Diagnostic performance of MRI of post-laminar optic nerve invasion detection in retinoblastoma: A systematic review and meta-analysis

Magnetic resonance features in advanced retinoblastoma - what a paediatric radiologist needs to know

Retinoblastoma is the most common intra-ocular malignancy in children worldwide. Current treatment modalities include focal therapy such as laser or cryotherapy, intravenous chemotherapy, intra-arterial chemotherapy and surgical enucleation. According to the current retinoblastoma staging systems, there are a few common features indicating advanced disease stage, which can subsequently predict the prognosis and the necessity of surgical enucleation and adjuvant chemotherapy. This review uses annotated magnetic resonance (MR) images to illustrate these features, which include (1) retinal detachment, (2) vitreous and subretinal seeding, (3) choroidal invasion, (4) anterior segment invasion and increase in intra-ocular pressure, (5) iris neovascularization, (6) vitreous haemorrhage, (7) aseptic orbital cellulitis with lens dislocation, (8) post-laminar optic nerve invasion and (9) distant and systemic metastasis. Prompt identification of these features is important for paediatric radiologists to guide the clinical team for appropriate management.

Management of Extraocular Retinoblastoma: ICMR Consensus Guidelines

Retinoblastoma (RB) is the most common intraocular malignancy of childhood. Advanced stage presentation of RB is common in low middle-income countries (LMICs) due to lack of awareness, social taboos associated with enucleation, seeking alternative conservative treatment options, and poor accessibility to health care. Over the last few decades, there have been significant advancements in the management of extraocular RB (EORB) which have improved outcomes and helped in minimizing treatment-related toxicities. The incorporation of multimodality approaches including chemotherapy, surgery, and radiotherapy (RT) has shown promising results; however, prognosis remains poor especially in LMICs. In this article, authors have discussed the ICMR consensus guidelines on the management of EORB, including metastatic RB.

Optic nerve thickening on high-spatial-resolution MRI predicts early-stage postlaminar optic nerve invasion in retinoblastoma

OBJECTIVES

To assess the diagnostic accuracy of nerve thickening on MRI to predict early-stage postlaminar optic nerve invasion (PLONI) in retinoblastoma. Furthermore, this study aimed to incorporate measurements into a multiparametric model for radiological determination of PLONI.

METHODS

In this retrospective multicenter case-control study, high-spatial-resolution 3D T2-weighted MR images were used to measure the distal optic nerve. Histopathology was the reference standard for PLONI. Two neuroradiologists independently measured the optic nerve width, height, and surface at 0, 3, and 5 mm from the most distal part of the optic nerve. Subsequently, PLONI was scored on contrast-enhanced T1-weighted and 3D T2-weighted images, blinded for clinical data. Optic nerve measurements with the highest diagnostic accuracy for PLONI were incorporated into a prediction model for radiological determination of PLONI.

RESULTS

One hundred twenty-four retinoblastoma patients (median age, 22 months [range, 0-113], 58 female) were included, resulting in 25 retinoblastoma eyes with histopathologically proven PLONI and 206 without PLONI. ROC analysis of axial optic nerve width measured at 0 mm yielded the best area under the curve of 0.88 (95% confidence interval: 0.79, 0.96; p < 0.001). The optimal width cutoff was ≥ 2.215 mm, with a sensitivity of 84% (95% CI: 64, 95%) and specificity of 83% (95% CI: 75, 89%) for detecting PLONI. Combining width measurements with the suspicion of PLONI on MRI sequences resulted in a prediction model with an improved sensitivity and specificity of respectively up to 88% and 92%.

CONCLUSION

Postlaminar optic nerve thickening can predict early-stage postlaminar optic nerve invasion in retinoblastoma.

CLINICAL RELEVANCE STATEMENT

This study provides an additional tool for clinicians to help determine postlaminar optic nerve invasion, which is a risk factor for developing metastatic disease in retinoblastoma patients.

KEY POINTS

• The diagnostic accuracy of contrast-enhanced MRI for detecting postlaminar optic nerve invasion is limited in retinoblastoma patients. • Optic nerve thickening can predict postlaminar optic nerve invasion. • A prediction model combining MRI features has a high sensitivity and specificity for detecting postlaminar optic nerve invasion.

Diagnostic Imaging for Retinoblastoma Cancer Staging: Guide for Providing Essential Insights for Ophthalmologists and Oncologists

Retinoblastoma is the most common cause of all intraocular pediatric malignancies. It is caused by the loss of RB1 tumor suppressor gene function, although some tumors occur due to MYCN oncogene amplification with normal RB1 genes. Nearly half of all retinoblastomas occur due to a hereditary germline RB1 pathogenic variant, most of which manifest with bilateral tumors. This germline RB1 mutation also predisposes to intracranial midline embryonal tumors. Accurate staging of retinoblastoma is crucial in providing optimal vision-, eye-, and life-saving treatment. The AJCC Cancer Staging Manual has undergone significant changes, resulting in a universally accepted system with a multidisciplinary approach for managing retinoblastoma. The authors discuss the role of MRI and other diagnostic imaging techniques in the pretreatment assessment and staging of retinoblastoma. A thorough overview of the prevailing imaging standards and evidence-based perspectives on the benefits and drawbacks of these techniques is provided. Published under a CC BY 4.0 license. Test Your Knowledge questions for this article are available in the supplemental material.

Preoperative Detection of Local Tumor Extent in Patients with Advanced Retinoblastoma: Predictive Value of MRI and Clinical Findings

OBJECTIVE

Before planned enucleation, local tumor extension in advanced retinoblastoma is routinely assessed preoperatively using high-resolution magnetic resonance imaging (MRI). The aim of our study was to analyse the predictive value of MRI and clinical characteristics for predicting tumor extent, as confirmed by histopathology postoperatively.

PATIENTS AND METHODS

All consecutive patients were included who underwent primary enucleation for advanced retinoblastoma after high-resolution MRI examination in our hospital between January 2011 and December 2021. The primary study endpoint was the evaluation of the predictability of histopathological risk factors on preoperative MRI examination. The sensitivity and specificity of the MRI examination with respect to clinically relevant optic nerve infiltration and choroidal infiltration were determined.

RESULTS

The mean age of the 209 included patients was 1.6 years (range 1 month to 4.7 years). MRI indicated optic nerve infiltration in 46 (22%) patients, extensive choroidal infiltration in 78 (40.2%) patients, and scleral infiltration in one patient (2.6%). Histopathological examination demonstrated postlaminar optic infiltration in 25 (12%) patients and extensive choroidal infiltration in 17 (8.1%) cases. Scleral infiltration was evident in 8 (3.8%) patients. In the final multivariate analysis, MRI findings of tumor infiltration and a preoperative intraocular pressure ≥ 20 mmHg were independently associated with histopathological evidence of clinically relevant optic nerve (p = 0.033/p = 0.011) and choroidal infiltration (p = 0.005/p = 0.029). The diagnostic accuracy of the prediction models based on the multivariate analysis for the identification of the clinically relevant optic nerve (AUC = 0.755) and choroidal infiltration (AUC = 0.798) was greater than that of purely MRI-based prediction (respectively 0.659 and 0.742). The sensitivity and specificity of MRI examination for determining histopathological risk factors in our cohort were 64% and 65% for clinically relevant optic infiltration and 87% and 64% for clinically relevant choroidal infiltration.

CONCLUSION

The local tumor extent of retinoblastoma with infiltration of the optic nerve and choroid can be well estimated based on radiological and clinical characteristics before treatment initiation. The combination of clinical and radiological risk factors supports the possibility of early treatment stratification in retinoblastoma patients.

Risk of secondary tumours in patients with non-metastatic and metastatic human retinoblastoma

BACKGROUND

Retinoblastoma is an intraocular cancer in children and infants. Despite all the available treatment options and high survival rates in children with retinoblastoma, exposure to secondary tumours in adulthood is one of the concerns that physicians face. In many cases, dysfunction of the RB1 gene is the main cause of secondary tumours due to retinoblastoma. Therefore, the aim of this study was to evaluate the incidence of other secondary tumours in children with retinoblastoma.

METHODS

In this regard, we performed continuous and integrated bioinformatics analyses to find genes, protein products, and signal pathways involved in other cancers.

RESULTS

1170 high-expression genes and 960 low-expression genes between non-invasive and invasive retinoblastoma were isolated. After examining the signal pathways, we observed bladder cancer and small cell lung cancer in the overexpressed genes. We also observed 5 cancers of endometriosis, prostate, non-small cell lung cancer, glioblastoma and renal cell carcinoma in low-expression genes. Based on the P-value index, non-small cell lung cancer, prostate and bladder cancers had the highest risk, and endometriosis cancer showed a lower probability of developing a secondary tumour in patients with retinoblastoma. In addition, the network between proteins also showed us that TP53, CDK2, SRC, MAPK1 proteins with high expression and JUN, HSP90AA1, and UBC proteins with low-expression play a significant role in candidate cancers.

CONCLUSION

Lastly, we used continuous bioinformatics analysis to show that seven cancers are strongly linked to retinoblastoma cancer. Of course, more research is needed to find the best way to care for children who have been treated for retinoblastoma.

Explainable AI for Retinoblastoma Diagnosis: Interpreting Deep Learning Models with LIME and SHAP

Retinoblastoma is a rare and aggressive form of childhood eye cancer that requires prompt diagnosis and treatment to prevent vision loss and even death. Deep learning models have shown promising results in detecting retinoblastoma from fundus images, but their decision-making process is often considered a "black box" that lacks transparency and interpretability. In this project, we explore the use of LIME and SHAP, two popular explainable AI techniques, to generate local and global explanations for a deep learning model based on InceptionV3 architecture trained on retinoblastoma and non-retinoblastoma fundus images. We collected and labeled a dataset of 400 retinoblastoma and 400 non-retinoblastoma images, split it into training, validation, and test sets, and trained the model using transfer learning from the pre-trained InceptionV3 model. We then applied LIME and SHAP to generate explanations for the model's predictions on the validation and test sets. Our results demonstrate that LIME and SHAP can effectively identify the regions and features in the input images that contribute the most to the model's predictions, providing valuable insights into the decision-making process of the deep learning model. In addition, the use of InceptionV3 architecture with spatial attention mechanism achieved high accuracy of 97% on the test set, indicating the potential of combining deep learning and explainable AI for improving retinoblastoma diagnosis and treatment.

Diffusion‐Weighted MR Imaging of the Optic Nerve Can Improve the Detection of Post‐Laminar Optic Nerve Invasion from Retinoblastoma

BACKGROUND

Post-laminar optic nerve invasion (PLONI) is a high-risk factor for the metastasis of retinoblastoma (RB). Unlike conventional MRI, diffusion-weighted imaging (DWI) reflects histopathological features, and may aid the assessment of PLONI.

PURPOSE

To determine the value of conventional MRI plus DWI in detecting PLONI in RB patients.

STUDY TYPE

Retrospective.

POPULATION

Eighty-three RB patients, including 28 with histopathologically proven PLONI and 55 without PLONI.

FIELD STRENGTH/SEQUENCE

3.0 T, precontrast axial T1-weighted and T2-weighted imaging, DWI, and postcontrast axial, coronal, and oblique-sagittal T1-weighted imaging.

ASSESSMENT

PLONI was assessed using post-enucleation histology and preoperative MRI features (optic nerve signal intensity, enlargement, and enhancement on conventional MRI, and apparent diffusion coefficient [ADC] of the optic nerve on DWI) evaluated by three observers.

STATISTICAL TESTS

MRI features suggesting the presence of PLONI were identified using univariable and multivariable analyses. Receiver operating characteristic (ROC) curve and area under the curve (AUC) were used to analyze diagnostic performance.

RESULTS

Optic nerve enhancement and low ADC of the optic nerve were significant indicators of PLONI. ROC curve analysis showed that the AUC of the combination of these two features for detecting PLONI was 0.87 (95% confidence interval [CI]: 0.78-0.93). The diagnostic performance of this model was significantly superior to that of optic nerve enhancement alone (0.76, 95% CI: 0.65-0.85) and marginally superior to that of the ADC of the affected optic nerve (0.78, 95% CI: 0.68-0.87, P = 0.051).

DATA CONCLUSION

Conventional MRI combined with DWI can improve the detection of PLONI in RB patients over conventional MRI alone.

EVIDENCE LEVEL

3 Technical Efficacy: Stage 2.

Correlation between clinical presentations, radiological findings and high risk histopathological features of primary enucleated eyes with advanced retinoblastoma at Queen Sirikit National Institute of Child Health: 5 years result

Purpose

To evaluate the correlation between clinical presentations, radiological findings and high risk histopathological features of primary enucleated eyes in patients with advanced retinoblastoma at a tertiary child hospital in Thailand.

Materials and methods

We retrospectively reviewed the medical records of patients who were treated with primary enucleation of tumor eyes between 2015–2020. Demographic data, radiological assessment, and histopathological findings were collected. The association between clinical presentations and high-risk histopathological features in the primary enucleated eyes were evaluated. The radiological findings, which showed characteristic of high risk features, were compared with the histopathological findings.

Results

Thirty-three enucleated eyes were enrolled in this study. The mean age at diagnosis was 23.12 months. Most patients had unilateral group E retinoblastoma, with no difference in sex and laterality of the eye. Leukocoria was the most common presentation, followed by proptosis and uveitis. Older age at presentation were statistically associated with post laminar cribrosa optic nerve invasion (P-value 0.0027) and high-risk histopathological features in enucleated eyes (P-value 0.0032). Clinical presentations with proptosis were statistically associated with post laminar cribrosa optic nerve invasion, while leukocoria and uveitis were statistically associated with anterior segment invasion. Unifocal intraocular mass with necrosis was the most common histopathological finding. High-risk features were found in 45% of primary enucleated eye. The sensitivity and specificity of magnetic resonance imaging (MRI) for detected optic nerve invasion in retinoblastoma patients were 75% and 54%, respectively.

Conclusion

Patients with unilateral retinoblastoma who presented with older age related to high-risk features after enucleation. Ophthalmic examination with slitlamp is the best way for detection of anterior segment invasion. Choroidal invasion was unable to predict with clinical presentation. MRI was the better imaging for detection of post laminar cribrosa optic nerve invasion.

MRI can reliably differentiate optic nerve inflammation from tumor invasion in retinoblastoma with orbital cellulitis

OBJECTIVE

To investigate prevalence and MRI phenotype of retinoblastoma-associated orbital cellulitis. Additionally, this study aimed to identify postlaminar optic nerve enhancement patterns differentiating between inflammation and tumor invasion.

DESIGN

A monocenter cohort study assessed the prevalence of orbital cellulitis features on MRI in retinoblastoma patients. A multicenter case-control study compared MRI features of the retinoblastoma-associated orbital cellulitis cases with retinoblastoma controls.

SUBJECTS

A consecutive retinoblastoma patient cohort of 236 patients (311 eyes) was retrospectively investigated. Subsequently, 30 retinoblastoma cases with orbital cellulitis were compared with 30 matched retinoblastoma controls without cellulitis.

METHODS

In the cohort study, retinoblastoma MRI scans were scored on presence of inflammatory features. In the case-control study MRI scans were scored on intraocular features and postlaminar optic nerve enhancement patterns. Postlaminar enhancement patterns were compared with histopathologic assessment of postlaminar tumor invasion. Interreader agreement was assessed and exact tests with Bonferroni-correction were adopted for statistical comparisons.

MAIN OUTCOME MEASURES

Prevalence of retinoblastoma-associated orbital cellulitis on MRI was calculated. Frequency of intra-ocular MRI features were compared between orbital cellulitis cases and controls. Sensitivity and specificity of postlaminar optic nerve patterns for detection of postlaminar tumor invasion was assessed.

RESULTS

The MRI prevalence of retinoblastoma-associated orbital cellulitis was 6.8% (16/236). Retinoblastoma with orbital cellulitis showed significantly more tumor necrosis, uveal abnormalities (inflammation, hemorrhage and necrosis), lens luxation (all P < 0.001), and a larger eye size (P = 0.012). The inflammatory pattern of optic nerve enhancement (strong enhancement similar to adjacent choroid) was solely found in orbital cellulitis cases, of which none (0/16) showed tumor invasion on histopathology. Of patients with invasive pattern enhancement, 50% (5/10) showed tumor invasion on histopathology. Considering these different enhancement patterns, i.e. suggestive for either inflammation or tumor invasion, increased specificity for detection of postlaminar tumor invasion within the context of orbital cellulitis from 32% (95%CI:16-52%) to 89% (95%CI:72-98%).

CONCLUSIONS

Retinoblastoma cases presenting with orbital cellulitis show MRI findings of a larger eye size, extensive tumor necrosis, uveal abnormalities, and lens luxation. MRI contrast enhancement patterns within the postlaminar optic nerve can differentiate between tumor invasion and inflammatory changes.

MRI-based radiomics model can improve the predictive performance of postlaminar optic nerve invasion in retinoblastoma

OBJECTIVES

To develop an MRI-based radiomics model to predict postlaminar optic nerve invasion (PLONI) in retinoblastoma (RB) and compare its predictive performance with subjective radiologists' assessment.

METHODS

We retrospectively enrolled 124 patients with pathologically proven RB (90 in training set and 34 in validation set) who had MRI scans before surgery. A radiomics model for predicting PLONI was developed by extracting quantitative imaging features from axial T2W images and contrast-enhanced T1W images in the training set. The Kruskal-Wallis test, least absolute shrinkage and selection operator regression, and recursive feature elimination were used for feature selection, where upon a radiomics model was built with a logistic regression (LR) classifier. The area under the curve (AUC) of the receiver operating characteristic (ROC) curve and the accuracy were assessed to evaluate the predictive performance in the training and validation set. The performance of the radiomics model was compared to radiologists' assessment by DeLong test.

RESULTS

The AUC of the radiomics model for the prediction of PLONI was 0.928 in the training set and 0.841 in the validation set. Radiomics model produced better sensitivity than radiologists' assessment (81.1% vs  43.2% in training set, 82.4vs 52.9% in validation set). In all 124 patients, the AUC of the radiomics model was 0.897, while that of radiologists' assessment was 0.674 (p < 0.001, DeLong test).

CONCLUSION

MRI-based radiomics model to predict PLONI in RB patients was shown to be superior to visual assessment with improved sensitivity and AUC, and may serve as a potential tool to guide personalized treatment.

Current Indications of Secondary Enucleation in Retinoblastoma Management: A Position Paper on Behalf of the European Retinoblastoma Group (EURbG)

Simple Summary

Although secondary enucleation (SE) is the treatment of choice for retinoblastoma eyes that did not respond favorably to conservative therapies, clear criteria for its indication are, however, currently missing. In this position paper on behalf of the European Retinoblastoma Group (EURbG), we discuss the available literature on SE, including its influence on metastases rate and survival, and propose guidelines to assist decision-making to interrupt eye-preserving therapies depending on the availabilities of advanced diagnostic and therapeutic modalities. Absolute indications to SE may be restricted to eyes with refractory tumor activity resisting all salvage treatments or eyes under apparent tumor control but no visual potential and irreducible complications. In contrast, eyes with an obscured optic nerve head and/or ocular complications amenable to specific surgical or medical management can be considered relative indications, provided that appropriate follow-up can be implemented and that parents are fully aware of a residual risk.

Abstract

Secondary enucleation (SE) puts an irreversible end to eye-preserving therapies, whenever their prolongation is expected to violate the presumed state of metastatic grace. At present, it must be acknowledged that clear criteria for SE are missing, leading to empiric and subjective indications commonly related to disease progression or relapse, disease persistence masking the optic nerve head or treatment-related complications obscuring the fundus view. This absence of evidence-based consensus regarding SE is explained by the continuously moving frontiers of the conservative management as a result of diagnostic and therapeutic advances, as well as by the lack of studies sufficiently powered to accurately stratify the risk of metastasis in conservatively treated patients. In this position paper of the European Retinoblastoma Group (EURbG), we give an overview of the progressive shift in the indications for SE over the past decades and propose guidelines to assist decision-making with respect to when SE becomes imperative or recommended, with corresponding absolute and relative SE indications. Further studies and validation of biologic markers correlated with the risk of metastasis are expected to set more precisely the frontiers of conservative management and thus consensual criteria for SE in the future.