摘要
目的 探讨MR扩散加权成像（DWI）早期预测胃肠间质瘤（GIST）靶向治疗疗效的价值.方法 前瞻性连续入组29例经病理证实,至少有1个长径＞1 cm的靶病灶,口服甲磺酸伊马替尼单药治疗的不可切除GIST患者.患者在治疗前、治疗后2周及治疗后3个月接受MRI检查,并对靶病灶进行连续监测.测量病灶整体ADC值（ADCentire值）、DWI显著高信号区域ADC值（ADCmin值）和T2WI肿瘤长径（LD）,计算治疗后2周LD值变化率（％ΔLD）、ADCmin值变化率（％ΔADCmin）及ADCentire值变化率（％ΔADCentire）.根据疗效综合评判标准将病灶分为治疗有效组和反应不良组.两组间各量化值及其早期变化率的比较采用t检验或Mann-Whitney U检验;以疗效判断综合标准为金标准,采用ROC曲线判断各参数值预测疗效的效能.结果 29例患者共82个病灶纳入观察,治疗有效组病灶57个,反应不良组病灶25个.治疗有效组治疗前基线ADCmin值和ADCentire值均低于反应不良组[ADCmin值分别为（0.90±0.27）×10-3和（1.15±0.29）× 10-3 mm2/s,t=3.673,P＜0.01;ADCentire值分别为（1.06±0.26）×10-3和（1.31 ±0.32）×10-3mm2/s,t=3.698,P＜0.01];两组病灶治疗前长径差异无统计学意义（P＞0.05）.GIST靶向治疗2周后,治疗有效组和反应不良组％ΔADCmin、％ΔADCentire和％ΔLD差异均有统计学意义（％ ΔADCmin分别为48％和3％,Z=-5.310,P＜0.01;％ΔADCentire分别为45％和2％,Z=-5.768,P＜0.01;％ΔLD分别为-6％和-1％,Z=-2.752,P＜0.01）.治疗前ADCmn值、ADCentire值及LD值预测疗效的曲线下面积分别为0.735、0.721和0.503,治疗2周后％ΔADCmin值、％ΔADCentire值及％ΔLD预测疗效的曲线下面积分别为0.859、0.877和0.694.以％ΔADCmin值≥13％为截点预测治疗有效,敏感度和特异度分别为0.842和0.760;以％ΔADCentire值≥22％为截点预测治疗有效,敏感度和特异度分别为0.862和0.800.结论 GIST靶向治疗后2周ADC值变化率对疗效的预测效能优于长径变化率和治疗前基线ADC值,反映肿瘤整体特性的ADCentire值预测效能优于仅反映肿瘤高活性部分的ADCmin值.
Objective To investigate the performance of DWI in the prediction of the response of gastrointestinal stromal tumor （GIST） to targeted therapy.Methods Twenty nine consecutive patients with unresectable GIST confirmed by pathology were enrolled in this prospective study.All patients were treated by imatinib mesylate single drug,and had at least one targeted lesion with longest diameter＞ 1 cm.MRI examinations were performed prior to and 2 weeks and 3 months afollowing treatment.The ADC of the whole lesion （ADCentire） and obviously high signal area on DWI （ADCmin） and the longest diameter （LD） on T2WIs were measured.The early percentage changes of the quantitative indicators （％ Δ LD,％ Δ ADCmin and ％ ΔADCentire） were calculated.All patients were divided into good response （GoodR） and poor response （PoorR） groups according to the combined response assessment criteria.The pre-therapy values and early percentage changes （％Δ） of the above quantitative parameters were compared using t test or Mann-Whitney U test.Receiver operating characteristic （ROC） curves were employed to compare the performance of the above quantitative parameters in the discrimination of response groups.Results A total of 82 lesions in 29 patients were included in the study.There were 57 lesions of GoodR and 25 lesions of PoorR.The pre-therapy baseline ADCmin and ADC in GoodR group were both lower than PoorR group [ADCmin：（0.90± 0.27） × 10-3 mm2/s vs.（1.15± 0.29） × 10-3 mm2/s,t=3.673,P＜0.01 ;ADCentire：（1.06 ± 0.26） × 10-3mm2/s vs.（1.31 ± 0.32） × 103mm2/s,t=3.698,P＜0.01].No statistical difference in baseline LD was displayed （P＞0.05）.The ％ ΔADC ％ΔADC and ％ΔLD all had statistical difference between GoodR and PoorR groups（％ΔADCmin：48％ vs.3％,Z=-5.310,P＜0.01;％ΔADCentire：45％ vs.2％,Z=-5.768,P＜0.01;％ΔLD：-6％ vs.-1％,Z=-2.752,P＜0.01）.Through ROC curves,the AUCs for pre-therapy ADCmin,ADCentire and LD were 0.735,0.721 and 0.503,respectively; and AUCs for ％ΔADCmin,％ΔADCentire and ％ΔLD after two weeks of therapy were 0.859,0.877 and 0.694,respectively,in the differentiation of responses.Using ％ ΔADCmin ≥ 13％ to predict GoodR,the sensitivity was 0.842 and specificity 0.760; Using ％ΔADCentire ≥ 22％ to predict GoodR,the sensitivity was 0.862 and specificity 0.800.Conclusions The percentage changes of the ADC after two weeks of therapy outperform the longest diameter and pre-therapy baseline ADC in the response predictionof GIST to targeted therapy.It is more effective of ADCentire than ADCmin values in the respo nse prediction.
出处
《中华放射学杂志》
CAS
CSCD
北大核心
2015年第3期186-190,共5页
Chinese Journal of Radiology
基金
国家自然科学基金（81201215,81371715）
北京市自然科学基金（7132039）
关键词
胃肠间质瘤
磁共振成像
靶向治疗
Gastrointestinal stromal tumor
Magnetic resonance imaging
Targeted therapy
