Imaging of Pediatric Bone Tumors: A COG Diagnostic Imaging Committee/SPR Oncology Committee White Paper

Abstract

Malignant primary bone tumors are uncommon in the pediatric population, accounting for 3–5% of all pediatric malignancies^1,2. Osteosarcoma and Ewing sarcoma comprise 90% of malignant primary bone tumors in children and adolescents³. This manuscript provides consensus-based recommendations for imaging in children with osteosarcoma and Ewing sarcoma at diagnosis, during therapy, and after therapy.

Introduction

Osteosarcoma (OS) is an aggressive, osteoblastic tumor accounting for about 55% of primary malignant bone tumors in children and adolescents³. The World Health Organization classifies OS into intramedullary and surface types, with multiple histologic subtypes; the most common is high grade conventional intramedullary OS^4,5. Osteosarcoma has a bimodal incidence with a primary peak during adolescence and a secondary peak in adults older than 60 years. There is a slight male predominance with the adolescent incidence peak occurring earlier in females than males, theorized to correspond to the pubescent growth spurt. It most commonly arises in the metaphyses of long bones with the distal femur, proximal tibia, and proximal humerus being the most frequent sites^1,3,6–10. It is associated with genetic abnormalities, including Li-Fraumeni syndrome, hereditary retinoblastoma, Rothmund-Thomson syndrome Type 2, Bloom syndrome, Werner syndrome, RAPADILINO syndrome, and Diamond Blackfan anemia¹¹.

Ewing sarcoma (EWS) is a high-grade, aggressive, small round blue cell tumor that most commonly arises in bone but also occurs in extra-skeletal soft tissue^12–14. It comprises 36% of malignant primary bone tumors in children and adolescents, with a peak incidence during adolescence. It is slightly more common in males and is more common in Whites than in other racial designations^8,15–17. Approximately 50% of skeletal EWS arise within the axial skeleton, commonly the pelvis, chest wall, ribs, and vertebra. Appendicular EWS typically arises in the diaphyses of the femur, tibia, fibula, and humerus^12,13. Ewing sarcoma is associated with a characteristic EWSR1/FLI1 fusion gene resulting from a t(11;22)(q24;q12) reciprocal chromosome translocation^12,18.

The most common presenting symptom for both OS and EWS is pain, which may be insidious, intermittent, worse at night, and mistaken for “growing pains”. Patients may present after trauma, which may exacerbate the pain or result in a pathologic fracture. A mass is palpable in 30–40% of patients. EWS may present with constitutional symptoms mimicking osteomyelitis or lymphoma in a minority of patients^12,19,20.

Staging Systems

The two primary staging systems used for bone and soft tissue tumors are the Musculoskeletal Tumor Society (MSTS) / Enneking system²¹ and the American Joint Committee on Cancer (AJCC) system²², however these systems remain poorly validated and inconsistently utilized for OS and EWS^23,24. A small study of OS patients compared the MSTS and current AJCC staging systems and found similar predictive accuracy between them²³. A recent study analyzed OS data from the National Cancer Registry and identified histologic grade and presence of metastases as the only strong, independent prognostic factors for staging²⁴. A staging system created using only grade and metastatic disease was then compared to the MSTS and AJCC systems using OS data from the Surveillance, Epidemiology, and End Results (SEER) Program and showed that the simplified system had similar predictive accuracy²⁴.

Imaging at Diagnosis

Radiographs of the symptomatic site are recommended as the initial imaging examination for malignant bone tumors (GRADE: A, SOR 1.00, Very Strong Recommendation).

Radiographs are typically obtained for evaluation of pain, a palpable mass, after injury, or when a bone lesion is incidentally detected on a prior study. Radiographs are accurate for initial identification and characterization of an osseous mass, assist generation of a differential diagnosis, and prompt further workup^9,20,25–30.

MRI without and with contrast is recommended for evaluation of a suspected malignant bone tumor (GRADE: A, SOR 1.00, Very Strong Recommendation).

Magnetic resonance imaging (MRI) of the primary site should be obtained after identification of a suspicious bone lesion or if radiographs do not provide an adequate assessment of symptoms^{9,20,25,28,30–32}. The multiplanar capabilities and superior soft tissue contrast of MRI allow accurate assessment of staging characteristics. The long-axis field of view must cover the entire bone and adjacent joints (“join-through-joint imaging”) for proper evaluation of skip metastases^28,32–37. MRI provides information vital to surgical planning such as intramedullary tumor extent, involvement of the physis and epiphysis, intraarticular extension, and involvement of adjacent neurovascular structures^{9,25–30,32–38}. MRI should be performed prior to biopsy to avoid confounding post-procedural changes^31,39.

Non-contrast enhanced chest CT is recommended at initial staging for pulmonary metastases (GRADE: A, SOR: 1.00, Very Strong Recommendation).

Pulmonary metastatic disease at diagnosis occurs in approximately 20% of OS and 25% of EWS patients with about 80% and 50% respectively involving the lungs^{10,12,13,30,40–42}. Computed tomography (CT) of the chest is more sensitive for pulmonary metastases than radiographs^43–46. The addition of maximum intensity projection (MIP) images improves the sensitivity, specificity, and accuracy of CT for detection of pulmonary metastases^46–49. Administration of intravenous iodinated contrast is not required unless there is concern for hilar, mediastinal, or chest wall involvement^28,44.

18F-FDG PET/CT or PET/MRI is recommended at initial staging for extrapulmonary metastases (GRADE: A, SOR: 1.22, Very Strong Recommendation).

Extrapulmonary metastases occur in both high-grade OS and EWS but are more common in EWS. Extrapulmonary metastases primarily involve the bone and bone marrow and portend a worse prognosis than localized disease or pulmonary metastases alone^{9,41,42,50,51}. 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) has been shown to be more accurate than whole body Tc99m-methylene diphosphonate (Tc99m-MDP) bone scintigraphy for detection of extrapulmonary metastases in both OS and EWS, particularly within the spine and pelvis^52–72. Use of 18F-FDG PET/CT for staging allows potential omission of bone marrow biopsy in Ewing sarcoma patients^73–76. 18F-FDG PET/CT has consistently been found to be less sensitive and accurate than chest CT for detection of pulmonary metastases and cannot replace it^52,59,70. The metabolic information provided by 18F-FDG PET/CT can determine high-yield biopsy targets, assess for tumor necrosis, and is being investigated as a prognostic biomarker^53,65,71,72. Whole body MRI (WB-MRI) is a promising technique for whole body evaluation for extrapulmonary metastases but has not been investigated or validated in OS or EWS to the same extent as 18F-FDG PET/CT^65,77–82. WB-MRI benefits from high tissue contrast, including the availability of diffusion weighted imaging (DWI), and does not utilize ionizing radiation. 18F-FDG PET/MRI combines the metabolic imaging of 18F-FDG PET with the advantages of WB-MRI and is an alternative to 18F-FDG PET/CT where available^{53,58,72,81,83,84}.

Follow Up Imaging Prior to Local Control

Follow up evaluation is recommended prior to local control, using the same imaging studies as initial staging (GRADE B, SOR 1.78, Strong Recommendation).

Follow-up is performed to assess changes to the primary tumor and metastatic disease, facilitate planning for local control, and to provide prognostic information^28,30.

Primary Tumor Response Assessment

The pathologic response of the primary tumor to neoadjuvant chemotherapy is the primary prognostic factor in OS and EWS with tumor necrosis > 90% correlating with a better outcome^85–89. Imaging has historically played a limited role in assessing tumor response. The neoplastic osteoid matrix of OS precludes significant decreases in tumor volume, even in tumors with good response to chemotherapy^88,90. Tumor volume reduction occurs in EWS but correlates poorly with percent tumor necrosis^88,91.

Assessment of primary tumor response to neoadjuvant chemotherapy using PET and MRI are current topics of research. Numerous studies analyzing parameters derived from 18F-FDG PET^{70–72,92–108}, MRI with DWI and apparent diffusion coefficient (ADC) maps^92,109–119, and dynamic contrast enhanced MRI^119–124 show the promise of imaging biomarkers for noninvasive assessment of primary tumor response and prognosis.

Follow Up Imaging During and at End of Adjuvant Therapy

Radiographs and MRI without and with contrast are recommended for follow up imaging of the primary site (GRADE B, SOR 2.11, Moderate Recommendation).

The timing and frequency of follow up imaging of the primary site during adjuvant therapy is variable and dependent on the site of the primary tumor, the methods of local control and adjuvant therapy, and the patient’s clinical status. Both radiographs and MRI of the local site should be obtained at the end of adjuvant therapy to serve as baseline studies for surveillance.

Radiographs are obtained to assess the surgical result, follow healing, assess for complications, provide MRI correlation, and may show evidence of local recurrence^{28,44,125–131}. MRI provides superior assessment of soft tissue reconstruction and healing, evaluation of complications, and identification of local recurrence^{44,129,131,132}. MRI is not required for follow up after limb-salvage surgery, though imaging adjacent to metallic hardware has improved with commercially available metallic artifact reduction sequences (MARS) and can be performed in the setting of new clinical symptoms or imaging findings warranting further assessment^{127,133–135}. 18F-FDG PET/CT is an alternative to MRI if there is clinical or imaging findings concerning for local recurrence near metallic hardware, showing comparable efficacy as well as the ability to assess for concurrent extrapulmonary metastases^{72,127,136–138}.

Non-contrast chest CT is recommended for pulmonary follow up imaging (GRADE B, SOR 1.78, Strong Recommendation).

The optimal timing and frequency of pulmonary follow up is unknown, though non-contrast chest CT is typically obtained roughly halfway through adjuvant therapy and at the end of adjuvant therapy, the latter serving as a baseline for surveillance. Chest CT allows for detection of new pulmonary metastases, reassessment of previously identified pulmonary nodules, and evaluation of sites of prior pulmonary nodule resection^44,46.

18F-FDG PET/CT or PET/MRI is recommended for whole-body follow up imaging at the end of adjuvant therapy (GRADE B, SOR 2.33, Moderate Recommendation)²⁸.

Routine whole-body imaging is not recommended during adjuvant therapy unless new clinical symptoms or imaging findings warrant further assessment or restaging²⁸.

Imaging for Surveillance

The purpose of surveillance imaging is to identify recurrent disease for which viable therapeutic options exist. Surveillance imaging is widely practiced for OS and EWS but remains a controversial subject. The data supporting a survival benefit from surveillance imaging is scarce^139–145 and the optimal sites, modalities, and frequency of surveillance imaging remains unknown.

Following therapy, about 40% of patients with high-grade OS and 30–40% with EWS develop local or metastatic recurrence^146–149. Recurrence in both high-grade OS and EWS occurs most commonly within the first 2 years after therapy^141,150. Most guidelines recommend more frequent imaging during this time reasoning that early detection may facilitate successful secondary therapy and improve survival, but this remains controversial as a survival advantage from early detection has not been well validated. Early recurrence is a known indicator of poor prognosis and may correspond to more aggressive tumors, mitigating a potential advantage of earlier detection^50,151–153. Additionally, a higher frequency of surveillance imaging results in a greater financial burden, increased radiation exposure, a higher risk of false positive exams, and may increase anxiety in patients^126,154.

The recommended surveillance intervals for high grade OS and EWS are every 3 months for the first 2 years, every 6 months for the next 3 years, and every year for the next 5 years^44,126,145 (GRADE C, SOR 2.89, Moderate Recommendation).

Due to the low risk of recurrence in low grade OS, surveillance is only recommended annually for the first 2 years¹⁴⁵.

Radiographs (GRADE B, SOR 1.78, Strong Recommendation) and MRI without and with contrast (GRADE C, SOR 2.78, Moderate Recommendation) are recommended for local surveillance.

Radiographs have value in longitudinal assessment of post-therapy healing, identifying hardware complications, and evaluating for local recurrence^{28,30,44,125–131}. MRI for local surveillance has recently been advocated, including in guidelines by the American College of Radiology (ACR) and National Comprehensive Cancer Network (NCCN)^{30,44,127,129,132}. Although MRI is not required after limb-salvage surgery, the use of MARS techniques allows MRI evaluation in the setting of new clinical symptoms or imaging findings warranting further assessment^{127,129,132–135}.

For high grade OS and EWS with pulmonary metastases at diagnosis, non-contrast chest CT is recommended for pulmonary surveillance for the first 5 years with radiographs recommended for the next 5 years (GRADE C, SOR 2.22, Moderate Recommendation). For non-metastatic EWS, radiographs are recommended for pulmonary surveillance (GRADE B, SOR 2.11, Moderate Recommendation).

The optimal modality and frequency of pulmonary surveillance remains unknown. Studies evaluating pulmonary surveillance in OS and EWS are sparse, often include pooled data from soft tissue sarcoma, and show mixed results^{141,144,155–159}. There is a survival advantage to pulmonary metastasectomy in high grade OS when complete metastatic resection is achievable, even with repeated resections, which has not been definitively demonstrated in EWS and soft tissue sarcoma^160–162. A prospective trial showed non-inferiority of chest radiograph to chest CT and 6 month follow up to 3 month follow up in a cohort of soft tissue and bone sarcomas^157,158 but this result is difficult to apply given the greater value of metastasectomy in OS. Additionally, a retrospective study showed a survival advantage when using chest CT for pulmonary surveillance compared to chest radiographs in patients with OS¹⁵⁶. In the absence of a clear survival advantage provided by metastasectomy in non-metastatic EWS, pulmonary surveillance with chest radiography is recommended, with non-contrast chest CT recommended if new clinical symptoms or imaging findings warrant further assessment.

18F-FDG PET/CT or PET/MRI is recommended for whole-body evaluation of extrapulmonary metastases if new clinical symptoms or imaging findings warrant further assessment and if additional interventions are feasible (GRADE C, SOR 1.89, Strong Recommendation)^28,44.

Routine whole-body surveillance is not recommended due to the low incidence of extra-pulmonary metastases, the lack of effective therapies, and the dismal prognosis^44,140,142.

Clinical evaluation and targeted diagnostic imaging for evaluation of long-term complications in OS and EWS survivors should be performed as clinically indicated. Therapy related health risks include cardiomyopathy, hypertension, and chronic kidney disease. Limb-salvage operations carry the risk of hardware complications and failure¹⁶³. Secondary neoplasms can occur as sequelae of chemotherapy and radiation therapy^164–166.

Future Advancements

These recommendations will continue to evolve with time. Continued validation and improved availability of WB-MRI and 18F-FDG PET/MRI may allow for decreased reliance on whole body 18F-FDG PET/CT, decreasing the cumulative radiation dose for imaging in OS and EWS.

The currently recommended surveillance protocols for OS and EWS remain controversial, with supporting data often limited by study size and confounded by combined patient cohorts of soft tissue and bone sarcoma. Future prospective, multi-institutional clinical trials specific to OS and EWS could bring clarity to this issue by comparing outcomes between different surveillance modalities and intervals.

Continued investigations into imaging biomarkers of tumor response, including features from PET and MRI^92,112, as well as artificial intelligence assisted segmentation and analysis tools^167–169, may allow creation of validated radiomics¹⁷⁰ that predict poor responders at an early timepoint when therapy could be altered or intensified for therapeutic benefit.

Novel radiopharmaceuticals are being investigated for their ability to target specific sarcomas, including pediatric bone tumors^72,171. This may increase the specificity of PET imaging and could lead to the development of powerful theragnostic tools^172,173.

Table 1:

Advantages and disadvantages of each modality for the evaluation of the primary tumor

Procedure Name	Timepoint(s)	Advantage(s)	Disadvantage(s)
Radiographs	Diagnosis Follow Up Surveillance	- Availability - Inexpensive - Initial lesion characterization - Post-therapy assessment of healing and hardware complications	- Low tissue contrast - Lacks multiplanar capability - Less sensitive for axial tumors - Limited sensitivity for local recurrence
MRI	Diagnosis Follow Up Surveillance	- Lacks ionizing radiation - Superior tissue contrast - Superior evaluation of local tumor extent - Superior assessment of tumoral involvement of adjacent critical structures - High sensitivity for local recurrence	- Expensive - Potential need for anesthesia - Artifacts from hardware, partially mitigated using artifact-reduction techniques.
18F-FDG PET/CT or PET/MRI	Staging Follow Up	- High value biopsy target assessment - May obviate need for bone marrow biopsy in EWS - Potential for tumor response assessment - Efficacious for assessment of local recurrence	- Availability

Table 2:

Suggested MRI protocol for primary tumor evaluation at all sites.

Plane	Sequence	Contrast phase	Coverage	Required/Optional	Comment
MRI Without and With Contrast
Coronal	T1W	Pre-contrast	Joint through joint (Appendicular)	Required	Best sequence for evaluation of tumor extent and skip metastases
Coronal	STIR or T2W Fat Suppressed	Pre-contrast	Joint through joint (Appendicular)	Required
Sagittal	STIR or T2W Fat Suppressed	Pre-contrast	Joint through joint (Appendicular)	Recommended	Added value for the vertebra and at many joints
Axial	T1W	Pre-contrast		Required
Axial	STIR or T2W Fat Suppressed	Pre-contrast		Required
Axial	DWI	Pre-contrast		Optional	b = 0, 900
Axial	T1W Fat Suppressed	Post-contrast		Required
Coronal	T1W Fat Suppressed	Post-contrast	Joint through joint (Appendicular)	Required
Sagittal	T1W Fat Suppressed	Post-contrast	Joint through joint (Appendicular)	Required

Table 3: Non-contrast CT chest protocol for pulmonary metastatic assessment.

Volumetric noncontrast CT of the chest is performed and reconstructed as contiguous 1–3 mm axial, sagittal, and coronal images and 10 mm axial and coronal MIP images.

Name	Coverage	Slice thickness	Contrast phase	Reformat planes/types/ reconstruction kernel	Comment
CT Chest Without Contrast	Thoracic inlet through the upper abdomen	3 mm	Non-contrast	Axial, sagittal, coronal (soft tissue and lung kernals) 10 mm axial and coronal MIP reconstruction (lung kernal)

Table 4:

Whole Body 18F-FDG PET/CT and PET/MRI protocol for whole-body assessment of extrapulmonary metastases.

Study name	Patient prep	Radiopharmaceutical	Dose range	Delivery	Time from dose to imaging	Imaging acquisition	Comment
18F-FDG PET/CT	- NPO for 4–6 hours to decrease serum glucose and insulin levels. - Stop all IV dextrose 4 hours prior to 18F-FDG administration - Warm the patient for 30–60 minutes prior to 18F-FDG administration to minimize uptake in brown adipose tissue.	18F-Fluorodeoxyglucose (FDG)	0.10–0.14 mCi/kg Minimum dose 0.7 mCi	Intravenous	60 ± 10 minutes	PET/CT CT performed with appropriate pediatric CT settings.	True whole-body field of view recommended (skull vertex to feet).
18F-FDG PET/MRI	Same	18F-Fluorodeoxyglucose (FDG)	0.08 mCi/kg Minimum dose 0.7 mCi	Intravenous	60 ± 10 minutes	PET/MR MR acquisition (dependent on available technology and local practice).	True whole-body field of view recommended (skull vertex to feet). Dedicated MRI of the primary tumor can be performed during the uptake phase.

Abbreviation

OS

Osteosarcoma

EWS

Ewing Sarcoma

MSTS

Musculoskeletal Tumor Society

AJCC

American Joint Committee on Cancer

SEER

Surveillance, Epidemiology, and End Results

MRI

Magnetic Resonance Imaging

CT

Computed Tomography

MIP

Maximum Intensity Projection

18F-FDG PET/CT

18F-fluorodeoxyglucose positron emission tomography/computed tomography

Tc99m-MDP

Tc99m-methylene diphosphonate

WB-MRI

Whole Body Magnetic Resonance Imaging

DWI

Diffusion Weighted Imaging

MARS

Metallic Artifact Reduction Sequences

ACR

American College of Radiology

NCCN

National Comprehensive Cancer Network