600, 900,1200s/mm2). As 1.31×10-3mm2/s was considered as a threshold (b=300s/mm2), the sensitivity, specificity and accuracy were 55.5%,82.7%,64.7%, respectively.3. ROC for the adenocarcinoma and benign with different sizes:With 2 cm (centimeters) as a boundary, there was significant difference between quadruplet overall (P=0.0018<0.05) AUCs were 0.749,0.622,0.609,0.545 (b=300,600,900,1200s/mm2).Opposite result was acquired under 2 cm (P=0.5233>0.05).AUCs were 0.708,0.664,0.648,0.617 (b=300, 600,900,1200s/mm2).The adenocarcinoma with diameter more than 2 cm had a highest AUC value of 0.749 (b=300s/mm2). AUCs of all groups decreased as b value increased.【Conclusion】Multi-b value diffusion-weighted MRI has moderate diagnosis potency to differentiate pulmonary malignant lesions from benigns. When b was 300s/mm2,which had the most potential value for differential diagnosis of pulmonary carcinoma. Part Three Correlation Between ADC Value and Tumorous Cellular Density and Prognosis Factor in NSCLC【Objective】To evaluate the pathologic and prognostic characteristics in DWI, to explore the correlation between ADCs and the tumorous cellular density and prognosis factor in lung cancer.【Materials and Methods】Four-six patients of lung carcinoma with surgery conformed by histopathology were performed in DWI. ADCs were calculated at multiple b values images. Histologic specimens were stained with hematoxylin and eosin. Tumor cellularity was analysed in Axioplan 2 imaging optical microscopy. Correlations between the ADCs and tumor cellularity, nuclear-cytoplasmic ratio, nuclear-kytoplasm ratio and prognosis factor were analysed by Spearman rank correlation.【Results】The ADCs of lung cancer correlated negatively with tumorous cellular density at four b values and as b was 300 s/mm2,which can achieve the best result (b=300 s/mm2, r=-0.792, p<.0001). There were inverse correlations between ADC values and nuclear-cytoplasmic ratio, nuclear-kytoplasm ratio(r=-0.739, p<.0001;-0.607, p=0.0187). As b value was 300 s/mm2, the ADCs of lung cancer correlated negatively with MF (multiplicaiton factor) Ki67(r=-0.565, p<.005). There was no significant statistical correlation in other MFs.【Conclusion】There were inverse correlations between ADCs and tumorous cellular density and proliferation index Ki67. DWI may become a new method for distinguishing different histopathology differentiation and grading of lung cancer. DWI has the potential to evaluate lung cancer behaviour and improve prognosis and survival.