The Value To Early Predict And Evaluate The Efficacy Of Breast Cancer Neoadjuvant Chemotherapy With 3.0T MRI

PartⅠThe Value to Evaluate the Lesion after Breast Cancer Neoadjuvant Chemotherapy and Predict the Efficacy with the change of Lesion's Maximum Diameter on 3.0T MRIObjective:To investigate the accuracy of evaluating the residual lesion's number and scope after NAC with 3.0T MRI; To investigate the ability that the changes in the maximum diameter between before NAC and post-first (or second) cycle treatment predict the efficacy of chemotherapy; To approach the correlation between efficacy of NAC and the change rate of the lesion's maximum diameter from before NAC to post-all cycle chemotherapyMaterials and Methods:90 patients pathology confirmed invasive breast cancer with core needle puncture biopsy undergone at least one time breast MR scans at before NAC, post-first cycle, post-second cycle.or after NAC.The scan series includes:routine plain scan, Diffusion weighted imaging(DWI) and Volume imaging of breast enhancemant(VIBRANT)。Referring to the fist editon MR Breast imaging reports and data systcm(BI-RADS), the enhancement modality of breast cancer in MRI is defined four categories:single masse enhancement. mutiple mass enhancement, non-mass enhancement, Mixing mass and non-mass enhancement. The single mass enhancement lesion is defined single focus, the rest categories lesions defined multi-focus. The maximum diameter of lesion is measured in VIBARANT early enhanced images. The pathology response is evaluated by methods of Miller & Payne(MP)system, MP1-3 are Non-major histological response(NMHR) group, MP4～5 are Major histological response(MHR) group. SPSS 13.0 software is used. The residual lesion's maximum diameter mesured by MRI and pathology are compared with Pearson correlation analysis; The lesion is single focus or multi-focus by MRI and patology compared with chi square test; The correlation between MP results and the change rates of lesion's maximum diameter is analyzed with analysis of variance. Receiver operating characteristic curve(ROC)is used to find the optimized value of lesion'predicting the MHR. P<0.05 considered statistical significant difference.Results:90 patients, including 3 cases of bilateral breast cancer, a total of 93 sides. MR show:56 sides of breast cancer as single focus.37 sides as multi-focus; pathology show: 78 sides as single focus,15 sides as multi-focus(P<0.001, Kappa=0.451). The MRI measured value of lesion's maximum diameters after NAC is(2.79±2.44)cm, pathology measured value is (1.22±0.89)cm, Pearson's correlation coefficient of both is 0.740(P<0.001).Conclusion:The sensitivity of multi-foci residula diagnosis with 3.0T MRI is 100%,the specificity of single focus diagnosis is 100%.3.0TMRI can accurately measure the residual lesion's maximum diamerer of post NAC. The change rate of lesion's maximum diamerer of post NAC can actively predict the efficacy of chemotherapy. the opimized predicting value is 42.29%, with sensitivity 91.7%, specificity 75%。 Part IIThe Value to Predict Efficacy of Breast Cancer NAC with Lesion's Apparent Diffusion CoefficientObjective:To investigate the predictive value of chemotherapy efficacy with the leisions'ADC and change of ADC at before NAC,during NAC(post-first or second cycle) and after NAC, using different diffusion sensitivity coefficient b(600 and 1000) and different methods to measure ADC value.Materials and Methods:90 patients pathology confirmed invasive breast cancer with core needle puncture biopsy undergone at least one time breast MR scans at before NAC, post-first cycle, post-second cycle, or after NAC. The scan series includes:routine plain scan, Diffusion weighted imaging(DWl) and Volume imaging of breast enhancemant(VIBRANT). Especially, twice DWI sacn are performed, respectively usingb=600 and b=1000. All Scan data is transmited toGE ADW4.4 workstation, DWI images postprocessing was completed using Functool. Referring to plain and contrast enhanced images. determine the lesion's position of breast cancer. using the lower ADC value methods and hand lesion ADC methods to measure ADC value of lesion. Using SPSS13.0 software. When b is certain, the lesion's ADC value with two measure methods were under Peareson's correlation analysis. when measure methods is certain, lesion's ADC value measured under b=600 and b=1000 was analyzed by mean compare. The correlation between efficacy of chemothrapy and ADC value (or change of ADC value) at NAC different time point was analyzed using analysis of variance. ROC is used to find the ADC optimal cutoff value of lesions predicting the MHR. P<0.05 considered statistical significant difference.Results:under b=600 and 1000, The pearson's correlation coefficient of the ADC value of lesion before NAC and after NAC was respectively 0.707,0.922 and 0.614,0.930 using two different ADC measure methods (all having P<0.001); using lower ADC value methods and hand ADC methods, The pearson's correlation coefficient of the ADC value of lesion before NAC and after NAC was respectively 0.675,0.885 and 0.919,0.915 under b=600 and 1000. MHR group, NMHR group's lesions ADC lower value and hand ADC value is repectively(1.741±0.226)mm2/s, (1.349±0.373) mm2/s, P=0.003 and (1.751±0.231)mm2/s, (1.456±0.308) mm2/s,P=0.008; b=600, the Area Under ROC of lesion's ADC value and MP evaluation at before NAC and after NAC is repectively 0.841 and 1.515mm2/s, the optimal cutoff value is 1.515 and 1.575mm2/s, sensitivity both are 88.9%, specificity both are 75.5%; b=1000, the Area Under ROC both are 0.775, the optimal cutoff value is 1.17 mm2/s and 1.265 mm2/s, sensitivity both are 88.9%, specificity is 65.4% and 71.2%.Conclusion:1. For the lesion before and after NAC, when b value is certain, The correlation coefficient is high(after NAC is higher) of the ADC value using two measure methods.2. For the lesion before and after NAC, when ADC measure methods is certain, The correlation coefficient is high(hand ADC methods is higher) of the ADC value under b=600 and 1000(b=1000, ADC value is lwer).3. For post NAC lesion, the ADC value in MHR group is higher than in NMHR group.4. For post NAC lesion, the ADC value measured by both lower ADC method and hand ADC method can actively predict MHR, specificity is higher when b=600.5. For the lesion, relative to the lesion at before NAC, the change of ADC at post-first cycle and after NAC is greater in MHR than NMHR. but not statistically significant. Part IIIThe Value to Predict the Efficacy of Breast Cancer NAC with Parameters of 3.0T Dynamic Contrast Enhanced Magnetic Resonance ImagingObjective:To study the correlation between the MP results and dynamic parameters, such as the time signal intensity curve (TIC), the signal enhancement value at second phase (SI2%), the maximum enhancement value of TIC(SI peak%); the maximum linear slope equivalent(Smax) at before NAC, post-first or second treatment and after NACMaterials and methods:90 patients pathology confirmed invasive breast cancer with core needle puncture biopsy undergone at least one time breast MR scans at before NAC. post-first cycle, post-second cycle, or after NAC.The scan series includes:routine plain scan, DWI and VIBRANT dynamic contast enhanced scan. All Scan data is transmited to GE ADW4.4 workstation. Referring to plain and contrast enhanced images, TIC was drawn in the VIBRANT early contast enhanced Images using Functool. Lesion's Region of ROI (ROI) was selected by two methods(the Maximum enhanced ROI methods and hand-drawing ROI methods). ROI>10 Voxel, parameters were recorded. Referring to Kuhl standard. TIC is divided into three categoryies. Using SPSS 13.0 perform data analysis. The correlation between MP efficacy evaluation and the TIC categories of lesion at before NAC and after NAC was analyzed using chi square test; the mean of parameters and change of parameters, such as SI2%, SI peak% and Smax of MHR group and NMHR group at before and after NAC was compared by analysis of variance. P<0.05 considered significant difference.Results:For lesions after NAC, using the Maximum enhanced ROI methods and hand-drawing ROI methods, the SI2% mean of MHR(14), NMHR(55) are respectively (50.450±25.379)%, (158.713±102.949)%(P<0.001) and (3.971±20.326)%, (103.854±61.868)%(P<0.001); The area under ROC of SI2% and MP categories are 0.840 and 0.873(both P<0.001), MHR is positive, the optimist SI2% cutoff value are 83% and 39.9%, sensitivity 70.9% and 87.3%, specificity 100% and 78.6%; The SIpeak% mean was repectively (126.057±46.036)%, (217.156±85.643)%(P<0.001) and (105.907±8.725)%, (194.238±122.248)(P=0.01); The area under ROC of SIpeak% and MP categories are 0.583 and 0.515 (P=0.353 and 0.874), MHR is positive, the optimist SIpeak% cutoff value are 229.4% and 192.5%, sensitivity are 71.4% and 61.5%, specificity are both 52.8%; the Smax mean of MHR(14), NMHR(55) are respectively 275.158±145.716,983.817±551.842(P< 0.001) and 7.022±123.108,653.001±360.746(P<0.001); The area under ROC of Smax and MP categories are 0.919 and 0.908(both P<0.001), MHR is positive, the optimist Smax cutoff value are 404.995%/s and 275.45%/s, sensitivity are 87.3% and 85.5%, specificity are 85.7% and 78.6%.Conclusion:Using two different ROI selected methods, for the lesion of breast cancer after NAC, the SI2%, SIpeak% and Smax in MHR are all lower than NMHR, and the three parameters all can activly predict the MHR. the Smax is the best one. For the lesion before NAC, only Smax can actively predict the MHR. for the lesion at post-first or second cycle treament, neither all the parameters nor their changes can not predict MHR. just there are a tendency. Part IVThe Predictive Value of NAC Efficacy with 3.0T MRSObjective:To investigate the MRS achievement ratio and the frequency of choline In the lesion of breast cancer before NAC, during NAC(post-first or second cycle) and after NAC(all cycles are over); To investigate the predictive value of NAC efficacy With choline/water hump height ratio and the area under hump ratio.Materials and Methods:90 patients pathology confirmed invasive breast cancer with core needle puncture biopsy undergone at least one time breast MR scans at before NAC. post-first cycle, post-second cycle, or after NAC. The scan series includes:routine plain scan. DWI and VIBRANT dynamic contast enhanced scan.23(of 90) patients voluntary accepted the braest MRS scan. All MRS scans were perfomed after VIBRANT 2-12 hours. Referring to the plain and contast enhanced iamges. The horizontal axis T2WI. VIBRANT enhanced sagittal image before the mask film as a positioning image, determining ROI. Using single voxel scan sequence. Scan time about 4.8 min. Scan direction is horizontal surface. The routine automatic pre-scan, including automatic shimming and water suppression, were performed before MRS data collection. line width of the study is 6 to 13, inhibition of the signal level of water reached 92% to 98%. After MRS scan, the data were postprocessing using SAGE software developed by GE company in ADW 4.2 workstation. Obtaining Spectral line and various metabolites at different position in Spectral line. The normal breast tissues's Spectral line including fat peak at 1-2 ppm and water peak (water) at 4.7ppm; The lesion's Spectral line Including choline peak at 3.2ppm (Cho) and water peak (water) at 4.7ppm. the water peak as a reference, choline/water hump height ratio and the area under hump ratio were caculated. Using SPSS13.0 statistical software for data entry and statistical analysis. CalculatedMRS achievement ratio and the frequency of choline; the correlation between MP categories and choline/water hump height ratio and the area under hump ratio was analyzed by one-factor analysis of variance. P<0.05 considered statistically different Significance.Results:For normal breast tissues and lesions before NAC, post-first cycle treatment, after NAC, the MRS achievement ratio are all 100%(13/13,5/5,7/7)and the frequency of choline are respectively 85%(11/13),60%(3/5),28.6%(2/7). For lesion before NAC, choline/water hump height ratio and the area under hump ratio in MHR group(4 sides) and NMHR(9sides) are respectively 0.113±0.080,0.236±0.283 (P=0.420) and 0.061±0.069,1.467±3.862(P=0.492). Conclusion:the MRS achievement ratio are all 100% at three check point and the frequency of choline decrease in turn. For the lesion, Neither choline/water hump height ratio nor the area under hump ratio can not actively predict the MP efficacy statisticly, just there are the tendengcy that choline/water hump height ratio and the area under hump ratio is higher in NMHR.