To establish and validate a multimodal MRI radiomics model based on intratumoral and peritumoral heterogeneity for prediction of spatial pattern of locally recurrent high-grade gliomas (HGGs).

A retrospective analysis was conducted on the clinical and imaging data of all HGGs patients who underwent maximum safe resection followed by postoperative radiotherapy combined with temozolomide treatment and experienced in local recurrence in Army Medical Center of PLA from 2012 to 2021. Two radiologists independently assessed the spatial patterns of locally recurrence HGGs through continuous follow-up MRI data, and primarily categorized the pattern into intra-resection cavity recurrence and extra-resection cavity recurrence. The subjected patients were randomly divided into a training set and a validation set in a 7∶3 ratio. In the training set, Pearson or Spearman correlation analysis and least absolute shrinkage and selection operator (LASSO) analysis were employed to screen radiomic features within the intratumoral and peritumoral regions, as well as to calculate radiomic scores. A radiomics model was established using logistic regression analysis. The performance of the model was assessed using calibration curves, Hosmer-Lemeshow goodness-of-fit test, and the area under the receiver operating characteristic curve (AUC). Validation of the model was performed in the validation set.

A total of 121 patients with locally recurrent HGGs were enrolled in this study, including 54 in intra-resection cavity recurrence group and 67 in extra-resection cavity recurrence group. Among them, 84 were assigned into the training set and 37 into the validation set. In the training set, the radiomics score for the extra-resection cavity recurrence group was 0.424 (0.278, 0.573), which was higher than that for the intra-resection cavity recurrence group [-0.030 (-0.226,0.248), P<0.001]. In the validation set, the radiomics score for the extra-resection cavity recurrence group was 0.369 (0.258, 0.487), which was higher than that for the intra-resection cavity recurrence group [0.277 (0.103, 0.322), P=0.033]. The established radiomics model exhibited good calibration and performed well in predicting spatial recurrence patterns, with an AUC value of 0.844 (95%CI: 0 749~0.914) in the training set and 0.706 (95%CI: 0.534~0.844) in the validation set.

Our multimodal radiomics model combined with intratumoral and peritumoral heterogeneity can predict the spatial pattern of locally recurrent HGGs, providing a basis for individualized treatment of HGGs.

To evaluate the predictive value of T2-fluid attenuated inversion recovery (FLAIR) signal suppression rate for the short arm of chromosome 1 and long arm of chromosome 19 (1p/19q) molecular features in lower-grade gliomas (LGG), and to construct and verify the predictive model based on magnetic resonance imaging (MRI) tumor features and T2-FLAIR signal suppression rate.

Clincal and imaging data of the patients with pathologically confirmed supratentorial LGG (WHO grade 2~3) in our medical center from 2017 to 2021 were collected and retrospectively analyzed. According to the results of postoperative molecular pathology, they were divided into 1p/19q-codeleted (1p/19q-Codel) and 1p/19q-noncodeleted (1p/19q-Noncodel) groups. MRI tumor features were blindly assessed by 2 neuroradiologists. Five circular regions of interest were respectively delineated in the tumor area and the normal-appearing white matter in contralateral semioval center using the hot-spot method in order to calculate the T2-FLAIR signal suppression rate. The differences of clinical features, MRI tumor features and T2-FLAIR signal suppression rate were analyzed between the 2 groups. Univariate and multivariate logistic regression analyses were used to screen independent predictors and constructa predictive model and nomogram. Receiver operating characteristic (ROC) curve, calibration curve and Hosmer-Lemeshow test were applied to assess the model performance, and the model was internally validated by bootstrap method.

A total of 146 supratentorial LGG patients were enrolled, including 68 being assigned into the 1p/19q-Codel group and 78 into the 1p/19q-Noncodel group. The T2-FLAIR signal suppression rate was 0.43 (0.28, 0.62) in the 1p/19q-Noncodel group, which was significantly higher than that in the 1p/19q-Codel group [0.29 (0.24, 0.35), P<0.001]. Multivariate logistic regression analysis showed that T2-FLAIR signal suppression rate >0.374 (P<0.001), cortex infiltration (P=0.001) and calcification (P=0.004) were independent predictors for 1p/19q status. The AUC value of T2-FLAIR signal suppression rate >0.374 in predicting 1p/19q-Noncodel was 0.720, the sensitivity was 60.26% and the specificity was 83.82%. DeLong test indicated that T2-FLAIR signal suppression rate >0.374 was more effective than T2-FLAIR mismatch sign in predicting 1p/19q molecular features (P<0.001). ROC curve analysis suggested that the predictive model established by T2-FLAIR signal suppression rate >0.374 combined with cortex infiltration and calcification had good performance, with an AUC value of 0.808, and the AUC value verified internally by bootstrap method was 0.807. At the same time, the calibration and goodness of fit of the model were good.

T2-FLAIR signal suppression rate can be used as a quantitative imaging marker to predict 1p/19q-Noncodel LGG. The predictive model with T2-FLAIR signal suppression rate >0.374 combined with cortex infiltration and calcification can effectively predict 1p/19q molecular features.

To evaluate the predictive value of apparent diffusion coefficient（ADC）histogram in preoperative magnetic resonance imaging（MRI）of the 2 cm peritumoral edema zone for spatial pattern of recurrence in IDH wild-type glioblastoma（GB）.

A case-control study was performed on 50 patients with IDH wild-type GB treated with standard protocol, diagnosed by pathology as tumor recurrence and accepted regular follow-up MRI in Department of Neurosurgery of Army Medical Center of PLA from January 2012 to December 2021. The postoperative spatial recurrence patterns were divided into the local recurrence group（n=28）and the non-local recurrence group（n=22）according to whether the distance between the recurrence foci and the operative cavity was more than 2 cm or not. According to the maximum level of tumor enhancement in axial planes of enhanced MRI, the corresponding level of ADC images were selected. Mazda software was used to outline the region of interest（ROI）along the edge of the 2 cm range of peritumoral edema for histogram analysis. The histogram parameters of the 2 groups with different spatial patterns of recurrence were statistically analyzed, and the area under the curve（AUC）was obtained by receiver operating characteristic curve to evaluate the diagnostic efficacy.

Among the 50 patients with recurrent GB, progression-free survival（PFS）and overall survival（OS）in the local recurrence group were longer than those in the non-local recurrence group（median PFS: 6.6 vs 4.6 months; median OS: 15.4 vs 12.4 months）, and the differences were statistically significant（PFS: log-rank Chi-square=4.325, P=0.038; OS: log-rank Chi-square=4.022, P=0.045）. Variance, kurtosis and Perc.90% of the 9 features extracted from ADC histogram were statistically significant differences between the 2 groups（P<0.05）. Variance had the best diagnostic efficiency（AUC: 0.804, sensitivity: 75.00%, specificity: 81.82%）. Among the 4 multivariate logistic regression models constructed based on the variance, kurtosis and Perc.90%, the model constructed by the combination of variance, kurtosis and Perc. 90% had the best diagnostic efficiency（AUC: 0.878, sensitivity: 78.57%, specificity: 86.36%）.

The characteristic parameters of the preoperative ADC histogram in peritumoral edema zone can be used as an imaging marker to predict the different spatial patterns of postoperative recurrence in patients with IDH wild-type GB.