2026 SIOPEN建议：高危神经母细胞瘤的影像学检查

ABSTRACT

Neuroblastoma is the most common extracranial solid tumor in early childhood. Its clinical behavior is highly variable, ranging from spontaneous regression to fatal outcome despite intensive treatment. The International Society of Pediatric Oncology Europe Neuroblastoma Group (SIOPEN) Radiology and Nuclear Medicine Specialty Committees developed consensus-based imaging recommendations for high-risk neuroblastoma at first presentation and throughout follow-up. These expert opinions aim to improve reproducibility and enhance diagnostic performance in order to advance the current standard of care and assist in the development of future trials.

1 | Introduction

Neuroblastoma is the most common extracranial solid malignancy of childhood [1], and about half of the patients have metastatic disease at diagnosis [2]. Within Europe, clinical trials

in neuroblastoma aiming to improve care are coordinated by the International Society of Pediatric Oncology Europe Neuroblastoma Group (SIOPEN). The current High Risk-Neuroblastoma (HR-NBL2) protocol (NCT04221035) includes all patients above one year of age with metastatic neuroblastomas and all

neuroblastomas with MYCN amplification (except INGRSS-stage L1). Histological confirmation and molecular biology classification are usually obtained by imaging-guided percutaneous biopsy [3]. Treatment consists of a combination of chemotherapy, surgery, high-dose chemotherapy and autologous stem cell rescue, radiotherapy, followed by immunotherapy. Multimodality imaging, incorporating anatomical and metabolic techniques, plays an essential role. Imaging is performed at diagnosis, end-of-induction chemotherapy, for preoperative planning and before radiation therapy, before maintenance, at the end of treatment, during follow-up after treatment, and in case of relapse.

The aim of this paper is to provide an expert statement on imaging acquisition and interpretation at diagnosis and during treatment for patients with high-risk neuroblastoma to allow standardization, improve reproducibility, and enhance diagnostic performance. These expert statements were developed to improve current care and assist in the development of future trials.

2 | Materials and Methods

Twelve members from the Nuclear Medicine and Physics (NM) and Radiology (Rad) SIOPEN Specialty Committees met both in person and online from January 2023 to June 2024 to discuss current evidence, clinical practices, challenges, and developments in the respective fields. The aim of this working group was to identify aspects of imaging in patients with high-risk neuroblastoma that are consistently relevant, yet exhibit significant variability in implementation and interpretation, variability that necessitates harmonization through expert consensus. After identifying these key areas, the working group was divided into smaller groups with expertise in specific subtopics to conduct a narrative literature search to identify the most relevant publications. Based on the available evidence, recommendations were formulated for each chapter. These recommendations were revised as needed to reflect a final consensus within the working group through open discussion during several online meetings. ASL and JMMR subsequently rated the quality of evidence for each final recommendation according to the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) system [4]. Finally, all members of each Specialty Committee were invited to vote on the recommendations within their respective specialty of radiology or nuclear medicine. The standard Strength of Recommendation (SOR) system was adopted [5]. Each participant was required to independently describe their own agreement with the recommendation using a 4-point Likert scale through an online form. A score of 1 was defined as a strong recommendation, while a score of 4 was defined as a weak recommendation. Overall, SOR was calculated as the arithmetic mean of the individual scores. As previously described, an overall SOR score of 1-1.49 was considered a very strong recommendation, a score of 1.5-1.99 was considered a strong recommendation, a score of 2-2.99 was considered a moderate recommendation, and a score of 3 and above was considered a weak recommendation [4, 5].

In total, 25 committee members (12 NM members, 10 Rad members, and 3 members from both Rad and NM) completed the online voting form.

3 | Recommendations

3.1 | Imaging Techniques

3.1.1 | Ultrasound

Within the SIOPEN framework, ultrasound (US) is the primary imaging investigation in the initial assessment of suspected neuroblastoma of the abdomen, pelvis, and neck. Cross-sectional imaging is required for the assessment and local tumor extent, relationship to adjacent tissues, and vascular encasement. The International Neuroblastoma Risk Group (INRG) staging system recommends contrast-enhanced computed tomography (CT) or magnetic resonance imaging (MRI) depending on the primary site of disease [6]. For assessment of distant metastases, functional imaging using nuclear medicine techniques is recommended as the first line.

3.1.2 | Computed Tomography

CT is a readily available modality. A main advantage of CT is its very short acquisition time, usually allowing successful acquisition without sedation or anesthesia. CT radiation dose must always be optimized, but it is much less alarming when put in context of radiation dose from subsequent radiotherapy treatment usually recommended in these high-risk patients. CT offers evaluation of the primary mass, tumoral morphology (e.g., calcification), and its relationship with adjacent organs and vessels. A single-phase post-contrast CT is often sufficient. When visualization of both arterial and venous anatomy is required, a "split bolus" administration of iodinated contrast can be employed, optimizing radiation dose [7]. In cases of cervicothoracic and thoracic neuroblastoma, CT may be preferred to evaluate the relationships with airways and major vessels, while MRI is superior to CT in assessing intraspinal invasion (so-called "dumbbell" tumors), chest wall invasion, and potential involvement of the brachial plexus [6].

3.1.3|MRI

Where available and feasible, MRI is preferred over CT because of its higher intrinsic soft-tissue contrast, reasonable spatial resolution, avoidance of radiation and potential for additional functional information (e.g., diffusion-weighted imaging (DWI)). MRI exhibits higher sensitivity compared with CT for assessing bone marrow disease, liver metastases, soft tissue, and intraspinal extension. Image-defined risk factors (IDRF) can be assessed using MRI during preoperative work-up and have been demonstrated to be non-inferior to CT [8]. The main drawback of MRI is its long acquisition time (around 45–60 min) and the associated need for sedation or anesthesia in younger patients. The value of gadolinium-based contrast agents to assess neuroblastoma extension at staging remains under active debate. Use of post-contrast images may improve assessment of infiltration into adjacent tissues and tumor vascularity [6]. However, recent studies demonstrate that gadolinium does not improve tumor size measurement or radiologist agreement on IDRF or metastatic disease [9].


FIGURE 1 | A four-year-old girl with stage M, MYCN amplified, 16q loss, 17q gain HR-NBL. Transverse T1w MR image after intravenous gadolinium contrast (a) and transverse ADC map (b) show a large heterogeneously enhancing upper abdominal mass lesion encasing retroperitoneal vessels and displacing the abdominal structures. The mean ADC value from the non-necrotic solid component of the lesion in the midline measures $0.7 \times 10^{-3}$ mm$^{2}$/s. MRI abdomen after chemotherapy induction (c and d) shows partial response to treatment; however, the large solid component in the midline has significantly reduced in size with the mean ADC from this non-necrotic solid component now measuring $1.6 \times 10^{-3}$ mm$^{2}$/s, indicating a decrease in cellularity of the residual tumor.

DWI should be included in the MRI protocol. Preliminary studies suggest DWI can provide additional information about tumor cellularity and differentiation. Retrospective studies have shown that apparent diffusion coefficient (ADC) values can reliably distinguish between different histopathological subtypes of neuroblastic tumors, with poorly differentiated and undifferentiated areas showing lower ADC values than those with differentiated composition [10–13]. Given the heterogeneity of these tumors, ADC mapping could help guide percutaneous biopsies by identifying the most malignant regions [3]. Furthermore, DWI has potential utility in detecting distant metastases, monitoring treatment response, and predicting outcome (Figure 1) [10, 14, 15]. These preliminary retrospective data do not include validated criteria for objective response assessment according to standardized measures such as Response Evaluation Criteria in Solid Tumors (RECIST).

Total-body MRI is able to assess the extent of bone marrow metastases. This may detect meta-iodobenzylguanidine (mIBG)-negative bone marrow lesions, highlighting cases that could benefit from additional imaging with a positron emission tomography (PET) tracer. However, total-body MRI is not universally available across Europe and is not part of the SIOPEN HR-NBL2 protocol. The additional anatomical coverage and sequences will

increase acquisition time, with implications for resource and tolerability by patients. Furthermore, the role of MRI in response assessment of bone marrow metastases is uncertain, as the reliability of ADC measurements in small lesions is questionable.

It is advisable to use the same MRI acquisition protocol for diagnosis and response assessment (Table 1).

Recommendation 1.1: Local tumor staging is performed with cross-sectional imaging, preferably with MRI. MRI as proposed is especially recommended for every tumor with spinal invasion, brachial plexus, and skull base involvement. (GRADE D, SOR 1.17 (very strong recommendation)

Recommendation 1.2: MRI-DWI is of additional value for the detection of small metastatic lesions. GRADE B, SOR 1.83 (strong recommendation)

Recommendation 1.3: MRI-DWI is of additional value to detect tumor heterogeneity. GRADE B, SOR 1.58 (strong recommendation)

Recommendation 1.4: MRI-DWI is of additional value to assess response to treatment. GRADE B, SOR 2.17 (moderate recommendation)

Recommendation 1.5: Whole-body MRI is advised as an addition to nuclear medicine modalities to assess distant metastases and identify potential mIBG-negative disease. GRADE B, SOR 2.09 (moderate recommendation), one "disagree" vote

4 of 14

Pediatric Blood & Cancer, 2026

TABLE 1 | MRI protocol.

|Basic Protocol Local Regional Disease|||||| |---|---|---|---|---|---| |Sequence|Orientation|Anatomical coverage|Pixel spacing (mm$^{2}$)|Slice thickness|Comment| |T2w|3D isotropic coronal|Primary tumor body compartment|0.625-0.8|0.625-0.8 or 3 (2D)|Can be substituted by the T2 sequence in 2 orthogonal planes. Option HASTE with respiratory triggering| |T2w TSE, FS|Axial|Primary tumor body compartment|1.25|4|With respiratory triggering. Option: radial K-space sampling| |T1w Vibe Dixon|Axial|Primary tumor body compartment|1.0|3|If Dixon is not available, T1 with and without fs. Option radial sampling| |Ce T1w Vibe Dixon|Axial|Primary tumor body compartment|1.0|3|If Dixon is not available, T1 with fs. The use of iv contrast is under debate| |DWI|Axial|Primary tumor body compartment|1.0|3|b value of 0 or 100 and 800 or 1000| |Total Body for Distant Metastases|||||| |T2w FS|Sagittal|Spine|0.7|3|DIXON or STIR| |T1w|Sagittal|Spine|0.5|3|DIXON or TSE| |Ce 3D T1w|Sagittal|Head|1.0|1|| |T2w TSE|Axial|Head|0.5|4|| |T2w FS (STIR or DIXON)|Coronal|Total body|0.8|4|For detection of distant metastases| |DWI|Axial or coronal|Total body|1.0|3|b value of 0 or 100 and 800 or 1000| |Optional|||||| |DTI|Axial|Brachial or sacral plexus|||In case of brachial or sacral plexus involvement|

3.2 | Nuclear Medicine Modalities

3.2.1 | [123I]mIBG

[123]Imeta-iodobenzylguanidine (mIBG) scintigraphy is currently the standard nuclear medicine examination for patients with neuroblastoma. As a norepinephrine analogue, it is actively absorbed by neuroblastoma cells via the norepinephrine transporter. Approximately 90% of neuroblastoma cases exhibit an affinity for mIBG, making it a valuable ligand for neuroblastomaspecific imaging and radionuclide therapy with [131]ImIBG [13]. [123]ImIBG scan comprises a planar whole-body scintigram 24 h after tracer injection and should always be followed by additional single photon emission computed tomography combined with CT (SPECT/CT) (Table 2). SPECT/CT should cover the entire trunk and be extended to other areas where further information is required.

Patient preparation, including thyroid blockade, the recommended tracer activity, image acquisition, and reconstruction, should follow procedural guidelines, e.g., those of the European Association of Nuclear Medicine (EANM) [17–19].

Scan durations are typically around 90 min for whole-body scintigraphy with trunk SPECT/CT. Appropriate use of anesthesia/sedation should be considered to ensure optimal scan quality.

Diagnostic quality contrast-enhanced CT can be acquired on SPECT/CT scanners if this will spare an additional anesthesia/sedation. Where both nuclear medicine imaging and cross-sectional imaging are required in patients needing sedation or anesthesia, coordination between radiology and nuclear medicine departments can streamline the acquisitions, reducing burden to medical/anesthetic staff, patients, comforters, and carers.

3.2.2| MIBG-negative Disease

The definition of mIBG-negative disease has not been precisely described in the SIOPEN HR-NBL2 protocol. A pragmatic approach for its definition can be recommended at a per-lesion level. Additionally, photopenia associated with necrotic tumor parts should not be regarded as mIBG-negative disease. If one or more mIBG-negative lesions are suspected, the recommendation for PET imaging should be guided by the potential clinical consequences. SIOPEN HR-NBL2 protocol recommends [18F] fluorodeoxylglucose (FDG) PET/CT in case of mIBG-negative disease. A major disadvantage of [18F]FDG PET is treatment-related reactive high bone marrow FDG uptake, hampering skeletal disease response evaluation in follow-up studies. The SIOPEN HR-NBL2 protocol recommends using one of the two modalities during treatment and follow-up (either mIBG scintigraphy or [18F]FDG PET/CT) based on their potential to influence clinical decision. In future trial protocols, more specific tracers (e.g., somotostatin receptor (SSR)-targeting tracers ([68Ga]Gallium-DOTATOC), [18F]mFBG, or [18F]Fluorodihydroxyphenylalanine (F-DOPA)) may result in a more accurate assessment of disease. However, it is currently not known whether and how additional

lesions identified with PET imaging affect patient outcomes and whether they should therefore alter the treatment.

3.3 | PET/CT Imaging

In recent years, there has been an increasing use of PET imaging in neuroblastoma. PET/CT has several advantages over planar scintigraphy and SPECT/CT, including superior spatial resolution and sensitivity, larger tomographic (3D) coverage, accurate quantification of tracer uptake, shorter acquisition times, which reduce the need for anesthesia/sedation. Imaging protocols for PET tracers often allow for single-day studies and are more convenient for patients and families than the two-day [123]mIBG scintigraphy. It is also recognized that 10% of neuroblastoma patients have mIBG-negative disease, highlighting the need to explore alternative tracers in these cases (Table 2).

The choice of PET tracer to be used may depend on factors such as:

- The location of lesions (skeletal lesions or liver lesions may especially benefit from a more specific tracer than FDG)

- Previous treatment (bone marrow activation after chemotherapy is especially problematic for bone marrow assessment with [$^{18}$F]FDG)

- Tracer availability

- The requirement for follow-up examinations (the tracer should also be adequate and available for whole-body restaging in future scans)

3.3.1 | [¹⁸F]FDG PET

[$^{18}$F]fluorodeoxyglucose (FDG) is a commonly available PET tracer with established use in oncological imaging. [$^{18}$F]FDG PET is the functional imaging technique of choice in current international guidelines for disease assessment and response evaluation in [$^{123}$I]mIBG-negative neuroblastoma cases. [$^{18}$F]FDG PET findings have been shown to be prognostic in patients with neuroblastoma [20].

FDG uptake can be observed in inflammatory lesions, normal and cytokine-stimulated bone marrow, and in other neoplasms than neuroblastoma. Therefore, careful interpretation of [18]FDG PET by an experienced reader is prudent, especially given the potential for confounding inflammatory uptake around the time of post-surgery and radiotherapy treatment. Detailed methodology for patient preparation, procedure, and imaging acquisition is available from international procedural guidelines and should be adhered to [21, 22].

3.3.2 | $^{68}$Ga-labeled Somatostatin Receptor (SSR)-targeting PET Tracers

Somatostatin receptor (SSR) PET can provide complementary information to [$^{123}$]mIBG imaging, due to heterogeneous distribution of the respective cellular targets between different

6 of 14

6 of 14

Pediatric Blood & Cancer, 2026

References

neuroblastoma lesions even in the same patient [23, 24]. Direct comparison of both modalities has been performed only in small case series. These indicate that SSR PET is at least as sensitive as [$^{123}$I]mIBG scintigraphy for the detection of neuroblastoma metastases [25–27]. SSR PET tracer use in neuroblastoma offers potential for molecular radiotherapy applications; e.g., the ther-anostic pairing of $^{68}$Ga-/$^{177}$Lutetium(Lu)-labeled SSR ligands, utilized in ongoing trials [28].

3.3.3 | [¹⁸F]F-DOPA

[$^{18}$F]F-DOPA is an amino acid analogue that is taken up into cells via dopamine transporters, with demonstrated use in neuroendocrine tumors, paragangliomas, and phaeochromocytomas. Prospective studies in neuroblastoma have demonstrated that [$^{18}$F]F-DOPA PET is more sensitive than [$^{123}$I]mIBG imaging in the detection of metastases [29–31]. However, [$^{18}$F]F-DOPA is not part of study protocols due to its limited availability and lack of theranostic applicability.

3.3.4 | mIBG Analogues PET Tracers

[18F]mFBG PET: Three recent pilot studies provide initial evidence that [18F]mFBG PET/CT achieves higher SIOPEN or Curie scores than [123]mIBG imaging [32–35]. In addition to the improved spatial resolution of ~3–4 mm offered by modern PET scanners, lesion contrast is significantly enhanced by the more rapid [18F]mFBG blood clearance (compared with [123]mIBG) [36]. The single-day [18F]mFBG-PET/CT protocol simplifies and accelerates workflows compared with the two-day [123]mIBG protocol. Furthermore, [18F]mFBG does not require thyroid blockage [33]. Similar to [123]mIBG, [18F]mFBG targets the norepinephrine transporter, where it demonstrates comparable affinity and cellular uptake [37, 38]. Imaging with [18F]mFBG PET could therefore maintain the theranostic potential in combination with [131]mIBG therapy.

[124I]mIBG PET: While availability is limited, data from a small number of patients with neuroblastoma suggest that [124I]mIBG PET is more sensitive than [123I]mIBG imaging for the detection of metastases but results in higher radiation exposure [39]. Being chemically identical to [131I]mIBG and with a suitably long half-life, there may be a theoretical advantage for theranostic applications, e.g., predictive personalized mIBG molecular radiotherapy dosing. Further studies, including validation, are needed.

Recommendation 1.6: [123]I]mIBG scintigraphy with SPECT/CT remains the first-line nuclear imaging modality due to its presence in treatment study protocols and its role in clinical decision-making. The procedure should be performed according to current guidelines [17]. GRADE B, SOR 1.07 (very strong recommendation).

Recommendation 1.7: An mIBG-negative lesion should be defined by the absence of discernible mIBG accumulation in a viable tumor lesion that fulfills both of the following criteria:

1. Sufficient size considering the limited spatial resolution of mIBG imaging

2. Unequivocal or probable lesion on cross-sectional imaging

GRADE B, SOR 1.71 (strong recommendation)

Recommendation 1.8: Additional PET imaging for (suspected) mIBG-negative disease is advised in case of staging or treatment consequences. GRADE B, SOR 1.64 (strong recommendation)

Recommendation 1.9: In case of mIBG-negative disease, a more specific PET tracer (non-hierarchical order: [$^{18}$F]mFBG, [$^{18}$F]F-DOPA or SSR ligand) should be preferred over [$^{18}$F]FDG if allowed within the treatment protocol. GRADE C, SOR 2.00 (moderate recommendation)

3.4 | Reporting

Cross-sectional and nuclear medicine imaging should be interpreted together to maximize accuracy and certainty of reports [40]. When imaging studies are performed on different dates, the subsequent report should reference the initial report, and the second reader should collaborate with the first to resolve any discrepancies.

Recommendation 1.10: Reports on nuclear medicine or radiological imaging for neuroblastoma should indicate the level of certainty and—if present—the sources of uncertainty.

Consider using these levels of certainty:

Consistent with (>90%), suspicious for (~75%), possible (50%), less likely (~25%), unlikely (<10%)

GRADE B, SOR 1.33 (very strong recommendation)

Recommendation 1.11: Whenever possible, nuclear medicine and radiological imaging should be interpreted together to maximize accuracy and certainty of reports. GRADE B, SOR 1.33 (very strong recommendation)

4 | Staging

4.1 | Local Tumor

Staging for neuroblastoma nowadays follows the INRG staging system (INRGSS) proposed in 2009 as a more consistent pretreatment staging system [41] to replace the older International Neuroblastoma Staging System (INSS). INRGSS uses a combination of clinical, imaging, and biological data to assign one of four stages to the neuroblastoma. The first two stages (L1 and L2) belong to a localized tumor [42]. IDRFs were developed to describe the relationship of the tumor with vital structures

associated with a high risk of surgical complications (major vessels, nerves, and airways) or risk of incomplete resection. Although IDRFs do not modify the stage in metastatic patients (stages M and MS), primary tumor surgery may be postponed due to expected complications in the presence of certain IDRFs. IDRFs should therefore be methodically assessed according to the criteria proposed by the INRG imaging committee [6], and all identified IDRFs should be listed and described in the imaging report.

[123]mIBG imaging, even with SPECT/CT, does not usually add to anatomical imaging in assessing the local tumor extent and locoregional lymph node metastases. However, areas of high tracer uptake can be preferred localizations for a guided biopsy [3].

Recommendation 2.1: IDRFs should be assessed and reported in patients with neuroblastoma. GRADE B, SOR 1.08 (very strong recommendation)

4.2 | Distant Metastases

At the time of diagnosis, 35% of patients have regional lymph node involvement, and 50% present with distant metastases [2]. Both lymphatic and hematogenous spread are observed. Bone marrow, non-regional lymph nodes, and liver are the most common metastatic sites [42]. The definition of locoregional lymph nodes according to the INRGSS differs from that of the TNM system, which restricts locoregional nodes to those directly draining the primary tumor site. In abdominal neuroblastoma, all abdominal lymph nodes are classified as locoregional, regardless of their drainage pattern. Additionally, lower medialt nodes are considered locoregional in upper abdominal tumors, and inguinal nodes are classified as locoregional in pelvic primaries [41]. While imaging often detects diffuse and/or focal bone marrow disease, bone marrow involvement, and its response to treatment, according to the INRGSS [41] and the International Neuroblastoma Response Criteria (INRC) [43] is determined by morphology on bone marrow smears and biopsies (accompanied by immunocytology or immunohistochemistry). Both classification systems categorize imaging-detected lesions in the skeleton as “bone lesions,” even though these findings are presumed to represent, from a pathophysiological standpoint, infiltration originating in the bone marrow compartment. In the current article, the term “bone marrow involvement” is used primarily in its imaging-based sense [44, 45]. Metastases to the central nervous system, lungs, pleura, or skin occur more rarely, whereas certain locations, such as pancreas, testis, ovaries, kidneys, or breast tissue, may be noted in exceptional cases [46].

4.3 | Image Interpretation on Cross-Sectional Imaging

In general, metastases will demonstrate similar imaging characteristics as the primary tumor on either CT or MRI. Fine calcifications could serve as a reliable diagnostic feature on US and CT. On MRI, DWI likely increases the detection rate

and assists in characterizing (especially small) lesions [47]. Physiological diffusion restriction occurs in, for example, lymph nodes, gonads, and spleen. Differentiating reactive from involved lymph nodes can be challenging; ancillary features such as size, distribution, calcifications, and signal heterogeneity may assist. In upper abdominal tumors, a common location for distant lymph node metastases is the left supraclavicular node (Virchow node).

With MRI, bone marrow metastases are usually seen as low T1, high T2/STIR signal lesions with heterogeneous enhancement and strong diffusion restriction [48]. In infants and young children, the differentiation between normal high cellular red bone marrow and metastases can be challenging. However, the T2 signal intensity of normal red marrow is usually less distinct than malignant infiltration, best seen on T2-weighted sequences with fat suppression. Furthermore, it can be helpful to relate the signal intensity of the bone marrow on T1-weighted sequences with adjacent muscles or vertebral disks [49]. In diffuse bone marrow involvement T1 signal is usually hypointense compared with muscles and intervertebral discs, in contrast to normal red bone marrow, which is mildly hyperintense relative to these tissues [50].

A definite bone marrow metastasis is defined as a focal lesion with low T1 signal (similar or lower than surrounding muscles), high T2 signal on fat-suppressed images, and strong diffusion restriction (similar to the primary tumor). However, INRGSS criteria require bone marrow aspirates and biopsies to be performed to assess metastatic disease to the bone marrow [41].

4.4 | Image Interpretation in Nuclear Medicine (NM) Imaging

Image interpretation in neuroblastoma can reveal equivocal findings that need to be assessed for their level of certainty (Table 3).

4.5 | Scoring Systems

4.5.1 | [$^{123}$I]mIBG Scintigraphy

There are two widely used semi-quantitative scoring systems for [123]I[mIBG: the Curie score and the SIOPEN score (Figures 2 and 3). The SIOPEN score evaluates skeletal mIBG uptake on a 0-6 scale in 12 anatomical regions [51]. The Curie score evaluates nine skeletal regions, primary tumor uptake, and soft-tissue metastatic disease using a 0-3 scale [52]. Both scores have been validated as prognostic scores for [123]I[mIBG planar scintigraphy in neuroblastoma preinduction and post-induction therapy and are valid for clinical use [52, 53, 54]. Notably, the prognostic significance of the post-induction Curie score has been externally validated using SIOPEN trial data, and conversely, the prognostic value of the post-induction SIOPEN score using Children's Oncology Group (COG) trial results [53, 55]. Furthermore, both scores are highly correlated.

High or very high interrater agreement has been reported for the SIOPEN score, both before and after induction chemotherapy [51, 55–57]. However, certain aspects are potential causes for interrater

TABLE 3 | Summary of specific NM findings and their potential causes of variability in reading.

|High certainty findings|Potential causes of false-positive findings|Potential causes of rater variability—SIOPEN score| |---|---|---| |Primary tumor uptake of any tracer|Any liver uptake on [123]mIBG scintigraphy without correlation on CT or MRI|Planar versus SPECT/CT scoring| |A skeletal lesion detected by a specific $tracer^{a}$ in typical intensity and shape|“Lesion” uptake similar to the general level of background noise|Focal (score 1) versus diffuse (score 4): lack of clear criteria: size or degree of delineation of its borders| |Skeletal uptake in [18]F]FDG PET that correlates with MRI or a specific $tracer^{a}$|Superimposition of activity from other organs, physiological uptake in soft tissues, brown fat tissue and central venous line|Any false-positive finding| |Unequivocal uptake of any NM tracer of a soft-tissue lesion with a correlate on CT or MRI|Physiological [18]F]FDG bone marrow uptake or increased [18]F]FDG uptake due to bone marrow activation [18]F]FDG and [68]Ga]Ga-DOTA uptake in inflammatory/infectious lesions or in postsurgical changes||

$^{a}$Specific tracer: [123]mIBG, [18]mFBG, [68]Ga]Ga-DOTA, and [18]F]F-DOPA.


FIGURE 2 | SIOPEN and Curie score. (A) The SIOPEN skeletal score per segment is defined as: 0, no lesion; 1, 1 distinct spot; 2, 2 distinct spots; 3, 3 distinct spots; 4, >3 distinct lesions or diffuse involvement <50% of the segment; 5, diffuse involvement of 50%-95% of the segment; 6, diffuse involvement of the segment. (B) The Curie score for the skeletal segments: 0, no involvement; 1, 1 lesion; 2, >1 lesion; 3, diffuse involvement >50% of an individual site. Soft-tissue lesion "segment": 0, no soft-tissue lesions; 1, 1 soft-tissue lesion; 2, >1 soft-tissue lesion; 3, a soft-tissue lesion that occupies >50% of the chest or abdomen.

or intrater variability (Table 3). Curie scores have also shown good or very good agreement between readers [56].

Within the SIOPEN HR NBL2 protocol, the SIOPEN score has to be evaluated exclusively based on the planar whole-body scintigram to avoid an overestimation of the bone marrow involvement.

4.6 | PET Tracers

There is no validated scoring system for PET tracers in neuroblastoma.

Recommendation 2.2: Scoring for [123] mIBG scintigraphy (SIOPEN or Curie score) should only include lesions that are detected by planar images to ensure adherence with established thresholds. GRADE B, SOR 2.08 (moderate recommendation), one "disagree" vote

Recommendation 2.3: The SIOPEN and Curie score based on planar [$^{123}$I]mIBG scintigraphy can both be used for prognostic evaluation in routine clinical care. The same scoring system should be used during follow-up. GRADE B, SOR 1.57 (strong recommendation)

Recommendation 2.4: Lesions detected by [123] I] mIBG SPECT or SPECT/CT but not by the planar images should be indicated in the report. GRADE B, SOR 1.31 (very strong recommendation), one "disagree" vote

5 | Response Assessments

Response assessment of neuroblastoma, according to the INRC, includes anatomical imaging for primary and metastatic soft-


FIGURE 3 | Imaging comparison in a four-year-old boy with stage M neuroblastoma. On the left, planar whole-body scintigraphy performed with [123]mIBG shows tracer uptake. On the right, a maximum intensity projection (MIP) image from a PET scan with [18]F]mFBG demonstrates the same tracer distribution pattern, with intense uptake in neuroblastoma lesions in the abdomen and multifocal uptake in the skeleton. The PET image offers significantly better resolution, allowing clear delineation of individual lesions compared with the planar mIBG scan. The PET scan was completed in approximately 10 min without sedation, while the mIBG scan required approximately 2 h, including additional SPECT/CT imaging and the need for sedation or general anesthesia. Another important difference is the timing: the PET scan was performed 1-hour post-injection on the same day, whereas the mIBG scan required a 2-day procedure with imaging performed 1 day after tracer injection.

tissue disease, functional imaging methods for measuring soft tissue and skeletal disease, and bilateral bone marrow aspirates and trephine biopsies for assessing bone marrow involvement [43]. The INRC is used in all trials to report response to treatment.

While tumor volume was previously calculated using three-dimensional measurements, the INRC now suggests simply measuring the “single longest dimension in any orthogonal plane” of the non-nodal target lesions as per RECIST 1.1 [43, 58, 59]. Discrete, pathological appearing lymph nodes are measured using a single short-axis diameter.

Within the HR-NBL2 protocol, ultrasound is performed during induction chemotherapy for interim evaluation of primary tumor response. Response assessment with anatomical imaging and

functional imaging is advised at the end-of-induction chemotherapy, for preoperative planning, for post-operative evaluation of residual disease prior to radiotherapy planning and at the end of treatment. Molecular imaging and anatomical imaging should be interpreted in conjunction for the assessment of primary tumor and metastatic response.

5.1 | Preoperative Assessment

Contrast-enhanced CT might be preferred for preoperative planning, as MRI may underestimate true disease extent due to extensive treatment-induced changes, such as fibrosis and calcifications, which result in low and less conspicuous MRI signals compared with MRI at diagnosis [60]. However, this can largely be overcome by adding a T1- or T2-weighted sequence without fat suppression. The residual mass is more easily depicted by the contrast of signal intensities between the high T1 signal of the retroperitoneal fat and the low signal of the residual mass [47]. Preoperative CT has been proposed in posterior mediastinal neuroblastoma (T8-L1) for assessment of the arterial blood supply to the spinal cord (the artery of Adamkiewicz) [61]. However, the risk of injury to the spinal cord is very low in children, probably because of the abundant collateral blood flow that develops in response to the rather slow growth of most tumors in this location. Thus, angiography is not mandatory [6].

The major advantage of preoperative MRI is the additional value of DWI in detecting viable and poorly differentiated residual disease, which can be critical for guiding surgical planning and decision-making.

5.2 | Assessment of Residual Disease for Radiotherapy

A major advantage of MRI over CT is the better discrimination between post-operative changes, fibrotic tissue, and viable tumor. Within the HR-NBL2 protocol, all patients will receive 21.6 Gy radiotherapy of the preoperative tumor bed. Patients with macroscopic residual disease will be randomized for an additional 14.4 Gy boost. Macroscopic residual disease after surgical resection is defined as (1) MRI showing definite residual tumor, (2) [123]mIBG scan showing residual tumor, or (3) surgical report mentioning residual tumor. Surgeons describe two options for residual disease: excision with minimal residual tumor (< 5 cm$^{3}$) or incomplete excision (> 5 cm$^{3}$). In both instances, it is advised to correlate these estimated volumes with post-operative cross-sectional imaging.

In clinical practice, the German group has proposed a uniform scoring system, considering the surgical estimation of residual disease, NM and MRI findings. Early postoperative MRI (similar to the practice with brain tumor management) is suggested to increase diagnostic confidence in the differentiation between scar, residual tumor, and recurrence; however, the additional value of this early post-operative MRI has not yet been scientifically proven.

Reliable interpretation of NM images of the primary tumor bed after surgery requires:

- knowledge of the prior uptake within the primary tumor (i.e., positive/negative) for the specific tracer used for post-surgical imaging

- comparison with radiological imaging, as a correlate on CT or MRI, is usually required to identify residual primary tumor uptake with high certainty

Notably, non-specific tracer uptake in normal organs (e.g., from bowel loops filling the adrenal space after adrenalectomy) can mimic residual primary tumor uptake. After right adrenalectomy, focal liver uptake in segment 6 or 7, especially if incorrectly co-registered on SPECT/CT, can be misinterpreted as residual uptake in the adrenal surgical bed. Moreover, physiological tracer uptake of the remaining (contralateral) adrenal gland can vary over time, especially after unilateral adrenalectomy.

Recommendation 3.1: Either contrast-enhanced CT or MRI can be used for preoperative assessment of residual tumor, depending on the location of the tumor and local preferences. GRADE C, SOR 2.17 (moderate recommendation)

Recommendation 3.2: MRI is advised for assessment of possible residual tumor after surgery before radiotherapy. GRADE C, SOR 1.83 (strong recommendation)

Recommendation 3.3: The same tracer that is known to be positive for the primary tumor in the individual patient based on preoperative images should be used to assess possible residual tumor after surgery. GRADE D, SOR 1.29 (very strong recommendation)

Recommendation 3.4: If [123]mIBG scintigraphy is used to assess residual tumor after surgery, SPECT/CT is mandatory. GRADE B, SOR 1.21 (very strong recommendation)

6| DISCUSSION

The recommendations outlined in this paper aim to harmonize image acquisition and interpretation in high-risk neuroblastoma. It is hoped that this would enhance diagnostic accuracy and clinical outcome for future patients. Several challenges became apparent in the processes of developing these recommendations.

Firstly, there is diversity in current clinical and research imaging management of these patients in participating countries, and in individual centers within each country, within the SIOPEN community. Main factors contributing to this could include variations in infrastructure and expertise, competing priorities imposed on existing resources in individual centers, and differences in commissioning and funding arrangements. Unfortunately, not all countries or centers are currently represented in the SIOPEN imaging specialty committees. This may result in underrepresentation of regions facing particular healthcare challenges.

Secondly, for some imaging questions considered, there is a lack of robust evidence. None of the recommendations achieved grade A status. This likely relates to the low incidence of high-risk neuroblastoma, heterogeneity of disease phenotypes, and variation in clinical practice, making robust pooling of clinical and research data difficult. In the hope of bridging the gap in areas where gold-standard evidence may not be available, these recommendations were created through amalgamating views of best practice by international experts currently active in imaging management of neuroblastoma. Areas with a relative paucity of research evidence are transparently presented.

Finally, recent advancements in new imaging technologies have markedly outpaced clinical validation. For instance, in MRI, developments such as multichannel coils and compressed sensing enhanced the feasibility of whole-body MRI. More recent advances in machine learning image reconstruction could reduce overall study time and improve image quality. Significant advances have also been seen in recent years in NM imaging, including the development of digital SPECT/CT and large field-of-view PET/CT. Advanced PET/CT scanners now offer unprecedented sensitivity, allowing for significant reductions in scan time and/or injected activity and enabling anesthesia-free imaging with low radiation exposure [32].

The authors of this publication envisage that novel PET tracers such as [$^{18}$F]mFBG will replace [$^{123}$I]mIBG in the future. With its enhanced lesion detection rate and improved diagnostic certainty, along with a reduction in anesthesia, [$^{18}$F]mFBG PET/CT clearly holds promise for neuroblastoma imaging [30–32].

However, it must be noted that the clinical consequences of this improved lesion detection and image characterization are not yet understood. In the context of high-risk neuroblastoma, assessment of adequate response after induction treatment is currently based on an absolute SIOPEN score of <3 on planar [123]mIBG scintigraphy [51]. Although this prognostic threshold has been validated, an equivalent threshold for the biologically similar but technically much more sensitive [18F]mFBG PET-CT is not presently known.

The determination of a new threshold applicable to [$^{18}$F]mFBG PET-CT with an independent imaging trial of sufficient statistical power and outcome for validity would require significant research effort, which would considerably delay the introduction of [$^{18}$F]mFBG PET in treatment protocols. A plausible and pragmatic approach could be to consider the use of a response assessment system that is less dependent on an absolute score. For example, harmonization with the COG and adoption of the INRC, which uses a relative decrease in mIBG skeletal score as a marker for response [62]. This approach would provide treating centers with flexibility to introduce novel PET tracers, while still supporting the use of conventional NM imaging.

A balance must be maintained between the adoption of promising new technologies, the need for robust clinical validation, and equitable access to imaging technologies and clinical treatments, including treatment optimization trials. In developing these recommendations, efforts were made to carefully consider and reconcile these factors. The resulting guidelines are intended to be

inclusive and adaptable to diverse clinical settings across Europe and beyond.

Acknowledgments

The authors would like to thank all members of the Radiology and Nuclear Medicine and Physics Speciality Committee who contributed by voting for the recommendations, including (alphabetical order) Michela Allocca, Julia Balaguer, Lise Borgwardt, Claudio Granata, Ricarda von Krüchten, Laure Michaud, Alice Monaci, Erika Pace, Anne Perrone, Paola Saletti, Raquel Sánchez Vañó and Tom A Watson. The authors gratefully acknowledge the reviewing and approval of this manuscript by the Principal Investigators of the HR-NBL2 trial, Executive Committee of SIOPEN, Surgery Specialty Committee of the SIOPEN, Radiotherapy Specialty Committee of the SIOPEN, Taskforce Oncology of the European Society of Pediatric Radiology (ESPR), the European Society for Pediatric Oncology (SIOPE) Imaging Working Group, and the European Association of Nuclear Medicine (EANM).

Conflicts of Interest

The authors declare no conflicts of interest.

References

1. R. L. Siegel, K. D. Miller, H. E. Fuchs, and A. Jemal, “Cancer Statistics,” CA: A Cancer Journal for Clinicians 71, no. 1 (2021): 7-33, https://doi.org/10.3322/caac.21654.

2. J. M. Maris, "Recent Advances in Neuroblastoma," New England Journal of Medicine 362 (2010): 2202-2211.

3. L. Pio, H. J. Brisse, R. Alaggio, et al., “Image-Guided Core-Needle or Surgical Biopsy for Neuroblastoma Diagnosis in Children: A Systematic Review and Meta-Analysis From the International Society of Pediatric Surgical Oncology (IPSO),” Pediatric Blood & Cancer 71, no. 2 (2024): e30789, https://doi.org/10.1002/pbc.30789.

4. G. R. Schooler, J. H. Squires, A. Alazraki, et al., “Pediatric Hepatoblastoma, Hepatocellular Carcinoma, and Other Hepatic Neoplasms: Consensus Imaging Recommendations From American College of Radiology Pediatric Liver Reporting and Data System (LI-RADS) Working Group,” Radiology 296, no. 3 (2020): 493–497, https://doi.org/10.1148/radiol.2020200751.

5. E. R. Tang, A. L. Alazraki, P. G. Thacker, M. B. McCarville, and A. J. Towbin, "Introduction to the COG Diagnostic Imaging Committee/SPR Oncology Committee White Papers: Rationale and Methods," Pediatric Blood & Cancer 70, no. Suppl 4 (2023): e30411, https://doi.org/10.1002/pbc.30411.

6. H. J. Brisse, M. B. McCarville, C. Granata, et al., "Guidelines for Imaging and Staging of Neuroblastic Tumors: Consensus Report From the International Neuroblastoma Risk Group Project," Radiology 261, no. 1 (2011): 243-257, https://doi.org/10.1148/radiol.11101352.

7. A. C. Rocha, L. Alamo, N. Ostojic, C. Chevallier, and E. Tenisch, “Development of a User-Friendly Calculator for a Pediatric Split-bolus Polytrauma Computed Tomography Protocol,” Pediatric Radiology 54, no. 12 (2024): 2077–2081, https://doi.org/10.1007/s00247-024-06082-5.

8. A. Delval, T. Touitou, C. Gondry-Jouet, C. Khanfar, and E. Haraux, "A Non-Inferiority Study of MRI versus CT for Staging and Image-Defined Risk Factor Assessment in the Preoperative Work-up of Abdominopelvic Neuroblastoma," European Journal of Radiology 177 (2024): 111580, https://doi.org/10.1016/j.ejrad.2024.111580.

9. C. E. Morin, “Gadolinium-Based Contrast Media Does Not Improve the Staging of Neuroblastoma Image-Defined Risk Factors at Diagnosis,” Pediatric Blood & Cancer 2024.

10. N. Gahr, K. Darge, G. Hahn, B. W. Kreher, M. von Buiren, and M. Uhl, “Diffusion-Weighted MRI for Differentiation of Neuroblastoma and Gan-

glioneuroblastoma/Ganglioneuroma," European Journal of Radiology 79, no. 3 (2011): 443-446, https://doi.org/10.1016/j.ejrad.2010.04.005.

11. D. Tambasco, M. Zlotnik, S. Joshi, et al., "Characterisation of Paediatric Neuroblastic Tumours by Quantitative Structural and Diffusion-Weighted MRI," Journal of Clinical Medicine 13, no. 22 (2024): 6660, https://doi.org/10.3390/jcm13226660.

12. S. Gassenmaier, I. Tsiflikas, J. Fuchs, et al., “Feasibility and Possible Value of Quantitative Semi-Automated Diffusion Weighted Imaging Volumetry of Neuroblastic Tumors,” Cancer Imaging 20 (2020): 89, https://doi.org/10.1186/s40644-020-00366-3.

13. H. Neubauer, M. Li, V. R. Müller, T. Pabst, and M. Beer, “Diagnostic Value of Diffusion-Weighted MRI for Tumor Characterization, Differentiation and Monitoring in Pediatric Patients With Neuroblastic Tumors,” Rofo 189, no. 7 (2017): 640-650, https://doi.org/10.1055/s-0043-108993. English.

14. S. Demir, N. Altinkaya, N. E. Kocer, A. Erbay, and P. Oguzkurt, "Variations in Apparent Diffusion Coefficient Values Following Chemotherapy in Pediatric Neuroblastoma," Diagnostic and Interventional Radiology 21, no. 2 (2015): 184–188, https://doi.org/10.5152/dir.2014.14187.

15. L. Privitera, P. W. Hales, L. Musleh, et al., "Comparison between Diffusion-Weighted MRI and $^{123}$ I-mIBG Uptake in Primary High-Risk Neuroblastoma," Journal of Magnetic Resonance Imaging 53, no. 5 (2021): 1486–1497, https://doi.org/10.1002/jmri.27458.

16. T. A. Vik, T. Pfluger, R. Kadota, et al., “(123)I-mIBG Scintigraphy in Patients With Known or Suspected Neuroblastoma: Results From a Prospective Multicenter Trial,” Pediatric Blood & Cancer 52, no. 7 (2009): 784-790, https://doi.org/10.1002/pbc.21932.

17. E. Bombardieri, F. Giammarile, C. Aktolun, et al., European Association for Nuclear Medicine, "13II/123I-metaiodobenzylguanidine (mIBG) Scintigraphy: Procedure Guidelines for Tumour Imaging," European Journal of Nuclear Medicine and Molecular Imaging 37, no. 12 (2010): 2436-2446, https://doi.org/10.1007/s00259-010-1545-7.

18. M. Lassmann, S. T. Treves, EANM/SNMMI Paediatric Dosage Harmonization Working Group, “Paediatric Radiopharmaceutical Administration: Harmonization of the 2007 EANM Paediatric Dosage Card (Version 1.5.2008) and the 2010 North American Consensus Guidelines,” European Journal of Nuclear Medicine and Molecular Imaging 41, no. 5 (2014): 1036–1041, https://doi.org/10.1007/s00259-014-2731-9.

19. Z. Bar-Sever, L. Biassoni, B. Shulkin, et al., "Guidelines on Nuclear Medicine Imaging in Neuroblastoma," European Journal of Nuclear Medicine and Molecular Imaging 45, no. 11 (2018): 2009–2024, https://doi.org/10.1007/s00259-018-4070-8.

20. R. Hu, Y. Zhang, S. Liu, P. Lee, C. Liu, and A. Liu, “Prognostic Prediction by $^{18}$F-FDG-PET/CT Parameters in Patients With Neuroblastoma: A Systematic Review and Meta-Analysis,” Frontiers in Oncology 13 (2023): 1208531, https://doi.org/10.3389/fonc.2023.1208531.

21. R. Vali, A. Alessio, R. Balza, et al., “SNMMI Procedure Standard/EANM Practice Guideline on Pediatric $^{18}$F-FDG PET/CT for Oncology 1.0,” Journal of Nuclear Medicine 62, no. 1 (2021): 99–110, https://doi.org/10.2967/jnumed.120.254110.

22. R. Boellaard, R. Delgado-Bolton, W. J. Oyen, et al., European Association of Nuclear Medicine (EANM), “FDG PET/CT: EANM Procedure Guidelines for Tumour Imaging: Version 2.0,” European Journal of Nuclear Medicine and Molecular Imaging 42, no. 2 (2015): 328–354, https://doi.org/10.1007/s00259-014-2961-x.

23. J. E. Gains, N. J. Sebire, V. Moroz, K. Wheatley, and M. N. Gaze, “Immunohistochemical Evaluation of Molecular Radiotherapy Target Expression in Neuroblastoma Tissue,” European Journal of Nuclear Medicine and Molecular Imaging 45, no. 3 (2018): 402–411, https://doi.org/10.1007/s00259-017-3856-4.

24. N. Alexander, P. Marrano, P. Thorner, et al., “Prevalence and Clinical Correlations of Somatostatin Receptor-2 (SSTR2) Expression in Neuroblastoma,” Journal of Pediatric Hematology/Oncology 41, no. 3 (2019): 222–227, https://doi.org/10.1097/MPH.000000000001326.

25. A. Kroiss, D. Putzer, C. Uprimny, et al., “Functional Imaging in Phaeochromocytoma and Neuroblastoma With 68Ga-DOTA-Tyr 3-octreotide Positron Emission Tomography and 123I-metaiodobenzylguanidine,” European Journal of Nuclear Medicine and Molecular Imaging 38, no. 5 (2011): 865-873, https://doi.org/10.1007/s00259-010-1720-x.

26. T. A. Hope, M. Allen-Auerbach, L. Bodei, et al., “SNMMI Procedure Standard/EANM Practice Guideline for SSTR PET: Imaging Neuroendocrine Tumors,” Journal of Nuclear Medicine 64, no. 2 (2023): 204–210, https://doi.org/10.2967/jnumed.122.264860.

27. R. AlSadi, A. U. R. Maaz, O. Bouhali, and M. Djekidel, “68 Ga-DOTATATE PET in Restaging and Response to Therapy in Neuroblastoma: A Case Series and a Mini Review,” Journal of Nuclear Medicine Technology 51, no. 2 (2023): 140–146, https://doi.org/10.2967/jnmt.122. 264694.

28. F. Sundquist, K. Georgantzi, K. B. Jarvis, et al., "A Phase II Trial of a Personalized, Dose-Intense Administration Schedule of 177 Lutetium-DOTATATE in Children with Primary Refractory or Relapsed High-Risk Neuroblastoma-LuDO-N," Frontiers in Pediatrics 10 (2022): 836230, https://doi.org/10.3389/fped.2022.836230.

29. M. Y. Lu, Y. L. Liu, H. H. Chang, et al., National Taiwan University Neuroblastoma Study Group, "Characterization of Neuroblastic Tumors Using 18F-FDOPA PET," Journal of Nuclear Medicine 54, no. 1 (2013): 42-49, https://doi.org/10.2967/jnumed.112.102772.

30. A. Piccardo, E. Lopci, M. Conte, et al., "Comparison of 18F-dopa PET/CT and 123I-MIBG Scintigraphy in Stage 3 and 4 Neuroblastoma: A Pilot Study," European Journal of Nuclear Medicine and Molecular Imaging 39, no. 1 (2012): 57-71, https://doi.org/10.1007/s00259-011-1938-2.

31. A. Piccardo, G. Morana, M. Puntoni, et al., "Diagnosis, Treatment Response, and Prognosis: The Role of $^{18}$F-DOPA PET/CT in Children Affected by Neuroblastoma in Comparison With $^{123}$I-mIBG Scan: The First Prospective Study," Journal of Nuclear Medicine 61, no. 3 (2020): 367-374, https://doi.org/10.2967/jnumed.119.232553.

32. N. Pandit-Taskar, P. Zanzonico, K. D. Staton, et al., “Biodistribution and Dosimetry of 18F-Meta-Fluorobenzylguanidine: A First-in-Human PET/CT Imaging Study of Patients With Neuroendocrine Malignancies,” Journal of Nuclear Medicine 59, no. 1 (2018): 147-153, https://doi.org/10.2967/jnumed.117.193169.

33. A. Samim, T. Blom, A. J. Poot, et al., "[$^{18}$F]mFBG PET-CT for Detection and Localisation of Neuroblastoma: A Prospective Pilot Study," European Journal of Nuclear Medicine and Molecular Imaging 50, no. 4 (2023): 1146–1157, https://doi.org/10.1007/s00259-022-06063-6. Epub 2022 Dec 12. Erratum in: Eur J Nucl Med Mol Imaging. 2023 Apr;50(5):1541.; PMCID: PMC9931849.

34. P. Wang, T. Li, Z. Liu, et al., "[$^{18}$F]MFBG PET/CT Outperforming [$^{123}$I]MIBG SPECT/CT in the Evaluation of Neuroblastoma," European Journal of Nuclear Medicine and Molecular Imaging 50, no. 10 (2023): 3097–3106, https://doi.org/10.1007/s00259-023-06221-4.

35. L. Borgwardt, J. Brok, K. F. Andersen, et al., “Performing [$^{18}$F]MFBG Long-Axial-Field-of-View PET/CT without Sedation or General Anesthesia for Imaging of Children With Neuroblastoma,” Journal of Nuclear Medicine 65, no. 8 (2024): 1286–1292, https://doi.org/10.2967/jnumed.123. 267256.

36. G. A. Prenosil, H. Sari, M. Fürstner, et al., “Performance Characteristics of the Biograph Vision Quadra PET/CT System With a Long Axial Field of View Using the NEMA NU 2–2018 Standard,” Journal of Nuclear Medicine 63, no. 3 (2022): 476–484, https://doi.org/10.2967/jnumed.121. 261972.

37. S. Turnock, D. R. Turton, C. D. Martins, et al., "18F-meta-fluorobenzylguanidine (18F-mFBG) to Monitor Changes in Norepinephrine Transporter Expression in Response to Therapeutic Intervention in Neuroblastoma Models," Scientific Reports 10, no. 1 (2020): 20918, https://doi.org/10.1038/s41598-020-77788-3.

38. H. Zhang, R. Huang, N. K. Cheung, et al., “Imaging the Norepinephrine Transporter in Neuroblastoma: A Comparison of [18F]-MFBG and 123I-MIBG,” Clinical Cancer Research 20, no. 8 (2014): 2182-2191, https://doi.org/10.1158/1078-0432.CCR-13-1153.

39. M. S. Aboian, S. Y. Huang, M. Hernandez-Pampaloni, et al., "124-1-MIBG PET/CT to Monitor Metastatic Disease in Children With Relapsed Neuroblastoma," Journal of Nuclear Medicine 62, no. 1 (2021): 43-47, https://doi.org/10.2967/jnumed.120.243139.

40. D. M. Panicek and H. Hricak, "How Sure Are You, Doctor? A Standardized Lexicon to Describe the Radiologist's Level of Certainty," American Journal of Roentgenology 207, no. 1 (2016): 2–3, https://doi.org/10.2214/AJR.15.15895.

41. T. Monclair, G. M. Brodeur, P. F. Ambros, et al., "The International Neuroblastoma Risk Group (INRG) Staging System: An INRG Task Force Report," Journal of Clinical Oncology 27, no. 2 (2009): 298–303, https://doi.org/10.1200/JCO.2008.16.6876.

42. D. A. Morgenstern, W. B. London, D. Stephens, et al., “Prognostic Significance of Pattern and Burden of Metastatic Disease in Patients With Stage 4 Neuroblastoma: A Study From the International Neuroblastoma Risk Group Database,” European Journal of Cancer 65 (2016): 1-10, https://doi.org/10.1016/j.ejca.2016.06.005.

43. J. R. Park, R. Bagatell, S. L. Cohn, et al., “Revisions to the International Neuroblastoma Response Criteria: A Consensus Statement from the National Cancer Institute Clinical Trials Planning Meeting,” Journal of Clinical Oncology 35, no. 22 (2017): 2580–2587, https://doi.org/10.1200/JCO.2016.72.0177.

44. K. Yoshino, M. Tanabe, N. Ohnuma, and H. Takahashi, “Histopathologic Analysis of Bone Marrow and Bone Metastasis in Murine Neuroblastoma,” Clinical & Experimental Metastasis 14, no. 5 (1996): 459–465, https://doi.org/10.1007/BF00128962.

45. B. L. Shulkin and B. Shapiro, “Current Concepts on the Diagnostic Use of MIBG in Children,” Journal of Nuclear Medicine 39, no. 4 (1998): 679-688.

46. S. G. DuBois, Y. Kalika, J. N. Lukens, et al., “Metastatic Sites in Stage IV and IVS Neuroblastoma Correlate With Age, Tumor Biology, and Survival,” Journal of Pediatric Hematology/Oncology 21, no. 3 (1999): 181–189, https://doi.org/10.1097/00043426-199905000-00005.

47. A. S. Littooij and B. de Keizer, “Imaging in Neuroblastoma,” Pediatric Radiology 53, no. 4 (2023): 783–787, https://doi.org/10.1007/s00247-022-05489-2.

48. J. S. Meyer, M. J. Siegel, S. O. Farooqui, D. Jaramillo, B. D. Fletcher, and F. A. Hoffer, “Which MRI Sequence of the Spine Best Reveals Bone-Marrow Metastases of Neuroblastoma?” Pediatric Radiology 35, no. 8 (2005): 778-785, https://doi.org/10.1007/s00247-005-1470-2.

49. S. M. Booth, T. H. Kapadia, V. Tang, and A. F. Maniyar, “Pictorial Review: Challenges in Distinguishing Bilateral Metaphyseal Marrow Abnormalities on Magnetic Resonance Imaging,” Pediatric Radiology 54, no. 8 (2024): 1247-1260, https://doi.org/10.1007/s00247-024-05896-7.

50. R. A. J. Nievelstein, L. Borgwardt, T. von Kalle, A. S. Littooij, L. S. Ording Muller, and N. Tolboom, “Multimodality Imaging of Bone Marrow Involvement in Paediatric Oncology,” EJC Paediatric Oncology 4 (2024): 100185.

51. V. Lewington, B. Lambert, U. Poetschger, et al., "123 I-mIBG Scintigraphy in Neuroblastoma: Development of a SIOPEN Semi-Quantitative Reporting Method by an International Panel," European Journal of Nuclear Medicine and Molecular Imaging 44, no. 2 (2017): 234-241, https://doi.org/10.1007/s00259-016-3516-0.

52. G. A. Yanik, M. T. Parisi, B. L. Shulkin, et al., “Semiquantitative mIBG Scoring as a Prognostic Indicator in Patients With Stage 4 Neuroblastoma: A Report From the Children’s Oncology Group,” Journal of Nuclear Medicine 54, no. 4 (2013): 541–548, https://doi.org/10.2967/jnumed.112.112334.

53. G. A. Yanik, M. T. Parisi, A. Naranjo, et al., "Validation of Postinduc- tion Curie Scores in High-Risk Neuroblastoma: A Children's Oncology

G

o

j 5 n C N

a

3 5 m N T n 5 B A

(

h 5 C W h 5 R M h 5 R h 6 C S

P

p 6 I T S

S

1 6 N C e

1

roup and SIOPEN Group Report on SIOPEN/HR-NBL1,” Journal f Nuclear Medicine 59, no. 3 (2018): 502–508, https://doi.org/10.2967/numed.117.195883 .

4 . K. A. Streby, M. T. Parisi, B. L. Shulkin, et al., “Impact of Diagostic and End-of-Induction Curie Scores With Tandem High-Dose hemotherapy and Autologous Transplants for Metastatic High-Risk euroblastoma: A Report From the Children’s Oncology Group,” Peditric Blood & Cancer 70, no. 8 (2023): e30418, https://doi.org/10.1002/pbc.

0418 .

5 . R. Ladenstein, B. Lambert, U. Pötschger, et al., “Validation of the IBG Skeletal SIOPEN Scoring Method in Two Independent High-Risk euroblastoma Populations: The SIOPEN/HR-NBL1 and COG-A3973 rials,” European Journal of Nuclear Medicine and Molecular Imaging 45,o. 2 (2018): 292–305, https://doi.org/10.1007/s00259- 017- 3829- 7 .

6 . J. Van Heerden, M. Kruger, T. M. Esterhuizen, et al., “The Association etween Tumour Markers and Meta-Iodobenzylguanidine Scans in South frican Children With High-Risk Neuroblastoma,” Clinical Oncology Royal College of Radiologists (Great Britain)) 33, no. 8 (2021): 517–526,ttps://doi.org/10.1016/j.clon.2021.03.003 .

7 . E. A. Sokol, R. Engelmann, W. Kang, et al., “Computer-Assisted urie Scoring for Metaiodobenzylguanidine (MIBG) Scans in Patients ith Neuroblastoma,” Pediatric Blood & Cancer 65, no. 12 (2018): e27417,ttps://doi.org/10.1002/pbc.27417 .

8 . R. Bagatell, K. McHugh, A. Naranjo, et al., “Assessment of Primary Site esponse in Children with High-Risk Neuroblastoma: An International ulticenter Study,” Journal of Clinical Oncology 34, no. 7 (2016): 740–746,ttps://doi.org/10.1200/JCO.2015.63.2042 .

9 . K. E. Orr and K. McHugh, “The New International Neuroblastoma esponse Criteria,” Pediatric Radiology 49, no. 11 (2019): 1433–1440,ttps://doi.org/10.1007/s00247- 019- 04397- 2 .

0 . K. Burnand, G. Barone, K. McHugh, and K. Cross, “Preoperative omputed Tomography Scanning for Abdominal Neuroblastomas Is uperior to Magnetic Resonance Imaging for Safe Surgical Planning,”ediatric Blood & Cancer 66, no. 11 (2019): e27955, https://doi.org/10.1002/bc.27955 .

1 . A. Zarfati, F. Guérin, M. Dioguardi Burgio, et al., “Preoperative dentification of Adamkiewicz Artery in Pediatric Posterior Thoracic umors: Fact or Fiction? A Systematic Review From the International ociety of Pediatric Surgical Oncology (IPSO),” Journal of Pediatric urgery 59, no. 12 (2024): 161985, https://doi.org/10.1016/j.jpedsurg.2024.

61985 .

2 . H. A. Lai, S. E. Sharp, A. Bhatia, et al., “Imaging of Pediatric euroblastoma: A COG Diagnostic Imaging Committee/SPR Oncology ommittee White Paper,” Pediatric Blood & Cancer 70, Suppl 4 (2023):29974, https://doi.org/10.1002/pbc.29974 .

4 of 14

Pediatric Blood & Cancer, 2026