目的 探讨局部进展期乳腺癌NAC前后MRI增强测量病变大小变化对于疗效评估的作用。方法 回顾性分析127例病理确 诊的局部进展期乳腺癌，接受4-6个周期NAC。分别于NAC前、NAC 2周期后、4周期后行动态增强扫描。术前临床反应 性评价按照RECIST分级。术后病理反应性评价按照Miller & Payne改良病理反应性分级比较术前临床反应性与术后 病理反应性的符合度。结果 以Miller & Payne病理反应性分组为金标准，其中MHR组69例，NMHR组58例。NAC 2、4 周期后RECIST标准判断术前化疗反应性与术后病理反应性分组差异具有显著性(P＜0.05)。ROC分析显示NAC 2、4周 期后最大径变化率曲线下面积(Az)分别为0.66、0.73。NAC 2、4周期后最大径变化率最佳诊断截断点分别为42.68% 及58.90%。结论 NAC 2周期后癌灶大小可早期预测局部进展期乳腺癌病理反应性，以4周期后癌灶最大径评估效能较 高，以NAC 4周期肿瘤最大径变化率大于58.90%作为判断治疗有效的标准可提高诊断准确性。

Objective To study the value of the maximum diameter of tumor in evaluating the therapeutic effect of neoadjuvant chemotherapy for locally advanced breast cancer (LABC). Methods 127 women with LABC underwent dynamic CEMRI examination before chemotherapy, after the 2nd cycle and the 4th cycle of chemotherapy. All patients were divided into major histological response group (MHR) and non-major histological response group (NMHR) according to the final pathologic response. The response to NAC before surgery evaluated by the variance of tumor size was compared to pathologic response after surgery. ROC curve analysis was performed to assess the effect of the maximum diameter and the variance of tumor size (⊿D %) in predicting the pathologic response to the NAC treatment. Results 127 cases of LABC patients were divided into a MHR group of 68 cases and a NMHR group of 59 cases. The response to NAC after 2nd and the 4th cycle of NAC before surgery evaluated by RECIST criterion was significant difference to pathologic response(P<0.05) after surgery. Az (area under curve) of The maximum diameter variance rate of the tumor after 2nd and 4th cycle were 0.66 and 0.73 respectively. The best diagnose cut off were variance rate of 42.68% and 58.90% respectively. Conclusion The maximum diameter variance rate of the tumor after 2nd cycles of NAC may be useful in predicting the final pathological response of breast cancer to NAC treatment, but which after the 4th cycle would be more efficiency. Take the variance rate greater than 58.90% as diagnostic criteria to definite the response level can improve the diagnose accuracy.