Clinical Imaging Study On The Regression Pattern Of Breast Cancer After Neoadjuvant Therapy

The first partCorrelation between regression patterns and clinicopathological features of breast cancer after neoadjuvant therapyObjective:To explore the correlation between clinical and pathological features and regression pattern of breast cancer after neoadjuvant chemotherapy.Methods:A total of 710 patients with breast cancer who received neoadjuvant chemotherapy(NAC)in Yunnan cancer hospital from August 2015 to June 2019 were retrospectively collected.According to the inclusion and exclusion criteria,215 patients were included in this study.According to the morphological changes of the lesions after NAC,they were divided into concentric shrinkage mode(CS)and non concentric shrinkage mode(NCS)groups.Rank sum test,chi square test or Fisher exact probability method were used to analyze the distribution differences of clinical and pathological features between the two groups,Then binary logistic regression was used to screen out independent variables related to withdrawal pattern.Intraclass correlation coefficient(ICC)was used to evaluate the consistency between the two physicians in measuring the size of residual lesions and judging the withdrawal pattern.Results:A total of 215 patients with invasive ductal carcinoma were enrolled in this study.There were 134 patients with concentric regression and 81 patients with noncentric regression after NAC.Univariate analysis showed that there were significant differences between the two groups in molecular classification,ER expression status,PR expression status,HER-2 expression status,pre NAC tumor T stage(CT),post NAC tumor T stage(YPT)and chemotherapy regimens(P<0.05).Multivariate logistic regression analysis showed that molecular typing,ER and PR expression status,and CT were independent factors influencing the regression pattern of breast cancer after NAC(P<0.05).Among them,the high expression of ER and PR,luminal a,CT3 and cT4 were risk factors of centripetal withdrawal,while negative or low expression of ER and PR,luminal B(HER-2 positive),HER-2 overexpression,TNBC,CT1 and CT2 were protective factors.Conclusions:The regression pattern of breast cancer after NAC is related to a variety of clinical and pathological features.Tumors with different molecular types,hormone receptor status and T stages may show different regression patterns after NAC.The second partCorrelation between regression pattern and residual tumor size after neoadjuvant therapy for breast cancerObjective:To explore the correlation between regression pattern and residual tumor size after neoadjuvant chemotherapy for breast cancer.Methods:A total of 710 patients with breast cancer who underwent NAC in Yunnan cancer hospital from August 2015 to June 2019 were retrospectively collected.According to the inclusion and exclusion criteria,215 patients were finally selected and included in this part of the study.0 cm and 0 cm respectively±25%was used as the judgment threshold of consistency to evaluate the consistency and absolute consistency between MRI measurement and pathological measurement.Then,the regression pattern,conventional MRI features and clinicopathological features of the lesions after NAC were included to explore the factors influencing the accuracy of MRI evaluation.Finally,the differences of the distribution of the characteristics between the overestimated group and the underestimated group were analyzed.Statistical tests included t test,rank sum test,chi square test,person correlation analysis,binary logistic regression and ICC test.Results:The sensitivity,specificity,positive predictive value and negative predictive value of MRI were 94.90%,53.10%,81.87%,75.76%and 83.93%,respectively.Pearson correlation analysis showed that the residual tumor size measured by MRI was highly positively correlated with pathological size(r=0.874,95%CI=0.817-0.920,P<0.001).The largest diameter of pathological tumor was the largest ± 25%as the cut-off point of consistency,the overall consistency rate of MRI and pathology was 73.49%.The inconsistency rate in NCS group was higher than that in CS group(45.68%vs 14.93%,P<0.05).Univariate analysis showed that there were significant differences in regression pattern,molecular classification,ER expression status,HER-2 expression status,histological grade,CT,residual tumor diameter,accompanying signs and chemotherapy regimen between the two groups(P<0.05).Regression pattern,molecular classification,histological grade and residual tumor diameter were independent factors affecting the accuracy of MRI measurement.The lesions with NCS,luminal a,histological grade Ⅲ and residual tumor diameter>5cm are easy to be underestimated,while the lesions with CS,luminal B(HER-2 negative),histological grade Ⅱ and residual tumor diameter<5cm are easy to be overestimated.Conclusions:Breast MRI is an accurate tool for evaluating residual invasive carcinoma after NAC,and there is a high consistency between MRI measurement and pathological measurement.The accuracy of MRI in measuring the lesions with concentric regression is higher than that of non concentric regression.The third partCorrelation between regression pattern and recurrence of breast cancer after neoadjuvant therapyObjective:To explore the correlation between regression pattern and recurrence of breast cancer after neoadjuvant chemotherapy.Methods:A total of 710 patients with breast cancer who underwent NAC in Yunnan cancer hospital from August 2015 to June 2019 were retrospectively collected.According to the inclusion and exclusion criteria,107 patients were selected to be included in this part of the study.All patients underwent MRI before and after NAC.All patients were divided into relapse group and no relapse group according to whether relapse or not.Univariate Cox regression analysis was used to analyze the distribution differences of clinical,pathological,regression mode and MRI features between the two groups.Then multivariate Cox regression analysis was used to screen out independent variables related to recurrence.Then all independent variables were combined into a multi parameter prediction model.ROC area was used to evaluate the discrimination between independent variables and prediction model;The calibration curve is used to evaluate the calibration degree of the model;The prediction model is represented by nomogram.Results:A total of 107 patients with invasive ductal carcinoma were enrolled in this study.There were 30 cases(28.04%)with NAC recurrence and 77 cases(71.96%)without recurrence.The median recurrence time was 21 months and the relapse free survival time was 42.5 months.The 3-year overall survival rate was 93.46%and the relapse free survival rate was 71.96%.Univariate analysis showed that the withdrawal pattern,molecular typing,ER expression status,and the expression of Er after NAC△%SER、△%ADC,histological grade,menopausal status,CT,lymph node status after NAC,peritumoral edema,RCR were associated with disease recurrence(P<0.05).Multivariate analysis showed that NCS,HR negative,no RCR,er negative/low expression,histological grade Ⅲ,peritumoral edema before NAC and positive lymph nodes after NAC were risk factors for recurrence,while CS,HR positive,obtaining RCR,er high expression,histological grade I,no peritumoral edema and negative lymph nodes after NAC were protective factors.Combined with the above independent variables,the combined forecasting model is established and expressed by nomogram.When the follow-up time was 36 months,the predictive ability of the model was good(C=0.897),the sensitivity and specificity were 90.84%and 95.67%respectively.Conclusions:The regression pattern of breast cancer after NAC is an imaging marker for predicting recurrence.The multi parameter model combined with clinical and pathological indexes can be used as an auxiliary tool for clinicians to evaluate the prognosis.