Introduction
Materials and methods
Search strategy
Selection of relevant studies
Inclusion criteria
Data extraction of included papers
Signaling questions and risk of bias | Patient selection | Signaling questions | Q1: Was a consecutive or random sample of patients enrolled? | YES/NO/UNCLEAR |
Q2: Was a case control design avoided? | YES/NO/UNCLEAR | |||
Q3: Did the study avoid inappropriate exclusions? | YES/NO/UNCLEAR | |||
Risk of bias | Could the selection of patients have introduced bias?a
| LOW/HIGH/UNCLEAR | ||
Index test/MRI | Signaling questions | Q4: Were the MRI test results interpreted without knowledge of the results of the reference standard? | YES/NO/UNCLEAR | |
Q5: If radiological criteria were used, were they (pro−/retrospectively) prespecified? | YES/NO/UNCLEAR | |||
Risk of bias | Could the conduct or interpretation of the MRI have introduced bias? a
| LOW/HIGH/UNCLEAR | ||
Reference tests | Signaling questions | Q6. Is the reference standard likely to correctly classify the target condition? b
| YES/NO/UNCLEAR | |
Q7. Were the reference test results interpreted without knowledge of the results of the MRI? | YES/NO/UNCLEAR | |||
Risk of bias | Could the conduct or interpretation of the reference standard have introduced bias? a
| LOW/HIGH/UNCLEAR | ||
Flow and timing | Signaling questions | Q8. Was there an appropriate interval between WB-MRI and reference standard (<1 month for biopsy and any other imaging techniques and <12 months for FU)? | YES/NO/UNCLEAR | |
Q9. Did all patients receive a reference standard? | YES/NO/UNCLEAR | |||
Q10. Were all patients included in the analysis (initially included based on inclusion and exclusion criteria)? | YES/NO/UNCLEAR | |||
Risk of bias | Could the patient flow have introduced bias? a
| LOW/HIGH/UNCLEAR | ||
Concerns regarding applicability | Patient selection | Is there concern that the included patients do not match the review question? c
| LOW/HIGH | |
Index test/MRI | Is there concern that the index test, its conduct or interpretation differs from the review question? d
| LOW/HIGH | ||
Reference tests | Is there concern that the reference test, its conduct or interpretation differs from the review question? e
| LOW/HIGH |
Results
Search and selection of relevant studies
Inclusion of relevant studies
Methodological quality characteristics
Author, year publication | Risk of bias a
| Concerns regarding applicabilityb
| |||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Patient selection | Index test: WB-MRI | Reference test | Flow and timing | Patient selectionc
| INDEX TEST: WB-MRId
| Reference teste
| |||||||||||
Q1 | Q2 | Q3 | Risk of bias a
| Q4 | Q5 | Risk of bias a
| Q6 | Q7 | Risk of bias a
| Q8 | Q9 | Q10 | Risk of bias a
| ||||
Daldrup-Link et al., 2001 [24] | Unclear | YES | Unclear | Unclear | YES | YES | Low | NO | Unclear | High | YES | YES | YES | Low | Low | High | High f
|
Mazumdar et al., 2002 [26] | Unclear | YES | Unclear | Unclear | YES | YES | Low | NO | Unclear | High | YES | YES | YES | Low | Low | High | High g
|
Goo et al., 2005 [25] | Unclear | YES | Unclear | Unclear | YES | YES | Low | NO | YES | High | YES | YES | YES | Low | High | Low | High |
Kumar et al., 2008 [27] | Unclear | YES | Unclear | Unclear | YES | YES | Low | NO | Unclear | High | YES | YES | YES | Low | Low | High | High |
Krohmer et al., 2010 [28] | Unclear | YES | Unclear | Unclear | YES | YES | Low | NO | YES | High | YES | YES | YES | Low | High | High | High |
Study design characteristics
Patient characteristics
Authors, year of publication | Selection criteria (inclusion and/or exclusion criteria) | Total number of patients included in the article | Age (mean or median, and range) | Gender ratio (male: female) | Distribution of patients (type of tumor) |
---|---|---|---|---|---|
Daldrup-Link et al., 2001 [24] | Inclusion criteria: children and young adults with primary tumors that potentially metastasize to bone. | 39 | Mean age: 12.9 Range: 2–19 years | 27:12 | Ewing sarcoma, 20 Osteosarcoma, 3 Rhabdomyosarcoma, 3 Lymphoma, 2 Myelosarcoma, 1 Melanoma, 1 Langerhans cell histiocytosis, 9 |
Mazumdar et al., 2002 [26] | Inclusion criteria: small cell tumor in child, new or recurrent, MRI, reference imaging within 10 days Exclusion criteria: patients with newly diagnosed tumors who have had chemotherapy or radiation therapy for longer than 48 h before the imaging examinations, contraindications to sedation, a history of major allergic reaction to IV contrast material, and the presence of a cardiac pacemaker or intracranial vascular clips. | 7 (5 newly diagnosed and 2 recurrences) | Mean age: 10.75 Range: 16–17 years | 5:2 | Ewing sarcoma, 2 (1 recurrent) Rhabdomyosarcoma, 4 (1 recurrent) Neuroblastoma, 1 |
Goo et al., 2005 [25] | Inclusion criteria: children who underwent WB-MRI and conventional oncological imaging within 15 days | 36 (11 prior to chemotherapy, 25 after chemotherapy) | Median age: 3. 5 Range: 4–12 year | 21:15 | Rhabdomyosarcoma, 6 Lymphoma, 10 Neuroblastoma/ganglioneuroblastoma, 11/2 Germ-cell tumors, 3 Wilms tumor, 1 Hepatoblastoma, 1 PNET, 1 Small round cell neoplasm, 1 |
Kumar et al., 2008 [27] | Inclusion criteria: children and adolescents with histopathologically proven small-cell neoplasms who underwent WB-MRI, SSC and FDG PET/CT and iliac crest biopsy. Exclusion criteria: contraindications to sedation and the presence of a cardiac pacemaker or intracranial vascular clips | 26 | Range: 7–16 years | 20:6 | Ewing sarcoma/PNET, 11 Rhabdomyosarcoma, 5 Ganglioneuroblastoma, 1 Neuroblastoma, 8 Granulocytic sarcoma, 1 |
Krohmer et al., 2010 [28] | Inclusion criteria: children with suspicion or histological confirmation of malignant disease (lymphoma or solid malignant tumors); a maximum age of 18 years; at most 14 days between PET and WB-MRI, at least one conventional cross-sectional imaging examination, and no therapeutic action related to the malignant disease (e.g. chemotherapy). Exclusion criteria: patients without conventional cross-sectional imaging examinations and patients with suspicion or diagnosis of recurrent malignant disease | 24 | Mean age: 14 5/12 Range: 5–18 years | 14:10 | Hodgkin lymphoma, 11 Ewing sarcoma, 5 Osteosarcoma, 3 Follicular lymphoma, 1 Fibrosarcoma, 1 Rhabdomyosarcoma, 1 Alveolar sarcoma, 1 Langerhans cell histiocytosis, 1 |
Technical characteristics regarding whole-body magnetic resonance imaging
Author, year publication | Magnetic field (T) | Coil | Sequences | Sedation | Examination time (min) | Imaging planes | Description of MRI techniquea
| Observers (number and experience defined and data analysis) | MRI criteria for skeletal metastasis | Interpretation described in detailc
|
---|---|---|---|---|---|---|---|---|---|---|
Daldrup-Link et al., 2001 [24] | 1.5 | Head coil (small children) Body coil (older children) Spine coil for the spine | T1-W SE | No information | 45 in young children, 55–60 in older children/ adolescents | AX, SAG, COR | Yes | 2 observers, experience not defined, consensus reading | On T1-W SE: metastatic bone or bone marrow lesion defined as focal or diffuse hypointense signal relative to adjacent (or, in the extremities, contralateral) normal bone marrow. | No |
Mazumdar et al., 2002 [26] | 1.5 | Head coil (small children), Body or phase array coil (older children) | COR T1 turbo SE, COR T2 turbo STIR | Nob
| Range: 15–20 | COR | Yes | 2 observers, experience not defined, consensus reading | On turbo STIR: skeletal metastasis defined as focal or diffuse hyperintensity of bone marrow relative to skeletal muscle or as destruction of cortical bone. On T1-W: skeletal metastases defined as areas of hypointensity | No |
Goo et al., 2005 [25] | 1.5 | Body coil | COR turbo STIR (in all patients), SAG FS T2 (51/95 exams), SAG turbo STIR (32/95), COR FS T1 pre-and post-contrast (12/95) | Yes | Approximately 30 | COR, SAG | Yes | 2 observers, 1 year and 2 years of experience with WB-MRI, consensus reading | Bone marrow lesion hyperintense to muscle and normal bone marrow | Yes |
Kumar et al., 2008 [27] | 1.5 | Total imaging matrix coil | SAG T1 of the spine, COR STIR | Yes | Mean: 50 (range 40–60) | COR, SAG | Yes | 2 observers, experience not defined, consensus reading | On turbo STIR: skeletal metastasis defined as focal or diffuse hyperintensity of marrow, ≥signal intensity of CSF; focal or heterogeneous marrow signal variations or destruction of cortical bone. On T1-W: bone marrow metastases defined as areas of hypointensity, ≤signal intensity of skeletal muscle or as heterogeneous bone marrow signal variations. | No |
Krohmer et al. 2010 [28] | 1.5 | Body coil | AX T2-STIR, COR T2-STIR, COR T1-TSE | Yes | Mean: 45 | AX, COR | Yes | 2 observers, experience not defined, consensus reading | Bone lesion: signal alteration (hyperintensity on T2-W) | No |
Interpretation of whole-body magnetic resonance imaging
Reference test and flow of timing
Study authors, year publication | Composition of reference standard | Interval between WB-MRI and reference standard | Proportion of study group undergoing reference standard |
---|---|---|---|
Daldrup-Link et al., 2001 [24] | Pathology (biopsy), PET, bone scintigraphy, clinical and imaging follow-up | Mean: 11 days, Range: 3–25 days maximum interval for imaging | 39/39 |
Mazumdar et al., 2002 [26] | Bone scintigraphy, chest and abdominal CT, iliac crest biopsy (some patients), PET, histology, clinical and radiologic follow-up | <10 days | 7/7 |
Goo et al., 2005 [25] | Pathology and other clinical results, including ultrasound, CT, MRI, scintigraphy, and clinical follow-up | ≤15 days | 36/36 |
Kumar et al., 2008 [27] | Bone scintigraphy, PET-CT, follow-up, dedicated MRI, iliac crest biopsy | ≤25 days | 26/26 |
Krohmer et al. 2010 [28] | Different cross-sectional imaging (MRI, CT and/or US) | Mean: 5 days Range: 0–13 days | 24/24 |
Data on per-lesion, per-region and per-patient
Study authors, year publication | Positives | Negatives | Sensitivity | PPVa
| Specificityb
| ||
---|---|---|---|---|---|---|---|
TP | FP | FN | TN | TP/(TP+FN) | TP/(TP+FP) | TN/(TN+FP) | |
Per lesion | |||||||
Daldrup-Link et al., 2001 [24] (n=51 lesions) | 42 | 3 | 9 | NA | 82.4% | 93.3% | |
Goo et al., 2005 [25] (n=112 lesions compared with bone scintigraphy)c
| 111 | 7 | 1 | NA | 99.1% | 94.1% | |
Goo et al., 2005 [25] (n=4 lesions compared with MIBG) c
| 4 | 48 | 0 | NA | 100% | 7.7% | |
Goo et al., 2005 [25] (n=76 lesions compared with CT) c
| 76 | 35 | 0 | NA | 100% | 68.5% | |
Krohmer et al., 2010 [28] (n=42 lesions) | 42 | NAd
| 0 | NA | 100% | ||
Per region | |||||||
Kumar et al., 2008 [27] (n=208 regions) | 39 | 1 | 1 | 167 | 97.5% | 99.4% | |
Per patient | |||||||
Daldrup-Link, et al., 2001 [24] (n=39 patients) | 16 | 0 | 5 | 18 | 76.2% | 100% | |
Mazumdar et al., 2002 [26] (n=7 patients) | 2 | 0 | 0 | 5 | 100% | 100% |