分享:
分享到微信朋友圈
X
综述
影像学评估局部进展期结直肠癌新辅助治疗后肿瘤退缩分级的研究进展
卢婷 王媛媛 周凤瑜 董文洁 杨海婷 周俊林

Cite this article as LU T, WANG Y Y, ZHOU F Y, et al. A review on imaging in evaluating tumor regression grade after neoadjuvant theatment for locally advanced colorectal cancer[J]. Chin J Magn Reson Imaging, 2024, 15(5): 209-215, 221.本文引用格式卢婷, 王媛媛, 周凤瑜, 等. 影像学评估局部进展期结直肠癌新辅助治疗后肿瘤退缩分级的研究进展[J]. 磁共振成像, 2024, 15(5): 209-215, 221. DOI:10.12015/issn.1674-8034.2024.05.034.


[摘要] 结直肠癌是常见的消化系统恶性肿瘤之一,约60%~70%的结直肠癌患者在确诊时已为局部进展期直肠癌(locally advanced rectal cancer, LARC),新辅助治疗(neoadjuvant treatment, NAT)是LARC患者的标准治疗方案,近年来,随着NAT在LARC中展现出良好的治疗效果,NAT在局部进展期结肠癌也得到了初步临床应用,但并非所有局部进展期结直肠癌(locally advanced colorectal cancer, LACRC)患者均能从NAT中获益。因此,术前精准评估NAT疗效,筛选NAT优势人群是一个亟待解决的临床问题。诸多证据表明影像学可在术前评估LACRC-NAT疗效。因此,本文就计算机体层成像(computed tomography, CT)、MRI、正电子发射计算机体层成像(positron emission tomography/computed tomography, PET/CT)、影像组学及深度学习在LACRC-NAT疗效评估中的应用现状、优势及局限性进行综述,旨在提升影像学评估LACRC-NAT疗效的准确性,为临床制订个体化的NAT方案提供全面、客观的影像学依据。
[Abstract] Colorectal cancer is one of the common malignancies of the digestive system, about 60%-70% of colorectal cancer patients have locally advanced rectal cancer (LARC) at the time of diagnosis, neoadjuvant therapy (NAT) is the standard treatment for patients with LARC, in recent years, with NAT showing good treatment response in LARC, NAT has also been initially used in locally advanced colorectal cancer, but not all patients with locally advanced colorectal cancer (LACRC) can benefit from NAT. Consequently, it is an urgent clinical problem to accurately evaluate the treatment response of NAT and screen the NAT dominant population before surgery. There is abundant evidence that imaging can be used to assess the response of LACRC-NAT preoperatively. Therefore, this article reviews the application status, advantages and limitations of computed tomography, MRI, positron emission tomography/computed tomography, radiomics and deep learning in the evaluation of the response of LACRC-NAT, aiming to improve the accuracy of imaging evaluation of the response of LACRC-NAT and provide a comprehensive and objective imaging basis for the clinical formulation of individualized NAT protocols.
[关键词] 结直肠癌;新辅助治疗;疗效评估;计算机体层成像;磁共振成像;人工智能
[Keywords] colorectal cancer;neoadjuvant treatment;treatment response assessment;computed tomography;magnetic resonance imaging;artificial intelligence

卢婷 1, 2, 3, 4   王媛媛 1, 2, 3, 4   周凤瑜 1, 2, 3, 4   董文洁 1, 2, 3, 4   杨海婷 1, 2, 3, 4   周俊林 1, 2, 3, 4*  

1 兰州大学第二医院放射科,兰州 730030

2 兰州大学第二临床医学院,兰州 730000

3 甘肃省医学影像重点实验室,兰州 730030

4 医学影像人工智能甘肃省国际科技合作基地,兰州 730030

通信作者:周俊林,E-mail:lzuzjl601@163.com

作者贡献声明::周俊林设计本研究的方案,对稿件重要内容进行了修改,获得了兰州大学第二医院“萃英科技创新”计划项目基金资助;卢婷起草和撰写稿件,获取、分析和解释本研究的数据;王媛媛、周凤瑜、董文洁、杨海婷获取、分析或解释本研究的数据,对稿件重要内容进行了修改;全体作者都同意发表最后的修改稿,同意对本研究的所有方面负责,确保本研究的准确性和诚信。


基金项目: 兰州大学第二医院“萃英科技创新”计划项目 CY2021-ZD-01
收稿日期:2024-01-17
接受日期:2024-03-21
中图分类号:R445.2  R735.3 
文献标识码:A
DOI: 10.12015/issn.1674-8034.2024.05.034
本文引用格式卢婷, 王媛媛, 周凤瑜, 等. 影像学评估局部进展期结直肠癌新辅助治疗后肿瘤退缩分级的研究进展[J]. 磁共振成像, 2024, 15(5): 209-215, 221. DOI:10.12015/issn.1674-8034.2024.05.034.

0 引言

       结直肠癌(colorectal cancer, CRC)是全球第三常见恶性肿瘤[1],约60%~70%的CRC患者为局部进展期直肠癌(locally advanced rectal cancer, LARC)[2],即临床分期为T3~T4期和(或)有淋巴结转移的直肠癌,其标准治疗方案为新辅助治疗(neoadjuvant treatment, NAT)联合全肠系膜切除术辅以术后化疗[3]。近期多项临床试验表明全程新辅助治疗(total neoadjuvant treatment, TNT)[4]和新辅助免疫治疗[5]等均能显著提高LARC的局部控制率与病理完全缓解(pathological complete response, pCR)率,并减少手术及放化疗所致并发症的发生[6]。FOxTROT Ⅲ期临床试验[7]也证实对于有望行手术切除的结肠癌患者,可通过NAT诱导肿瘤退缩和降期,从而降低肿瘤负荷与术中肿瘤细胞脱落风险,提高肿瘤完全切除率。然而,并非所有LACRC患者均能从NAT中获益,部分局部进展期结直肠癌(locally advanced colorectal cancer, LACRC)患者可在NAT后出现毒副作用、肿瘤进展、肛门功能、性功能及排尿功能损伤等不良事件。此外,由于计算机体层成像(computed tomography, CT)评价结肠癌分期的准确性不佳,使得局部进展期结肠癌(locally advanced colon cancer, LACC)患者面临着过度治疗与延迟手术导致疾病进展的风险[8, 9, 10]。因此,迫切需要一种可靠的方法评估LACRC-NAT疗效。

       目前用于评估NAT疗效的方法是肿瘤降期和肿瘤退缩分级(tumor regression grade, TRG)评价系统[11]。肿瘤降期是通过比较NAT期间临床分期的变化来得出,由于影像学或内镜检查得出的临床分期易受医师主观评价与临床经验的影响,导致肿瘤降期评价的准确性与一致性较差,因而在临床上应用受限。此外,由美国癌症联合委员会(American joint committee on cancer, AJCC)依据NAT后残余肿瘤与纤维化比例所制订的TRG系统在临床上较为常用,其中TRG 0为pCR,即原发病灶内无肿瘤细胞残留,而TRG 1~3为非病理完全缓解(non-pathological complete response, npCR)[12],然而,TRG仅能在术后进行评价,存在一定的滞后性,不能指导NAT方案的及时调整。影像学作为一种无创的NAT疗效评估方法,可以实现LACRC患者NAT疗效的早期及动态评估。目前国内外的研究多基于MRI及MRI影像组学评估LARC患者NAT疗效[13, 14, 15],而对LACC患者NAT疗效评估与其他影像学方法在NAT疗效评估中的研究较为少见。因此,本文通过综合分析CT、MRI、影像组学与深度学习(deep learning, DL)在术前评估NAT疗效的研究,旨在提高影像学评估NAT疗效的准确性,为临床优化NAT方案提供客观的影像学依据。

1 CT评估LACRC-NAT后TRG

       CT已广泛应用于CRC患者的诊断和随访,其中CT增强扫描在评估CRC局部复发、淋巴结转移、远处转移、CRC原发灶及转移瘤NAT或转化治疗中应用广泛[3]。既往研究表明CT图像特征,如肿瘤体积、增强模式和异常血管改变等,可能与肿瘤的生物学特性相关[16],在食管癌NAT疗效评估中得到了初步应用[17, 18]。JE等[19]研究发现CT-TRG分级和肿瘤体积变化与LACC-NAT后的病理TRG之间存在相关性。LUO等[20]发现基于增强CT的细胞外体积分数(extracellular volume fraction, ECV)可以预测LARC患者NAT后的pCR,其中NAT后pCR组的ECV(17.05%±2.36% vs. 29.94%±1.20%;P<0.001)和NAT前后ECV变化值(-17.01%±3.01% vs. 0.44%±1.45%;P<0.001)显著低于npCR组。袁文静等[21]通过分析52例LARC患者行NAT前后的能谱CT参数,发现NAT前动脉期标准化碘浓度、静脉期能谱曲线斜率,NAT后静脉期标准化碘浓度和静脉期能谱曲线斜率等多个参数在治疗反应良好组(TRG 0~1)与反应不良组(TRG 2~3)患者间存在显著差异(P均<0.05),其预测NAT反应的AUC为0.655~0.807,联合NAT后静脉期标准化碘浓度及NAT前、后动脉期与静脉期标准化碘浓度变化率预测NAT反应的AUC为0.869。此外,淋巴结受累和直肠系膜筋膜侵犯程度也是LARC患者NAT反应的影响因素,DILEK等[22]通过计算88例LARC患者经NAT后的肠系膜周围脂肪组织体积(mesorectal fat tissue volume, MRV),发现无治疗反应患者组MRV低于有治疗反应患者组MRV [(69.6±31.0)mL vs.(105.8±47.5)mL;P<0.05],MRV预测NAT反应的AUC为0.757。

       部分LACRC患者会在NAT期间出现毒副反应,例如食欲不振、恶心和呕吐等,这往往会导致患者出现体重减轻和肌肉消耗,其中骨骼肌的质量减少和功能减退被定义为肌肉减少症[23]。近年来,诸多研究表明肌肉减少症与CRC患者预后不良及化疗耐受程度差相关[24, 25]。为了结合CT在定量机体骨骼肌含量及密度方面的优势,即通过测得第3腰椎水平的腰大肌面积除以患者身高的平方得出骨骼肌指数,再将其与诊断阈值进行比较判断患者是否患有肌肉减少症。基于此,OLMEZ等[26]研究发现非肌肉减少症组LARC患者NAT后出现pCR的比例明显高于肌肉减少症组(21.4% vs. 3.0%;P=0.025)。BEDRIKOVETSKI等[27]在一项多中心研究中得出了不同的结论,即肌肉减少症不是LARC患者NAT疗效不佳的预测指标,而在BEDRIKOVETSKI等[28]另一项分析118例经TNT的LARC患者时发现肌肉减少症是LARC患者达到临床完全缓解的独立危险因素,造成上述研究结论不一致的原因可能与判断患者是否患有肌肉减少症所选择的诊断阈值不一致及术前NAT方案不同相关。总之,基于CT分析LACRC患者肌肉特征的易受患者性别、种族等多种因素的影响,而且基于CT分析肌肉特征的方法和肌肉减少症的诊断阈值目前尚不一致,有望未来深入研究肌肉特征与NAT后TRG之间的潜在联系,并制订CT分析肌肉特征的评价标准和诊断阈值,为LACRC-NAT疗效评估提供新的研究方向。

       综上,CT是CRC患者诊断、疗效评估的常规影像学检查方法,也是结肠癌影像学分期的首选方法,随着LACC-NAT的开展,CT在LACC-NAT疗效评估中具有广泛的应用前景。而且对于存在MRI禁忌证不能进行MRI检查的LARC患者,可以使用CT的增强模式、异常血管改变、ECV、能谱CT参数、MRV与基于CT的肌肉特征等评估NAT疗效。但目前基于CT评估NAT疗效多为小样本研究,且常规CT评估NAT疗效的参数较为单一,主要集中于肿瘤及瘤周的密度和形态学改变,容易受到感兴趣区(region of interest, ROI)选择、部分容积效应与医师主观评价的影响,特别是对于NAT后出现pCR的患者,ROI的勾画尤为困难。此外,CT的软组织分辨率不及MRI,因而在识别NAT病变早期细微结构改变的敏感度不佳。随着能谱CT与灌注CT多参数在消化道肿瘤中NAT疗效评估的应用,未来有望开展能谱CT联合灌注CT、光子计数CT评估LACRC-NAT疗效的前瞻性、多中心研究,从而全面认识CT在LACRC-NAT疗效评估中的独特价值。

2 MRI评估LARC-NAT后TRG

       MRI是监测LARC复发和转移的首选影像学方法,近期一项Meta分析发现MRI裂痕征[29]预测LARC-pCR的AUC为0.83。2011年MERCURY研究组[30]建立了一个类似于病理TRG的MRI-TRG系统,然而,其与病理TRG之间的一致性较差(加权Kappa系数为0.240)[31]。为了改良MRI-TRG系统,PANG等[32]通过在高空间分辨率T2WI和扩散加权成像(diffusion-weighted imaging, DWI)上量化LARC-NAT后肿瘤残余与肿瘤纤维化的比例,建立了一种新MRI-TRG系统,在1033例LARC患者的TRG评价中,新MRI-TRG系统与术后病理AJCC TRG系统的一致性良好(Kendall's tau-b相关性系数为0.671)。

       常规MRI不易鉴别NAT导致的肿瘤纤维化、细胞水肿与肿瘤残余,而DWI与扩散峰度成像(diffusion kurtosis imaging, DKI)以水分子扩散为基础,可以反映组织内部的微环境。OUYANG等[33]通过分析71例行TNT后LARC患者的MRI参数,发现巩固化疗后肿瘤平均表观扩散系数(apparent diffusion coefficient, ADC)和DWI上的肿瘤最大横截面积是LARC患者TNT后完全缓解(包括pCR和临床完全缓解)的最佳预测因子。BORASCHI等[34]通过分析41例LARC患者NAT前后的多b值DWI,得出LARC患者行NAT前、后ADC值及其变化值(ΔADC),发现NAT后ADC及ΔADC与病理TRG存在相关性。LIAN等[35]分析了63例LARC患者NAT前的定量合成MRI,与BORASCHI等的结论一致,NAT前ADC无法预测NAT反应,而NAT前T1弛豫时间和T2弛豫时间值能够区分pCR与npCR,显示了定量MRI参数在预测NAT反应中的潜在优势。然而IAFRATE等[36]在36例经TNT的LARC患者中得出了不同的结论,pCR和npCR组间NAT前的ADC存在显著性差异,即NAT前ADC可以区分pCR和npCR。总之,NAT前、后的ADC及ΔADC可以用于评估NAT疗效,但是其在上述研究中的诊断性能并不一致,原因可能是MRI设备、扫描协议不同致使测量的ADC存在差异,因此未来有望提出标准化的MRI扫描方案,进一步提升MRI评价LARC-NAT后TRG的准确性。

       传统DWI基于水分子的高斯分布,而真实的生物组织中水分子扩散为非高斯分布,水分子的扩散受周围环境的限制程度越大,扩散的非高斯性越显著,在DWI基础上衍生的DKI可以在一次扫描中获得通过平均扩散系数和平均峰度系数等指标评估水分子的扩散程度和量化扩散差异。YANG等[37]通过分析42例LARC患者NAT前后体素内不相干运动参数(真扩散系数、假扩散系数和灌注分数)、DKI参数(平均扩散系数和平均峰度)和ADC,发现pCR组NAT后ADC、假扩散系数、灌注分数、平均扩散系数均明显高于npCR组,其中NAT后平均扩散系数区分pCR和npCR的AUC最高,AUC为0.788。动态对比增强MRI(dynamic contrast enhanced MRI, DCE-MRI)通过显示不同区域肿瘤成分的强化特点,反映肿瘤内部血供分布特点和真实的肿瘤负荷状态。FUSCO等[38]发现基于DCE-MRI的标准化形状指数可以用于评估NAT后TRG,其中标准化形状指数识别NAT反应不良组的灵敏度、特异度及准确度分别为95.90%,84.70%和91.80%。CIOLINA等[39]发现NAT前的速率常数和容积转运常数可用于预测LARC患者NAT反应,说明DCE-MRI可以通过量化细胞密度和血流特征以评估NAT后TRG,但目前关于DCE-MRI半定量分析的研究较少。

       综上,高空间分辨率T2WI能够清晰显示直肠肠壁层次、肿瘤浸润范围、直肠系膜筋膜及毗邻结构侵犯状态,功能MRI序列及DCE-MRI可在形态学变化出现之前反映出由NAT所致的病变局部功能及代谢改变,在LARC-NAT评估具有重要的价值。然而,LARC-NAT后病变局部可产生多种病理学改变(肿瘤细胞坏死、炎症细胞浸润、纤维组织增生及黏液池形成)等,因而在鉴别DWI、DCE-MRI异常信号区域为组织炎症反应、黏蛋白成分还是肿瘤残余方面存在困难,且使用MRI评估NAT后TRG易受医师主观性及图像质量的影响,难以实现NAT后TRG的精准定量评估,有望未来进行大样本、多中心的临床研究验证MRI参数在评估NAT疗效的价值,为LARC患者NAT疗效评估提供更为可靠的影像学支持。此外,随着研究者们对MRI征象的深入解读,基于MRI评价的肿瘤沉积、壁外血管侵犯和环周切缘累积等与LARC患者的预后密切相关,有望未来探索上述征象联合基于MRI评估的TRG在LARC患者预后预测的价值,为LARC患者诊疗策略的选择及术后随访方案的制订提供指导。

3 正电子发射计算机体层成像评估LARC-NAT后TRG

       18F脱氧葡萄糖正电子发射断层扫描(18F-fluoro-deoxy-glucose positron emission tomography, 18F-FDG PET)是临床上常用的代谢显像方式。正电子发射计算机体层成像(positron emission tomography/computed tomography, PET/CT)是将PET和CT结合起来的一种显像设备,不同于CT的解剖结构成像,在一次PET/CT显像中不仅能获得病灶的解剖结构与功能代谢信息,也能实现患者全身的解剖和代谢断层显像(有助于诊断患者是否存在远处转移灶,并显示远处转移灶的位置与代谢状态),在肿瘤疗效评估中具有灵敏度高与定位精准度高的优势。标准化摄取值(standardized uptake value, SUV)是最常用的PET/CT半定量参数,SCHURINK等[40]通过分析基线水平的MRI和PET/CT参数在LARC患者NAT疗效中的应用时,发现结合MRI定量成像特征联合MRI-T分期的模型能够识别对NAT表现出良好反应的患者(Mandard TRG 1~2),而在加入PET/CT定量参数后模型性能未得到提升,提示NAT前PET/CT参数评估TRG的价值有限。然而,CERNY等[41]发现NAT后SUV平均值和SUV最大值与Mandard TRG独立相关(P<0.001),说明SUV可作为量化LARC中残留肿瘤负荷和纤维化的指标。SAKIN等[42]通过分析51例行NAT前LARC患者的肿瘤代谢体积(metabolic tumor volume, MTV)后发现,当MTV的临界值为12时,MTV能够区分治疗反应良好组(Ryan TRG 1~2)与治疗反应不良组(Ryan TRG 3~5)。ŞEN等[43]发现LARC患者行NAT前的MTV和淋巴结转移状态与LARC患者NAT反应相关,联合MTV和淋巴结转移状态构建的模型在训练集与测试集中的AUC分别为0.714和0.838。LOVINFOSSE等[44]在分析66例行NAT的LARC患者RAS基因突变状态、MTV、总病变糖酵解(total lesion glycolysis, TLG)和纹理特征与Dworak TRG之间的关系时,发现只有RAS基因突变状态(OR=0.22,P=0.021)和TLG(OR=0.12,P=0.004)与治疗反应良好(Dworak TRG 3~4)显著相关。VUIJK等[45]也在一项前瞻性试验(在NAT前、NAT后2周和NAT后6~8周内对患者行MRI和PET/CT检查,获取各期MRI和PET/CT参数)发现NAT前MTV和TLG,NAT前到NAT后2周内最大标准化摄取值(SUVmax)的变化值与瘦体质量标准化摄取值峰值的变化值可以区分治疗反应良好组(Mandard TRG 1~2)与反应不良组(Mandard TRG 3~5)。

       综上,PET/CT不是CRC分期的常规影像学检查方法,当临床怀疑CRC患者出现远处转移(特别是肿瘤标志物持续上升)或患者行CT和MRI检查后仍无法明确病变是否为远处转移灶时,PET/CT将有助于LACRC远处转移灶的显示及病变性质分析。PET/CT对LACRC患者NAT疗效具有一定的预测价值,但目前基于PET/CT评估NAT疗效的研究较少,且研究结果一致性欠佳,使得其在LACRC患者NAT疗效评估中的应用仍存有争议,有待后续研究进一步探索PET/CT在LACRC患者NAT疗效评估中的应用潜力。

4 影像组学及影像病理联合组学评估LACRC-NAT后TRG

4.1 影像组学评估LACRC-NAT后TRG

       随着CRC治疗方案的改变以及多学科诊疗模式的普及,促进了影像学精准评价的快速发展,影像组学通过高通量技术提取并整合医学影像大数据中蕴含的海量特征,对特征进行降维和建模构建肿瘤疗效评估模型。影像组学的一阶统计学特征通过基于直方图分析法描述图像中像素的灰度分布情况,反映肿瘤内部的差异性及不均质性[46]。PHAM等[47]采用全肿瘤体素技术分别生成基于DWI和DCE-MRI的ADC和容积转运常数直方图,在异质性分析中发现NAT后第75百分位数ADC和第90百分位数ADC能够区分NAT反应。JIMÉNEZ DE LOS SANTOS等[48]通过计算48例LARC患者NAT前后的8个全病灶ADC直方图参数,发现pCR组LARC患者NAT后的峰度和偏度均显著低于npCR组,而且pCR组NAT前后峰度、偏度的变化幅度均显著高于npCR组。LI等[49]在基于DKI的直方图分析中发现除峰度和偏度外,NAT前后校正扩散系数与TRG 3(LARC患者对NAT治疗存在耐药性)之间存在相关性,其中NAT前后校正扩散系数的变化率评估LARC-NAT耐药性的诊断性能最优,AUC为0.939。SCHURINK等[50]通过分割61例LARC患者经过NAT前的MRI和PET/CT图像,计算肿瘤整体特征和肿瘤局部直方图特征并与临床参数相结合构建多变量预测模型,发现多变量预测模型区分NAT反应良好组(Mandard TRG 1~2)与反应不良组(Mandard TRG 3~5)的AUC最高,AUC=0.83。既往研究证实影像组学特征中一阶直方图特征的可重复性最高,而高阶影像组学特征的可重复性较差[51]。KURATA等[52]在一项基于LARC患者基线MRI预测术后病理TRG的多中心研究中发现基于临床信息与影像学分期的联合模型预测性能最优,其预测pCR和NAT反应良好(Mandard TRG 1~2)的AUC分别为0.60和0.65,而从T2WI和ADC中提取的一阶直方图特征并不能提升模型性能,说明影像组学特征的可重复性可能会对其在多中心研究中的应用产生负面影响。

       KURATA等[53]研究发现增强CT图像的纹理分析和肿瘤分形维数、偏度和峰度可以用于识别pCR和npCR,而且pCR组的峰度明显高于npCR组。BORDRON等[54]使用合成少数类过采样技术(synthetic minority over-sampling technique, SMOTE)调整了不平衡和异质性的数据后,发现基于T2WI、DWI以及增强CT的影像组学模型和影像组学联合临床参数的模型在SMOTE后预测pCR的准确度为68.20%和85.50%。NIE等[55]基于NAT前LARC患者的T1WI、T2WI、DCE-MRI与DWI建立MRI影像组学模型,其预测LARC患者出现NAT反应良好(改良Ryan TRG 0~1)的AUC为0.89。WEN等[56]使用NAT前后的T2WI,建立LARC-NAT前、后以及Delta影像组学TRG评估模型,并与2位影像科医师的TRG评估结果进行比较,发现NAT前、后以及Delta影像组学评估TRG的AUC分别为0.717,0.805和0.724,2位影像科医师评估TRG的AUC分别为0.606和0.621,其中以基于NAT前、后影像组学及NAT前临床T分期的联合模型AUC最高(AUC=0.852)。QIN等[57]研究证实LARC瘤内和肿瘤周围系膜血管和淋巴结影像组学特征的整合,可作为LARC患者NAT疗效评估的有效预测因子,有望通过多中心的验证,为提升NAT评估准确性提供新的影像学标志物。目前影像组学评估CRC的NAT疗效研究多集中于LARC的NAT疗效评估,由于NAT在LACC患者中的应用较少且尚存有争议[8, 9],导致了LACC患者NAT期间的影像学数据不充足,进一步限制了影像组学在LACC-NAT疗效评估中的应用,因而使用影像组学评估LACC-NAT疗效的研究较少。许汝鑫等[58]研究发现基于增强CT门静脉期影像组学模型、临床T分期及肿瘤分化程度构建的联合模型在训练集与验证集中评估NAT后LACRC患者出现pCR的AUC分别为0.850和0.824。

4.2 影像病理联合组学评估LARC-NAT后TRG

       研究发现CRC组织病理学图像特征与临床分子变异(如微卫星不稳定性和PTEN基因缺失等)之间存在相关性,说明病理组学在预测CRC患者基因表达、组织病理学模式、治疗反应及患者总生存期、无进展生存期等方面具有广阔的应用前景[59]。ZHANG等[60]在151例LARC患者经苏木精伊红(hematoxylin and eosin, H&E)染色的肠镜活检组织全视野切片图像(digital whole slide images, WSIs)中筛选出17个特征,上述特征在训练集与验证集中预测LARC患者经过NAT后pCR的AUC为0.930和0.879。为了结合病理组学在评价LARC-NAT疗效方面的优越性,研究者们提出了MRI影像病理联合组学模型,SHAO等[61]通过从术前MRI和内镜活检标本的WSI中提取图像特征,并构建影像病理联合组学模型,研究结果表明影像病理联合组学模型可以预测NAT后TRG。FENG等[62]在一项多中心研究中通过对LARC患者的盆腔多参数MRI以及经H&E染色肠镜活检组织WSIs进行注释和特征提取,建立了一个由9个影像组学MRI特征、12个病理组学细胞核特征和18个病理组学肿瘤微环境特征组成的影像病理联合组学预测模型在训练集、验证集1、验证集2及前瞻性研究队列中预测pCR的AUC为0.868、0.860、0.872与0.812。

       综上,随着人工智能领域的飞速发展,影像组学在消化系统恶性肿瘤NAT疗效预测、评估及预后预测中展现出了巨大的潜力。近年来,基于影像组学评估LACRC-NAT的研究数量在逐步增加,但是影像组学的工作流程尚未实现标准化,因而在构建影像组学模型的过程中(如特征筛选和模型验证等)仍存在较大差异,部分研究所构建的影像组学模型能否稳定地应用于临床值得商榷。首先,建立影像组学模型所采集的图像易受扫描设备、扫描协议及重建算法等多种因素的影响,进而导致所提取特征的一致性较差,而且部分研究缺乏外部验证,使得模型的可重复性及泛化性欠佳。其次,影像组学中可提取到海量特征,如何去除冗余特征、降低特征共线性与减少模型过拟合仍是当下不可避免的问题,而且影像组学特征不同于传统的影像学征象(可通过病理学变化及组织学特点解释征象含义),影像组学特征由数字变换所得,其与病理生理机制之间的联系尚不明确。最后,影像组学研究采用手动勾画病灶,不仅费时费力,还存在观察者间差异的影响,继而导致模型的再现性和稳定性下降。未来有望建立多任务DL模型,实现LACRC病灶的自动勾画及NAT疗效评估,同时结合设计严谨的多中心、大样本的临床研究和前瞻性验证,为影像组学向临床转化提供支持。

5 DL评估LARC-NAT后TRG

       近年来,基于DL可以挖掘图像中隐藏信息、避免人工标记、特征提取以及在原始图像实现肿瘤自动分割和治疗疗效预测的优势,在CRC筛查、诊断、疗效评估和预后预测等方面的得到了广泛应用。LI等[63]开发的DL工具通过使用MRI图像特征预测LARC患者pCR,其预测经NAT后pCR的准确度、特异度及敏感度分别为78.90%、72.50%和81.00%。ZHANG等[64]通过开发一种基于DKI和T2WI的DL模型预测LARC患者的NAT反应,在测试集中DL模型预测pCR的AUC为0.99。HU等[65]通过使用CE-Net以端对端的形式提取MRI和CT图像特征并自动分割图像,再结合卷积神经网络实现pCR预测的AUC为0.833,相比单独基于CT或MRI的DL模型预测pCR的性能更优。JIN等[66]通过使用基于NAT前后的纵向多参数MRI(T1WI、T2WI及DWI),构建了具有深度神经网络架构的多任务学习模型3D RP-Net,该模型由用于特征提取和肿瘤分割的卷积编码子网和用于NAT反应预测的多流连体子网组成,3D RP-Net在2个独立验证集中预测pCR的AUC分别为0.95和0.92。KE等[67]在一项多中心研究中通过对比LARC患者NAT前后的MRI(T1WI、T2WI及DWI)变化构建模型,在内部和3个外部验证集中预测pCR的AUC分别为0.969、0.946、0.943和0.919。OUYANG等[68]通过使用逻辑回归和DL方法,构建了基于T2WI的肿瘤病灶、瘤周特征及周围阳性淋巴结特征联合临床特征的LARC-TNT反应预测模型,结果显示逻辑回归模型预测pCR的AUC为0.853,DL模型的AUC为0.829。ZHOU等[69]通过使用注意力机制构建基于T2WI、增强T1WI及ADC 3个序列的DL和影像组学模型,并构建2个未使用注意力机制的DL和影像组学模型进行比较,研究发现使用注意力机制构建的DL模型预测LARC患者NAT反应的性能最优,其在内部与外部验证集中的AUC分别为0.898和0.873,而且研究还发现使用注意力机制的模型较未使用注意力机制的模型性能更好。JANG等[70]发现基于LARC患者行NAT6周后T2WI的DL模型在预测pCR和治疗反应良好组(Dworak TRG 3~4)的性能优于影像科医师的预测性能。

       综上,DL模型在预测LARC患者pCR方面具有良好的准确性,但是DL模型在LACRC患者NAT疗效评估的临床应用中存在透明度和可重复性的限制,有望未来开展大型前瞻性研究,降低DL模型过拟合的风险并提升DL模型的稳健性,再将DL模型与影像科医师的NAT疗效评估相结合,提升影像学评估LACRC患者NAT疗效的可靠性和准确性。

6 总结与展望

       随着LACRC患者NAT的重要性日益增加,术前精准评估LACRC-NAT疗效有助于临床制订个性化的治疗策略。NAT前后基于CT、MR与PET/CT的参数能够评估NAT疗效,但易受到ROI勾画、医师主观评价的影响而导致应用受限。影像组学可以提取高通量特征,但其提取特征费时费力,DL可以直接处理原始数据并自动开发模式识别信息,具有广泛的应用前景,病理学评估在LACRC患者的NAT评估中有着不可或缺的作用,除了影像学和病理组学,肿瘤微环境组成成分(如肿瘤相关成纤维细胞、肿瘤浸润T细胞和肿瘤相关巨噬细胞等)与LARC患者NAT耐药性及肿瘤退缩程度相关。因此,影像学联合病理组学或者肿瘤微环境评估LACRC-NAT疗效是未来的研究方向。此外,影像学评估NAT疗效所选择的时间节点多为NAT前和术前(最后一次NAT时),有望未来进一步探讨NAT期间其他时间节点获取的影像学指标能否用于NAT疗效评估,为LACRC患者NAT疗效评估提供更精准、全面和个体化的影像学依据。

[1]
SIEGEL R L, MILLER K D, FUCHS H E, et al. Cancer statistics, 2022[J]. CA A Cancer J Clin, 2022, 72(1): 7-33. DOI: 10.3322/caac.21708.
[2]
白中元, 赵欣烜, 徐钧. 新辅助放化疗治疗局部进展期直肠癌效果及预测指标研究进展[J]. 肿瘤研究与临床, 2023, 35(5): 397-400. DOI: 10.3760/cma.j.cn115355-20221125-00742.
BAI Z Y, ZHAO X X, XU J. Progress of therapeutic efficacy and predictive indicators of neoadjuvant chemoradiotherapy for locally advanced rectal cancer[J]. Cancer Res Clin, 2023, 35(5): 397-400. DOI: 10.3760/cma.j.cn115355-20221125-00742.
[3]
国家卫生健康委员会医政司, 中华医学会肿瘤学分会. 国家卫生健康委员会中国结直肠癌诊疗规范(2023版)[J]. 中华胃肠外科杂志, 2023, 26(6): 505-528. DOI: 10.3760/cma.j.cn441530-20230525-00182.
Medical Administration Department of the National Health Commission, Oncology Branch of the Chinese Medical Association. Chinese protocol of diagnosis and treatment of colorectal cancer of the National Health Commission (2023 edition)[J]. Chin J Gastrointest Surg, 2023, 26(6): 505-528. DOI: 10.3760/cma.j.cn441530-20230525-00182.
[4]
DONNELLY M, RYAN O K, RYAN É J, et al. Total neoadjuvant therapy versus standard neoadjuvant treatment strategies for the management of locally advanced rectal cancer: network meta-analysis of randomized clinical trials[J]. Br J Surg, 2023, 110(10): 1316-1330. DOI: 10.1093/bjs/znad177.
[5]
CHEN G, JIN Y, GUAN W L, et al. Neoadjuvant PD-1 blockade with sintilimab in mismatch-repair deficient, locally advanced rectal cancer: an open-label, single-centre phase 2 study[J]. Lancet Gastroenterol Hepatol, 2023, 8(5): 422-431. DOI: 10.1016/S2468-1253(22)00439-3.
[6]
WANG L, ZHANG X Y, ZHAO Y M, et al. Intentional watch and wait or organ preservation surgery following neoadjuvant chemoradiotherapy plus consolidation CAPEOX for MRI-defined low-risk rectal cancer: findings from a prospective phase 2 trial (PKUCH-R01 trial, NCT02860234)[J]. Ann Surg, 2023, 277(4): 647-654. DOI: 10.1097/SLA.0000000000005507.
[7]
MORTON D, SEYMOUR M, MAGILL L, et al. Preoperative chemotherapy for operable colon cancer: mature results of an international randomized controlled trial[J]. J Clin Oncol, 2023, 41(8): 1541-1552. DOI: 10.1200/JCO.22.00046.
[8]
ALISEDA D, ARREDONDO J, SÁNCHEZ-JUSTICIA C, et al. Survival and safety after neoadjuvant chemotherapy or upfront surgery for locally advanced colon cancer: meta-analysis[J/OL]. Br J Surg, 2024, 111(2): znae021 [2024-03-12]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10881053. DOI: 10.1093/bjs/znae021.
[9]
VIMALACHANDRAN D. Neoadjuvant chemotherapy for locally advanced colonic cancer is not ready to be the standard of care[J]. Br J Surg, 2023, 110(12): 1677-1678. DOI: 10.1093/bjs/znad184.
[10]
孙雪峰, 管真, 闫铄, 等. 结肠癌原发灶周围脂肪组织异常影像学表现与预后的关系[J]. 中华放射学杂志, 2023, 57(8): 904-911. DOI: 10.3760/cma.j.cn112149-20230413-00276.
SUN X F, GUAN Z, YAN S, et al. Relationship between abnormal imaging findings of adipose tissue around primary colon cancer and prognosis[J]. Chin J Radiol, 2023, 57(8): 904-911. DOI: 10.3760/cma.j.cn112149-20230413-00276.
[11]
孙应实, 张晓燕. 直肠癌新辅助治疗疗效评估: 现状与挑战[J]. 中国医学影像学杂志, 2022, 30(9): 875-880. DOI: 10.3969/j.issn.1005-5185.2022.09.002.
SUN Y S, ZHANG X Y. Evaluation of neoadjuvant therapy for rectal cancer: current situation and challenges[J]. Chin J Med Imag, 2022, 30(9): 875-880. DOI: 10.3969/j.issn.1005-5185.2022.09.002.
[12]
NICHOLLS R J, ZINICOLA R, HABOUBI N. Extramural spread of rectal cancer and the AJCC Cancer Staging Manual 8th edition, 2017[J]. Ann Oncol, 2019, 30(8): 1394-1395. DOI: 10.1093/annonc/mdz147.
[13]
李戟玭, 张文凯, 庄乐怡, 等. 基于ADC诊断局部晚期直肠癌新辅助治疗后完全缓解准确性的Meta分析[J]. 磁共振成像, 2022, 13(8): 36-42, 54. DOI: 10.12015/issn.1674-8034.2022.08.007.
LI J P, ZHANG W K, ZHUANG L /Y)Y, et al. Value of different ADC to access complete response of rectal cancer after neoadjuvant therapy: a Meta analysis[J]. Chin J Magn Reson Imag, 2022, 13(8): 36-42, 54. DOI: 10.12015/issn.1674-8034.2022.08.007.
[14]
梁程, 安小霞, 李瑞, 等. MRI影像组学预测局部进展期直肠癌新辅助放化疗疗效的研究进展[J]. 磁共振成像, 2024, 15(1): 224-228. DOI: 10.12015/issn.1674-8034.2024.01.038.
LIANG C, AN X X, LI R, et al. Research progress of radiomics in predicting the efficacy of neoadjuvant chemoradiotherapy for locally advanced rectal cancer[J]. Chin J Magn Reson Imag, 2024, 15(1): 224-228. DOI: 10.12015/issn.1674-8034.2024.01.038.
[15]
胡鸿博, 赵升, 姜昊, 等. 基于原发癌灶T2WI的影像组学特征预测局部进展期直肠癌新辅助治疗疗效及多种分类器效能比较[J]. 磁共振成像, 2023, 14(11): 77-83. DOI: 10.12015/issn.1674-8034.2023.11.013.
HU H B, ZHAO S, JIANG H, et al. Predicting the efficacy of neoadjuvant therapy for locally advanced rectal cancer based on 3.0 T MRI and comparing the effectiveness of multiple classifiers[J]. Chin J Magn Reson Imag, 2023, 14(11): 77-83. DOI: 10.12015/issn.1674-8034.2023.11.013.
[16]
CHEN W, WANG Y T, BAI G J, et al. Can lymphovascular invasion be predicted by preoperative contrast-enhanced CT in esophageal squamous cell carcinoma?[J/OL]. Technol Cancer Res Treat, 2022, 21: 15330338221111229 [2024-03-12]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9340382. DOI: 10.1177/15330338221111229.
[17]
CAO X L, LI X D, WANG X Y, et al. Use CT imaging to predict the short-term outcome of concurrent chemoradiotherapy in patients with locally advanced esophageal squamous cell carcinoma[J/OL]. Dose Response, 2019, 17(4): 1559325819897175 [2024-03-12]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6937540. DOI: 10.1177/1559325819897175.
[18]
ZHOU H Y, GUO W W, OU J, et al. A CT-based novel model to predict pathological complete response of locally advanced esophageal squamous cell carcinoma to neoadjuvant PD-1 blockade in combination with chemotherapy[J/OL]. Eur J Radiol, 2023, 167: 111065 [2024-03-12]. https://www.ejradiology.com/article/S0720-048X(23)00379-0/abstract. DOI: 10.1016/j.ejrad.2023.111065.
[19]
JE H J, CHO S H, OH H S, et al. Response prediction after neoadjuvant chemotherapy for colon cancer using CT tumor regression grade: a preliminary study[J]. J Korean Soc Radiol, 2023, 84(5): 1094-1109. DOI: 10.3348/jksr.2022.0124.
[20]
LUO Y S, LIU L L, LIU D H, et al. Extracellular volume fraction determined by equilibrium contrast-enhanced CT for the prediction of the pathological complete response to neoadjuvant chemoradiotherapy for locally advanced rectal cancer[J]. Eur Radiol, 2023, 33(6): 4042-4051. DOI: 10.1007/s00330-022-09307-z.
[21]
袁文静, 黄芷晴, 贾子琪, 等. 双层探测器光谱CT定量参数评估局部进展期直肠癌新辅助放化疗治疗反应的价值[J]. 中华放射学杂志, 2024, 58(2): 194-200. DOI: 10.3760/cma.j.cn112149-20230825-00127.
YUAN W J, HUANG Z Q, JIA Z Q, et al. Value of dual-layer detector spectral CT quantitative parameters in evaluating treatment response of neoadjuvant chemoradiotherapy in patients with locally advanced rectal cancer[J]. Chin J Radiol, 2024, 58(2): 194-200. DOI: 10.3760/cma.j.cn112149-20230825-00127.
[22]
DILEK O, AKKAYA H, PARLATAN C, et al. Can the mesorectal fat tissue volume be used as a predictive factor in foreseeing the response to neoadjuvant chemoradiotherapy in rectum cancer? A CT-based preliminary study[J]. Abdom Radiol, 2021, 46(6): 2415-2422. DOI: 10.1007/s00261-021-02951-x.
[23]
安悦铭, 范金蕾, 左立平, 等. CT定量肌肉特征在结直肠癌分期及预后中的应用研究进展[J]. 中华放射学杂志, 2024, 58(1): 110-113. DOI: 10.3760/cma.j.cn112149-20230920-00215.
AN Y M, FAN J L, ZUO L P, et al. Advances in the application of CT quantitative muscle characteristics in the staging and predicting prognosis of colorectal cancer[J]. Chin J Radiol, 2024, 58(1): 110-113. DOI: 10.3760/cma.j.cn112149-20230920-00215.
[24]
KIM M, LEE S M, SON I T, et al. Prognostic value of artificial intelligence-driven, computed tomography-based, volumetric assessment of the volume and density of muscle in patients with colon cancer[J]. Korean J Radiol, 2023, 24(9): 849-859. DOI: 10.3348/kjr.2023.0109.
[25]
NIE T, WU F H, HENG Y X, et al. Influence of skeletal muscle and intermuscular fat on postoperative complications and long-term survival in rectal cancer patients[J/OL]. J Cachexia Sarcopenia Muscle, 2024 [2024-03-12]. https://onlinelibrary.wiley.com/doi/epdf/10.1002/jcsm.13424. DOI: 10.1002/jcsm.13424.
[26]
OLMEZ T, OFLUOGLU C B, SERT O Z, et al. The impact of sarcopenia on pathologic complete response following neoadjuvant chemoradiation in rectal cancer[J]. Langenbecks Arch Surg, 2020, 405(8): 1131-1138. DOI: 10.1007/s00423-020-01983-z.
[27]
BEDRIKOVETSKI S, TRAEGER L, VATHER R, et al. Does sarcopenia predict local response rates after chemoradiotherapy for locally advanced rectal cancer?[J]. Dis Colon Rectum, 2023, 66(7): 965-972. DOI: 10.1097/DCR.0000000000002451.
[28]
BEDRIKOVETSKI S, TRAEGER L, PRICE T J, et al. Can sarcopenia predict complete response after total neoadjuvant therapy in advanced rectal cancer? A multicentre observational cohort study[J]. J Surg Oncol, 2023, 128(1): 75-84. DOI: 10.1002/jso.27251.
[29]
TORRI G B, WIETHAN C P, LANGER F W, et al. Split scar sign to predict complete response in rectal cancer after neoadjuvant chemoradiotherapy: systematic review and meta-analysis[J/OL]. Eur Radiol, 2023 [2024-03-12]. https://link.springer.com/article/10.1007/s00330-023-10447-z. DOI: 10.1007/s00330-023-10447-z.
[30]
PATEL U B, TAYLOR F, BLOMQVIST L, et al. Magnetic resonance imaging-detected tumor response for locally advanced rectal cancer predicts survival outcomes: mercury experience[J]. J Clin Oncol, 2011, 29(28): 3753-3760. DOI: 10.1200/JCO.2011.34.9068.
[31]
SCLAFANI F, BROWN G, CUNNINGHAM D, et al. Comparison between MRI and pathology in the assessment of tumour regression grade in rectal cancer[J]. Br J Cancer, 2017, 117(10): 1478-1485. DOI: 10.1038/bjc.2017.320.
[32]
PANG X L, XIE P Y, YU L, et al. A new magnetic resonance imaging tumour response grading scheme for locally advanced rectal cancer[J]. Br J Cancer, 2022, 127(2): 268-277. DOI: 10.1038/s41416-022-01801-x.
[33]
OUYANG G L, YANG X B, DENG X B, et al. Predicting response to total neoadjuvant treatment (TNT) in locally advanced rectal cancer based on multiparametric magnetic resonance imaging: a retrospective study[J]. Cancer Manag Res, 2021, 13: 5657-5669. DOI: 10.2147/CMAR.S311501.
[34]
BORASCHI P, CERVELLI R, DONATI F, et al. Response assessment of locally advanced rectal cancer after neoadjuvant chemoradiotherapy: is apparent diffusion coefficient useful on 3T magnetic resonance imaging?[J]. Colorectal Dis, 2023, 25(5): 905-915. DOI: 10.1111/codi.16483.
[35]
LIAN S S, LIU H M, MENG T B, et al. Quantitative synthetic MRI for predicting locally advanced rectal cancer response to neoadjuvant chemoradiotherapy[J]. Eur Radiol, 2023, 33(3): 1737-1745. DOI: 10.1007/s00330-022-09191-7.
[36]
IAFRATE F, CICCARELLI F, MASCI G M, et al. Predictive role of diffusion-weighted MRI in the assessment of response to total neoadjuvant therapy in locally advanced rectal cancer[J]. Eur Radiol, 2023, 33(2): 854-862. DOI: 10.1007/s00330-022-09086-7.
[37]
YANG L Q, XIA C C, ZHAO J, et al. The value of intravoxel incoherent motion and diffusion kurtosis imaging in the assessment of tumor regression grade and T stages after neoadjuvant chemoradiotherapy in patients with locally advanced rectal cancer[J/OL]. Eur J Radiol, 2021, 136: 109504 [2024-03-12]. https://www.sciencedirect.com/science/article/pii/S0720048X2030694X?via%3Dihub. DOI: 10.1016/j.ejrad.2020.109504.
[38]
FUSCO R, GRANATA V, SANSONE M, et al. Validation of the standardized index of shape tool to analyze DCE-MRI data in the assessment of neo-adjuvant therapy in locally advanced rectal cancer[J]. Radiol Med, 2021, 126(8): 1044-1054. DOI: 10.1007/s11547-021-01369-1.
[39]
CIOLINA M, CARUSO D, SANTIS D D, et al. Dynamic contrast-enhanced magnetic resonance imaging in locally advanced rectal cancer: role ofperfusion parameters in the assessment of response to treatment[J]. Radiol Med, 2019, 124(5): 331-338. DOI: 10.1007/s11547-018-0978-0.
[40]
SCHURINK N W, MIN L A, BERBEE M, et al. Value of combined multiparametric MRI and FDG-PET/CT to identify well-responding rectal cancer patients before the start of neoadjuvant chemoradiation[J]. Eur Radiol, 2020, 30(5): 2945-2954. DOI: 10.1007/s00330-019-06638-2.
[41]
CERNY M, DUNET V, REBECCHINI C, et al. Response of locally advanced rectal cancer (LARC) to radiochemotherapy: DW-MRI and multiparametric PET/CT in correlation with histopathology[J]. Nuklearmedizin, 2019, 58(1): 28-38. DOI: 10.1055/a-0809-4670.
[42]
SAKIN A, SAHIN S, KARYAGAR S S, et al. The predictive value of baseline volumetric PET/CT parameters on treatment response and prognosis in locally advanced rectal cancer treated with neoadjuvant chemoradiotherapy[J]. J Gastrointest Cancer, 2022, 53(2): 341-347. DOI: 10.1007/s12029-021-00608-y.
[43]
ŞEN N P K, AKSU A, KAYA G Ç. Value of volumetric and textural analysis in predicting the treatment response in patients with locally advanced rectal cancer[J]. Ann Nucl Med, 2020, 34(12): 960-967. DOI: 10.1007/s12149-020-01527-x.
[44]
LOVINFOSSE P, POLUS M, VAN DAELE D, et al. FDG PET/CT radiomics for predicting the outcome of locally advanced rectal cancer[J]. Eur J Nucl Med Mol Imaging, 2018, 45(3): 365-375. DOI: 10.1007/s00259-017-3855-5.
[45]
VUIJK F A, FESHTALI SHAHBAZI S, NOORTMAN W A, et al. Baseline and early digital [18F] FDG PET/CT and multiparametric MRI contain promising features to predict response to neoadjuvant therapy in locally advanced rectal cancer patients: a pilot study[J]. Nucl Med Commun, 2023, 44(7): 613-621. DOI: 10.1097/MNM.0000000000001703.
[46]
KUROKAWA R, HAGIWARA A, KUROKAWA M, et al. Diffusion histogram profiles predict molecular features of grade 4 in histologically lower-grade adult diffuse gliomas following WHO classification 2021[J]. Eur Radiol, 2024, 34(2): 1367-1375. DOI: 10.1007/s00330-023-10071-x.
[47]
PHAM T T, LINEY G, WONG K, et al. Multi-parametric magnetic resonance imaging assessment of whole tumour heterogeneity for chemoradiotherapy response prediction in rectal cancer[J]. Phys Imaging Radiat Oncol, 2021, 18: 26-33. DOI: 10.1016/j.phro.2021.03.003.
[48]
JIMÉNEZ DE LOS SANTOS M E, REYES-PÉREZ J A, DOMÍNGUEZ OSORIO V, et al. Whole lesion histogram analysis of apparent diffusion coefficient predicts therapy response in locally advanced rectal cancer[J]. World J Gastroenterol, 2022, 28(23): 2609-2624. DOI: 10.3748/wjg.v28.i23.2609.
[49]
LI D D, CUI Y F, HOU L N, et al. Diffusion kurtosis imaging-derived histogram metrics for prediction of resistance to neoadjuvant chemoradiotherapy in rectal adenocarcinoma: preliminary findings[J/OL]. Eur J Radiol, 2021, 144: 109963 [2024-03-12]. https://www.sciencedirect.com/science/article/pii/S0720048X21004447?via%3Dihub. DOI: 10.1016/j.ejrad.2021.109963.
[50]
SCHURINK N W, VAN KRANEN S R, BERBEE M, et al. Studying local tumour heterogeneity on MRI and FDG-PET/CT to predict response to neoadjuvant chemoradiotherapy in rectal cancer[J]. Eur Radiol, 2021, 31(9): 7031-7038. DOI: 10.1007/s00330-021-07724-0.
[51]
SCHURINK N W, VAN KRANEN S R, ROBERTI S, et al. Sources of variation in multicenter rectal MRI data and their effect on radiomics feature reproducibility[J]. Eur Radiol, 2022, 32(3): 1506-1516. DOI: 10.1007/s00330-021-08251-8.
[52]
SCHURINK N W, VAN KRANEN S R, VAN GRIETHUYSEN J J M, et al. Development and multicenter validation of a multiparametric imaging model to predict treatment response in rectal cancer[J]. Eur Radiol, 2023, 33(12): 8889-8898. DOI: 10.1007/s00330-023-09920-6.
[53]
KURATA Y, HAYANO K, OHIRA G, et al. Computed tomography-derived biomarker for predicting the treatment response to neoadjuvant chemoradiotherapy of rectal cancer[J]. Int J Clin Oncol, 2021, 26(12): 2246-2254. DOI: 10.1007/s10147-021-02027-2.
[54]
BORDRON A, RIO E, BADIC B, et al. External validation of a radiomics model for the prediction of complete response to neoadjuvant chemoradiotherapy in rectal cancer[J/OL]. Cancers, 2022, 14(4): 1079 [2024-03-12]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8870201. DOI: 10.3390/cancers14041079.
[55]
NIE K, SHI L M, CHEN Q, et al. Rectal Cancer: assessment of Neoadjuvant Chemoradiation Outcome based on Radiomics of Multiparametric MRI[J]. Clin Cancer Res, 2016, 22(21): 5256-5264. DOI: 10.1158/1078-0432.CCR-15-2997.
[56]
WEN L, LIU J, HU P S, et al. MRI-based radiomic models outperform radiologists in predicting pathological complete response to neoadjuvant chemoradiotherapy in locally advanced rectal cancer[J]. Acad Radiol, 2023, 30(Suppl 1): S176-S184. DOI: 10.1016/j.acra.2022.12.037.
[57]
QIN S Y, LU S Y, LIU K, et al. Radiomics from mesorectal blood vessels and lymph nodes: a novel prognostic predictor for rectal cancer with neoadjuvant therapy[J/OL]. Diagnostics, 2023, 13(12): 1987 [2024-03-12]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10297215. DOI: 10.3390/diagnostics13121987.
[58]
许汝鑫, 崔艳芬, 杨晓棠. 基于增强CT门静脉期影像组学模型对局部进展期结直肠癌新辅助化疗疗效的预测研究[J]. 中华解剖与临床杂志, 2022, 27(7): 449-457. DOI: 10.3760/cma.j.cn101202-20211129-00360.
XU R X, CUI Y F, YANG X T. Predictive of a radiomic model based on enhanced CT portal vein images for predicting neoadjuvant chemotherapy in patients with locally advanced colorectal cancer[J]. Chin J Anat Clin, 2022, 27(7): 449-457. DOI: 10.3760/cma.j.cn101202-20211129-00360.
[59]
TSAI P C, LEE T H, KUO K C, et al. Histopathology images predict multi-omics aberrations and prognoses in colorectal cancer patients[J/OL]. Nat Commun, 2023, 14(1): 2102 [2024-03-12]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10102208. DOI: 10.1038/s41467-023-37179-4.
[60]
ZHANG F, YAO S, LI Z, et al. Predicting treatment response to neoadjuvant chemoradiotherapy in local advanced rectal cancer by biopsy digital pathology image features[J/OL]. Clin Transl Med, 2020, 10(2): e110 [2024-03-12]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7403709. DOI: 10.1002/ctm2.110.
[61]
SHAO L Z, LIU Z Y, FENG L L, et al. Multiparametric MRI and whole slide image-based pretreatment prediction of pathological response to neoadjuvant chemoradiotherapy in rectal cancer: a multicenter radiopathomic study[J]. Ann Surg Oncol, 2020, 27(11): 4296-4306. DOI: 10.1245/s10434-020-08659-4.
[62]
FENG L L, LIU Z Y, LI C F, et al. Development and validation of a radiopathomics model to predict pathological complete response to neoadjuvant chemoradiotherapy in locally advanced rectal cancer: a multicentre observational study[J/OL]. Lancet Digit Health, 2022, 4(1): e8-e17 [2024-03-12]. https://www.sciencedirect.com/science/article/pii/S2589750021002156?via%3Dihub. DOI: 10.1016/S2589-7500(21)00215-6.
[63]
LI L L, XU B, ZHUANG Z K, et al. Accurate tumor segmentation and treatment outcome prediction with DeepTOP[J/OL]. Radiother Oncol, 2023, 183: 109550 [2024-03-12]. https://www.sciencedirect.com/science/article/pii/S0167814023000889?via%3Dihub. DOI: 10.1016/j.radonc.2023.109550.
[64]
ZHANG X Y, WANG L, ZHU H T, et al. Predicting rectal cancer response to neoadjuvant chemoradiotherapy using deep learning of diffusion kurtosis MRI[J]. Radiology, 2020, 296(1): 56-64. DOI: 10.1148/radiol.2020190936.
[65]
HU Y H, LI J, ZHUANG Z K, et al. Automatic treatment outcome prediction with DeepInteg based on multimodal radiological images in rectal cancer[J/OL]. Heliyon, 2023, 9(2): e13094 [2024-03-12]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9918765. DOI: 10.1016/j.heliyon.2023.e13094.
[66]
JIN C, YU H, KE J, et al. Predicting treatment response from longitudinal images using multi-task deep learning[J/OL]. Nat Commun, 2021, 12(1): 1851 [2024-03-12]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7994301. DOI: 10.1038/s41467-021-22188-y.
[67]
KE J, JIN C, TANG J H, et al. A longitudinal MRI-based artificial intelligence system to predict pathological complete response after neoadjuvant therapy in rectal cancer: a multicenter validation study[J/OL]. Dis Colon Rectum, 2023, 66(12): e1195-e1206 [2024-03-12]. https://journals.lww.com/dcrjournal/abstract/2023/12000/a_longitudinal_mri_based_artificial_intelligence.6.aspx. DOI: 10.1097/DCR.0000000000002931.
[68]
OUYANG G L, CHEN Z B, DOU M, et al. Predicting rectal cancer response to total neoadjuvant treatment using an artificial intelligence model based on magnetic resonance imaging and clinical data[J/OL]. Technol Cancer Res Treat, 2023, 22: 15330338231186467 [2024-3-12]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10338728. DOI: 10.1177/15330338231186467.
[69]
ZHOU X Z, YU Y, FENG Y R, et al. Attention mechanism based multi-sequence MRI fusion improves prediction of response to neoadjuvant chemoradiotherapy in locally advanced rectal cancer[J/OL]. Radiat Oncol, 2023, 18(1): 175 [2024-03-12]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10612200. DOI: 10.1186/s13014-023-02352-y.
[70]
JANG B S, LIM Y J, SONG C, et al. Image-based deep learning model for predicting pathological response in rectal cancer using post-chemoradiotherapy magnetic resonance imaging[J]. Radiother Oncol, 2021, 161: 183-190. DOI: 10.1016/j.radonc.2021.06.019.

上一篇 磁共振T1 mapping及ECV技术在扩张型心肌病中的应用进展
下一篇 磁共振弹性成像在腹盆部肿瘤中的研究进展
  
诚聘英才 | 广告合作 | 免责声明 | 版权声明
联系电话:010-67113815
京ICP备19028836号-2