Study Of MRI Diffusion-weighted Imaging In Differential Diagnosis And Treatment Of Lymphadenopathy

1. BackgroundLymph node metastasis is a most common malignant tumor migration path, through the imaging evaluation of the nature of lymph node lesions is conducive to the development and prognosis of treatment options. MRI technology, with its excellent soft tissue resolution, can identify the quality of lymph nodes diseases more accurately than CT and ultrasound. But the signal intensity on Tl-weighted imaging and T2-weighted imaging of benign and malignant lymph nodes have some overlaps in conventional MRI, and can not determine the nature of the lymph nodes based on them. Up to now, conventional imaging techniques play important roles in assessment of the lymph node status, which rely on lymph node morphology such as shape, size, necrosis and extracapsular spread. The shortest axis of malignant lesions is larger than 1cm and the shape is similarly round. Conversely, the shortest axis of benign lesions is smaller than 1 cm and the shape is oval, in which the ratio of large and short axis is approximately 2. Necrosis of malignant lymph nodes is more common, and often shows extracapsular spread. However, many studies have found that these criteria can not completely distinguish benign from malignant lymph nodes, especially for small lymph node lesions, in which may already exist micro-metastases, meanwhile morphological diagnosis becomes lagging.With the development of radiological imaging technology, the functional imaging technologies are used to reveal the functional changes in different way. MRI diffusion-weighted imaging can reveal micro-movement of water molecules in vivo, and is used to detecte early physiological and functional changes related to tissue water content. A few researches are mainly concentrated in cervical lymph nodes. Razek et al. found that diffusion-weighted imaging can found lesions with the smallest diameter of 0.9cm, while the malignant nodules ADC value are lower than benign nodules. Sumi etc al. found that ADC values of cervical lymph node metastasis were significantly higher than benign lymph node lesions. Although DWI can be more sensitive and more accurate to diagnosis, the results between different studies had some differences. DWI imaging is a potent technology with short examination time, non-invasive, no radiation, and no injection of contrast agents, so a comprehensive and systematic evaluation of DWI in the diagnosis and treatment of benign and malignant lymph nodes plays an important role in its clinical practice.2. Objectives1. To evaluate the value of ADC values and relative ADC values in the differentiation of benign and malignant lymph nodes in animal models with MRI diffusion-weighted. imaging, and estimate its clinical application value by comparison with PET imaging.2. To analysis of the value of diffusion-weighted imaging and PET in discrimination of benign and malignant lymph nodes by compared of the ADC values, relative values, early SUV and delayed SUV.3. To evaluate the value of DWI in assessment of radiation effect by comparison with pre-and pro-radiation ADC values of lymph nodes, combined with pathological results.4. To collect ADC values of metastatic lymph node in patients before and after irradiation, and verify the value of DWI in assessment of radiation effect.3. Material and methodsPart I Value of MRI diffusion-weighted imaging in the differential diagnosis of benign and malignant lymph nodesTwo groups of animal models, including metastatic lymph node and inflammatory lymph nodes, were established. Some morphological indicators of lymph nodes were measured in conventional MRI, including shortest diameter, border (smooth, lobulated or indistinct), and necrosis. ADC values of lymph nodes also were measured in MRI DWI. The contralateral quadriceps femoris was selected as the reference site for rADC values. rADC was calculated by ADCiesion/ADCreference site.The benefit of rADC over ADC was validated for identifying metastatic lymph nodes. SUVs of benign and malignant lymph nodes of PET were measured. By comparing the values between ADC values and SUVs in identifying the nature of lymph nodes, the clinical value of DWI was evaluated.PartⅡStudy of different imaging methods in discrimination of benign and malignant lymph nodesThe animal models of metastatic lymph node and inflammatory lymph nodes were established respectively. MRI diffusion-weighted imaging and dual-phase PET examination were both made. All ADC values, rADC values and the value of early and delayed SUV of benign and malignant lymph nodes were measured. ROC curve analysis was made to evaluate the ability of four sets of data to identify benign and malignant lymph nodes.PartⅢValue of DWI in evaluation of metastatic lymph nodes response to radiation therapyThe animal models of metastatic lymph nodes were established. Then they were divided into control group and therapy group. And the therapy group received radiotherapy. MRI and DWI were performed respectively pretreatment,1,3,7 days after radiotherapy. The volumns of lymph nodes and ADC values were measured. The difference between the two groups was compared. In 1 day and 3days, one rabbit were sacrificed every day. The remaining rabbits in 7 days all were killed. Correlation between pathological changes and ADC values of lymph nodes after radiotherapy was explored.PartⅣMR diffusion-weighted imaging for evaluation of therapeutic responds in neck metastastic lymph nodesBy follow-up of nasopharyngeal carcinoma, according to the size of lymph nodes, all patients were divided into an effective group and an ineffective group. All patients were examined with MRI and diffusion imaging. The ADC values of cervical metastastic lymph nodes were measured before treatment and the 7th and 14th days after treatment. Repeated measures analysis was made to evaluate ADC values at different time points, and to explore the possibility of diffusion-weighted imaging in predicting therapeutic effect. 4. ResultsPart I The mean shortest diameter of 16 metastatic lymph nodes was 8.96±1.67mm, and of inflammatory lymph nodes was 7.23±0.66mm, that there is no statistically significant difference (P= 0.005). Two groups of lymph nodes were clear boundary, and iso-intensity on T1WI and slight hyperintensity on T2WI. Necrosis was observed in eight metastatic lymph nodes, in which one is an entirely cysti. There was no necrosis in the ten inflammatory lymph nodes. All lymph nodes were hyperintensity in DWIs, hypointensity in ADC maps; necrotic areas in DWI showed low signal, hyperintensity in ADC maps. Under the conditions of the low b values (0～500s/mm2), the ADC and rADC value between benign and malignant lymph nodes had no significant statistical difference; under conditions of high b values (0～1000s/mm2), the ADC and rADC value between benign and malignant lymph nodes had statistically difference. ROC curve analysis of benign and malignant lymph nodes was made. The area under the ROC curve (AUC) of ADC value was 0.817. When the threshold was 0.88×10-3mm2/s, the sensitivity was 86.67%and the specificity was 80%. The AUC of rADC value was 0.973. When the threshold value was 0.640, the sensitivity was 93.33%and the specificity was 90%.Part II The boundaries of all lymph nodes were clear. They are iso-intensity on T1WI and slightly hyperintensity on T2WI. Necrosis was seen in three metastatic lymph nodes. The signals of all inflammatory lymph nodes were homogenous. Parechymal parts of all lymph nodes revealed hyperintensity on DWI, and hypointensity on ADC maps. Necrotic areas showed hypointensity on DWI and hyperintensity on ADC maps. The ADC and rADC values were significant differences between two groups. SUVearly value and SUVdelay of malignant lymph nodes were significantly greater than those of benign nodes, and there was statistical difference. ROC curve analysis was done, and only ADC values and rADC values revealed statistical difference.PartⅢBefore treatment, the sizes of lymph node were from 56.12-103.22mm2. Lymph nodes were iso-intenstiy on TiWI and hyperintensity on T2WI. Necrosis was seen in ten lymph nodes, in which eight was belong to therapy group and two lay in control group. In 7 days after treatment, signals of four lymph nodes begun uneven. In control group, signals had no significant change. There were no significant differences between the two groups at each time point before and after radiotherapy in lymph node size. There were significant difference of ADC values between the third and seventh day after treatment and pretreatment and once day after treatment. Repeated measurements indicated significantly difference between the two groups. In the third day after radiotherapy, ADC values of the treatment group were significantly higher than before and the first day after treatment. The control group at all time points before and after radiotherapy revealed no significant difference in ADC values.PartIV Seventy-five lymph nodes were found in 24 patients before radiotherapy. After 4 weeks of the end of radiotherapy,51 lymph nodes in effective group completely disappeared, seven lymph nodes were partial response and seven lymph nodes were stable. A total of 10 lymph nodes were no response. Metastastic lymph nodes in 12 cases located in left necks,6 cases located in right necks, and 6 cases located in bilateral necks. All lymph nodes were iso-intensity on T1WI and hyperintensity on T2WI. After enhancement, all lymph nodes can be enhanced. The 14th days after treatment, within seven lymph nodes necrotic foci can be seen. Two groups of lymph nodes showed high signals on DWI,and low signals on ADC maps. There were significant difference between effective group and ineffective group before and after radiotherapy in lymph node sizes. There was no significant difference of ADC values between the effective group and ineffective group before (t=0.582, P=0.562), the 7th and 14th days after treatment(t=-2.219, P=0.30; t=-2.597, P=0.011). Effective group at each time point ADC values gradually increased, there were obvious differences between them; in ineffective group, ADC values in the 14th day after treatment revealed significant difference between before treatment and the 7th day after treatment (P=0.004, P=0.013).5. Conclusion1. DWI with high b value can better distinguish between benign and malignant lymph nodes than those with a low b values. The ADC values and rADC values of malignant lymph nodes were significantly higher than benign lymph nodes. rADC values can better distinguish between benign and malignant lymph nodes than the ADC value.2. DWI can effectively distinguish between benign and malignant lymph nodes, and is similar to PET imaging. rADC values are better than ADC values in diagnostic capabilities. DWI is an effective method and has important clinical value.3. DWI can reflect response indirectly of foci early after radiotherapy, and help to monitor radiation efficacy of lymph nodes.4. After treatment in nasopharyngeal carcinoma metastatic cervical lymph nodes, diffusion imaging can reflect differences between effective and ineffective group at different time points of early response. And it is useful to monitor the treatment responds.