Tumor grading of soft tissue sarcomas using MRI-based radiomics

Abstract

a b s t r a c tBackground: Treatment decisions for multimodal therapy in soft tissue sarcoma (STS) patients greatly depend on the differentiation between low-grade and high-grade tumors. We developed MRI-based radiomics grading models for the differentiation between low-grade (G1) and high-grade (G2/G3) STS. Methods: The study was registered at ClinicalTrials.gov (number NCT03798795). Contrast-enhanced T1-weighted fat saturated (T1FSGd), fat-saturated T2-weighted (T2FS) MRI sequences, and tumor grading following the French Federation of Cancer Centers Sarcoma Group obtained from pre-therapeutic biopsies were gathered from two independent retrospective patient cohorts. Volumes of interest were manually segmented. After preprocessing, 1394 radiomics features were extracted from each sequence. Features unstable in 21 independent multiple-segmentations were excluded. Least absolute shrinkage and selection operator models were developed using nested cross-validation on a training patient cohort (122 patients). The influence of ComBatHarmonization was assessed for correction of batch effects. Findings: Three radiomic models based on T2FS, T1FSGd and a combined model achieved predictive performances with an area under the receiver operator characteristic curve (AUC) of 0.78, 0.69, and 0.76 on the independent validation set (103 patients), respectively. The T2FS-based model showed the best reproducibility. The radiomics model involving T1FSGd-based features achieved significant patient stratification. Combining the T2FS radiomic model into a nomogram with clinical staging improved prognostic performance and the clinical net benefit above clinical staging alone. Interpretation: MRI-based radiomics tumor grading models effectively classify low-grade and high-grade soft tissue sarcomas.

Intensity harmonization techniques influence radiomics features and radiomics-based predictions in sarcoma patients

“…Our results also deepened that intra-tumoral heterogeneous SIs on T2-WI is predictive of MFS in a quantitative manner and other studies have also correlated this parameter with overall and/or metastatic-relapse free survivals in STS patients with relatively close and similar performances to ours 6,7,20 . Indeed, Peeken et al used an equivalent of IHTstd and applied ComBat to correct for multicenter effect.…”

“…In this case, lowering these performances with a particular IHT would mean that this IHT caused noise and inappropriate deviation in the data. However, as already stated, prior studies converged towards same results regarding the relationship between MRI-based radiomics features, heterogeneity on T2-WI and outcomes of sarcoma patients 6,7,20,35 . Alternative study designs could have been proposed in the absence of such relationship, (i) either by using a phantom made of compartments with various degrees of heterogeneity, (ii) or by using MRIs of healthy volunteers covering organs with different textures and investigating which IHT enables the best radiomics-based classification of these organs (by analogy with the study by Orlhac et al) 19 .…”

“…Indeed, Peeken et al used an equivalent of IHTstd and applied ComBat to correct for multicenter effect. They also provided the sarcoma histological type as a biological covariable (which slightly improved the performances) 6 . Their best model relied on radiomics features from Fat Sat T2 weighted imaging and showed a concordance-index of 0.74 in the validation cohort.…”

“…Thus, our aim was to investigate how the IHT could influence MRI-based radiomics analyses in a uniformly-treated cohort of soft-tissue sarcomas (STS) patients with which the presence of intra-tumor heterogeneity on initial T2-weighted-imaging (-WI) has been previously correlated with metastatic-relapse free survival (MFS) 4 , 6 , 20 . To do so, to comprehensively assess the impact of IHT on radiomics analyses, we investigated its influence on: (i) the RFs values; (ii) the prognostic value of radiomics-based unsupervised classifications; and (iii) the performances of supervised classifiers to predict early metastatic relapses.…”

Magnetic Resonance Imaging-based Radiomics Nomogram for Prediction of the Histopathological Grade of Soft Tissue Sarcomas: A Two-center Study

“…Spraker et al [20] found that radiomics features alone or combined with age and grade were respectively independent predictors of over survival, and the latter achieved best performance. Peeken et al [37] demonstrated that the radiomics model based on T1FSGd MRI sequences showed signi cant patient strati cation performance for overall survival, and improved prognostic performance was found with the combination of a T2FS radiomics model with the AJCC system. Conversely, we evaluated survival prediction by incorporating a nomogram-predicted grade model that included clinical stage and MRI features, and it showed superior patient risk strati cation performance.…”