Potential value of detection of minimal residual disease in colorectal cancer following radical resection

Although there has been significant advancement in the identification and management of colorectal cancer(CRC)in recent years,there is still room for improvement in the current standard treatment regimen.One area of concern is the lack of reliable tumor markers to predict treatment efficacy and guide tailored care.Due to its dynamic,effective,and non-invasive benefits over tissue biopsy,the detection of minimal or molecular residual lesions(MRD)based on circulating tumor DNA(ctDNA)is beneficial to the clinical development of drugs for patients with CRC after radical treatment,as well as for continuous monitoring of tumor recurrence and malignancy molecular gene evolution.The detection of ctDNA can currently be used to guide individual postoperative auxiliary treatment decisions(upgrade or downgrade treatment)in CRC,stratify the risk of clinical recurrence more precisely,and predict the risk of recurrence in advance of imaging examination,according to a large number of observational or prospective clinical studies.With increasing clarity comes the possibility of selecting a regimen of treatment based on postoperative ctDNA,which also improves the accuracy of clinical recurrence risk assessment for CRC.Therefore,it is anticipated that the identification of ctDNA would alter the current framework for dealing with CRC and lead to individualized,stratified precision therapy;however,additional confirmation will require subsequent high-quality,prospective,large-scale randomized controlled studies.This article will provide an overview of the definition and clinical significance of MRD,the primary indications and technological challenges for MRD detection,along with the advancement in clinical research about ctDNA detection following radical resection of the CRC.

Over-expression of VEGF and MMP-9 in Residual Tumor Cells of Hepatocellular Carcinoma after Embolization with Lipidol

The expression and implication of vascular endothelial growth factor （VEGF） and matrix metalloproteinase-9 （MMP-9） in residual hepatic tumor cells after lipiodol embolization were investi- gated. Two weeks after transplantation of VX2 tumor cells into the livers of rabbits, a xenograft model of the human hepatic neoplasm was successfully established. Forty rabbits were randomly divided into control group （n=20） and lipiodol group （n=20）. For the control group, 1 mL normal saline was injected through the gastroduodenal artery, whereas 0.3 mL/kg lipiodol was applied for the lipiodol group. One week after embolization, the expression level of VEGF in the plasma was measured by using en- zyme-linked immunosorbent assay （ELISA）. A three-step immunohistochemieal technique （ABC） was employed to detect the protein levels of VEGF and MMP-9 and the quantitative PCR for their mRNA levels was performed in the residual tumor cells. The VEGF in the plasma was significantly higher in the lipiodol group （1.42~0.29 ng/mL） than in the control group （1.12~0.21 ng/mL） （P〈0.01）. Moreover, the positive rate of VEGF protein in the residual tumor cells was significantly higher in the lipiodol group （62.13%~7.69%） than in the control group （53.16%~9.17%） （P〈0.05）. Similarly, the MMP-9 ex- pression in the residual ~mor cells was higher in the lipiodol group. The mRNA levels of VEGF （2.9313~2.4231） and MMP-9 （3.5721~1.6107） in the lipiodol group were significantly higher than those in the control group （1.5728~0.9453 and 1.7573~1.0641, respectively, P〈0.05）. Therefore, it was rea- sonable to speculate that the increased expression of VEGF and MMP-9 in residual hepatic tumor cells and tumor angiogenesis post-embolization would be responsible for the increased metastatic potentiality and invasiveness of these cells.

Risk factors for residual tumor after resection of hepatocellular carcinoma

AIM:To identify the clinicopathological risk factors correlated with residual tumor in hepatocellular carcinoma (HCC) patients after resection. METHODS:From January 2001 to April 2007,766 HCC patients who had undergone resection were included in this research. Lipiodol angiography was performed within 2 mo after surgery and followed by post-Lipiodol computed tomography (CT) 4 wk later for all 766 patients to monitor tumor in the remnant liver. Tumor detected within the first 3-mo postoperative period was defined as residual tumor. Patients were divided into 2 groups:disease or disease-free within the first 3 mo after surgery. Risk factors for residual tumor were investigated among various clinicopathological variables. RESULTS:A total of 63 (8.22%) patients were found to have residual tumor after surgery. Three independent factors associated with residual tumor were identified by multivariate analysis:preoperative serum α-fetoprotein (AFP) level [odds ratio (OR) = 1.68 (95% confidence interval (CI):1.20-2.36)],tumor size [OR = 1.73 (95% CI:1.29-2.31)] and microvascular invasion [OR = 1.91 (95% CI:1.12-3.24)]. CONCLUSION:Residual tumor is related to AFP level,tumor size and microvascular invasion. Patients at high risk should undergo closer follow-up and could be candidates for multimodality therapy.

MRI Brain Tumor Segmentation Using 3D U-Net with Dense Encoder Blocks and Residual Decoder Blocks

The main task of magnetic resonance imaging (MRI) automatic brain tumor segmentation is to automaticallysegment the brain tumor edema, peritumoral edema, endoscopic core, enhancing tumor core and nonenhancingtumor core from 3D MR images. Because the location, size, shape and intensity of brain tumors vary greatly, itis very difficult to segment these brain tumor regions automatically. In this paper, by combining the advantagesof DenseNet and ResNet, we proposed a new 3D U-Net with dense encoder blocks and residual decoder blocks.We used dense blocks in the encoder part and residual blocks in the decoder part. The number of output featuremaps increases with the network layers in contracting path of encoder, which is consistent with the characteristicsof dense blocks. Using dense blocks can decrease the number of network parameters, deepen network layers,strengthen feature propagation, alleviate vanishing-gradient and enlarge receptive fields. The residual blockswere used in the decoder to replace the convolution neural block of original U-Net, which made the networkperformance better. Our proposed approach was trained and validated on the BraTS2019 training and validationdata set. We obtained dice scores of 0.901, 0.815 and 0.766 for whole tumor, tumor core and enhancing tumorcore respectively on the BraTS2019 validation data set. Our method has the better performance than the original3D U-Net. The results of our experiment demonstrate that compared with some state-of-the-art methods, ourapproach is a competitive automatic brain tumor segmentation method.

Residual tumor and central lymph node metastasis after thermal ablation of papillary thyroid carcinoma: A case report and review of literature

BACKGROUND Debate exists regarding the use of thermal ablation(TA)to treat papillary thyroid carcinoma(PTC).Some studies have recommended TA as a new,ef
cient and safe technology for PTC.In this article,we report one case of a residual tumor and central lymph node metastasis(CLNM)after TA for PTC.CASE SUMMARY A 63-year-old female underwent bilateral ultrasound(US)-guided radiofrequency ablation for PTC.Three months later,she was diagnosed as thyroid cancer with suspected CLNM by US and contrast-enhanced computed tomography.The subsequent fine-needle aspiration(FNA)biopsies were negative.Due to her strong personal preference,she underwent total thyroidectomy and central lymph node dissection.Local tissue adhesion and a difficult dissection were noted during the operation.The pathology of the frozen sections during the operation was still negative.The final pathology results of paraffin-embedded sections revealed residual tumor cells at the edge of the PTC and CLNM.CONCLUSION TA may lead to a residual tumor in patients with PTC.Follow-up using US and FNA biopsy may not be adequate to evaluate the residual tumor.TA should be carefully considered in PTC treatment.

Prognostic relevance of minimal residual disease in colorectal cancer

Presence of occult minimal residual disease in patients with colorectal cancer(CRC)has a strong prognostic impact on survival.Minimal residual disease plays a major role in disease relapse and formation of metastases in CRC.Analysis of circulating tumor cells(CTC)in the blood is increasingly used in clinical practice for disease monitoring of CRC patients.In this review article the role of CTC,disseminated tumor cells(DTC)in the bone marrow and micrometastases and isolated tumor cells(ITC)in the lymph nodes will be discussed,including literature published until September 2013.Occult disease is a strong prognostic marker for patient survival in CRC and defined by the presence of CTC in the blood,DTC in the bone marrow and/or micrometastases and ITC in the lymph nodes.Minimal residual disease could be used in the future to identify patient groups at risk,who might benefit from individualized treatment options.

Coexisting opportunities and challenges:In which scenarios can minimal/measurable residual disease play a role in advanced non-small cell lung cancer?

Curative therapy was not previously available for patients with advanced non-small cell lung cancer(NSCLC);thus,the concept of minimal/measurable(or molecular)residual disease(MRD)was not applicable to these patients.However,advances in targeted and immunotherapy have revolutionized the treatment landscape for patients with advanced NSCLC,with emerging evidence of long-term survival and even the hope of complete remission(CR)by imaging examination.The latest research shows that patients with oligometastatic lung cancer can benefit from local treatment.After removing the lesions,the choice of follow-up therapy and monitoring of the lesions could remain uncertain.MRD plays a role in identifying early-stage NSCLC patients with high risks of recurrence and determining adjuvant therapy after radical treatment.In recent years,evidence has been accumulating regarding the use of circulating cell-free tumor DNA(ctDNA)to assess MRD in solid tumors.This study discussed the possible applications of ctDNA-based MRD monitoring in advanced NSCLC and described the current challenges and unresolved problems in the application of MRD in advanced NSCLC.

Why is endosonography insufficient for residual diagnosis after neoadjuvant therapy for esophageal cancer?Solutions using muscle layer evaluation

BACKGROUND The diagnosis of residual tumors using endoscopic ultrasound(EUS)after neoadjuvant therapy for esophageal cancer is considered challenging.However,the reasons for this difficulty are not well understood.AIM To investigate the ultrasound imaging features of residual tumors and identify the limitations and potential of EUS.METHODS This exploratory prospective observational study enrolled 23 esophageal squamous cell carcinoma patients receiving esophagectomy after neoadjuvant therapy[15 patients after neoadjuvant chemotherapy(NAC)and 8 patients after chemoradiotherapy(CRT)]at the Department of Surgery,Chiba University Hospital,between May 2020 and October 2021.We diagnosed the T stage for specimens using ultrasound just after surgery and compared ultrasound images with the cut surface of the fixed specimens of the same level of residual tumor.The ratio of esophageal muscle layer defect measured by ultrasound was compared with clinicopathological factors.Furthermore,the rate of reduction for the muscle layer defect was evaluated using EUS images obtained before and after neoadjuvant therapy.RESULTS The accuracy of T stage rate was 61%(n=14/23),which worsened after CRT(38%,n=3/8)than after NAC(73%,n=11/15)because of overstaging.Moreover,pT0 could not be diagnosed in all cases.The detection rate of residual tumor for specimens using ultrasound retrospectively was 75%(n=15/20).There was no correlation between after-NAC(79%,n=11/14)and after-CRT(67%,n=4/6)detection rate.The detection of superficial and submucosal types was poor.The pathologic tumor size and pathological response were correlated.Tumor borders were irregular and echogenicity was mixed type after CRT.There was a correlation between the pT stage(pT0/1 vs pT2/3)and the length of muscle layer circumference(P=0.025),the length of muscle layer defect(P<0.001),and the ratio of muscle layer defect(P<0.001).There was also a correlation between the pT stage and the rate of muscle layer defect reduction measured by EUS(P=0.001).CONCLUSION Compared to pathological images,some tumors are undetectable by ultrasound.Focusing on the esophageal muscle layer might help diagnose the depth of the residual tumor.

融合CNN与Transformer的MRI脑肿瘤图像分割

为解决卷积神经网络(convolutional neural network,CNN)在学习全局上下文信息和边缘细节方面受到很大限制的问题,提出一种同时学习局语义信息和局部空间细节的级联神经网络用于脑肿瘤医学图像分割。首先将输入体素分别送入CNN和Transformer分支,在编码阶段结束后,采用一种双分支融合模块将2个分支学习到的特征有效地结合起来以实现全局信息与局部信息的融合。双分支融合模块利用哈达玛积对双分支特征之间的细粒度交互进行建模,同时使用多重注意力机制充分提取特征图通道和空间信息并抑制无效的噪声信息。在BraTS竞赛官网评估了本文方法,在BraTS2019验证集上增强型肿瘤区、全肿瘤区和肿瘤核心区的Dice分数分别为77.92%,89.20%和81.20%。相较于其他先进的三维医学图像分割方法,本文方法表现出了更好的分割性能,为临床医生做出准确的脑肿瘤细胞评估和治疗方案提供了可靠依据。

血清TNF-α、IL-17表达预测腰椎间盘突出症患者术后残留疼痛的价值

目的探讨血清肿瘤坏死因子(TNF-α)、白细胞介素-17(IL-17)表达预测腰椎间盘突出症患者术后残留疼痛的价值。方法回顾性收集2020年1月至2023年1月郑州大学第二附属医院收治的165例腰椎间盘突出症患者的临床资料,根据患者术后6个月时残留疼痛情况分为残留疼痛组(19例)和未残留疼痛组(146例)。术前检测患者血清TNF-α、IL-17水平,分析血清TNF-α、IL-17对腰椎间盘突出症患者术后残留疼痛的预测价值。结果两组纤维环损伤、TNF-α、IL-17水平差异有统计学意义(P<0.05);经logistic回归分析,结果显示,血清TNF-α、IL-17水平、纤维环损伤是腰椎间盘突出症患者术后残留疼痛发生的危险因素(OR>1,P<0.05);绘制受试者工作特征(ROC)曲线,结果显示,血清TNF-α、IL-17以及联合预测腰椎间盘突出症患者术后残留疼痛的曲线下面积(AUC)均>0.7,具有一定预测价值,且联合预测价值更高。结论血清TNF-α、IL-17高表达是腰椎间盘突出症患者术后残留疼痛的影响因素,且具有一定预测价值。

肿瘤检测个体化和精准化的探索与思考——评早期肺癌头对头对比MRD检测策略的MEDAL研究

1文献来源Chen K,Yang F,Shen H,et al.Individualized tumor⁃informed circulating tumor DNA analysis for postoperative monitoring of non⁃small cell lung cancer[J].Cancer Cell,2023,41(10):1749-1762.e6.

LRAE-Unet:轻量级MRI脑肿瘤全自动分割网络

提出一种轻量级脑肿瘤全自动分割网络,即轻量级残差注意力增强网络(LRAE-Unet)。首先采用轻量级残差模块解决网络层数增加时出现的梯度消失和网络退化问题;其次采用轻量级自注意力模块抑制输入图像中的不相关区域,同时突出特定局部区域的显著特征;最后通过增强视野平均池化模块减少特征图的空间,节省计算资源,控制网络过拟合现象。在BraTS 2019数据集的测试结果显示LRAE-Unet在完整肿瘤、肿瘤核心与增强肿瘤区域的Dice相似系数为91.24%、88.64%与88.32%,证明使用LRAE-Unet进行脑瘤分割具有可行性与有效性。

基于有效感受野和注意力融合机制的脑肿瘤全自动分割

深度学习在医学图像分割领域取得了显著成果,但其在脑肿瘤分割任务中,仍面临感受野不足、冗余信息过多、信息丢失等问题;为此,本研究提出一种基于编-解码结构的脑肿瘤分割网络模型(EAU-Net)。EAU-Net采用有效感受野拓展模块和注意力融合模块改善脑肿瘤分割网络感受野不足与冗余信息过多带来的不利影响;同时,引入基于倒残差结构的瓶颈重采样模块,有效避免上下采样时造成的信息损失,并采用深度卷积降低网络的计算量。在BraTS2020数据集上的实验结果表明,EAU-Net获得最优的分割精度,验证了其在脑肿瘤分割任务中的可行性和有效性。

循环肿瘤DNA液态活检检测微小残留病在实体肿瘤诊疗中的研究进展

基于液态活检的微小残留病(minimal residual disease,MRD)检测在各肿瘤疾病的应用中具有巨大的潜力,随着基因测序技术和生物信息学的发展,二代测序(next-generation sequencing,NGS)作为一种新兴的MRD检测技术,可以对循环肿瘤DNA(circulation tumor DNA,ctDNA)进行定量和定性分析,目前在多种实体瘤中的应用均具有高灵敏度和高特异度。多项研究已证实基于ctDNA的MRD检测结果与患者复发预测、预后判断、疗效评估及个体化分层治疗均有临床相关性,且相较于传统临床手段优势明显。但目前ctDNA MRD的检测技术应用于常规临床实践仍需要大量临床数据的支持,检测方法的统一标准仍未建立。

一种结合注意力残差的肝脏及肝肿瘤分割算法

长时间的肝脏医学图像人工诊断容易使医生产生疲劳,导致误诊和漏诊情况发生。针对以上现象提出一种改进的Unet网络用于肝脏和肝肿瘤自动分割。改进Unet模型,引入注意力残差结构和特征复用结构,提高输入图像中特征信息的利用效率;对损失函数进行改进,在Dice系数中加入欠分割和过分割惩罚因子,提高模型的预测能力。在公开数据集上的实验结果表明:该算法对肝脏和肝肿瘤的分割相似系数分别达到了0.962和0.713,优于现有的分割模型且具有较强的鲁棒性。

动态监测ctDNA⁃MRD预测局部晚期NSCLC放化疗疗效

1文献来源Pan Y,Zhang JT,Gao X,et al.Dynamic circulating tumor DNA during chemoradiotherapy predicts clinical outcomes for locally advanced non⁃small cell lung cancer patients[J].Cancer Cell,2023,41(10):1763-1773.e4.2证据水平2b。

血清miR-155、miR-24检测对宫颈癌同步放化疗后肿瘤残留的早期预测价值

目的分析血清miR-155、miR-24表达水平对宫颈癌同步放化疗(CCRT)后肿瘤残留的早期预测价值。方法回顾性选取行CCRT治疗的宫颈癌患者300例为研究对象,按照7∶3的分配比例分为建模组210例和验证组90例。建模组患者根据术后MRI结果分为残留组81例与非残留组129例。收集建模组患者CCRT前临床基线资料与实验室检验指标,筛选独立影响因素,构建宫颈癌CCRT后早期肿瘤残留列线图预测模型,并通过验证组资料收集配合完成预测模型的验证与价值分析。结果宫颈癌患者血清miR-155、宫旁浸润、FIGO分期是影响患者CCRT后早期肿瘤残留的独立危险因素,血清miR-24是影响患者CCRT后早期肿瘤残留的独立保护因素(OR分别=1.77、3.22、8.55、0.18,P均<0.05)。基于以上独立影响因素构建了宫颈癌CCRT后早期肿瘤残留预测模型。建模组ROC曲线下面积(AUC)为0.84(95%CI:0.79~0.89),区分度良好;验证组ROC曲线AUC为0.90(95%CI:0.81~0.98)。内、外部校准曲线与标准曲线均有较好贴合度。内、外部决策曲线显示模型能提供显著的临床净收益。结论血清miR-155、miR-24表达水平对宫颈癌CCRT后早期肿瘤残留具有良好的预测价值,结合宫旁浸润、FIGO分期等临床特征构建列线图预测模型可为早期肿瘤残留高风险患者的有效筛选提供数据支持。

鼻咽癌放疗后残留的研究进展

鼻咽癌(nasopharyngeal carcinoma,NPC)是我国常见的头颈部恶性肿瘤,在我国华南地区较为高发,尤其是广东省,又被称为“广东癌”。由于鼻咽特殊的解剖位置及良好的放射敏感性,使得放射治疗成为鼻咽癌的主要治疗方式。然而,患者在鼻咽癌放疗结束后经常会出现肿瘤残留。目前对于肿瘤残留的诊断、治疗以及肿瘤残留对预后的影响尚存在很多争议。对于肿瘤残留的诊断,有多种方式,如各种不同的影像学、病理组织学活检等,我们应充分发挥各种检查手段的优势、提高诊断的准确性。另一方面,需及时、有效地判定肿瘤残留,避免错过最佳干预时机。针对残留病灶的处理,需要结合病灶的位置、大小、浸润范围等来制定治疗策略。既往的多个研究表明,肿瘤残留患者预后更差。本文就鼻咽癌放疗后肿瘤残留的有关问题研究做一综述。

基于联邦学习与改进CBAM-ResNet18的脑肿瘤分类

针对联邦学习框架下,基于卷积注意力模块的多客户端脑肿瘤分类方法对于MRI图像中肿瘤区域细节提取能力不足、通道注意力与空间注意力相互干扰的问题,以及针对多点医疗肿瘤数据分类准确性低的问题,提出了一种融合联邦学习框架和改进的CBAM-ResNet18网络的脑肿瘤分类方法.利用联邦学习特性联合多点脑肿瘤数据,采用Leaky ReLU激活函数代替ReLU激活函数以减轻神经元死亡,将卷积注意力模块中的通道注意力模块由先降维再升维改成先升维再降维,充分提高网络对图像细节的提取能力,将卷积注意力模块中的通道注意力模块与空间注意力模块由级联结构改为并联结构,使得网络的特征提取能力不会受到二者先后顺序的影响.通过在Kaggle公开的脑肿瘤MRI数据集上的进行实验,该方法的准确率、精准度、召回率与F1值分别为97.78%、97.68%、97.61%与97.63%,比基准模型分别高6.54%、4.78%、6.80%、7.00%.实验结果证明,该方法不仅能够打破数据孤岛,实现多点数据融合,而且比多数现有主流模型的性能更好.

切缘肿瘤细胞残留对局部进展期胃癌复发模式和远期预后的影响

目的探究切缘肿瘤细胞残留(R1切除)局部进展期胃癌辅助化疗后的复发模式及远期预后。方法收集440例经D2根治性切除的局部进展期胃癌患者。比较R0切除(414例)和R1切除(26例)患者的临床特征、复发转移模式及预后。结果倾向性匹配前,R0切除和R1切除局部进展期胃癌患者N分期、脉管侵犯情况比较,差异均有统计学意义(P﹤0.05);倾向性匹配后,各临床特征比较,差异均无统计学意义(P﹥0.05)。R0切除和R1切除局部进展期胃癌患者复发模式比较,差异无统计学意义(P﹥0.05)。倾向性匹配前,R0切除和R1切除患者总生存情况和无复发生存情况比较,差异均有统计学意义(P﹤0.01)。倾向性匹配后,R0切除和R1切除患者总生存情况和无复发生存情况比较,差异均无统计学意义(P﹥0.05)。倾向性匹配后,不同化疗方案R1切除局部进展期胃癌患者的生存情况比较,差异无统计学意义(P﹥0.05)。结论R1切除局部进展期胃癌患者通常伴随着更晚的分期、更差的生物学行为和预后。R1切除胃癌患者术后辅助治疗应以局部治疗不影响全身系统性药物治疗为原则。