Experiment Study To Differentiate Inflammatory And Metastatic Lymph Nodes In Axillary Fossa With Diffusion Weighted Imaging And CT Perfusion Imaging

Objective To establish rabbit models of axillary inflammatory hyperplastic and metastatic lymph nodes with VX2 implanted breast carcinoma. All the lymph nodes performed magnetic resonance imaging(MRI), diffusion-weighted imaging (DWI) and CT perfusion imaging(CTPI) examination.To analysis the distinction of relative apparent diffusion coefficient (rADC) value which was the quantizing index of DWI, to compare the difference of time-density curve(TDC) and CT perfusion parameters including blood flow(BF), blood volume(BV), mean transit time(MTT), and permeability surface(PS) after CTPI. On the basis of pathology, the correlation between rADC value and the cell density was assessed. The correlation of BF, BV with expression of vascular endothelial growth factor(VEGF) also was tried to assesse. The diagnostic value of DWI and CTPI differentiating axillary benign and malignant lymph nodes was explored.Materials and Methods1. Forty female new Zealand white rabbits were divided randomly into 2 equal groups, inflammatory hyperplastic group and tumor-bearing group, body weight 2.0-2.5 Kg. In our study,0.5ml egg yolk emul was injected into the lateral upper quadrant of the rabbit's second mammaris respectively so as to set up inflammatory hyperplastic model, while the tumor-bearing model was injected 0.5ml VX2 tissue mass suspension into the same location of mammary as the inflammatory hyperplastic model. All the successful models were performed MRI scanning. The other 14 female new Zealand white rabbits were also divided randomly into 2 equal groups, the way to establish model of whose was the same as the former. All the successful models accepted CTPI examination.2. Of the 40 rabbits,19 inflammatory hyperplastic models and 20 tumor-bearing models were established successfully respectively. MRI scanning were performed with bilateral axillary fossa using GE 3.0T HDx superconducting magnet,8-channel head and neck joint coil. After obtaining the routine localization images, the axial T1WI, T2WI and coronal T2WI were performed, respectively. Then after the ASSET CAT checking sequence scanning, axial DWI scanning was performed, using periodically rotated overlapping parallel lines enhanced reconstruction (PROPELLER) fast spin-echo(FSE) sequence, with two different b values(0 and 800s/mm2), the diffusion sensitive gradient was applied to X, Y and Z directions at the same time. After DWI examination performed, ADC maps were obtained by postprocessing. Signal intensities of DWI, the values of ADC with b value of 800s/mm2 were measured in axillary inflammatory hyperplastic lymph nodes, tumor-bearing lymph nodes and dorsal muscle at the same slice of lymph nodes, respectively, then the corresponding relative signal intensities(SLN/M) and values of rADC were calculated. The difference between the rADC values of the two groups was statistically analyzed.Using receiver operating characteristic curve(ROC) analysis the diagnostic value to differentiate benign and malignant lymph nodes, and definite the diagnostic threshold. The maximum longitudinal diameter(L), maximum transverse diameter(S), the ratios of maximum longitudinal to maximum transverse diameter(L/S), signal intensities and corresponding SLN/M of the lymph nodes were measured and compared on the selected axial or the coronal T2WI. Serial sections of the entire lymph nodes in the axillary fossa of all rabbits were obtained according to the direction of MR scanning rightly after MR examination, subjected to hematoxylin-eosin (HE) staining and measured cell density. The correlation between the value of rADC and cell density was analyzed.3. The other 14 rabbits which were all established successfully (7 inflammatory hyperplastic models and 7 tumor-bearing models) underwent CTPI. CTPI was performed with axial shuttle scanning mode,using a 64-slice multi-detector CT scanner(Lightspeed VCT XT),double tube and dual pathways of CT dedicated high pressure syringe(Nemoto),table position at the level of the bilateral axillary fossa.A preliminary non-contrast 5-mm-thick CT of the axillary region (from the collar to the diaphragmatic muscle cacumen) was performed to guarantee a panoramic view of the entire anatomical region and thus to help the recognition of anatomical landmarks so as to overlap the bilateral axillary fossa and the bilateral second mammary when performing CTPI.4ml non-ionic iodinated contrast medium (Omnipaque,350mgl/L) was injected by CT automatic high pressure syringe at the rate of 0.8ml/s via an ear vein, with delay time 5s, scanning time 0.4s, time gap 2.4s,16 slices one time scanning, scanning phase 18, total scanning time 55.4s.576 reconstructing images of 2.5-mm slices were acquired. Images and datas obtained were transferred to an image processing workstation (Advantage Windows 4.4, General Electric Company). Commercially available software (CT perfusion4; General Electric Company) was used for perfusion CTPI analysis, using service pack body tumor software package with undeconvolution method. The TDC was recorded and the perfusion parameters including BF, BV, MTT, PS were measured automatically. There are three types of TDC:quickly rise and fall type(typeⅠ), flat type(typeⅡ) and slowly rise type(typeⅢ). The types of TDC between inflammatory and metastatic lymph nodes were compared.The differences between BF, BV, MTT and PS value of the two group lymph nodes were statistically analyzed. Serial sections of the entire lymph nodes in the axillary fossa of all rabbits were obtained according to the direction of CT scanning rightly after CTPI examination, subjected to hematoxylin-eosin(HE) staining and immunohistochemical staining to detect expression degree of VEGF. The correlation between the expression degree of VEGF and BF, BV was analyzed.Results1. Inflammatory hyperplastic and tumor-bearing lymph nodes appeared iso-signal intensity on T1WI and high signal intensity on T2WI, compared with adjacent muscle. On the DWI, all the lymph nodes were homogeneous high signal intensity. The differences of L,S,L/S and SLN/M between the two groups were not statistically significant (p>0.05).2. With b value of 800s/mm2, on the condition of using dorsal muscle at the same slice of lymph nodes as the reference organ, the rADC value of tumor-bearing lymph nodes was inferior to inflammatory hyperplastic lymph nodes, with statistically significance (t=3.879, p=0.03), but there was overlap between the rADC value of two groups. The diagnostic efficacy of rADC value to differentiate inflammatory hyperplastic and tumor-bearing lymph nodes was evaluated through using ROC. When choosing 1.67 as diagnostic threshold of rADC value, the Youden index was maximum, and its sensibility and specificity was 86.2% and 74.4%, respectively.3. The correlation between rADC value and cell density of inflammatory hyperplastic and tumor-bearing lymph nodes was inverse, whoes correlation coefficient was -0.5323.4. The BF value of tumor-bearing lymph nodes was distinctly superior to inflammatory hyperplastic lymph nodes,150±70 ml.min-1100g-1 and 75±40 ml.min-1100g-1 respectively, the difference was statistically significant; the BV and PS value was a little superior to inflammatory hyperplastic lymph nodes, 9.00±5.12 ml.100g-1 and 7.96±3.52 ml.100g-1,25.42±9.26 ml.minn-1100g-1 and 21.63±8.12 ml.min-1100g-1, respectively, the difference was not statistically significant; the MTT value was inferior to inflammatory hyperplastic lymph nodes,5.31±3.69s and 6.78±2.16s respectively, the difference was statistically significant.5. The TDC of all the tumor-bearing lymph nodes presented as typeⅠ. The TDC of inflammatory hyperplastic lymph nodes presented as diversity, TypeⅢ,ⅡandⅠwere 11,2 and 1, respectively.6. The correlation between the expression degree of VEGF and BF, BV value of inflammatory hyperplastic and tumor-bearing lymph nodes was direct correlation, and BF value presented much more stronger correlation with the expression degree of VEGF. The correlation coefficient was statistically significant (r=0.821,p= 0.000; r=0.519,p=0.001).Conclusions1. Routine MRI examination is limited to differentiate benign and malignant lymph nodes through morphology criteria, including size, shape, margin.2. rADC value acquired from DWI can help to differentiate benign and malignant lymph nodes with high differential diagnostic efficacy, but it can't be perfectly considered as the absolutely differential criteria because of the partial overlap between the both.3. The correlation between rADC value and cell density of inflammatory hyperplastic and tumor-bearing lymph nodes is inverse.Cell density may influence the rADC value of the tumor.DWI is expected to a new safe, non-invasive and effective method of evaluating the density of tumor in vivo.4. The BF value of tumor-bearing lymph nodes is superior to inflammatory hyperplastic lymph nodes, and the difference is statistically significant.The TDC of all the tumor-bearing lymph nodes present as typeⅠ,and the TDC of a inflammatory hyperplastic lymph nodes primarily present as typeⅢ. CTPI gives information on the level of blood capillary of lymph nodes reflecting the status of blood perfusion, helping to differentiate benign and malignant lymph nodes.5. The BF, BV value of benign and malignant lymph nodes are related to expression of VEGF. CTPI is expected to provide a new technique for understanding vascularization in vivo.