Application of radiomics-based multiomics combinations in the tumor microenvironment and cancer prognosis

Kang, Wendi; Qiu, Xiang; Luo, Yingen; Luo, Jianwei; Liu, Yang; Xi, Junqing; Li, Xiao; Yang, Zhengqiang

doi:10.1186/s12967-023-04437-4

Journal of Translational Medicine

Table 1 Radiomics study for predictive evaluation of tumor-infiltrating lymphocytes

From: Application of radiomics-based multiomics combinations in the tumor microenvironment and cancer prognosis

Tumor type	Sample size	TILs evaluation	TILs evaluation method	Imaging modality	Feature selection	Model	Training set performance	Validation set performance	References
N/A	254	CD8⁺TILs	RNA-seq	CT	Elastic-net	Linear regression	AUC 0.74	AUC 0.67	[43]
UPS	14	CD8⁺TILs	RNA-seq	MRI	N/A	N/A	ACC 93%	N/A	[54]
TNBC	139	TILs	RNA-seq	MRI	Elastic net	LR	AUC 0.868	AUC 0.790	[46]
TNBC	43	TILs	IHC	Mammography	Mann–Whitney U-test/PCC	N/A	N/A	N/A	[55]
TNBC	80	TILs	HE	MRI	N/A	N/A	AUC 0.752	N/A	[51]
BC	133	TILs	HE	MRI	LASSO	LR	AUC 0.934	AUC 0.872	[47]
BC	172	TILs	IHC	MRI	LASSO	Linear regression	AUC 0.742	AUC 0.718	[56]
BC	121	TILs	HE	Mammography	RFE	LR	AUC 0.83	AUC 0.79	[57]
BC	154	TILs	NA	MRI	LASS0	LR	AUC 0.86	AUC 0.83	[58]
RC	141	CD8⁺TILs	IHC	MRI	LASSO	Linear regression	AUC 0.760	AUC 0.729	[42]
RC	133	T cells	IHC	MRI	GBDT	LR	AUC 0.770	AUC 0.768	[59]
CCLM	103	T cells	IHC	CT	N/A	SVM	N/A	N/A	[60]
HGG	51	T cells	FCM	MRI	sPLS	sPLS-DA	AUC 0.986	N/A	[61]
LGG	107	B/T cells	RNA-seq	MRI	LASSO	COX	R correlation coefficient		[62]
							0.975 (B cell)	0.429
							0.474 (CD8 T cell)	0.552
NSCLC	100	TILs	IHC	CT	Mann–Whitney U	Cox model	AUC 0.91	N/A	[52]
NSCLC	103	CD8⁺TILs	IHC	PET/CT	LASSO	LR	AUC 0.800	AUC 0.794	[63]
NSCLC	290	TILs	IHC	CT	N/A	COX	N/A	N/A	[64]
NSCLC	117	CD8⁺TILs	IHC	CT	LASSO	N/A	AUC 0.83	AUC 0.68	[65]
NSCLC	60	TILs	IHC	CT	PCA	N/A	N/A	N/A	[66]
NSCLC	149	CD8⁺T cells	RNA-seq	CT	N/A	RF	AUC 0.681 (RF)	N/A	[67]
						LDA	AUC 0.674 (LDA)
						CART	AUC 0.647 (CART)
NSCLC	97	CD8⁺TILs	FACS	CT	N/A	Neural network	0.788	0.753	[68]
NSCLC	91	CD8⁺TILs	IHC	PET-CT	LASSO	LR	0.818	N/A	[69]
NSCLC	44	CD8⁺TILs	IHC	PET	N/A	COX	0.9	N/A	[70]
NSCLC	221	CD8⁺T cells	IHC	PET/CT	LASSO	LR	0.907	0.883	[71]
PDAC	184	CD8⁺TILs	IHC	CT	LASSO	XGBoost	AUC 0.75	AUC 0.67	[48]
PDAC	156	CD20B cells	IHC	MRI	LASSO	LR	AUC 0.79	AUC 0.79	[72]
PDAC	114	CD8⁺T cells	IHC	MRI	LASSO	LDA	AUC 0.85	AUC 0.76	[73]
PDAC	183	TILs	IHC	CT	NA	XGBoost	AUC 0.93	AUC 0.79	[74]
PDAC	156	TILs	IHC	MRI	LASSO	XGBoost	AUC 0.86	AUC 0.79	[49]
HCC	142	CD8⁺T cells	IHC	CT	Elastic-net	linear regression	AUC 0.751	0.705	[50]
HCC	207	T cells	IHC	MRI	Randomized tree	Randomized tree	AUC 0.904	AUC 0.823	[75]
HNSCC	160	CD8⁺T cells	RNA-seq	CT	Consensus clustering	RF	ACC 65.7%	N/A	[76]
HNSCC	71	CD8⁺T cells	IHC	CT	LASSO	LR	0.786	N/A	[77]
ESCC	220	CD8⁺T cells	IHC	CT	LASSO	LR	0.764	0.728	[78]

UPS undifferentiated pleomorphic sarcomas, HGG high-grade gliomas, TNBC triple-negative breast cancer, BC breast cancer, NSCLC non-small cell lung cancer, PDAC pancreatic ductal adenocarcinoma, HCC hepatocellular carcinoma, LGG lower-grade gliomas, CCLM colorectal cancer lung metastasis, HNSCC head and neck squamous cell carcinoma, ESCC esophageal squamous cell carcinoma, CART classification and regression tree, H&E hematoxylin and eosin, IHC immunohistochemistry, PET Positron emission tomography, LASSO least absolute shrinkage and selection operator, ACC accuracy, sPLS-DA sparse partial least squares discriminant analysis, RFE recursive feature elimination, LR logistic regression, GBDT gradient boosting decision tree, PCC Pearson correlation coefficient, PCA principal component analysis, RNA-seq RNA-sequencing, LDA linear discriminative analysis, SVM support vector machine, RF random forest, FACS fluorescence-activated cell sorting, FCM flow cytometry, AUC area under the curve, N/A not applicable

Back to article page

ISSN: 1479-5876

Contact us

Submission enquiries: Access here and click Contact Us
General enquiries: info@biomedcentral.com