瘤内及瘤周CT影像组学特征预测软骨肉瘤预后的研究
Intratumoral and Peritumoral CT Radiomics in Predicting Prognosis in Patients with Chondrosarcoma
摘要: 目的:建立基于CT影像组学的软骨肉瘤瘤内、瘤周以及瘤内瘤周联合的影像组学模型,并评价其对患者无进展生存期(PFS)预测的效能。方法:回顾性收集2009年1月至2023年1月期间诊断为软骨肉瘤的患者,并将来自两个中心的212例软骨肉瘤患者分为训练组(n = 101)和验证组(n = 111)。从CT图像中提取瘤内和瘤周的影像组学特征,分别构建瘤内、瘤周和联合的影像组学模型,并计算其影像组学评分(Rad-score)。通过C指数、受者工作特征曲线下时间依赖面积(AUC)和时间依赖C指数来评估瘤内、瘤周和联合影像组学特征在预测软骨肉瘤患者无进展生存期中的作用。结果:分别使用了11、7和16个影像组学特征来构建瘤内、瘤周和联合影像组学模型。联合影像组学模型表现出最好的预测效能。该模型在训练组中C指数为0.788 (95%置信区间0.711~0.861),验证组中C指数为0.750 (95%置信区间0.623~0.867)。结论:联合瘤内和瘤周的CT影像组学特征可以更好地预测软骨肉瘤患者的无进展生存期,有助于临床医生为软骨肉瘤患者选择个性化的监测和治疗方案。
Abstract: Objective: To establish radiomic models based on intratumoral, peritumoral, and combined intratumoral-peritumoral features derived from CT imaging for chondrosarcoma, and to evaluate their efficacy in predicting progression-free survival (PFS) in patients. Methods: A retrospective collection of patients diagnosed with chondrosarcoma from January 2009 to January 2023 was conducted. A total of 212 patients with chondrosarcoma from two centers were divided into a training cohort (n = 101) and a validation cohort (n = 111). Radiomic features from intratumoral and peritumoral regions were extracted from CT images to construct separate intratumoral, peritumoral, and combined radiomic models, and to calculate their radiomic scores (Rad-score). The roles of intratumoral, peritumoral, and combined radiomic features in predicting PFS in chondrosarcoma patients were assessed using the C-index, time-dependent area under the receiver operating characteristic curve (AUC), and time-dependent C-index. Results: Eleven, seven, and sixteen radiomic features were used to construct the intratumoral, peritumoral, and combined radiomic models, respectively. The combined radiomic model demonstrated the best predictive performance. The C-index for this model was 0.788 (95% confidence interval 0.711~0.861) in the training cohort and 0.750 (95% confidence interval 0.623~0.867) in the validation cohort. Conclusion: The combined intratumoral and peritumoral CT radiomic features can better predict PFS in patients with chondrosarcoma, aiding clinicians in selecting personalized monitoring and treatment plans for these patients.
文章引用:孙丽, 李琪媛, 王瑶, 聂佩, 高传平. 瘤内及瘤周CT影像组学特征预测软骨肉瘤预后的研究[J]. 临床医学进展, 2025, 15(3): 2279-2288. https://doi.org/10.12677/acm.2025.153864

1. 引言

软骨肉瘤是起源于软骨细胞或间胚叶组织的原发性骨恶性肿瘤,占所有原发性骨肿瘤的30%,是仅次于骨肉瘤的第二常见实性骨肿瘤[1]。由于软骨肉瘤缺乏血供,其对化疗的敏感性较低[2] [3]。大多数患者术后预后较好[4],但约有15%~25%的患者会出现局部复发[5]-[7]。在精准医疗的时代,在术前做出准确的预后预测对于患者的个体化管理和临床抉择指导是至关重要的。然而,目前仍缺乏可靠的软骨肉瘤预后预测模型。

近年来,多位学者对软骨肉瘤的预后进行了研究。Yan等人比较了基于深度学习的算法和传统方法在预测软骨肉瘤患者总生存期(OS)方面的性能。他们的模型选择了组织学类型、原发部位、肿瘤大小和其他临床因素等9个特征。结果显示,将cox比例风险模型与神经网络相结合的深度生存模型预测效能最高,验证组的C指数为0.832 [8]。Song等人基于Surveillance、流行病学、End Result数据库确定了6个独立的预后因素并构建列线图,包括年龄、组织学亚型、分级、手术、肿瘤大小和远处转移。他们发现列线图预测生存率与实际生存率之间存在良好的一致性,验证组中预测软骨肉瘤患者总生存期和肿瘤特异性生存率的C指数分别为0.803和0.829 [9]。但以往的大多数软骨肉瘤预后研究大都是基于临床或病理结果,没有影像学特征的研究。

医学影像学在软骨肉瘤患者的诊治中起着重要的作用,例如检测、诊断、分期和预后预测[10]。部分学者认为影像学观察到的溶骨性病变和明显的骨质疏松症可能提示患者的不良预后,但相关研究较少[11] [12]。常规影像学依赖于肉眼所见的征象,提供的信息有限,可能会丢失大量与肿瘤异质性相关的信息[13]。在大数据时代,影像组学可以无创地将病变异质性转化为高维图像特征。目前影像组学已成功应用于预测消化系统[14]、神经系统[15]、生殖系统[16]肿瘤的预后,从而利于癌症的精准管理和临床决策[17]

软骨肉瘤具有高度空间异质性,包括肿瘤本身与瘤周区域。瘤周区域即紧连肿瘤微环境(TME)或肿瘤生存环境。TME是由肿瘤细胞、细胞外基质、血管以及免疫细胞等共同构成复杂的生态系统[18]。肿瘤周围基质的变化决定了肿瘤生长和扩散、逃避机体免疫保护和抵抗治疗干预的能力,进而导致肿瘤患者不同的临床治疗及预后结局[19]。影像组学通过高通量提取影像特征,能够评估肉眼无法识别的微环境及肿瘤的病理生理变化,为临床选择科学的治疗方案提供重要依据。虽然瘤内和瘤周的影像组学特征代表不同的空间异质性信息,但它们不是独立的,而是互补的。如今,许多研究将瘤周影像组学特征整合到瘤内影像组学模型中,或在临床模型中用于生存分析,如食管癌[20],乳腺癌[21],肝细胞癌[22]以及乳腺[10]等。目前来说,大多数软骨肉瘤影像组学研究都集中在鉴别诊断和病理分级上[23]-[26],并且大多是基于瘤内影像组学特征,但瘤内和瘤周影像组学对软骨肉瘤预后的预测效能高低尚未得到评估。

本研究的目的是评估瘤内、瘤周和联合影像组学特征在预测软骨肉瘤患者预后方面的效能。

2. 材料与方法

2.1. 研究对象

本回顾性多中心研究在机构审查委员会下进行,并放弃知情同意。本研究回顾性收集了2009年1月至2023年1月机构的病理数据库中手术切除或活检标本中选择诊断为软骨肉瘤的患者。纳入标准如下:1) 经手术或活检病理证实的软骨肉瘤患者;2) 在手术或活检病理前2周内行非增强CT检查。排除标准包括其他恶性肿瘤、图像质量不足(有金属或运动伪影的图像)和随访资料不完整的患者。最后,青岛大学附属医院(训练组,n = 101)和山东第一医科大学附属山东省立医院(验证组,n = 111)共212例患者纳入本研究(图1)。

所有患者均进行非增强CT扫描,两个中心共使用6台扫描仪进行轴向扫描。

从电子病历中收集病例的临床和病理数据,包括年龄、性别、身高、体重、肿瘤大小、肿瘤部位、病理分级、治疗方案等。

2.2. 随访

患者在前两年至少每6~12个月随访一次,然后每年随访一次。最后一次随访日期为2023年12月28日。随访数据的获取是通过医疗记录、影像学表现(X射线、CT、核磁共振)或电话。本研究的随访终点是无进展生存期或最后一次随访的时间。无进展生存期的定义为从诊断之日到局部复发、远处转移、任何原因死亡的时间。

2.3. 兴趣区勾画及影像组学特征提取

影像组学的工作流程如图2所示。由两名分别具有6年和8年的肌肉骨骼系统影像诊断经验的放射科医生使用ITK-SNAP软件(版本3.8.0,http://www.itksnap.org/)在轴向CT图像上勾画瘤内感兴趣区(ROI)。本研究中,瘤周ROI的获取是利用RIAS软件的“ROI操作”模块[27],自动向肿瘤外扩展3 mm并切除瘤内区域而生成。

考虑到患者来自不同中心,我们对CT图像进行重采样,并进行灰度离散和归一化后再进行特征提

Figure 1. Flow diagram depicting the patient selection process. *The Affiliated Hospital of Qingdao University. **Shandong Provincial Hospital Affiliated to Shandong First Medical University

1. 描述选择患者的流程图。*青岛大学附属医院。**山东第一医科大学附属山东省立医院

取。影像组学特征提取通过RadCloud平台(惠英医疗科技有限公司)进行。从瘤内和瘤周ROI中共提取2818 (1409 + 1409)个影像组学特征,包括一阶统计特征、形状和尺寸的特征、纹理特征和高阶统计特征。

2.4. 观察者内部和观察者之间的再现性

使用类内和类间相关系数(ICCs)来评估观察者内部和观察者之间的可重复性。随机选择40张CT图像,由观察者1和观察者2独立进行ROI勾画,以评估观察者间的再现性。观察者1在2周后重复勾画,以评估观察者内部的可重复性。观察者内和观察者间ICC > 0.75表明可重复性好,并排除ICC < 0.75的影像组学特征[28]。剩余的图像勾画由观察者1进行。

2.5. 瘤内、瘤周和联合影像组学特征的筛选

影像组学特征的筛选分为3个步骤,瘤内和瘤周的影像组学特征使用同样的方法。首先,使用Pearson相关分析,减少冗余的影像学特征。然后采用单因素Cox比例回归分析选择影像组学特征,P < 0.05。最后,采用最小绝对收缩和选择算子(LASSO) Cox回归模型选择最优特征。λ为LASSO回归的正则化参数,在交叉验证误差最小时选取。采用Cox比例风险回归模型计算每位患者的影像组学评分(Rad-score)。

2.6. 瘤内、瘤周和联合影像组学模型的评估

本研究采用Harrel的C指数和HR来评估瘤内、瘤周和联合三组影像组学模型预测训练组PFS的准确性,并在验证组中进行验证。使用X-tile软件确定Rad-score的最佳阈值,并用Kaplan-Meier生存分析来分析两组患者的PFS,以评估三组影像组学模型的预后意义。为了比较特征预测的预后与实际预后,生成校准曲线。然后使用受者工作特征曲线下时间依赖面积(AUC)和时间依赖C指数来评估三组影像组学模型的预测效能。最后计算净重新分类指数(NRI)来评估联合特征相对于瘤内、瘤周特征的优势。

2.7. 统计数据

我们使用SPSS 26.0软件(IBM, Chicago, IL, USA)进行独立样本t检验或Mann-Whitney U检验、卡方(χ2)检验或Fisher精确检验。再使用R统计软件(4.2.1版,https://www.r-project.org/)进行ICC、Pearson相关分析、单变量Cox比例回归分析、LASSO Cox回归分析、校正图、Kaplan-Meier生存分析、C指数、时间C指数、时间AUC分析和NRI分析,P < 0.05被认为具有统计学差异。模型构建的详细过程如图2所示。

Figure 2. The workflow of the multicenter study. The tumor was segmented to determine the intratumoral ROI (red) and peritumoral ROI (yellow) from non-contrast-enhanced CT images. More related images of this case can be found in the supplementary material. ICC, intra-/Inter-class correlation coefficient; LASSO, least absolute shrinkage and selection operator; AUC, area under the curve

2. 多中心研究工作流程。对肿瘤进行分割,以确定CT平扫图像的瘤内ROI (红色)和瘤周ROI (黄色)。本病例的更多相关图像可在补充资料中找到。ICC,类内/类间相关系数;LASSO,最小绝对收缩和选择算子;AUC,曲线下面积

3. 结果

3.1. 瘤内、瘤周和联合影像组学特征的选择和模型构建

将提取的2818 (1409 + 1409)个影像组学特征(包括一阶统计特征、形状和尺寸的特征、纹理特征和高阶统计特征)通过ANOVA筛选留下了735个特征(P > 0.05),最后经过LASSO算法10倍交叉验证筛选,最终分别选择了11、7和16个影像组学特征分别构建瘤内、瘤周和联合影像组学特征模型(RS瘤内、RS瘤周、RS联合),筛选后保留的影像组学特征所构建的三种模型预测软骨肉瘤患者PFS效能良好,在训练组中的C指数为0.788 (95%置信区间0.711~0.861),验证组中C指数为0.750 (95%置信区间0.623~0.867)。

3.2. 瘤内、瘤周和联合影像组学模型的评估

表1显示了不同模型中每个特征与复发相关的C指数和HR估计值。RS联合的预测能力最好,C指数为0.750 (95%CI: 0.623~0.867)。

Kaplan-Meier生存分析显示三组模型计算的Rad评分与PFS呈正相关(图3)。每个特征的校准曲线如图3所示。与RS瘤内与RS瘤周相比,RS联合显示出更高的时间AUC和时间相关C指数(图4图5)。RS瘤内相比,联合模型的NRI为0.386 (95%CI: 0.074~0.847, P < 0.1)。与RS瘤周相比,RS联合提供的NRI为0.296 (95%CI: 0.168~0.759, P = 0.2)。

Table 1. The performance of the RSregion, RSperi, and RScombine in predicting PFS of the patients with chondrosarcoma

1. RS瘤内、RS瘤周和RS联合预测软骨肉瘤患者FPS的效能

模型

C指数(95%CI)

训练组

验证组

RS瘤内

0.696 (0.609~0.785)

0.678 (0.538~0.815)

RS瘤周

0.674 (0.571~0.770)

0.625 (0.494~0.751)

RS联合

0.788 (0.711~0.861)

0.750 (0.623~0.867)

RS:影像组学模型;C指数:一致性指数;CI:置信区间

Figure 3. Kaplan-Meier survival curves of three models in training set (A~C) and verification set (D~F), in which A and D are intra-tumor models, B and E are peritumor models, and C and F are joint models

3. 训练集(A~C)和验证集(D~F)中三种模型的Kaplan-Meier生存曲线,A、D为瘤内模型,B、E为瘤周模型,C、F为联合模型

Figure 4. AUC curves of three models in training set (A) and verification set (B). The green, blue and red curves represent the intra-tumor model, peritumor model and combined model respectively

4. 三种模型在训练集(A)和验证集(B)中的AUC曲线。绿色、蓝色和红色曲线分别代表瘤内模型、瘤周模型和联合模型

Figure 5. C-exponential curves of three models in training set (A) and verification set (B). The green, blue and red curves represent the intra-tumor model, peritumor model and combined model respectively

5. 三种模型在训练集(A)和验证集(B)中的C-指数曲线。绿色、蓝色和红色曲线分别代表瘤内模型、瘤周模型和联合模型

4. 讨论

在这项回顾性多中心研究中,我们开发并验证了基于CT的软骨肉瘤瘤内、瘤周以及瘤内瘤周联合的影像组学模型。我们评估了三个模型的预测效能,发现这三个模型都具有筛选高风险患者与低风险患者的能力,而联合模型对软骨肉瘤患者预后的预测价值高于其他两种单一影像组学模型。

软骨肉瘤是第二常见的原发性骨恶性肿瘤。超过90%的软骨肉瘤为中低级别肿瘤,5%~10%的软骨肉瘤为高转移倾向的高级别侵袭性肿瘤[29] [30]。因此,准确预测软骨肉瘤患者的预后非常重要,它可以帮助临床医生选择合适的监测和治疗策略,从而改善软骨肉瘤患者的预后。目前我们已经确定了几个与复发相关的因素,如肿瘤大小、分级和分期[1] [31] [32]。骨肿瘤术前评估通常需要结合临床特征、影像学表现和组织病理学结果。随着人工智能的快速发展,图像大数据在肿瘤学领域的潜力得到越来越多的认可[13]。Yin等人开发了一种基于3D MR影像组学和临床特征的列线图,用来评估骨盆软骨肉瘤早期复发。他们发现联合列线图优于临床模型(AUC: 0.891 vs. 0.625)。Rad评分是预测骨盆软骨肉瘤早期复发最重要的危险因素(OR = 3, P < 0.01)。我们的研究还证实了基于影像组学的软骨肉瘤患者肿瘤异质性与PFS之间的关联。瘤内影像组学模型验证组的C指数为0.678,这表明瘤内影像组学特征对软骨肉瘤患者具有良好的预测价值。

肿瘤周围基质的变化决定了肿瘤生长和扩散、逃避机体免疫保护和抵抗治疗干预的能力[33]。此外,有研究表明,肿瘤周围区域反映了肿瘤周围免疫细胞的浸润情况[34] [35]。因此,肿瘤周围的异质性和微环境也与肿瘤的侵袭性有关。然而,肿瘤周围区域的特征不能通过肿瘤实质的影像组学分析来得到有效体现。如今,许多研究将瘤周影像组学特征整合到瘤内影像组学模型中,或在临床模型中用于生存分析,如食管癌[36]、乳腺癌[37]、肝细胞癌[38]、肺癌[39]等。Hu等人证明由瘤内和瘤周影像组学特征组成的联合模型可以预测食管鳞状细胞癌患者新辅助放化疗后的病理完全缓解情况,AUC为0.852 (95%CI: 0.753~0.951),在验证组中准确率为84.3% [36]。Khorrami等人发现,从CT图像中提取的瘤内和瘤周的形状和纹理特征可以识别非小细胞肺癌患者对新辅助放化疗的病理反应,训练组和验证组的AUC分别为0.90和0.86。此外,影像组学特征也与OS (HR: 11.18%; 95%CI: 3.17~44.1)和PFS (HR: 2.78; 95%CI: 1.11~4.12)显著相关[40]。Chong等人利用肝胆特异性Gd-EOB-DTPA MRI建立了肝细胞癌患者的瘤周影像组学模型,并发现该模型提供了最佳的临床净效益(NRI: 35.9~66.1%, P < 0.01)。该模型在预测肝细胞癌早期复发方面比现有的临床算法更有效[41]

充分考虑到瘤内及瘤周的异质性和微环境有助于精确预测软骨肉瘤患者的预后。虽然瘤内和瘤周的影像组学特征代表不同的空间异质性信息,但它们不是冗余的,而是互补的。在这项研究中,基于CT的瘤周影像组学模型预测验证组PFS的C指数为0.625。通过将瘤周影像组学特征与瘤内影像组学特征相结合,联合模型在验证组中获得了更高的C指数(0.750)、时间AUC及NRI。因此,瘤内和瘤周联合的影像组学模型可以为软骨肉瘤患者的生存预测提供更高的价值。在联合特征的指导下,如果患者被分类为高风险组,则建议进行更密切的复查监测和全身辅助治疗。反之,低风险组的患者我们提倡只进行定期复查随访即可。

不可避免的是这项研究也存在一些局限性。首先,由于患者来自不同中心,在扫描参数上难免存在差异。为了避免这种差异的影响,图像采集、特征提取、数据处理等操作应进行标准化。第二,在目前的研究中,肿瘤的3D分割是手动进行的,这具有不稳定性,也比较费时。我们要学习和推进半自动或全自动肿瘤分割的先进方法,这很可能会在影像组学领域得到广泛应用。第三,根据纳入和排除标准,只有约30%的软骨肉瘤患者被纳入本研究中。这是因为我们的研究是回顾性研究,无法对患者进行同质性管理。在未来的应用中,应该对患者进行更好的管理,以解决这一问题。未来的研究中我们将扩大纳入患者的病例数量,并分别基于CT、MR和X线片建立预后模型,以适用于不同的患者群体。

总而言之,我们的研究表明,瘤内和瘤周影像组学特征是软骨肉瘤患者潜在的预后标志物。结合瘤内和瘤周的联合影像组学模型可以作为预测软骨肉瘤患者预后的新型影像学工具。这可以提高患者的个体化治疗和管理,进一步促进精准医疗。

声 明

本研究已获得我院伦理委员会的批准。所有患者或其监护人均已签署知情同意书。

NOTES

*第一作者。

#通讯作者。

参考文献

[1] Tong, Y., Cui, Y., Jiang, L., Pi, Y., Gong, Y. and Zhao, D. (2022) Clinical Characteristics, Prognostic Factor and a Novel Dynamic Prediction Model for Overall Survival of Elderly Patients with Chondrosarcoma: A Population-Based Study. Frontiers in Public Health, 10, Article ID: 901680.
https://doi.org/10.3389/fpubh.2022.901680
[2] Italiano, A., Mir, O., Cioffi, A., Palmerini, E., Piperno-Neumann, S., Perrin, C., et al. (2013) Advanced Chondrosarcomas: Role of Chemotherapy and Survival. Annals of Oncology, 24, 2916-2922.
https://doi.org/10.1093/annonc/mdt374
[3] de Jong, Y., Ingola, M., Briaire-de Bruijn, I.H., Kruisselbrink, A.B., Venneker, S., Palubeckaite, I., et al. (2019) Radiotherapy Resistance in Chondrosarcoma Cells; a Possible Correlation with Alterations in Cell Cycle Related Genes. Clinical Sarcoma Research, 9, Article No. 9.
https://doi.org/10.1186/s13569-019-0119-0
[4] Laitinen, M.K., Parry, M.C., Le Nail, L., Wigley, C.H., Stevenson, J.D. and Jeys, L.M. (2019) Locally Recurrent Chondrosarcoma of the Pelvis and Limbs Can Only Be Controlled by Wide Local Excision. The Bone & Joint Journal, 101, 266-271.
https://doi.org/10.1302/0301-620x.101b3.bjj-2018-0881.r1
[5] Björnsson, J., McLeod, R.A., Unni, K.K., Ilstrup, D.M. and Pritchard, D.J. (1998) Primary Chondrosarcoma of Long Bones and Limb Girdles. Cancer, 83, 2105-2119.
https://doi.org/10.1002/(sici)1097-0142(19981115)83:10<2105::aid-cncr9>3.0.co;2-u
[6] Bruns, J., Elbracht, M. and Niggemeyer, O. (2001) Chondrosarcoma of Bone: An Oncological and Functional Follow-Up Study. Annals of Oncology, 12, 859-864.
https://doi.org/10.1023/a:1011162118869
[7] Lee, F.Y., Mankin, H.J., Fondren, G., Gebhardt, M.C., springfield, D.S., Rosenberg, A.E., et al. (1999) Chondrosarcoma of Bone: An Assessment of Outcome. The Journal of Bone & Joint Surgery, 81, 326-338.
https://doi.org/10.2106/00004623-199903000-00004
[8] Yan, L., Gao, N., Ai, F., Zhao, Y., Kang, Y., Chen, J., et al. (2022) Deep Learning Models for Predicting the Survival of Patients with Chondrosarcoma Based on a Surveillance, Epidemiology, and End Results Analysis. Frontiers in Oncology, 12, Article ID: 967758.
https://doi.org/10.3389/fonc.2022.967758
[9] Song, K., Shi, X., Wang, H., Zou, F., Lu, F., Ma, X., et al. (2018) Can a Nomogram Help to Predict the Overall and Cancer-Specific Survival of Patients with Chondrosarcoma? Clinical Orthopaedics & Related Research, 476, 987-996.
https://doi.org/10.1007/s11999.0000000000000152
[10] Liu, C., Xi, Y., Li, M., Jiao, Q., Zhang, H., Yang, Q., et al. (2017) Dedifferentiated Chondrosarcoma: Radiological Features, Prognostic Factors and Survival Statistics in 23 Patients. PLOS ONE, 12, e0173665.
https://doi.org/10.1371/journal.pone.0173665
[11] Soldatos, T., McCarthy, E.F., Attar, S., Carrino, J.A. and Fayad, L.M. (2011) Imaging Features of Chondrosarcoma. Journal of Computer Assisted Tomography, 35, 504-511.
https://doi.org/10.1097/rct.0b013e31822048ff
[12] Jurik, A., Jensen, O., Keller, J., Nielsen, O., Lundorf, E., Daugaard, S., et al. (1995) Imaging of Chondrosarcoma with Histopathological and Prognostic Correlation. An Analysis of 49 Cases Mainly Based on Plain Film Radiography. Rofo, 163, 372-377.
https://doi.org/10.1055/s-2007-1016011
[13] Bi, W.L., Hosny, A., Schabath, M.B., Giger, M.L., Birkbak, N.J., Mehrtash, A., et al. (2019) Artificial Intelligence in Cancer Imaging: Clinical Challenges and Applications. CA: A Cancer Journal for Clinicians, 69, 127-157.
https://doi.org/10.3322/caac.21552
[14] Qiu, H., Xu, M., Wang, Y., Wen, X., Chen, X., Liu, W., Zhang, N., Ding, X. and Zhang, L. (2022) A Novel Preoperative MRI-Based Radiomics Nomogram Outperforms Traditional Models for Prognostic Prediction in Pancreatic Ductal Adenocarcinoma. American Journal of Cancer Research, 12, 2032-2049.
[15] Liu, X., Li, Y., Qian, Z., Sun, Z., Xu, K., Wang, K., et al. (2018) A Radiomic Signature as a Non-Invasive Predictor of Progression-Free Survival in Patients with Lower-Grade Gliomas. NeuroImage: Clinical, 20, 1070-1077.
https://doi.org/10.1016/j.nicl.2018.10.014
[16] Rizzo, S., Manganaro, L., Dolciami, M., Gasparri, M.L., Papadia, A. and Del Grande, F. (2021) Computed Tomography Based Radiomics as a Predictor of Survival in Ovarian Cancer Patients: A Systematic Review. Cancers, 13, Article No. 573.
https://doi.org/10.3390/cancers13030573
[17] Mayerhoefer, M.E., Materka, A., Langs, G., Häggström, I., Szczypiński, P., Gibbs, P., et al. (2020) Introduction to Radiomics. Journal of Nuclear Medicine, 61, 488-495.
https://doi.org/10.2967/jnumed.118.222893
[18] de Visser, K.E. and Joyce, J.A. (2023) The Evolving Tumor Microenvironment: From Cancer Initiation to Metastatic Outgrowth. Cancer Cell, 41, 374-403.
https://doi.org/10.1016/j.ccell.2023.02.016
[19] Jiang, K., Wu, J., Wang, Q., Chen, X., Zhang, Y., Gu, X., et al. (2024) Nanoparticles Targeting the Adenosine Pathway for Cancer Immunotherapy. Journal of Materials Chemistry B, 12, 5787-5811.
https://doi.org/10.1039/d4tb00292j
[20] Dercle, L., Lu, L., Schwartz, L.H., Qian, M., Tejpar, S., Eggleton, P., et al. (2020) Radiomics Response Signature for Identification of Metastatic Colorectal Cancer Sensitive to Therapies Targeting EGFR Pathway. JNCI: Journal of the National Cancer Institute, 112, 902-912.
https://doi.org/10.1093/jnci/djaa017
[21] Afshar, P., Mohammadi, A., Tyrrell, P.N., Cheung, P., Sigiuk, A., Plataniotis, K.N., et al. (2020) DRTOP: Deep Learning-Based Radiomics for the Time-to-Event Outcome Prediction in Lung Cancer. Scientific Reports, 10, Article No. 12366.
https://doi.org/10.1038/s41598-020-69106-8
[22] Hosny, A., Parmar, C., Coroller, T.P., Grossmann, P., Zeleznik, R., Kumar, A., et al. (2018) Deep Learning for Lung Cancer Prognostication: A Retrospective Multi-Cohort Radiomics Study. PLOS Medicine, 15, e1002711.
https://doi.org/10.1371/journal.pmed.1002711
[23] Gitto, S., Cuocolo, R., van Langevelde, K., van de Sande, M.A.J., Parafioriti, A., Luzzati, A., et al. (2022) MRI Radiomics-Based Machine Learning Classification of Atypical Cartilaginous Tumour and Grade II Chondrosarcoma of Long Bones. eBioMedicine, 75, Article ID: 103757.
https://doi.org/10.1016/j.ebiom.2021.103757
[24] Li, L., Wang, K., Ma, X., Liu, Z., Wang, S., Du, J., et al. (2019) Radiomic Analysis of Multiparametric Magnetic Resonance Imaging for Differentiating Skull Base Chordoma and Chondrosarcoma. European Journal of Radiology, 118, 81-87.
https://doi.org/10.1016/j.ejrad.2019.07.006
[25] Pan, J., Zhang, K., Le, H., Jiang, Y., Li, W., Geng, Y., et al. (2021) Radiomics Nomograms Based on Non‐Enhanced MRI and Clinical Risk Factors for the Differentiation of Chondrosarcoma from Enchondroma. Journal of Magnetic Resonance Imaging, 54, 1314-1323.
https://doi.org/10.1002/jmri.27690
[26] Yin, P., Mao, N., Liu, X., Sun, C., Wang, S., Chen, L., et al. (2019) Can Clinical Radiomics Nomogram Based on 3D Multiparametric MRI Features and Clinical Characteristics Estimate Early Recurrence of Pelvic Chondrosarcoma? Journal of Magnetic Resonance Imaging, 51, 435-445.
https://doi.org/10.1002/jmri.26834
[27] Li, M., Li, X., Guo, Y., Miao, Z., Liu, X., Guo, S., et al. (2020) Development and Assessment of an Individualized Nomogram to Predict Colorectal Cancer Liver Metastases. Quantitative Imaging in Medicine and Surgery, 10, 397-414.
https://doi.org/10.21037/qims.2019.12.16
[28] Koo, T.K. and Li, M.Y. (2016) A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research. Journal of Chiropractic Medicine, 15, 155-163.
https://doi.org/10.1016/j.jcm.2016.02.012
[29] Heck, R.K., Peabody, T.D. and Simon, M.A. (2006) Staging of Primary Malignancies of Bone. CA: A Cancer Journal for Clinicians, 56, 366-375.
https://doi.org/10.3322/canjclin.56.6.366
[30] Chow, W.A. (2007) Update on Chondrosarcomas. Current Opinion in Oncology, 19, 371-376.
https://doi.org/10.1097/cco.0b013e32812143d9
[31] Frezza, A.M., Cesari, M., Baumhoer, D., Biau, D., Bielack, S., Campanacci, D.A., et al. (2015) Mesenchymal Chondrosarcoma: Prognostic Factors and Outcome in 113 Patients. A European Musculoskeletal Oncology Society Study. European Journal of Cancer, 51, 374-381.
https://doi.org/10.1016/j.ejca.2014.11.007
[32] Wang, Z., Chen, G., Chen, X., Huang, X., Liu, M., Pan, W., et al. (2019) Predictors of the Survival of Patients with Chondrosarcoma of Bone and Metastatic Disease at Diagnosis. Journal of Cancer, 10, 2457-2463.
https://doi.org/10.7150/jca.30388
[33] Braman, N., Prasanna, P., Whitney, J., Singh, S., Beig, N., Etesami, M., et al. (2019) Association of Peritumoral Radiomics with Tumor Biology and Pathologic Response to Preoperative Targeted Therapy for HER2 (ERBB2)-Positive Breast Cancer. JAMA Network Open, 2, e192561.
https://doi.org/10.1001/jamanetworkopen.2019.2561
[34] Polyak, K., Haviv, I. and Campbell, I.G. (2009) Co-Evolution of Tumor Cells and Their Microenvironment. Trends in Genetics, 25, 30-38.
https://doi.org/10.1016/j.tig.2008.10.012
[35] Faget, J., Groeneveld, S., Boivin, G., Sankar, M., Zangger, N., Garcia, M., et al. (2017) Neutrophils and Snail Orchestrate the Establishment of a Pro-Tumor Microenvironment in Lung Cancer. Cell Reports, 21, 3190-3204.
https://doi.org/10.1016/j.celrep.2017.11.052
[36] Hu, Y., Xie, C., Yang, H., Ho, J.W.K., Wen, J., Han, L., et al. (2020) Assessment of Intratumoral and Peritumoral Computed Tomography Radiomics for Predicting Pathological Complete Response to Neoadjuvant Chemoradiation in Patients with Esophageal Squamous Cell Carcinoma. JAMA Network Open, 3, e2015927.
https://doi.org/10.1001/jamanetworkopen.2020.15927
[37] Braman, N.M., Etesami, M., Prasanna, P., Dubchuk, C., Gilmore, H., Tiwari, P., et al. (2017) Intratumoral and Peritumoral Radiomics for the Pretreatment Prediction of Pathological Complete Response to Neoadjuvant Chemotherapy Based on Breast DCE-MRI. Breast Cancer Research, 19, 57.
https://doi.org/10.1186/s13058-017-0846-1
[38] Shan, Q., Hu, H., Feng, S., Peng, Z., Chen, S., Zhou, Q., et al. (2019) CT-Based Peritumoral Radiomics Signatures to Predict Early Recurrence in Hepatocellular Carcinoma after Curative Tumor Resection or Ablation. Cancer Imaging, 19, Article No. 11.
https://doi.org/10.1186/s40644-019-0197-5
[39] Akinci D'Antonoli, T., Farchione, A., Lenkowicz, J., Chiappetta, M., Cicchetti, G., Martino, A., et al. (2020) CT Radiomics Signature of Tumor and Peritumoral Lung Parenchyma to Predict Nonsmall Cell Lung Cancer Postsurgical Recurrence Risk. Academic Radiology, 27, 497-507.
https://doi.org/10.1016/j.acra.2019.05.019
[40] Khorrami, M., Jain, P., Bera, K., Alilou, M., Thawani, R., Patil, P., et al. (2019) Predicting Pathologic Response to Neoadjuvant Chemoradiation in Resectable Stage III Non-Small Cell Lung Cancer Patients Using Computed Tomography Radiomic Features. Lung Cancer, 135, 1-9.
https://doi.org/10.1016/j.lungcan.2019.06.020
[41] Chong, H., Gong, Y., Pan, X., Liu, A., Chen, L., Yang, C., et al. (2021) Peritumoral Dilation Radiomics of Gadoxetate Disodium-Enhanced MRI Excellently Predicts Early Recurrence of Hepatocellular Carcinoma without Macrovascular Invasion after Hepatectomy. Journal of Hepatocellular Carcinoma, 8, 545-563.
https://doi.org/10.2147/jhc.s309570

Baidu
map