合成磁共振预测局部进展期直肠癌患者行新辅助放化疗疗效的价值

摘要/Abstract

摘要：

目的：探索合成MRI序列预测局部进展期直肠癌（locally advanced rectal cancer， LARC）患者接受新辅助放化疗（neoadjuvant chemoradiotherapy treatment， nCRT）疗效的效能。方法：收集2023年8月至2024年6月就诊于开云网页登录医学院附属瑞金医院的经活检证实为直肠腺癌的51例患者，所有患者基线MRI评估为LARC，且接受nCRT治疗联合根治手术治疗。受试者在接受nCRT治疗前2周内完成合成MRI扫描以及高分辨率T2WI序列扫描。由放射科医师基于高分辨率T2WI图像，评估受试者基线状态下壁外血管侵犯（extramural vascular invasion， mrEMVI）；在合成MRI序列扫描完成后，通过Synthetic MR后处理软件生成T1 Mapping、T2 Mapping以及PD Mapping的合成图像，并采用python软件提取受试者基线状态下直方图特征量化参数，包括肿瘤原发灶及瘤周脂肪量化参数［T1弛豫时间（T1 relaxation time， T1RT）、T2弛豫时间（T2 relaxation time， T2RT）、质子密度（proton density， PD）］。以手术病理结果为金标准，将受试者分别［按原发灶病理缓解状态，分为病理完全缓解（pathological complete response， pCR）组和非pCR（non-pCR）组2组；按原发灶肿瘤退缩分级（tumor regression grade，TRG）分为TRG 0-1级组和TRG 2-3级组2组；按系膜淋巴结转移状态分为淋巴结转移阳性组和淋巴结转移阴性组2组。采用Student'st检验、Mann-WhitneyU检验和Chi-Square检验，比较以上分组间患者基线状态下mrEMVI状态的差异性以及肿瘤原发灶、瘤周脂肪的量化参数差异，通过二元逻辑回归筛选预测TRG分级、pCR状态以及系膜淋巴结状态的独立危险因素，并基于筛选的危险因素，建立逻辑回归模型，采用受试者操作特性曲线（receiver operating characteristic， ROC）评价量化参数、mrEMVI状态和回归模型预测TRG分级、pCR状态以及系膜淋巴结状态的能力。结果：基线mrEMVI（P=0.03）阳性、瘤周脂肪组织定量参数［T2RT_Fat的最大值（139.53 ms比129.60 ms，P=0.03）、90%分位数（P₉₀）（189.18 ms比174.00 ms，P=0.03）和均方根（120.09 ms比115.48 ms，P=0.04），更低的T2RT_Fat均匀性（0.54比0.61，P=0.04）］向提示nCRT治疗后的淋巴结转移阳性状态。所有观察指标与原发灶无相关性。Logistic回归分析显示，mrEMVI与升高的T2RT_Fat_P₉₀是预测系膜淋巴结的独立危险因素。mrEMVI（AUC=0.667）联合T2RT_Fat_P₉₀（AUC=0.692）构建的逻辑回归模型表现出良好的预测效能（AUC=0.747）吗，但差异无统计学意义。结论：基于MAGiC提取的基线瘤周脂肪量化参数T2RT_Fat_P₉₀是预测nCRT后系膜淋巴结转移的无创性影像学标志物，T2RT_Fat_P₉₀结合基线mrEMVI可以作为预测LARC患者nCRT后系膜淋巴结转移状态的辅助手段。

关键词:直肠癌,淋巴结转移,磁共振成像,新辅助放化疗

Abstract:

ObjectiveTo explore the effectiveness of synthetic MRI sequences in predicting the treatment response of patients with locally advanced rectal cancer (LARC) undergoing neoadjuvant chemoradiotherapy treatment (nCRT).MethodsA total of 51 patients with biopsy-confirmed rectal adenocarcinoma were enrolled at Ruijin Hospital from August 2023 to June 2024. All patients were assessed as having LARC by baseline MRI and received nCRT followed by radical surgery. All subjects completed synthetic MRI and high-resolution T2-weighted imaging (T2WI) scans within two weeks before receiving nCRT treatment. Based on high-resolution T2WI images, radiologists assessed extramural vascular invasion (mrEMVI) at baseline in the subjects. After the synthetic MRI sequence scanning was completed, synthetic images of T1 mapping, T2 mapping, and proton density (PD) mapping were generated using Synthetic MR post-processing software. Histogram-based quantitative parameters at baseline were extracted using python software, including quantitative parame-ters of the primary tumor and peritumoral fat region: T1 relaxation time (T1RT), T2 relaxation time (T2RT), and proton density (PD). Using postoperative pathological results as the gold standard, patients were grouped according to: (1) primary tumor response: pathological complete response (pCR) vs. non-pCR; (2) tumor regression grade: (TRG) 0-1vs.TRG 2-3; and (3) mesorectal lymph node metastasis status: positive (ypN+) vs. negative (ypN-). Differences in baseline mrEMVI status and quantitative parameters of the primary tumor and peritumoral fat among different groups were compared using Student's t-test, Mann-WhitneyUtest, and Chi-square test. Binary logistic regression was used to identify independent risk factors for predicting TRG grade, pCR status, and mesorectal lymph node status. Based on the selected risk factors, logistic regression models were established. The predictive performance of the quantitative parameters, mrEMVI status, and the regression models for TRG grade, pCR status, and mesorectal lymph node status was evaluated using receiver operating characteristic (ROC) curves.ResultsBaseline mrEMVI positivity (P=0.03) and quantitative parameters of peritumoral fat tissue-including the maximum T2RT_Fat (139.53 msvs.129.60 ms,P=0.03), 90th percentile (189.18 msvs. 174.00 ms,P=0.03), root mean square (120.09 msvs. 115.48 ms,P=0.04), and lower T2RT_Fat uniformity (0.54vs. 0.61,P=0.04)—were indicative of positive mesorectal lymph node status after nCRT. None of the observed indicators were correlated with the primary tumor response. Logistic regression analysis showed that mrEMVI and elevated T2RT_Fat_P₉₀were independent risk factors for predicting mesorectal lymph node metastasis. The logistic regression model combining both mrEMVI (AUC=0.667) and T2RT_Fat_P₉₀(AUC=0.692) demonstrated good predictive performance (AUC=0.747), although the improvement was not statistically significant.ConclusionsT2RT_Fat_P₉₀, extracted from baseline MAGiC synthetic MRI serves as a non-invasive imaging biomarker for predicting mesorectal lymph node metastasis after nCRT. The combination of T2RT_Fat_P₉₀and baseline mrEMVI can be used as an auxiliary tool for predicting the mesorectal lymph node metastasis status in LARC patients following nCRT.

Key words:Rectal cancer,Lymph node metastasis,Magnetic resonance imaging,Neoadjuvant chemoradiotherapy

中图分类号:

R445.2

王康宁, 朱兰, 冯威铭, 夏益涵, 石博文, 张欢. 合成磁共振预测局部进展期直肠癌患者行新辅助放化疗疗效的价值[J]. 诊断学理论与实践, 2025, 24(02): 170-177.

WANG Kangning, ZHU Lan, FENG Weiming, XIA Yihan, SHI Bowen, ZHANG Huan. Value of synthetic MRI in predicting treatment response to neoadjuvant chemoradiotherapy in patients with locally advanced rectal cancer[J]. Journal of Diagnostics Concepts & Practice, 2025, 24(02): 170-177.

图/表6

表1

患者一般特征

Parameter	Value
Age (y)	60.08(±10.4)
Gender
Male	37
Female	14
Baseline mrEMVI
negative	27
positive	24
Baseline mrT stage
2	8
3	22
4a	13
4b	8
Baseline mrN stage
0	9
1	22
2	20
TRG grade
0	6
1	9
2	19
3	17
ypT
0	6
1	1
2	13
3	28
4a	3
ypN
0	33
1	12
2	6

表1

图1

表2

各量化参数一致性检验

	ROI	ICC	P₁₀	P₉₀	Entropy	IQR	Mean	Median	Kurtosis	Maximum	MAD	Min	RMS	RMAD	Uniformity
T1RT	Fat	intra	0.990	0.788	0.950	0.953	0.949	0.984	0.763	0.801	0.831	0.901	0.897	0.937	0.958
		inter	0.971	0.840	0.927	0.865	0.918	0.946	0.658	0.643	0.807	0.751	0.874	0.866	0.949
	Tumor	intra	0.987	0.991	0.972	0.987	0.991	0.993	0.931	0.934	0.986	0.833	0.990	0.994	0.967
		inter	0.857	0.982	0.928	0.983	0.976	0.982	0.827	0.867	0.970	0.589	0.978	0.990	0.936
T2RT	Fat	intra	0.935	0.982	0.894	0.948	0.831	0.972	0.991	0.927	0.968	0.690	0.973	0.958	0.958
		inter	0.901	0.904	0.894	0.905	0.863	0.958	0.977	0.905	0.930	0.618	0.913	0.920	0.939
	Tumor	intra	0.987	0.990	0.973	0.965	0.993	0.995	0.844	0.839	0.971	0.946	0.993	0.974	0.977
		inter	0.954	0.948	0.833	0.903	0.969	0.977	0.524	0.721	0.872	0.793	0.966	0.906	0.902
PD	Fat	intra	0.951	0.984	0.894	0.927	0.984	0.984	0.851	0.956	0.911	0.797	0.985	0.920	0.900
		inter	0.959	0.985	0.901	0.922	0.985	0.976	0.768	0.932	0.915	0.613	0.987	0.925	0.906
	Tumor	intra	0.994	0.996	0.982	0.977	0.997	0.997	0.774	0.954	0.977	0.917	0.997	0.988	0.986
		inter	0.982	0.978	0.906	0.911	0.987	0.988	0.424	0.945	0.911	0.688	0.987	0.940	0.921

表2

Quantitative parameter	Post-CRT pN stage	P
T2RT_Fat_P₉₀	129.60 (125.50, 135.00)	139.53 (131.28, 144.48)	0.025
T2RT_Fat_RMS	115.48 ± 7.16	120.09 ± 7.22	0.033
T2RT_Fat_Maximum	174.00 (150.70, 188.20)	189.18 (165.71, 230.32)	0.036
T2RT_Fat_Uniformity	0.61 ± 0.12	0.54 ± 0.10	0.035

表3

Quantitative parameter	AUC	Sensitivity	Specificity	PPV	NPV	Cutoff
T2RT_Fat_Maximum	0.680	0.944	0.364	0.447	0.923	154.30
T2RT_Fat_Uniformity	0.673	0.889	0.121	0.356	0.667	0.437
T2RT_Fat_P₉₀	0.692	0.778	0.606	0.519	0.833	131.15
T2RT_Fat_RMS	0.678	0.611	0.788	0.611	0.788	119.95
mrEMVI	0.667	0.667	0.667	0.522	0.786	/
Model	0.747	0.778	0.667	0.560	0.846	0.32

表4

Quantitative parameter	T2RT_Fat_P₉₀	mrEMVI	Model
T2RT_Fat_P₉₀	\	0.808	0.260
mrEMVI	0.808	\	0.710
Model	0.260	0.710	\

表5

参考文献26

[1]	SUNG H, FERLAY J, SIEGEL R L, et al. Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J].CA Cancer J Clin,2021,71(3): 209-249.
[2]	范伯男, 李岩. 全球主要疾病负担及健康状况趋势分析——1990年至2021年全球疾病和伤害负担报告解读 [J].诊断学理论与实践,2024,23(5):474-483.
	FAN B, LI Y. Trends in global major disease burden and health conditions—interpretation of the Global Burden of Disease Study 1990-2021[J].J Diagn Concepts Pract,2024,23(5):474-483.
[3]	BENSON A B, VENOOK A P, ADAM M, et al. Colon Cancer, Version 3.2024, NCCN Clinical Practice Guidelines in Oncology [J].J Natl Compr Canc Netw,2024,22(2 D).
[4]	HABR-GAMA A, PEREZ R O, NADALIN W, et al. Operative versus nonoperative treatment for stage 0 distal rectal cancer following chemoradiation therapy: long-term results[J].Ann Surg,2004,240(4):711-717; discussion 7-8.
[5]	MAAS M, NELEMANS P J, VALENTINI V, et al. Long-term outcome in patients with a pathological complete response after chemoradiation for rectal cancer: a pooled analysis of individual patient data[J].Lancet Oncol,2010,11(9):835-44. doi:10.1016/S1470-2045(10)70172-8pmid:20692872
[6]	PAHLMAN L, BOHE M, CEDERMARK B, et al. The Swedish rectal cancer registry[J].Br J Surg,2007,94(10):1285-1292. pmid:17661309
[7]	TULCHINSKY H, SHMUELI E, FIGER A, et al. An interval >7 weeks between neoadjuvant therapy and surgery improves pathologic complete response and disease-free survival in patients with locally advanced rectal cancer [J].Ann Surg Oncol,2008,15(10):2661-2667.
[8]	VAN DER PAARDT M P, ZAGERS M B, BEETS-TAN R G, et al. Patients who undergo preoperative chemoradiotherapy for locally advanced rectal cancer restaged by using diagnostic MR imaging: a systematic review and meta-analysis [J].Radiology,2013,269(1):101-112. doi:10.1148/radiol.13122833pmid:23801777
[9]	SEO N, KIM H, CHO M S, et al. Response assessment with MRI after chemoradiotherapy in rectal cancer: current evidences[J].Korean J Radiol,2019,20(7):1003-1018. doi:10.3348/kjr.2018.0611pmid:31270972
[10]	ZHANG S, YU M, CHEN D, et al. Role of MRI-based radiomics in locally advanced rectal cancer (Review)[J].Oncol Rep,2022,47(2):34.
[11]	RAFAELSEN S R, VAGN-HANSEN C, SORENSEN T, et al. Ultrasound elastography in patients with rectal cancer treated with chemoradiation[J].Eur J Radiol,2013,82(6):913-917. doi:10.1016/j.ejrad.2012.12.030pmid:23410908
[12]	HAGIWARA A, FUJIMOTO K, KAMAGATA K, et al. Age-related changes in relaxation times, proton density, myelin, and tissue volumes in adult brain analyzed by 2-dimensional quantitative synthetic magnetic resonance imaging[J].Invest Radiol,2021,56(3):163-172.
[13]	WARNTJES J B, LEINHARD O D, WEST J, et al. Rapid magnetic resonance quantification on the brain: optimization for clinical usage[J].Magn Reson Med,2008,60(2): 320-329. doi:10.1002/mrm.21635pmid:18666127
[14]	KUMAR N M, FRITZ B, STERN S E, et al. Synthetic MRI of the knee: phantom validation and comparison with conventional MRI[J].Radiology,2018,289(2):465-477. doi:10.1148/radiol.2018173007pmid:30152739
[15]	CUI Y, HAN S, LIU M, et al. Diagnosis and grading of prostate cancer by relaxation maps from synthetic MRI [J].J Magn Reson Imaging,2020,52(2):552-564. doi:10.1002/jmri.27075pmid:32027071
[16]	LIAN S, LIU H, MENG T, et al. Quantitative synthetic MRI for predicting locally advanced rectal cancer response to neoadjuvant chemoradiotherapy[J].Eur Radiol,2023,33(3):1737-1745.
[17]	SMITH N J, SHIHAB O, ARNAOUT A, et al. MRI for detection of extramural vascular invasion in rectal cancer[J].Am J Roentgenol,2008,191(5):1517-1522. doi:10.2214/AJR.08.1298pmid:18941094
[18]	ZHANG L N, XIAO W W, XI S Y, et al. Pathological Assessment of the AJCC tumor regression grading system after preoperative chemoradiotherapy for chinese locally advanced rectal cancer[J].Medicine (Baltimore),2016,95(3):e2272.
[19]	RINGEL A E, DRIJVERS J M, BAKER G J, et al. Obesity shapes metabolism in the tumor microenvironment to suppress anti-tumor immunity[J].Cell,2020,183(7):1848-1866. doi:10.1016/j.cell.2020.11.009pmid:33301708
[20]	JENSEN M D. Visceral fat: culprit or canary? [J].Endocrinol Metab Clin North Am,2020,49(2):229-237.
[21]	LINO-SILVA L S, SALCEDO-HERNANDEZ R A, GAMBOA-DOMINGUEZ A. Tumour budding in rectal cancer. A comprehensive review[J].Contemp Oncol (Pozn),2018,22(2):61-74.
[22]	多中心直肠癌真实世界数据库建设与数据质量控制策略[J].中华消化外科杂志,2025,24(1):77-81.
	Multicenter rectal cancer real-world database construction and data quality control strategies[J].Chin Dig Surg,2025,24(1):77-81.
[23]	李珂璇, 肖体先, 汪晓东, 等. 中低位直肠癌初诊及新辅助治疗后评估完成度分析：全国多中心真实世界研究[J].中华消化外科杂志,2025,24(1):113-119.
	LI K X, XIAO T X, WANG X D, et al. Analysis of completion rate of tumor evaluation at initial assessment and after neoadjuvant therapy for mid and low rectal cancer : a national multicenter real-world study[J].Chin J Dig Surg,2025,24(1):113-119.
[24]	毛翌皓, 冯青阳, 许剑民. 人工智能时代机器人结直肠癌手术的现状及进展[J].中华消化外科杂志,2024,23(4): 573-578.
	MAO Y H, FENG Q Y, XU J M. Current status and progress of robotic colorectal cancer surgery in the era of artificial intelli-gence[J].Chin J Dig Surg,2024,23(4):573-578.
[25]	SHIROUZU K, ISOMOTO H, KAKEGAWA T, et al. A prospective clinicopathologic study of venous invasion in colorectal cancer[J].Am J Surg,1991,162(3):216-222. pmid:1928581
[26]	KRASNA M J, FLANCBAUM L, CODY R P, et al. Vascular and neural invasion in colorectal carcinoma. Incidence and prognostic significance[J].Cancer,1988,61(5):1018-1023. pmid:3338045