Quantitative MRI in Childhood Neuroblastoma: Beyond the Assessment of Image-defined Risk Factors

Abstract

Neuroblastoma commonly occurs in children. MRI is a radiation-free imaging modality and has played an important role in identifying image-defined risk factors of neuroblastoma, providing necessary guidance for surgical resection and treatment response evaluation. However, image-defined risk factors are limited to providing structural information about neuroblastoma. With the evolution of MRI technologies, quantitative MRI can not only help assess image-defined risk factors but can also quantitatively reflect the biologic features of neuroblastoma in a noninvasive manner. Therefore, compared with anatomic imaging, these emerging quantitative MRI technologies are expected to provide more imaging biomarkers for the management of neuroblastoma. This review article discusses the current applications of quantitative MRI technologies in evaluating childhood neuroblastoma.

Keywords: Pediatrics, MR–Functional Imaging, Children, MRI, Neuroblastoma, Quantitative Imaging

© RSNA, 2024

Summary

Compared with anatomic imaging, quantitative MRI offers the potential to provide more imaging biomarkers for the diagnosis, risk stratification, treatment response evaluation, and prognosis prediction of childhood neuroblastoma.

Essentials

• ■ Image-defined risk factors derived from anatomic imaging of neuroblastoma are limited to providing only structural information about the tumor.

• ■ Quantitative MRI demonstrates the ability to help predict prognosis, including overall survival, in patients with neuroblastoma and may provide more imaging biomarkers compared with anatomic imaging.

• ■ Quantitative MRI-derived indicators and image-defined risk factors could serve as complementary biomarkers for neuroblastoma.

Introduction

Neuroblastoma is a common childhood solid tumor that primarily occurs in the peripheral sympathetic nerve chain, especially in the adrenal and retroperitoneal paravertebral regions. It accounts for a substantial proportion of pediatric cancer cases. Originating from neural crest cells, neuroblastoma presents diverse clinical manifestations, ranging from localized growths that can be surgically resected to widespread metastatic disease (1). Given the highly variable prognosis of patients with neuroblastoma, it is crucial to develop advanced diagnostic and prognostic tools to guide effective treatment strategies.

MRI has emerged as an important imaging modality in the management of neuroblastoma. Unlike CT, MRI offers the advantage of being free from ionizing radiation, making it particularly suitable for repeated use in pediatric patients. By utilizing MRI, radiologists can determine the anatomic relationship between neuroblastoma and adjacent structures, thereby providing critical guidance for the surgical resection and treatment response evaluation of neuroblastoma (2). However, although anatomic imaging can provide some quantitative information, such as percentage change in tumor size and attenuation measurements, as well as qualitative information about the morphology and structure of neuroblastoma, it is insufficient for prognostic evaluation of the disease. A previous study indicated that percentage change in tumor size was not a predictor of high-risk neuroblastoma prognosis (3).

Quantitative MRI techniques, such as multi-b-value diffusion-weighted imaging (DWI) and amide proton transfer-weighted (APTw) imaging, can not only present the morphologic and structural characteristics of lesions but can also provide insights into the biologic features of tumors beyond anatomic visualization, in a noninvasive manner (4). By applying these emerging quantitative MRI techniques, we can identify more imaging biomarkers associated with the prognosis of neuroblastoma, thereby facilitating the development of personalized medicine for these patients. For example, a previous study found that a low baseline apparent diffusion coefficient (ADC) value was associated with tumor progression and relapse in patients with neuroblastoma (5). This review discusses the current applications of quantitative MRI technologies in evaluating childhood neuroblastoma (Table 1).

Table 1:

Summary of Principles and Current Applications of Quantitative MRI Techniques in Neuroblastoma

Epidemiology and Pathophysiology of Neuroblastoma

Neuroblastoma is the most frequently diagnosed malignancy during infancy, accounting for about 8%–10% of all childhood cancers. A report from the Netherlands indicated that 45% of the 593 newly diagnosed neuroblastoma cases between 1990 and 2014 were in children younger than 18 months of age, and 52% had stage IV disease (6). The overall age-adjusted incidence rate of neuroblastoma in Southeast Europe registries between 1990 and 2016 was 10.1 per million persons (7). A comprehensive analysis of neuroblastoma incidence in the United States between 2001 and 2019 revealed that the age-adjusted incidence of 11 543 primary neuroblastoma cases was 8.3 per million persons, with an average annual percent change of 0.4% (8).

Neuroblastoma originates from neural crest cells, which are embryonic cells that normally develop into the peripheral sympathetic nervous system. This tumor can occur anywhere along the sympathetic chain from the neck to the pelvis (Fig 1), particularly in the adrenal gland and paraspinal ganglia, and occasionally occurs in the central nervous system. Neuroblastoma frequently metastasizes to the bone, bone marrow, liver, lymph nodes, and skin. Metastasis can be either single- or multiple-site. Liu et al (9) found that bone was the most common site of metastasis (42.6%), followed by the liver (10.9%), lung (1.3%), and brain (0.2%), among patients with single-organ metastasis. They also found that the combination of bone and liver metastases was the most common form of multiple-site metastasis, accounting for 9.5% of cases.

Figure 1:

Examples of lesions at different anatomic distributions. (A–C) Coronal T2-weighted images of lesions located at different locations in the mediastinum. (D, E) Coronal and sagittal T2-weighted images of lesions located at adrenal gland and pelvis, respectively. These localized lesions (white arrow) do not exhibit image-defined risk factors and are classified as L1 stage.

Genetic and molecular abnormalities in neuroblastoma include MYCN amplification, chromosomal alterations, and mutations in the ALK gene. While most neuroblastoma cases are sporadic, approximately 1%–2% are familial. Familial cases are often associated with mutations in the PHOX2B and ALK genes (10).

Staging and Treatment of Neuroblastoma

Due to the heterogeneity of neuroblastoma, patient prognosis varies widely. The overall survival rate for low-risk patients can exceed 90% but remains around 50% for high-risk patients (11). Therefore, to improve the prognosis and quality of life for patients with neuroblastoma, it is essential to perform risk stratification and implement tailored treatment strategies for each risk group. Over the past 15 years, the International Neuroblastoma Risk Group has developed an imaging-based staging system to stratify neuroblastoma preoperatively (12). This system introduced 20 image-defined risk factors (IDRFs) (Table 2) to stage neuroblastoma into L1, L2, M, and MS stages (Table 3, Fig 2). These IDRFs, combined with clinical features, are used to classify patients into very-low-, low-, intermediate-, and high-risk groups (13).

Table 2:

Description of Image-defined Risk Factors of Neuroblastoma

Table 3:

International Neuroblastoma Risk Group Staging System

Figure 2:

A case of abdominal neuroblastoma in a 1-month-old male infant. (A) Coronal and (B) axial T2-weighted images indicate that the primary lesion is confined to the right adrenal gland without image-defined risk factors. (C) Axial T2-weighted image shows that metastatic lesions are confined to the liver. This case is classified as MS stage.

The treatment of neuroblastoma varies across different risk groups. For low-risk patients, particularly those without IDRFs, upfront surgery can completely remove the primary tumor in most cases. Conversely, high-risk patients require a combined treatment approach that includes surgical removal, induction and postoperative chemotherapy, radiation therapy, and hematopoietic stem cell transplantation (14). Additionally, immunotherapy with anti-GD2 monoclonal antibodies is beneficial for high-risk patients (15). Due to the propensity of neuroblastoma to encase vascular structures and invade adjacent organs, upfront surgical removal is typically not performed for intermediate- and high-risk patients. Instead, induction chemotherapy is used to shrink the tumor before surgery.

Role of Anatomic Imaging in Neuroblastoma

Anatomic imaging can be used to determine tumor size and identify IDRFs in neuroblastoma (Fig 3), which helps guide surgical resection. As mentioned earlier, the identification of IDRFs is integral to neuroblastoma staging, which helps assess disease extent and stratify risk. The IDRF-based staging system has been widely validated in the management of neuroblastoma, regardless of surgical resection or prognosis evaluation (16,17). For surgeons, detailed anatomic imaging is crucial for planning tumor resection approaches. In cases without IDRFs, laparoscopic surgery can be performed safely and effectively, as noted in a previous study (18). Zenitani et al (19) have also found that laparoscopic resection of abdominal neuroblastoma in children is a feasible and safe procedure for tumors less than 60 mm in diameter without IDRFs. Additionally, identifying IDRFs can help predict surgical complications in neuroblastoma (20).

Figure 3:

(A) Coronal T2-weighted image shows an example of a lesion located at the cervicothoracic junction compressing the trachea (white arrow) in a 36-month-old female child, classified as L2 stage. (B) Axial and (C) sagittal T2-weighted images show an example of a tumor infiltrating the costovertebral junction and expanding into the vertebral canal in a 26-month-old female child diagnosed with high-risk neuroblastoma located in the thorax. (B) Axial T2-weighted image shows the spinal cord (red arrow) compressed to the left by the tumor (white arrowhead). The tumor only contacts the aorta, indicating it envelops less than 50% of the aorta’s perimeter (white arrows). (C) Sagittal T2-weighted image shows the tumor infiltrating the costovertebral junction between T4 and T12 (white arrows).

Similarly, anatomic imaging can also help evaluate treatment response in neuroblastoma. Some studies have found that induction chemotherapy helps reduce the number of IDRFs in neuroblastoma, although some vascular-related IDRFs respond less effectively to this treatment (21). In induction chemotherapy, assessing tumor size is helpful for evaluating the primary site treatment response. Previously, the International Neuroblastoma Response Criteria often used the ellipsoid formula to calculate tumor volume. However, Bagatell et al (3) have shown that there is no evidence of a difference in prognosis prediction between using the ellipsoid formula and the single longest dimension to assess treatment response in high-risk neuroblastoma. Therefore, the latest revised International Neuroblastoma Response Criteria have adopted the single longest dimension to evaluate treatment response at the primary site (22). For the primary tumor, a complete response is defined as the presence of less than 10 mm of residual soft tissue at the primary site and complete resolution of metaiodobenzylguanidine (MIBG) or fluorodeoxyglucose PET uptake (for MIBG-nonavid tumors) at the primary site. A partial response for the primary tumor is defined as a decrease of 30% or more in the longest diameter of the primary site, along with MIBG or fluorodeoxyglucose PET uptake at the primary site that is stable, improved, or resolved.

Limitations of Anatomic Imaging in Neuroblastoma

Although anatomic imaging is validated for neuroblastoma surgical planning, it remains inadequate for prognostic evaluation of the disease, especially in high-risk neuroblastoma. A previous study showed that changes in tumor size do not predict high-risk neuroblastoma prognosis (3). Additionally, the role of primary tumor resection in high-risk neuroblastoma remains controversial. Yeung et al (23) found no evidence of differences in terms of 5-year event-free survival and overall survival among the complete gross tumor resection, gross total resection, and subtotal tumor resection groups in high-risk stage IV neuroblastoma.

When investigating the ability of pretreatment IDRFs to predict MYCN amplification and overall survival in neuroblastoma, a previous study reported area under the receiver operating characteristic curve (AUC) values of 0.74 and 0.70, respectively, indicating only moderate diagnostic performance (17). These results suggest that anatomic imaging alone is insufficient to effectively predict the prognosis of neuroblastoma.

Another investigation suggested that anti-GD2 immunotherapy can reduce the number of IDRFs in neuroblastoma; however, there was no evidence of a difference in the reduction of IDRFs between the groups receiving immunotherapy and those not receiving it (24). Furthermore, routine imaging examinations cannot effectively detect the treatment response of high-risk neuroblastoma to anti-GD2 immunotherapy, highlighting certain limitations of these imaging methods (25). In this clinical context, the emerging quantitative MRI technologies are expected to provide more quantitative imaging biomarkers, improving the ability to stratify neuroblastoma risk and predict prognosis (Table 4).

Table 4:

Summary of Reports Related to Quantitative MRI Techniques in Neuroblastoma Discussed in This Article

DWI in Neuroblastoma

Pathologic Correlation

DWI can track the Brownian motion of water molecules within tissues at a microscopic level. One of the important indicators of DWI is ADC, which quantitatively reflects the microstructural status of cellular tissues. Neuroblastoma, due to its high nuclear-cytoplasmic ratio and dense arrangement of tumor cells, restricts the movement of water molecules (26). This leads to elevated signal intensity of neuroblastoma at a high b value at DWI compared with tissues with free diffusion of water. This restricted diffusion corresponds to a low ADC value. However, ADC values vary among different histopathologic subtypes of neuroblastic tumors and exhibit an increasing trend in neuroblastoma, ganglioneuroblastoma, and ganglioneuroma (27). Wen et al (28) explored the value of DWI in evaluating the degree of tumor cell differentiation in patients with ganglioneuroblastoma-nodular and ganglioneuroblastoma-intermixed subtypes. The results showed that the ADC values of ganglioneuroblastoma-nodular were lower than those of ganglioneuroblastoma-intermixed. These imaging findings could be explained by the fact that ganglioneuroblastoma-nodular is more malignant than ganglioneuroblastoma-intermixed, and the mitosis-karyorrhexis index in the former is also higher than that in the latter. Additionally, the authors found that the ADC values of undifferentiated or poorly differentiated neuroblastoma were lower than those of their differentiated counterparts. Therefore, DWI can be helpful for the noninvasive assessment and initial evaluation of childhood neuroblastic tumors (29).

Genetic Relevance

In addition to providing pathologic information, DWI quantitative indicators can be used as imaging biomarkers for revealing the genetic characteristics of neuroblastoma. The MYCN oncogene is mainly expressed in embryonic tissues and developing nervous tissues, playing an important role in embryonic and nervous system development. Amplification of the MYCN oncogene is associated with the occurrence and progression of neuroblastoma, particularly within high-risk subgroups, and correlates with a poor prognosis for patients (10). Neubauer et al (30) found that MYCN-amplified neuroblastoma had lower ADC values compared with nonamplified neuroblastoma, and the average signal intensity ratio of the tumor to surrounding tissue was relatively lower in MYCN-amplified neuroblastoma. However, these findings require further validation with a larger sample size. By comparing the ADC histogram indicators of MYCN-amplified and nonamplified neuroblastoma groups in 62 patients, Ghosh et al (31) found that the maximum ADC value of MYCN-amplified neuroblastoma was significantly higher and the minimum ADC value was significantly lower than that of the control group. Additionally, the entropy, variance, energy, and uniformity of the ADC histogram also showed significant statistical differences between MYCN-amplified and nonamplified groups. Although there was a significant association between IDRFs and MYCN amplification in neuroblastoma, the performance of IDRFs in detecting MYCN amplification was moderate (17). Therefore, whether the combination of IDRFs and ADC values can further improve performance for predicting MYCN amplification in neuroblastoma is worthy of further study.

Differential Diagnosis

Another potential application of DWI is in the differential diagnosis of neuroblastoma from other pediatric tumors. Aslan et al (32) found that the ADC values of abdominal neuroblastoma were significantly lower than those of Wilms tumor, suggesting that DWI can help distinguish between the two most common retroperitoneal solid tumors in children. In the aforementioned studies, researchers typically selectively delineate the regions of interest of tumors on typical planes and often use mean or minimum ADC values to assess the utility of DWI in the diagnosis and differential diagnosis of neuroblastoma. However, considering the heterogeneity of neuroblastoma, relying solely on selectively delineated regions of interest and a single ADC value may not fully reveal the biologic characteristics of neuroblastoma. Through analysis of the ADC histogram based on whole tumor lesions located in the pediatric abdomen, Meeus et al (33) found that ADC histogram parameters such as kurtosis, skewness, and entropy were significantly higher, while the mean, median, and percentile ADC values were significantly lower in malignant tumors compared with benign tumors. Figure 4 illustrates the DWI-derived parameter maps of a grade IV neuroblastoma. Despite the assistance these histogram parameters provide in distinguishing between benign and malignant abdominal tumors, there are still limitations in their ability to differentiate between other types of malignant tumors such as neuroblastoma and Wilms tumor.

Figure 4:

A case of grade IV neuroblastoma. (A, B) Axial T2-weighted and b value of 150 sec/mm² images, respectively. (C–F) Maps for the apparent diffusion coefficient, D, D*, and f, respectively. The median values calculated for the apparent diffusion coefficient, D, D*, and f from the entire tumor region of interest were 1155 × 10⁻⁶ mm²/sec, 703 × 10⁻⁶ mm²/sec, 17 762 × 10⁻⁶ mm²/sec, and 23%, respectively. (Adapted, under a CC BY 4.0 license, from reference 33.)

Chemotherapy and Prognosis Evaluation

Traditionally, imaging techniques have been used to assess the effect of chemotherapy in neuroblastoma primarily by judging changes in tumor size and volume. In a study evaluating the efficacy of chemotherapy in neuroblastoma using DWI, researchers compared the ADC values of abdominal-pelvic neuroblastoma before and after chemotherapy (34). They found a significant increase in ADC values after chemotherapy. Neubauer et al (30) compared the signal intensity ratio of DWI and contrast-enhanced T1-weighted imaging before and after chemotherapy in neuroblastoma. Although the signal intensity ratio decreased in both modalities, DWI exhibited a greater signal intensity ratio compared with contrast-enhanced T1-weighted imaging, suggesting that DWI not only quantifies the efficacy of chemotherapy in neuroblastoma but also more effectively reflects lesion characteristics. In another study, Privitera et al (35) investigated the correlation between ADC values and MIBG uptake values in high-risk neuroblastoma before and after chemotherapy, and they discovered that the 25th percentile ADC value after chemotherapy accurately identified areas of inactive tumor cells. Figure 5 shows the correlation between ADC map features and iodine 123–MIBG uptake after chemotherapy for high-risk neuroblastoma. These findings suggest that calculating ADC values of neuroblastoma after chemotherapy can aid in providing guidance for surgical resection.

Figure 5:

Images in a 20-month-old male patient with high-risk neuroblastoma before and after chemotherapy. (A, B) Axial T2-weighted image and apparent diffusion coefficient map of the primary tumor (outlined in yellow) before chemotherapy, respectively. (C, D) Axial T2-weighted image and axial apparent diffusion coefficient map registered with SPECT iodine 123 metaiodobenzylguanidine (¹²³I-mIBG) after chemotherapy, respectively. The orange liver region of interest was used for normalization, while different regions of interest within the tumor indicate varying levels of ¹²³I-mIBG uptake. The green, pink, blue, and red regions of interest represent high uptake, moderate uptake, low uptake, and no uptake, respectively. (Adapted, under a CC BY 4.0 license, from reference 35.)

In terms of prognosis prediction, Peschmann et al (5) found that lower baseline ADC values were associated with malignant progression and recurrence of the neuroblastoma through comparison of tumor ADC values at initial presentation and at 3-month follow-up. Additionally, during the course of treatment, there was a correlation between ADC values and patients’ event-free survival, with a decrease in ADC values being an indicator of poor prognosis. This indicates the potential of ADC values to predict patient prognosis. Figure 6 shows baseline MR images of nonrelapsing neuroblastoma. These reports suggest that DWI quantitative indicators provide potential support in assessing the treatment response and prognosis of patients with neuroblastoma.

Figure 6:

Images of a nonrelapsing neuroblastoma lesion in a 5-week-old male infant. The images include (A) a baseline axial T2-weighted image, (B) an axial diffusion-weighted image with b value of 800 sec/mm², and (C) an axial apparent diffusion coefficient map. The baseline apparent diffusion coefficient from the large region of interest was 0.883 × 10⁻³ mm²/sec. (D) Locally enlarged image of C shows the different delineated regions of interest (ROIs) for measuring apparent diffusion coefficient values. (Adapted, under a CC BY 4.0 license, from reference 5.)

Distant Metastasis Detection and Skeletal Scoring

In the quantitative analysis of neuroblastoma, DWI exhibits a level of performance comparable to nuclear medicine imaging techniques and can serve as a complementary imaging biomarker. Whole-body DWI with background body signal suppression (DWIBS) can reduce the visibility of nonrelevant background signals, allowing for clearer visualization of lesions by minimizing interference from surrounding tissues. Compared with PET imaging, DWIBS demonstrates similar sensitivity but lower specificity and overall accuracy in helping detect bone and lymph node metastases (36). For skeletal scoring of high-risk neuroblastoma, the use of radiolabeled MIBG has prognostic value and can be used for risk stratification of neuroblastoma. Through investigating the correlation between MIBG skeletal scoring and whole-body DWI quantitative analysis, Gassenmaier et al (37) found that the International Society of Pediatric Oncology European Neuroblastoma Research Network skeletal scoring system was equally applicable to whole-body MRI compared with MIBG scintigraphy; however, whole-body MRI could result in slightly higher skeletal scores, potentially due to its higher spatial resolution.

Intravoxel Incoherent Motion and Diffusion Kurtosis Imaging in Neuroblastoma

Although DWI based on a monoexponential model can quantitatively assess the movement of water molecules within microstructures, it can be influenced by microcirculation blood perfusion and cannot fully exploit the abundant data information extracted from DWI. In contrast, intravoxel incoherent motion DWI (IVIM-DWI) based on a biexponential model calculates the true diffusion coefficient, pseudodiffusion coefficient, and perfusion fraction using different b values, which can more accurately reflect the diffusion and perfusion status within tissues and thus compensate for the shortcomings of conventional DWI (38). Although the application of IVIM-DWI has been reported in adult diseases, investigation on its use in childhood neuroblastoma is relatively limited. Meeus et al (33) indicated that the pseudodiffusion coefficient and perfusion fraction of neuroblastoma were significantly higher than those of Wilms tumor. This difference may stem from the tendency of neuroblastoma to encase surrounding blood vessels, leading to increased microcirculation perfusion within the tumor tissue. Additionally, histogram-derived parameters of pseudodiffusion coefficient and perfusion fraction (such as skewness, kurtosis, and entropy) were also helpful in distinguishing between neuroblastoma and Wilms tumor. Specifically, the skewness of pseudodiffusion coefficient distribution was higher in Wilms tumor, while the kurtosis and entropy of perfusion fraction value distribution were relatively higher in neuroblastoma, possibly related to the heterogeneity and irregularity of vascularization within neuroblastoma.

Conventional DWI assumes that water molecules diffuse according to a Gaussian distribution, and with increasing b values, the DWI signal exhibits a monoexponential decay. However, due to the substantial heterogeneity of tumor tissue, the diffusion of water molecules does not follow a simple Gaussian distribution (39). Therefore, diffusion kurtosis imaging (DKI) has been proposed based on non-Gaussian diffusion of water molecules. In DKI, the key measurement indicators usually include kurtosis and diffusion coefficient. Although the application of DKI has not been reported in pediatric thoracoabdominal neuroblastoma, it has been demonstrated to be significantly correlated with histopathologic characteristics in olfactory neuroblastoma. Xiao et al (40) found that the kurtosis of olfactory neuroblastoma was significantly higher than that of nasal squamous cell carcinoma, based on DKI. The AUC and accuracy of the kurtosis for distinguishing between these two tumors were 0.87 and 80%, respectively. Future investigations are needed to further explore whether IVIM-DWI and DKI can provide additional clinical value for the diagnosis and risk stratification of neuroblastoma compared with conventional DWI.

MR Spectroscopy in Neuroblastoma

MR spectroscopy (MRS) utilizes the spin properties of specific atomic nuclei (such as hydrogen 1 [¹H], phosphorus 31, carbon 13, and fluorine 19) to noninvasively quantify the metabolite content within tissues. In comparison to traditional anatomic imaging, MRS enables early and quantitative analysis of metabolic changes in lesions during biochemical processes, such as amino acids, lipids, lactate, and so forth, before morphologic changes occur (41). It holds potential applications in tumor diagnosis and differential diagnosis, early monitoring of treatment response, assessment of tumor cell activity, and identification of potential biomarkers. ¹H-MRS has been widely used in both adult and pediatric tumors, particularly in brain, breast, and prostate tumors.

Due to its sensitivity to motion artifacts, MRS is less commonly used in pediatric extracranial tumors such as neuroblastoma. An investigation conducted by Kohe et al (42) suggested that neuroblastoma exhibited higher levels of myoinositol and glutamate-glutamine ratio, indicating metabolic differences in embryonic tumors located outside the central nervous system. Additionally, Peet et al (43) found that neuroblastoma cell lines with MYCN amplification showed significantly elevated phosphocholine to total choline and taurine to total choline ratios, while glycerophosphocholine to total choline ratio decreased significantly, suggesting the potential value of MRS metabolite analysis in identifying MYCN amplification status. Although an animal experiment showed that ¹H-MRS could predict the response or resistance of neuroblastoma to standard chemotherapeutic agents such as platinum-based drugs (44), the clinical utility of MRS in neuroblastoma still requires further research and validation.

Arterial Spin Labeling Imaging in Neuroblastoma

Arterial spin labeling (ASL) imaging can noninvasively assess tissue perfusion by measuring tissue blood flow without the need for injection of contrast agents. Because gadolinium-based contrast agents have the potential risk of causing brain deposition in pediatric patients, and a previous study shows that contrast-enhanced MRI does not provide additional diagnostic value for identifying IDRFs in neuroblastoma (45), the application of contrast-enhanced MRI in pediatric neuroblastoma requires further research and more evidence. Compared with contrast-enhanced MRI using gadolinium-based contrast agents, ASL MRI does not require exogenous contrast agents. The basic principle of ASL MRI involves utilizing the MR pulse to label the spin state in arterial blood flow to obtain a labeled image. Subsequently, this labeled image is subtracted from the unlabeled control image to acquire a perfusion-weighted difference image, thereby enabling the measurement of perfusion parameters (46).

Harteveld et al (47) explored the feasibility of using multidelay pseudocontinuous ASL MRI in pediatric patients with common abdominal solid tumors. They found that the perfusion-weighted signals of neuroblastoma and Wilms tumor were lower than those of the kidneys, and there were differences in perfusion-weighted signals within different regions of the tumors. Figure 7 illustrates the correlation between high intratumoral signal intensity on perfusion-weighted images and low intratumoral signal intensity on T2-weighted images. Additionally, the authors compared ASL MRI with contrast-enhanced T1-weighted imaging and found that the perfusion-weighted signals within the tumors correlated with the degree of enhancement with gadolinium-based contrast agents. This suggests that ASL MRI may assist in noninvasive perfusion measurements of pediatric abdominal solid tumors without introducing gadolinium-based contrast agents. A previous study demonstrated that improved tumor perfusion can enhance drug uptake in neuroblastoma, thereby improving treatment efficacy and reducing the likelihood of systemic toxicity from chemotherapy drugs (48). Therefore, ASL MRI may help monitor changes in blood flow perfusion within neuroblastoma to determine the optimal timing for treatment.

Figure 7:

(A, B) Perfusion-weighted signal (PWS) and T2-weighted (T2W) images in an 8.3-year-old female patient and a 2.2-year-old female patient, respectively, with Wilms tumor. The images show that the high perfusion-weighted signal areas within the tumor correspond to the low signal intensity areas on the T2-weighted image (white arrowheads). (C) Perfusion-weighted signal and T2-weighted images in a 3.8-year-old male patient with neuroblastoma shows that the marked high-signal descending aorta and right renal artery on the perfusion-weighted signal image correspond to the regions on the T2-weighted image (white arrowheads). (Adapted, under a CC BY 4.0 license, from reference 47.)

APTw Imaging in Neuroblastoma

APTw imaging is a common technique of chemical exchange saturation transfer (CEST) imaging. CEST is a relatively new molecular imaging technique used to measure chemical exchange between endogenous molecules and water protons. Although the proton resonance frequencies of these endogenous molecules differ from those of water molecules, their proton content is relatively low, making them difficult to be detected directly in conventional MRI. CEST imaging indirectly helps detect the presence of these molecules by suppressing water proton signals and enhancing endogenous molecule proton signals (49). APTw imaging describes the distribution and metabolic information of biologic molecules by exploiting the chemical shift of protons in amide compounds. In APTw imaging, the excitation of protons in amide compounds is first performed using MRI techniques, and then the distribution and concentration of amide compounds in tissues can be inferred by observing changes in signal intensity at different chemical shifts (50). Therefore, APTw imaging holds potential applications in tumor imaging and could aid in the diagnosis and treatment monitoring of tumors.

APTw imaging indicators have been shown to correlate with the pathologic characteristics and risk stratification of neuroblastoma. In an animal experiment, Tanoue et al (51) found that APTw imaging effects not only reflect the proliferative potential of tumor cells in neuroblastoma but also indicate the presence of hemorrhage within the tumor. Jia et al (52) analyzed 57 cases of common solid tumors in the pediatric abdomen, including neuroblastoma, hepatoblastoma, and Wilms tumor, and analyzed the differences in APTw imaging quantitative indicators among different risk groups. They found significant statistical differences in APTw imaging quantitative indicators between high-risk neuroblastoma and low-risk neuroblastoma, but not in hepatoblastoma and Wilms tumor. The AUC for distinguishing low-risk neuroblastoma and high-risk neuroblastoma based on APTw imaging quantitative indicators reached 0.93. Figure 8 shows that neuroblastoma lesions in different risk groups have different APTw imaging signals, with the intermediate-risk and high-risk lesions displaying higher APTw imaging signals. Although neuron-specific enolase has clinical diagnostic value for neuroblastoma, the study by Jia et al (4) showed that there was no evidence of a difference in neuron-specific enolase between high-risk neuroblastoma and non–high-risk neuroblastoma; however, APTw imaging quantitative indicators showed significant value in neuroblastoma risk stratification. Given the potential for more IDRFs in high-risk neuroblastoma, the comparison between anatomic imaging and APTw imaging for differentiating high-risk from non–high-risk neuroblastoma remains a valuable area for future exploration.

Figure 8:

Axial T1-weighted (T1W), T2-weighted (T2W), and amide proton transfer (APT) images in patients with (A) very-low-risk, (B) low-risk, (C) intermediate-risk, and (D) high-risk neuroblastoma. (Adapted, under a CC BY-NC-ND 4.0 license, from reference 4.)

Native T1 Mapping in Neuroblastoma

Native T1 mapping is used to measure tissue T1 relaxation time, which is the time required for a substance to return to its equilibrium after being stimulated by a pulse signal. Because different tissues have distinct T1 relaxation times, native T1 mapping can quantitatively characterize tissue properties. During treatment, tissue T1 relaxation time may change. The treatment response can be monitored by performing native T1 mapping, allowing for adjustments to treatment plans. Zormpas-Petridis et al (53) found that native T1 mapping indicators correlated with the differentiation degree of tumor cells in a mouse model of neuroblastoma. Their study showed that regions with higher T1 values had an abundance of proliferating undifferentiated neuroblastoma cells, while regions with lower T1 values were richer in apoptotic or undifferentiated neuroblastoma cells. Additionally, native T1 mapping could detect the response of neuroblastoma to MYCN-targeted small molecule inhibitors. These results suggest that native T1 mapping not only helps assess the histologic characteristics of neuroblastoma but also assists in evaluating its response to targeted therapy.

Wang et al (54) discovered that the CT-based extracellular volume fraction of neuroblastoma was correlated with histologic characteristics, MYCN amplification, and risk stratification in neuroblastoma. Additionally, there was a correlation between CT-based extracellular volume fraction and the primary tumor response to treatment in neuroblastoma (55). Other studies have demonstrated that native T1 mapping indicators help quantitatively reflect the extracellular matrix (56). Thus, native T1 mapping may play a more important role in the diagnosis and risk stratification of neuroblastoma. However, further investigations are needed to confirm these hypotheses.

Conclusion

Although anatomic imaging aids in identifying IDRFs and evaluating treatment response in neuroblastoma, it has limitations in predicting prognosis. Specifically, the ability of IDRFs to predict overall survival in patients with neuroblastoma is moderate, highlighting the need for additional quantitative imaging biomarkers to enhance predictive tools. While quantitative MRI is promising, there is relatively less research and clinical validation in the context of neuroblastoma. More studies are needed to establish standardized protocols and validate its clinical utility in predicting prognosis in these patients. It is also worth noting that current studies on quantitative MRI in neuroblastoma do not include comparisons with anatomic imaging. Therefore, further investigation is needed to determine whether quantitative MRI can provide independent imaging biomarkers with additional clinical value beyond anatomic imaging in neuroblastoma.