Lymphoma
Hodgkin lymphoma and non-Hodgkin lymphoma together are the third most common form of malignancy in children and adolescents [
39]. Non-Hodgkin lymphoma is most frequent in children younger than 15 years, whereas Hodgkin lymphoma is predominantly diagnosed in teenagers. Pediatric lymphomas are staged using the Lugano classification for Hodgkin lymphoma [
40] and the International Pediatric Non-Hodgkin Lymphoma Staging System for non-Hodgkin lymphoma [
41].
Standard imaging procedure: In Hodgkin lymphoma, [F-18]2-fluoro-2-deoxyglucose (FDG) positron emission tomography (PET)/CT remains the first-line modality for staging and response assessment, providing both structural and functional metabolic information [
40]. In non-Hodgkin lymphoma the current guidelines propose to perform chest radiograph, ultrasonography (US) of the abdomen and cranial/spinal MRI if indicated, except for the B cell non-Hodgkin lymphoma, where FDG PET or whole-body MRI is recommended.
Review and comments on the literature: Because both PET and contrast-enhanced CT involve substantial radiation exposure and children often undergo several PET/CT examinations during the course of treatment, there is an increasing interest in the use of whole-body MRI as a good radiation-free alternative. Several studies have shown that whole-body MRI is feasible even in children [
2,
8,
29,
42]. Table
1 shows proposed whole-body MRI sequences for use in children with lymphoma.
Table 1
Proposed magnetic resonance imaging protocol at 1.5 T in lymphoma (please also refer to Technical Considerations)
Orientation | Coronal | Coronal | Axial | Axial | Axial |
Respiratory motion compensation | Breath hold (thorax and abdomen) | Respiratory triggering (thorax and abdomen) | Free breathing | Respiratory triggering (thorax and abdomen) | Breath holdb |
Anatomical coverage | Head to groin | Head to groin | Head to groin | Head to groin | Head to groin |
Staging: In a study by Punwani et al. [
29], the authors reported very good agreement for nodal and extranodal disease involvement between whole-body MRI compared to an FDG PET/CT reference standard, despite only using STIR for whole-body MRI. Because of the clear visualization of lymphoid tissue, diffusion-weighted imaging (DWI) is a potentially interesting additional sequence to use for evaluating lymphoma. Regacini et al. [
30] showed an excellent sensitivity for staging pediatric lymphoma compared to contrast-enhanced CT using coronal STIR and DWI sequences. However, other studies could not demonstrate the additional value of DWI to conventional MRI sequences in staging pediatric lymphoma [
6,
8]. This could be related to the fact that both benign and malignant nodes demonstrate impeded diffusion. Therefore, the detection of lymph nodes in whole-body DWI is still mainly based on size criteria. Overall, in pediatric Hodgkin lymphoma, whole-body MRI with DWI agreed with an FDG PET/CT-based reference for disease stage in 82–85% of cases [
8,
42]. Interestingly, a study performed by Latifoltojar et al. [
42] used a highest b value of 500 s/mm
2. The authors acknowledged that using higher b values could decrease perceptual errors for extranodal disease assessment. Indeed, higher b values are preferred to optimize background body signal suppression and improve lesion conspicuity, in particular in organs/regions with normal intrinsic diffusion restriction.
Response assessment: Early recognition of therapy response or failure to chemotherapy enables better selection of children who need more or less intensive therapy. The concept of early response assessment with FDG PET/CT in lymphoma has received considerable attention in the last few years, although it is still not officially recommended outside clinical trials [
40]. The role of whole-body MRI in response assessment in pediatric lymphoma is still under investigation. Mayerhoefer et al. [
43] showed in 64 adults with lymphoma that whole-body MRI with DWI could serve as a feasible alternative for FDG PET/CT during follow-up and treatment response assessment. Several, mostly pilot, studies compared the quantitative data from FDG PET/CT (standardized uptake value [SUV]) with DWI (apparent diffusion coefficient [ADC] values) for early response assessment, with inconclusive results. They reported presence or absence of an inverse correlation between ADC and SUV [
22,
23,
29,
44]. Latifoltojar et al. [
42] recently published their results of a prospective study in 55 children with Hodgkin lymphoma and showed that whole-body MRI was correct in 66% and underestimated response in 26% of cases during interim response assessment.
According to the currently available literature, whole-body MRI is a good alternative for contrast-enhanced CT in staging pediatric lymphoma. However, FDG PET with low-dose CT for attenuation correction remains vital for interim response assessment and therefore cannot be omitted during staging or interim response assessment in lymphoma.
Key recommendation in pediatric lymphoma: Whole-body MRI is a good alternative for contrast-enhanced CT in staging pediatric lymphoma. However, for FDG-avid lymphomas, 18F-FDG PET remains vital for staging and response assessment.
Neuroblastoma
Neuroblastoma is a neoplasm arising from primordial neural crest cells and is the second most common solid extracranial tumor in children, accounting for about 6% of all childhood cancers [
45]. It can occur anywhere along the sympathetic chain from neck to pelvis, with the adrenal medulla as the most common site at presentation. In about 70% of cases, metastatic disease is observed at diagnosis, including the liver, lymph nodes, bone marrow, cortical bone and skin. About 90% of cases are diagnosed before the age of 6 years. The outcome is variable because neuroblastoma can spontaneously regress in children younger than 1 year, whereas in older children it can cause death despite aggressive treatment with surgery, chemotherapy, bone marrow transplantation and radiotherapy [
46].
Standard imaging procedure: Accurate staging is pivotal for planning treatment. Furthermore, a systematic assessment of the relationship between the primary tumor and the adjacent structures according to a series of imaging-defined risk factors is mandatory to evaluate tumor resectability [
47]. Typically, CT or MRI, iodine-123 metaiodobenzylguanidine (I-123 MIBG) scintigraphy and bone marrow biopsies are used to evaluate local and metastatic disease at diagnosis and during treatment. Further imaging modalities (e.g., FDG PET in MIBG-negative tumors) and frequency of imaging in follow-up are also related to risk groups.
Review and comments on the literature: The availability of studies on the role of whole-body MRI in neuroblastoma is still very limited. Furthermore, some of them included children not only with neuroblastoma, but also with other common pediatric tumors. In 2003, Pfluger et al. [
48] retrospectively studied 28 people with neuroblastoma who had 50 I-123 MIBG scans in combination with 50 MRI studies. MRI studies included fast spin-echo T1-weighted and T2-weighted images, STIR images and contrast-enhanced T1-weighted images of suspected lesions. They concluded that MRI showed a higher sensitivity and I-123 MIBG a higher specificity, but that integrated imaging with both I-123 MIBG and MRI allowed an increase in both sensitivity and specificity.
In 2013, Siegel et al. [
7] compared whole-body MRI with conventional imaging for detecting distant metastases in children with common malignant tumors. Sixty-six children with newly diagnosed lymphoma, neuroblastoma or soft-tissue sarcoma were selected for image review and analysis [
7]. The authors concluded that the non-inferior accuracy for diagnosing distant metastases (e.g., pulmonary metastases) was not established for the use of whole-body MRI compared with conventional methods. However, improved accuracy was seen with whole-body MRI in children with non-lymphomatous tumors [
7]. In 2014, Kembhavi et al. [
49] assessed the diagnostic accuracy of whole-body MRI for metastatic disease in people with small round cell tumors including neuroblastoma, primitive neuroectodermal tumor and rhabdomyosarcoma by comparison with routine staging procedures. Whole-body MRI studies included coronal T1-W and STIR sequences. They concluded that whole-body MRI had high diagnostic accuracy for evaluating metastatic disease to the marrow [
49]. On the contrary, the detection rate of nodal metastases was less when whole-body MRI was compared with conventional staging, and chest CT was still essential for accurate evaluation of lung metastases.
Diffusion-weighted imaging can be effectively integrated in MRI studies in children with neuroblastic tumors: Peschmann et al. [
50] observed in 19 people that ADC values at diagnosis differed significantly between malignant and benign neuroblastic tumors. Furthermore, low baseline ADC was predictive of tumor progression and relapse. With therapy, increasing ADC appeared to predict relapse-free survival, whereas a decreasing ADC during therapy was an indicator of poor prognosis [
50]. Very recently, Ishiguchi et al. [
19] studied the role of whole-body DWI with background body suppression using only singular high b value imaging in detection of lymph node and bone metastases from neuroblastoma. Thirteen people underwent both 18F-FDG PET/CT and whole-body DWI with background body suppression. According to the results of this study, whole-body DWI with background body suppression showed a similar level of sensitivity for detecting lymph node metastases to that of FDG PET/CT [
19]. However, without ADC mapping, the specificity was poor in this study [
19].
The studies on the role of whole-body MRI in neuroblastoma are still scarce, based on a limited number of patients, and in some cases with contradictory results. Nevertheless, whole-body MRI could be a promising, radiation-free modality for detecting bone and bone marrow metastases (Table
2 indicates proposed sequences). Furthermore, the ADC value based on DWI of the primary tumor could be a good indicator of outcome. However, there is still a strong need for prospective large cohort studies to validate the role of whole-body MRI in children with neuroblastoma.
Table 2
Proposed magnetic resonance imaging protocol in neuroblastoma (please also refer to Technical Considerations)
Orientation | Coronal | Sagittal | Axial | Axial | Axial |
Respiratory motion compensation | Respiratory triggering (thorax and abdomen) | Free breathing | Free breathing | Respiratory triggering (thorax and abdomen) | Breath holdb |
Anatomical coverage | Whole-body | Spine | Whole-body or affected regions | Head to groin | Head to groin |
Key recommendations in neuroblastoma: Whole-body MRI could play a role as an ancillary study for detecting bone and bone marrow disease, while performing MRI for evaluation of the primary tumor. Prospective multicenter studies are needed to validate the role of whole-body MRI versus the current reference standards in pediatric neuroblastoma.
Sarcomas
Pediatric sarcomas are a heterogeneous group of rare tumors, most of which are highly malignant. They account for approximately 10–15% of solid malignancies in childhood and adolescence. Soft-tissue sarcomas are divided into two groups: rhabdomyosarcomas and non-rhabdomyosarcoma soft-tissue sarcomas. The most common primary malignant bone tumors are osteosarcomas and Ewing sarcomas.
Standard imaging procedure: A multidisciplinary diagnostic and therapeutic approach is mandatory in all cases and should be carried out by a reference center [
51,
52]. In soft-tissue sarcomas, conventional imaging and staging as recommended by international study groups consist of high-resolution MRI for local disease and loco-regional lymph nodes, CT for lung metastases, and bone scintigraphy for skeletal metastases. Imaging with FDG PET/CT is optional [
51,
52] — depending on local availability (European Paediatric Soft-tissue Sarcoma Study Group rhabdomyosarcoma [EpSSG RMS2005] protocol, German Cooperative Weichteilsarkom Studiengruppe [CWS] guidance 2012) — or is recommended at baseline and in cases of suspected tumor recurrence (Children’s Oncology Group Soft Tissue Sarcoma Committee). Bone sarcomas additionally require radiographs of the primary tumor and the search for skip lesions in the same extremity by MRI.
Review and comments on the literature: Apart from bone scintigraphy, whole-body imaging was not generally recommended [
51,
52] until the most recent EpSSG guideline, which recommends performing FDG PET/CT (MR) for staging at baseline and after three courses of chemotherapy. Its application is especially important in suspected disseminated disease because a curative therapeutic approach requires the local control of each single lesion [
52,
53]. The complete depiction of the whole body from head to toe is mandatory, be it to detect all skeletal metastases [
26], distant metastases in unexpected localizations [
54], or a small primary of a disseminated alveolar rhabdomyosarcoma in a hand or foot [
55] (Table
3).
Table 3
Proposed magnetic resonance imaging protocol in sarcoma (please also refer to Technical Considerations)
Orientation | Coronal | Axial | Axial | Axial | 2 planes, depending on tumor site | Multiple planesb |
Respiratory motion compensation | Respiratory triggering (thorax and abdomen) | Respiratory triggering (thorax and abdomen) | Free breathing | Breath holda | Free breathing | Free breathing |
Anatomical coverage | Whole bodyc | Head to groin | Whole-body or affected regions | Head to groin | Primary tumor with dedicated coil | Primary tumor with dedicated coil |
Given the young patient age and the high number of follow-up examinations, a modality without exposure to ionizing radiation such as whole-body MRI would be preferable. Whole-body MRI allows for excellent contrast resolution of soft tissue and bone marrow. It provides versatile options such as DWI and contrast enhancement. With a thoroughly planned imaging protocol and a thoughtful choice of MR coils, one examination might simultaneously give an overview of the tumor spread as well as high-resolution images of a tumor site.
Although most experts employ coronal STIR sequences, there is, however, neither a universally accepted standard protocol nor a clearly defined set of assessment criteria. Comparative evaluation of whole-body MRI is therefore difficult. Several studies found that for sarcomas and other solid tumors, whole-body imaging conducted as PET/CT or whole-body MRI outperforms conventional imaging in the detection of metastases [
1,
7,
15,
54,
56]. Other results are, however, inconsistent. In older studies, whole-body MRI has been found to be equally sensitive [
27] as well as less sensitive [
1] as compared to FDG PET. It is also known to detect fewer lung metastases than conventional imaging [
7]. PET/CT has, because of its metabolic information, higher sensitivity for detecting nodal disease [
7]. An early study showing the possible impact of whole-body MRI on patient outcome was limited by small numbers [
5]. A meta-analysis on the detection of skeletal metastases in children with primary solid tumors found data to be too scarce and heterogeneous to recommend whole-body MRI as an alternative. However, 3/5 of these studies included PET as a reference standard, which is also not generally recommended [
26]. Qualified studies on the role of diffusion-weighted imaging are also rare, especially in children with rhabdomyosarcoma [
57]. DWI might be helpful in staging and therapy monitoring, although diffusion restriction is not specific for malignant lesions and is also typical of normal lymphatic tissue. In osteosarcoma, DWI helps to predict tumor response to neo-adjuvant therapy [
58], but larger prospective studies need to be performed.
It seems reasonable that in highly malignant sarcomas the most effective search for metastases and relapses should be preferred even if associated with radiation exposure. However, bone scintigraphy is the only whole-body modality that is mandatory in basic imaging, whereas FDG PET/CT is only optional. To detect relapse, the use of cross-sectional imaging has not been demonstrated to be more beneficial or cost-effective than clinical assessment and chest radiographs alone. Thus, prospective clinical studies are needed, first to define patients who would benefit from whole-body imaging, and second to identify the most effective of the available modalities (whole-body MRI, PET/CT or PET/MRI). Whole-body MRI might, at present, be considered in addition to basic imaging recommendations for children who would probably benefit from whole-body imaging without increasing their cumulative radiation dose.
Key recommendations in sarcomas: Whole-body MRI could be considered in addition to imaging recommendations of international oncology groups. Children with sarcoma who have disseminated disease might benefit from whole-body MRI without increasing their cumulative radiation dose.
Langerhans cell histiocytosis
Langerhans cell histiocytosis is characterized by accumulation of clonal CD1a-positive immature dendritic cells, so-called Langerhans cell histiocytosis cells, together with eosinophils, macrophages, lymphocytes and osteoclast-like giant cells. In children younger than 15 years, incidence of Langerhans cell histiocytosis is 4–5 cases per million per year [
59]. Langerhans cell histiocytosis can affect many organs, including the skeleton, skin, lymph nodes, liver, lungs, spleen, hematopoiesis and central nervous system. Previously, Langerhans cell histiocytosis included three entities: eosinophilic granuloma, Hand–Schüller–Christian disease and Letterer–Siwe disease, but at present it is classified as unifocal, single-system multifocal, or multifocal multi-system disease [
59]. Children with single-system or organ Langerhans cell histiocytosis, such as skeleton, skin or lymph nodes, have an excellent prognosis and need minimal or sometimes no treatment at all. The outcome of children with multisystem Langerhans cell histiocytosis can range from spontaneous resolution to fatal outcome despite treatment. Therefore, it is of the utmost importance to stratify these patients according to unifocal versus multifocal disease.
Standard imaging procedure: Diagnostic evaluation and treatment are based on the ongoing LCH-IV International Collaborative Treatment Protocol for Children and Adolescents with Langerhans cell histiocytosis (EudraCT Nr.: 2011–001699-20). According to this protocol, diagnostic imaging at onset should include an abdominal US study for evaluating size and structure of the liver and spleen, a chest radiograph and a radiologic skeletal survey. Functional imaging like bone scan or FDG PET is optional and can be performed in addition to the skeletal survey. Chest CT is needed in case of lung involvement, whereas head CT or MRI should be performed in case of craniofacial lesions or mastoid involvement. Neurologic abnormalities or suspected endocrine abnormalities require a head MRI, and MRI of the spine is necessary in cases of suspected vertebral lesions.
Review and comments on the literature: Langerhans cell histiocytosis can be a multisystem and multifocal disease. Therefore, whole-body MRI could be an excellent method for evaluating the whole body in one examination. Nevertheless, very few studies with preliminary results on the role of whole-body MRI in Langerhans cell histiocytosis are available. The number of children included in these studies is limited (range 2 to 46) [
3,
25,
60‐
62]. Three of five available studies investigated the role of whole-body MRI at diagnosis for primary staging or follow-up; in two studies only the role of whole-body MRI at diagnosis was investigated [
3,
25,
60‐
62]. STIR sequence was performed in all studies; in one of them it was the only sequence performed [
61]. T1-W fast spin-echo (FSE) sequences with and without contrast enhancement were performed in two studies [
60,
62], T1-W FSE without contrast enhancement in one study [
25], and T1-W FSE with just contrast enhancement in another study [
3]. All studies included coronal and sagittal images, whereas only one study also included axial images [
62].
The standard of reference for comparison with whole-body MRI was skeletal survey or plain radiographs in four studies [
3,
25,
60,
61], in combination with skeletal scintigraphy in three of them [
3,
60,
61]. Histopathology or follow-up imaging was the standard of reference in another study [
62]; this study compared the performances of FDG PET with whole-body MRI in Langerhans cell histiocytosis. Sensitivity and specificity of whole-body MRI for lesion detection in Langerhans cell histiocytosis (81% and 47%, respectively) were reported only in the study by Mueller et al. [
62], whereas Kim et al. [
60] reported a sensitivity of 99%.
According to the results of their study, Steinborn et al. [
25] concluded that whole-body MRI had a higher detection rate of bony lesions than a skeletal survey, and they therefore suggested that whole-body MRI be the imaging modality of choice for assessing skeletal involvement in Langerhans cell histiocytosis. Similarly, Goo et al. [
3] and Laffan et al. [
61] reported that whole-body MRI can be an excellent imaging tool for assessing skeletal involvement. Furthermore, Goo et al. [
3] concluded that whole-body MRI can also detect extraskeletal disease, and they suggested that changes in signal intensity on STIR and T1-W contrast-enhanced images could demonstrate response to treatment in active lesions. Mueller et al. [
62] studied the diagnostic value of FDG PET and MRI in pediatric Langerhans cell histiocytosis. They concluded that MRI showed higher sensitivity than FDG PET in lesion detection, whereas FDG PET showed higher specificity. Interestingly, they also reported that FDG PET was more accurate than MRI in evaluating disease activity after chemotherapy because they observed persisting residual contrast enhancement and T2 hyperintensity in lesions with no residual FDG uptake on PET. Probably the most interesting results — based on the largest case series so far — come from a recent study by Kim et al. [
60]. They concluded that whole-body MRI had higher detectability for Langerhans cell histiocytosis lesions than skeletal survey and bone scintigraphy, with no significant differences in the number of false-positives per patient, while the three modalities had comparable accuracy in the initial staging. Table
4 presents a summary of these studies’ findings. Table
5 presents a whole-body MRI protocol for Langerhans cell histiocytosis.
Table 4
Summary of published data on whole-body magnetic resonance imaging in Langerhans cell histiocytosis
Number of patients | 9 | 2 | 14 | 15 | 46 |
Staging at diagnosis | Yes | Yes | Yes | Yes | Yes |
Follow-up | Yes | No | Yes | Yes | No |
Total number of lesions observed | NA | NA | NA | 53 | 105 |
Number of lesions observed on primary staging | NA | NA | NA | 25 | 105 |
Number of lesions observed on follow-up | NA | NA | NA | 28 | NA |
Sequences | STIR, T1-W FSE CE | STIR | T1-W FSE, STIR | T1-W FSE, STIR, T1-W FSE CE + dedicated study of the brain with T1-W FSE, T2-W FSE, FLAIR, T1-W FSE CE | STIR, T1-W FSE, T1-W FSE CE |
Acquisition planes | Coronal, sagittal limited to the trunk | Coronal and sagittal | Coronal and sagittal | Axial, coronal, sagittal | Coronal, sagittal |
Standard of reference | RX skeletal survey, bone scintigraphy | Skeletal scintigraphy and/or plain radiographs | RX skeletal survey | Histopathology and/or follow-up of lesions. Whole-body MRI performances were also compared with 18F-FDG PET | Histopathology and/or follow-up of lesions |
Sensitivity | NA | NA | NA | 81% | 99% |
Specificity | NA | NA | NA | 47% | NA |
Table 5
Proposed whole-body magnetic resonance imaging protocol in Langerhans cell histiocytosis (please also refer to Technical Considerations)
Orientation | Coronal | Coronal | Sagittal | Axial | Axial |
Respiratory motion compensation | Breath hold (thorax and abdomen) | Respiratory triggering (thorax and abdomen) | Free breathing | Free breathing | Free breathing |
Anatomical coverage | Head to toe | Head to toe | Spine | Head | Head |
According to the scientific literature available, whole-body MRI could be the imaging modality of choice for assessing skeletal involvement at onset, thus replacing radiologic skeletal survey and bone scintigraphy. Conventional radiologic studies could be performed only of bones with positive findings on whole-body MRI. Whole-body MRI should include at least T1-W FSE and STIR images in coronal and sagittal planes, whereas it is unclear whether T1-W contrast-enhanced images improve the accuracy of the study. None of the published studies assessed the role of diffusion-weighted imaging. The role of whole-body MRI during follow-up seems to be questionable because persisting signal abnormalities could be caused by post-therapy tissue reorganization. In addition, the role of whole-body MRI in evaluating extraskeletal disease is promising but not finally approved. New and larger prospective multicenter studies are therefore needed.
Key recommendations in Langerhans cell histiocytosis: Whole-body MRI could replace skeletal survey and bone scintigraphy for assessing skeletal involvement at onset. Its role in assessment of extraosseous disease and follow-up is still unclear. Thanks to its high sensitivity and accuracy, whole-body MRI should always be considered in children with Langerhans cell histiocytosis at onset, especially if sedation is not needed.