Prediction of MYCN Amplification, 1p and 11q Aberrations in Pediatric Neuroblastoma via Pre-therapy 18F-FDG PET/CT Radiomics

18F-fluorodeoxyglucose (18F-FDG) positron emission tomography/computed tomography (PET/CT) imaging of pediatric neuroblastoma: a multi-omics parameters method to predict MYCN copy number category

Background

The MYCN copy number category is closely related to the prognosis of neuroblastoma (NB). Therefore, this study aimed to assess the predictive ability of 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography/computed tomography (PET/CT) radiomic features for MYCN copy number in NB.

Methods

A retrospective analysis was performed on 104 pediatric patients with NB that had been confirmed by pathology. To develop the Bio-omics model (B-model), which incorporated clinical and biological aspects, PET/CT radiographic features, PET quantitative parameters, and significant features with multivariable stepwise logistic regression were preserved. Important radiomics features were identified through least absolute shrinkage and selection operator (LASSO) and univariable analysis. On the basis of radiomics features obtained from PET and CT scans, the radiomics model (R-model) was developed. The significant bio-omics and radiomics features were combined to establish a Multi-omics model (M-model). The above 3 models were established to differentiate MYCN wild from MYCN gain and MYCN amplification (MNA). The calibration curve and receiver operating characteristic (ROC) curve analyses were performed to verify the prediction performance. Post hoc analysis was conducted to compare whether the constructed M-model can distinguish MYCN gain from MNA.

Results

The M-model showed excellent predictive performance in differentiating MYCN wild from MYCN gain and MNA, which was better than that of the B-model and R-model [area under the curve (AUC) 0.83, 95% confidence interval (CI): 0.74-0.92 vs. 0.81, 95% CI: 0.72-0.90 and 0.79, 95% CI: 0.69-0.89]. The calibration curve showed that the M-model had the highest reliability. Post hoc analysis revealed the great potential of the M-model in differentiating MYCN gain from MNA (AUC 0.95, 95% CI: 0.89-1).

Conclusions

The M-model model based on bio-omics and radiomics features is an effective tool to distinguish MYCN copy number category in pediatric patients with NB.

A narrative review of radiomics and deep learning advances in neuroblastoma: updates and challenges

Neuroblastoma is an extremely heterogeneous tumor that commonly occurs in children. The diagnosis and treatment of this tumor pose considerable challenges due to its varied clinical presentations and intricate genetic aberrations. Presently, various imaging modalities, including computed tomography, magnetic resonance imaging, and positron emission tomography, are utilized to assess neuroblastoma. Nevertheless, these conventional imaging modalities have limitations in providing quantitative information for accurate diagnosis and prognosis. Radiomics, an emerging technique, can extract intricate medical imaging information that is imperceptible to the human eye and transform it into quantitative data. In conjunction with deep learning algorithms, radiomics holds great promise in complementing existing imaging modalities. The aim of this review is to showcase the potential of radiomics and deep learning advancements to enhance the diagnostic capabilities of current imaging modalities for neuroblastoma.

Predicting MYCN amplification in paediatric neuroblastoma: development and validation of a 18F-FDG PET/CT-based radiomics signature

OBJECTIVES

To develop and validate an 18F-FDG PET/CT-based clinical-radiological-radiomics nomogram and evaluate its value in the diagnosis of MYCN amplification (MNA) in paediatric neuroblastoma (NB) patients.

METHODS

A total of 104 patients with NB were retrospectively included. We constructed a nomogram to predict MNA based on radiomics signatures, clinical and radiological features. The multivariable logistic regression and the least absolute shrinkage and selection operator (LASSO) were used for feature selection. Radiomics models are constructed using decision trees (DT), logistic regression (LR) and support vector machine (SVM) classifiers. A clinical-radiological (C-R) model was developed using clinical and radiological features. A clinical-radiological-radiomics (C-R-R) model was developed using the C-R model of the best radiomics model. The prediction performance was verified by receiver operating characteristic (ROC) curve analysis, calibration curve analysis and decision curve analysis (DCA) in the training and validation cohorts.

RESULTS

The present study showed that four radiomics signatures were significantly correlated with MNA. The SVM classifier was the best model of radiomics signature. The C-R-R model has the best discriminant ability to predict MNA, with AUCs of 0.860 (95% CI, 0.757-0.963) and 0.824 (95% CI, 0.657-0.992) in the training and validation cohorts, respectively. The calibration curve indicated that the C-R-R model has the goodness of fit and DCA confirms its clinical utility.

CONCLUSION

Our research provides a non-invasive C-R-R model, which combines the radiomics signatures and clinical and radiological features based on 18F-FDGPET/CT images, shows excellent diagnostic performance in predicting MNA, and can provide useful biological information with stratified therapy.

CRITICAL RELEVANCE STATEMENT

Radiomic signatures of 18F-FDG-based PET/CT can predict MYCN amplification in neuroblastoma.

KEY POINTS

• Radiomic signatures of 18F-FDG-based PET/CT can predict MYCN amplification in neuroblastoma. • SF, LDH, necrosis and TLG are the independent risk factors of MYCN amplification. • Clinical-radiological-radiomics model improved the predictive performance of MYCN amplification.

Association of Computed Tomography Radiomics Signature with Progression-free Survival in Neuroblastoma Patients

AIMS

To investigate the association of computed tomography radiomics signature with progression-free survival (PFS) in neuroblastoma patients.

MATERIALS AND METHODS

We retrospectively included 167 neuroblastoma patients who were divided into a training set and a test set through stratified sampling at a ratio of 7:3. Regions of interest of the primary tumours were delineated on pretreatment contrast-enhanced computed tomography images and radiomics features were extracted from them. The intraclass correlation coefficient, Pearson correlation coefficient, and least absolute shrinkage and selection operator Cox regression algorithm were applied to select radiomics features and construct the radiomics signature. The effectiveness of the signature in predicting PFS was evaluated using the concordance index (C-index) and 95% confidence interval in both the training and the test sets. The time-dependent receiver operator characteristic curve of the radiomics signature was plotted and the area under the curve (AUC) was calculated. A calibration curve was used to assess the difference between the predicted probability of the radiomics signature and the observed probability at different time points.

RESULTS

The radiomics signature was composed of six features, which achieved a C-index of 0.733 (95% confidence interval 0.664-0.803) in the training set and 0.734 (95% confidence interval 0.608-0.861) in the test set. In the training set, the radiomics signature yielded an AUC of 0.707, 0.737, 0.788, 0.859 and 0.829 for 1-, 2-, 3-, 4- and 5-year PFS, respectively. Similarly, the radiomics signature exhibited an AUC of 0.738, 0.807, 0.761, 0.787 and 0.818 for 1-, 2-, 3-, 4- and 5-year PFS, respectively, in the test set. The calibration curves showed no significant difference between the predicted probability of the radiomics signature and the observed probability for up to 5 years.

CONCLUSIONS

Computed tomography radiomics features exhibit a significant correlation with the PFS of neuroblastoma patients, particularly in terms of long-term outcomes.

Role of MRI radiomics for the prediction of MYCN amplification in neuroblastomas

OBJECTIVES

We evaluate MR radiomics and develop machine learning-based classifiers to predict MYCN amplification in neuroblastomas.

METHODS

A total of 120 patients with neuroblastomas and baseline MR imaging examination available were identified of whom 74 (mean age ± standard deviation [SD] of 6 years and 2 months ± 4 years and 9 months; 43 females and 31 males, 14 MYCN amplified) underwent imaging at our institution. This was therefore used to develop radiomics models. The model was tested in a cohort of children with the same diagnosis but imaged elsewhere (n = 46, mean age ± SD: 5 years 11 months ± 3 years 9 months, 26 females and 14 MYCN amplified). Whole tumour volumes of interest were adopted to extract first-order histogram and second-order radiomics features. Interclass correlation coefficient and maximum relevance and minimum redundancy algorithm were applied for feature selection. Logistic regression, support vector machine, and random forest were employed as the classifiers. Receiver operating characteristic (ROC) analysis was performed to evaluate the diagnostic accuracy of the classifiers on the external test set.

RESULTS

The logistic regression model and the random forest both showed an AUC of 0.75. The support vector machine classifier obtained an AUC of 0.78 on the test set with a sensitivity of 64% and a specificity of 72%.

CONCLUSION

The study provides preliminary retrospective evidence demonstrating the feasibility of MRI radiomics in predicting MYCN amplification in neuroblastomas. Future studies are needed to explore the correlation between other imaging features and genetic markers and to develop multiclass predictive models.

KEY POINTS

• MYCN amplification in neuroblastomas is an important determinant of disease prognosis. • Radiomics analysis of pre-treatment MR examinations can be used to predict MYCN amplification in neuroblastomas. • Radiomics machine learning models showed good generalisability to external test set, demonstrating reproducibility of the computational models.

Development and validation of a CT-based radiomics signature for identifying high-risk neuroblastomas under the revised Children's Oncology Group classification system

BACKGROUND

To develop and validate a radiomics signature based on computed tomography (CT) for identifying high-risk neuroblastomas.

PROCEDURE

This retrospective study included 339 patients with neuroblastomas, who were classified into high-risk and non-high-risk groups according to the revised Children's Oncology Group classification system. These patients were then randomly divided into a training set (n = 237) and a testing set (n = 102). Pretherapy CT images of the arterial phase were segmented by two radiologists. Pyradiomics package and FeAture Explorer software were used to extract and process radiomics features. Radiomics models based on linear discriminant analysis (LDA), logistic regression (LR), and support vector machine (SVM) were constructed, and the area under the curve (AUC), 95% confidence interval (CI), and accuracy were calculated.

RESULTS

The optimal LDA, LR, and SVM models had 11, 12, and 14 radiomics features, respectively. The AUC of the LDA model in the training and testing sets were 0.877 (95% CI: 0.833-0.921) and 0.867 (95% CI: 0.797-0.937), with an accuracy of 0.823 and 0.804, respectively. The AUC of the LR model in the training and testing sets were 0.881 (95% CI: 0.839-0.924) and 0.855 (95% CI: 0.781-0.930), with an accuracy of 0.823 and 0.804, respectively. The AUC of the SVM model in the training and testing sets were 0.879 (95% CI: 0.836-0.923) and 0.862 (95% CI: 0.791-0.934), with an accuracy of 0.827 and 0.804, respectively.

CONCLUSIONS

CT-based radiomics is able to identify high-risk neuroblastomas and may provide additional image biomarkers for the identification of high-risk neuroblastomas.

Prognostic value of texture analysis of the primary tumour in high-risk neuroblastoma: An 18 F-DOPA PET study

PURPOSE

To evaluate the prognostic value of texture analysis of the primary tumour with ¹⁸ fluorine-dihydroxyphenylalanine positron emission tomography/X-ray computed tomography (¹⁸ F-DOPA PET/CT) in patients affected by high-risk neuroblastoma (HR-NBL).

METHODS

We retrospectively analysed 18 patients with HR-NBL, which had been prospectively enrolled in the course of a previous trial investigating the diagnostic role of ¹⁸ F-DOPA PET/CT at the time of the first onset. Texture analysis of the primary tumour was carried out on the PET images using LifeX. Conventional indices, histogram parameters, grey level co-occurrence (GLCM), run-length (GLRLM), neighbouring difference (NGLDM) and zone-length (GLZLM) matrices parameter were extracted; their values were compared with the overall metastatic load, expressed by means of whole-body metabolic burden (WBMB) score and the progression-free/overall survival (PFS and OS).

RESULTS

There was a direct correlation between WBMB and radiomics parameter describing uptake intensity (SUV_mean : p = .004) and voxel heterogeneity (entropy: p = .026; GLCM_Contrast: p = .001). Conversely, texture indices of homogeneity showed an inverse correlation with WBMB (energy: p = .026; GLCM_Homogeneity: p = .006). On the multivariate model, WBMB (p < .01) and the first standardised uptake value (SUV) quartile (p < .001) predicted PFS; OS was predicted by WBMB and the N-myc proto-oncogene protein (MYCN) amplification (p < .05) for both.

CONCLUSIONS

Textural parameters describing heterogeneity and metabolic intensity of the primary HR-NBL are closely associated with its overall metastatic burden. In turn, the whole-body tumour load appears to be one of the most relevant predictors of progression-free and overall survival.

Improved risk stratification by PET-based intratumor heterogeneity in children with high-risk neuroblastoma

PURPOSE

The substratification of high-risk neuroblastoma is challenging, and new predictive imaging biomarkers are warranted for better patient selection. The aim of the study was to evaluate the prognostic role of PET-based intratumor heterogeneity and its potential ability to improve risk stratification in neuroblastoma.

METHODS

Pretreatment 18F-FDG PET/CT scans from 112 consecutive children with newly diagnosed neuroblastoma were retrospectively analyzed. The primary tumor was segmented in the PET images. SUVs, volumetric parameters including metabolic tumor volume (MTV) and total lesion glycolysis (TLG), and texture features were extracted. After the exclusion of imaging features with poor and moderate reproducibility, the relationships between the imaging indices and clinicopathological factors, as well as event-free survival (EFS), were assessed.

RESULTS

The median follow-up duration was 33 months. Multivariate analysis showed that PET-based intratumor heterogeneity outperformed clinicopathological features, including age, stage, and MYCN, and remained the most robust independent predictor for EFS [training set, hazard ratio (HR): 6.4, 95% CI: 3.1-13.2, p < 0.001; test set, HR: 5.0, 95% CI: 1.8-13.6, p = 0.002]. Within the clinical high-risk group, patients with a high metabolic heterogeneity showed significantly poorer outcomes (HR: 3.3, 95% CI: 1.6-6.8, p = 0.002 in the training set; HR: 4.4, 95% CI: 1.5-12.9, p = 0.008 in the test set) compared to those with relatively homogeneous tumors. Furthermore, intratumor heterogeneity outran the volumetric indices (MTVs and TLGs) and yielded the best performance of distinguishing high-risk patients with different outcomes with a 3-year EFS of 6% vs. 47% (p = 0.001) in the training set and 9% vs. 51% (p = 0.004) in the test set.

CONCLUSION

PET-based intratumor heterogeneity was a strong independent prognostic factor in neuroblastoma. In the clinical high-risk group, intratumor heterogeneity further stratified patients with distinct outcomes.