Comparison of the Different Imaging Modalities Used to Image Pediatric Oncology Patients: A COG Diagnostic Imaging Committee/SPR Oncology Committee White Paper

Abstract

Diagnostic imaging is essential in the diagnosis and management, including surveillance, of known or suspected cancer in children. The independent and combined roles of the various modalities, consisting of radiography, fluoroscopy, ultrasonography (US), computed tomography (CT), magnetic resonance imaging (MRI), and nuclear medicine (NM) are both prescribed through protocols but also function in caring for complications that may occur during or subsequent to treatment such as infection, bleeding, or organ compromise. Use of a specific imaging modality may be based on situational circumstances such as a brain CT or MR for a new onset seizure, chest CT for respiratory signs or symptoms, or US for gross hematuria. However, in many situations there are competing choices that do not easily lend themselves to a formulaic approach as options; these situations depend on the contributions of a variety of factors based on a combination of the clinical scenario and the strengths and limitations of the imaging modalities. Therefore, an improved understanding of the potential influence of the imaging decision pathways in pediatric cancer care can come from comparison among the individual diagnostic imaging modalities. The purpose of the following material to is to provide such a comparison. To do this, pediatric imaging content experts for the individual modalities of radiography and fluoroscopy, US, CT, MRI and NM will discuss the individual modality strengths and limitations.

Introduction:

Diagnostic imaging is essential in the diagnosis and management, including surveillance, of known or suspected cancer in children. The independent and combined roles of the various modalities, consisting of radiography, fluoroscopy, ultrasonography (US), computed tomography (CT), magnetic resonance imaging (MRI), and nuclear medicine (NM) are both prescribed through protocols but also function in caring for complications that may occur during or subsequent to treatment such as infection, bleeding, or organ compromise. Use of a specific imaging modality may be based on situational circumstances such as a brain CT or MR for a new onset seizure, chest CT for respiratory signs or symptoms, or US for gross hematuria. However, in many situations there are competing choices that do not easily lend themselves to a formulaic approach as options; these situations depend on the contributions of a variety of factors based on a combination of the clinical scenario and the strengths and limitations of the imaging modalities. Therefore, an improved understanding of the potential influence of the imaging decision pathways in pediatric cancer care can come from comparison among the individual diagnostic imaging modalities.

The purpose of the following material to is to provide such a comparison. To do this, pediatric imaging content experts for the individual modalities of radiography and fluoroscopy, US, CT, MRI and NM will discuss the individual modality strengths and limitations. For clarity, each modality section will be arranged to first summarize the basic principles of how images are obtained; this information is intended to add clarity to subsequent material such as limitations, real and potential biological effects and other safety considerations. Equally important modality factors that impact decision pathways include, cost, availability, imaging performance and interpretation expertise, and broader healthcare community culture (i.e., personal or practice experience and expectation). The information is summarized in Table 1.

Table 1:

Comparisons of Imaging Modalities in Childhood Cancer

	RG	US	CT	MRI	NM	PET
ROLE
Detection	++	+++	+++	+++	+	+++
Staging			+++	+++	++	+++
Response *			+++	+++	+	+ − ++
Restaging			++	++	++	+++
Surveillance	++	+++	+++	+++	+++	++
Cancer Predisposition	++	++	++	+++	++	++
Portability	+++	+++	+
Exam Time **	Subsecond	0.5 hr	Second(s)	0.5–1.0 hr	0.5–1.0 hr	0.5–1.0 hr
Operator Dependence ***	+	+++	+	+	+	+
Relative Cost	Low	Moderate	High	High	High	High
Adverse Safety Consideration (for Radiation ****)	Radiation	None	Radiation; Contrast media; Anesthesia/sedation	Magnetic field; Gradients; Noise; Contrast media; Anesthesia/sedation	Radiation, Anesthesia/sedation	Radiation; Anesthesia/sedation

RG: radiography; US: ultrasound; CT: computed tomography; MRI: magnetic resonance imaging; NM: nuclear medicine; PET: positron emission tomography

“+” number: overall increasing relative value or impact on study (operator dependence)

^*(+) number value depends on relevance across all tumor types

^**Exam times vary; general modality representations provided. Does not include patient preparation/examination set up: values are length of all image acquisition times. NM does not include radiopharmaceutical administration phase.

^***Throughout image acquisition. Pediatric-specific techniques/protocols for operators (sonographers and technologists) are understood for all modalities.

^****Relative magnitudes not expressed (e.g., radiation exposure amount is very low in radiography compared to computed tomography)

Other: CT and MRI may require breath hold; movement can substantially degrade MRI and NM study quality

It is important to clarify some aspects of the following material. First, the information provided is not intended to necessarily represent an individual institution or professional body. However, contributions do represent many decades of expertise in both clinical practice as well as the academic domain of the authorship. In this respect, information is provided, including comparisons, but the material is intended to serve as guidance, short of recommendations, for one’s imaging approach to pediatric cancer care. This information, too, is based on current technology, recognizing that advances in imaging protocols (e.g., evidence supporting less frequent or alternate pathways), and technical advances such as with artificial intelligence or equipment itself (e.g., new MR sequences, or photon counting CT) will almost certainly alter the relative impact of the comparisons provided here. The scope of the discussion does not deal with image-guided intervention.

This manuscript was funded in part by the National Clinical Trials Network Operations Center Grant U10CA180886. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health

Radiography and Fluoroscopy

Radiography is very often simply referred to as an x ray, even within the radiology community. The technology underlying both fluoroscopy and radiography is a source, the x-ray tube, and a detector. In the past, for both radiography and fluoroscopy, the detector was film-based. The vast majority of current imaging is all digital in the United States and developed countries. This has promoted the potential immediate, “off-site”, even global availability of studies and significant advantage over the prior film based technology. Film loss, in addition, can be essentially eliminated. Moreover, the ability to manipulate the digital information obviates in most all circumstances the retakes that were not infrequent with prior technology. Advances in detector efficiency continue, and radiography currently has a 95% dose reduction compared with technology in the 1950s (1). There are as with most all imaging, unique challenges (including gaps) and special considerations for pediatric fluoroscopy (2, 3). More familiar advancements in fluoroscopy include improved system efficiency, low frame rates (e.g., two frames per second vs continuous or 30 frames per second), and last image store (allowing for prolonged assessment when fluoro is off) (3–5).

Radiography still comprises an estimated 66% of all imaging examinations in children in the Unites States (6). While the initial detection of malignancy in children can occur with radiography such as with mediastinal adenopathy with lymphoma or more typically with skeletal tumors, this is nearly always followed by subsequent modalities providing more specific information. The choice of such modalities may follow initial x-ray findings, such as MRI for an aggressive bone process, or chest CT for a mediastinal or parenchymal mass, or US for suspected abdominal or pelvic mass. Use of radiography is dependent on individual oncologic protocols. A chest x ray for metastatic lesions in tumors with propensity for lung involvement is migrating almost exclusively to chest CT for diagnosis and to a large extent surveillance, although exceptions exist. (7).While bone surveys have played a traditional role in assessment of disorders with diffuse skeletal propensity (e.g. skeletal involvement by Langerhans cell histiocytosis), whole body MRI, as well as PET/CT or PET/MR by virtue of increased sensitivity and specificity in multiple organ systems, are assuming more prevalent role staging and surveillance (8–11). Of course for acute respiratory or abdominal circumstances, such as infection in the immunocompromised child, for evaluation of bowel abnormalities such as obstruction, or assessment of support apparatus positions, radiography still occupies an important, and often sufficient role.

Fluoroscopy has a very limited role in the evaluation of suspect malignancies. There may be an unexpected luminal or extrinsic finding during evaluation of the gastroenteric or genitourinary system when assessing some nondescript clinical concerns. Fluoroscopy can be used during care for assessment of operative interventions for leaks or obstructions, or for placement of devices such as feeding tubes.

The safety concerns and biological considerations for radiography and fluoroscopy are mostly limited to radiation. The exposure for most radiography is quite low (12, 13). A chest x ray for example is the equivalent of about 2 days of normal background exposure. For a distal extremity series such as a three-view foot in the setting of Ewing sarcoma, the background equivalent radiation exposure just about 40 minutes. The radiation dose (estimated from exposure values) for modern fluoroscopy will depend on the individual application but is also in general quite low. The contributions to the estimated annual per capita collective effective dose for children in the U.S. from radiography are 6%, and for fluoroscopy 3%; this is compared with CT where the contribution is 84% due to a relatively higher radiation dose per examination. To complete the comparison of those imaging modalities that use ionizing radiation, the contribution of nuclear imaging in children is 3% and interventional procedures is 4% (13). Enteral (also known as oral) and contrast through various catheters (14), including rectal tubes, biliary catheters, or bladder catheters, has a much lower contrast reaction profile when administered intravenously (see CT modality section). Barium contrast is essentially inert, with an exceedingly uncommon contrast reaction profile. For many years, care has been taken however, to use an iodinated agent when there is a possibility of a gastroenteric or urinary collecting system leak as barium may persist for many years in the sequestered parts of the cavity. Previously held challenges to morbidity such as bowel obstruction resulting from this persistence of barium have been challenged (15), however the use of iodinated material in the setting of possible leak is still standard practice.

Radiography and fluoroscopy are widely available, portable (“c-arm” type devices for fluoroscopy), the lowest relative cost when considering the menu of exams representing the other imaging modalities, and are relatively consistently preformed, although some variation based on expertise can occur with fluoroscopic studies. Radiography does not require immobilization; fluoroscopy immobilization is relatively easily performed for those who cannot hold still, especially preschool children. In addition, movement, unlike with a CT, NM, or MRI examinations is immediately evident and simply addressed by waiting for a time the child is still to perform.

Ultrasound

After radiography, ultrasound is the most frequently used imaging technique in children. Compared to the other cross-sectional imaging modalities (CT and MRI), ultrasound is relatively inexpensive, widely available, does not expose the patient to radiation (unlike CT and PET), and does not require sedation (as is often needed for MRI). Advances in transducer technology and computing power have improved the image quality of ultrasound greatly over the last 20 years, and the cost of quality imaging equipment has markedly decreased allowing great access to ultrasound imaging. In resource constrained settings, ultrasound is arguably one of the more valuable imaging methods. In follow-up imaging for recurrence or surveillance in patients predisposed to malignancy, ultrasound is often the modality of choice for given its availability, lesser cost, and patient comfort and safety. The usual criticism, that ultrasound quality is “user dependent,” should be taken less as a critique of the modality and more as a reminder that we owe it to our patients to be proficient with this powerful tool.

As the name indicates, ultrasound uses high-frequency sound waves to create images of internal body structures. A coupling medium (usually a commercial ultrasound gel) is used to create air free contact between the patient’s skin and the ultrasound transducer. The transducer both transmits the ultrasound pulses and receives the return pulses (or echoes) from internal structures. The returning pulses are then converted into detailed anatomic images. Color and spectral (waveform display) Doppler evaluation can determine internal blood flow and assess adjacent or encased blood vessels. The introduction of intravenous and intracavitary ultrasound contrast agents beyond echocardiographic settings has enabled greater specificity for some diagnoses (such as characterizing liver lesions) and the evaluation of tumor perfusion kinetics that may aid in tumor diagnosis and response assessment. Since images are obtained in real time, ultrasound is an excellent problem-solving tool and imaging protocols can be immediately adjusted if needed.

In the initial evaluation of a suspected mass in a child, ultrasound is often the first modality used. Ultrasound can often help determine if a mass is developmental in origin, is solid or cystic, where it is arising, and its involvement of adjacent structures. For many pediatric inquiries, ultrasound can be a definitive diagnostic test in determining the presence or absence of a neoplasm: a palpable flank mass may be shown to be a congenitally hydronephrotic kidney; a neck mass may prove to be simple lymphadenitis, a thyroglossal duct cyst, or fibromatosis colli; a soft tissue mass may reveal characteristics of a venolymphatic malformation rather than a neoplasm. Such definitive diagnoses can provide rapid reassurance to patients and families and can help to speed the next steps in evaluation if a tumor is found.

For some tumors, ultrasound may be the only imaging needed for diagnosis informing subsequent clinical management, such as with thyroid and testicular malignancies. The differentiation of some hepatic masses can be made by the pattern of enhancement and washout of ultrasound contrast agents, adding diagnostic specificity to ultrasound imaging (16). However, for most cancers ultrasound is rarely sufficient as a sole imaging method (17). As ultrasound cannot image structures deep within aerated lung, evaluation for potential lung metastases requires CT (although in the future MRI may be used). In larger patients the visibility of deeper structures is more limited and the involvement of adjacent organs more difficult to discern, necessitating CT or MRI. Ultrasound can assess blood flow but cannot asses the metabolic activity of a lesion, which requires nuclear medicine and PET imaging, or advanced MRI techniques. Despite these limitations, ultrasound remains valuable by helping to direct what the next most likely beneficial imaging or intervention might be.

Once a tumor is detected, ultrasound continues to be valuable in diagnosis and treatment. Tissue sampling is almost always needed to confirm a specific diagnosis, and this can often be performed with ultrasound-guided interventional biopsy that provides real-time radiation-free imaging guidance to allow reliably safe and diagnostic procedures (18). Secure central venous access is required for most chemotherapeutic regimens, and ultrasound-guided venous access has been proven to be safer and more reliable than non-image-guided access procedures (19).

Once therapy is initiated, ultrasound is useful in in diagnosing treatment complications and side effects. Immunosuppression often accompanies antineoplastic therapies, predisposing patients to infectious and inflammatory complications. Since ultrasound can be brought to the patient’s bedside, it is particularly useful for critically ill or isolated patients. Neutropenic enterocolitis (typhlitis) that begins with intestinal mucosal injury from cytotoxic chemotherapy complicated by subsequent bowel invasion of microorganisms and infection. Ultrasound can readily diagnose the abnormally thickened and inflamed bowel, and can identify, characterize, and guide aspiration of fluid collections that may indicate bowel perforation and abscess formation (20). Pleural effusions and ascites are not uncommon during chemotherapy and are effectively diagnosed with ultrasound. Cancer and its treatment may produce thrombophilia, leading to a variety of thrombotic complications and ultrasound is the first line imaging modality for detection (21). Deep vein thrombosis may arise due to inactivity and stasis increasing the risk of pulmonary embolism. Thrombosis may be associated with central venous line placement and lead to catheter malfunction or symptoms such as superior vena cava syndrome. Small and microvascular thrombotic complications include sinusoidal obstruction syndrome (veno-occlusive disease) of the liver that may complicate bone marrow transplantation. While the clinical scenario and presentation are usually suggestive of the diagnosis, Doppler ultrasound can confirm by showing characteristic changes in hepatic vascular waveforms and can be used to monitor progression or improvement of disease (22). As important as ultrasound’s role in “ruling in” a diagnosis is its ability to exclude abnormalities. Cancer therapies can produce hepatic and renal dysfunction, and ultrasound can demonstrate that biliary and renal collecting systems are not dilated and that hepatic and renal blood flow is normal, and thus avoid unnecessary interventions or changes in treatment regimens.

Most management protocols employ CT, MRI, and PET to monitor and evaluate the effect of treatment. However, ultrasound is commonly used to monitor for tumor recurrence and metastatic disease (occasionally alternating with other cross-sectional imaging). After Wilms tumor resection ultrasound is the modality of choice for evaluating the contralateral kidney (23), and after testicular cancer treatment ultrasound is valuable in monitoring the remaining testis. In conjunction with tumor markers, ultrasound is used to monitor for hepatoblastoma recurrence. Similarly, ultrasound is useful in patients with cancer predisposition syndromes or other conditions for routine surveillance (24–26). Ultrasound is the primary modality for monitoring patients with Beckwith-Wiedemann syndrome for renal and liver tumors, Denys-Drash and Von-Hippel Lindau syndrome patients for renal tumors, patients with multiple endocrine neoplasia syndrome and patients with DICER 1 mutations for thyroid and other cancers. Patients with Li-Fraumeni syndrome are predisposed to develop a variety of tumors and also have greater sensitivity to the effects of ionizing radiation, making ultrasound (and MRI) especially valuable for their imaging care (27).

Computed Tomography

Computed tomography (CT) is a noninvasive cross-sectional imaging modality, utilizing ionizing radiation and cutting-edge computer reconstruction algorithms, which can rapidly create detailed multiplanar images of the body from the acquired volume data set. The images then can be reconstructed from the volume; think of this volume data set as a loaf of bread that can be cut (reconstructed) in any way. The basic technology consists of a radiation source (x-ray tube) opposite a set of detectors that together rotate around the patient on the table (bed) during movement of the patient through the scanner gantry. IV contrast media, when given for a contrast-enhanced scan, provides opportunities for different phases of organ and other structure enhancement depending on when the scanning is performed.

CT plays an indispensable role in the diagnosis, staging, and care of pediatric oncologic disorders. In addition to accurate and reproducible anatomic localization and characterization of disease, CT can monitor cancer therapy, identify acute and chronic complications related to oncologic disorders and treatment, and define the presence and extent of disease relapse (28). Although CT has been criticized for its use of ionizing radiation, the potential risks are thought to be at most very small, and in general, the benefits of a clinically-indicated CT will far outweigh the, potential risks. When considering cancer risks from ionizing radiation, it is crucial for one to first consider the immediate benefits of a given CT examination, in addition to the risks of less conclusive diagnostic examinations, more invasive procedures, or foregoing diagnostic imaging entirely (29). Indeed, the ionizing radiation associated with CT has driven a movement to create and develop strategies to minimize radiation dose in pediatric patients, including performing only clinical relevant CT examinations, and adjusting scan parameters to clinical indication, the anatomic region imaged, and patient size (30).

One of the primary virtues of CT for pediatric oncologic imaging is its speed of image acquisition, which has improved substantially since its inception and early advancements in the 1970s and 1980s. Modern CT scanners can image pediatric patients in a matter of seconds or even less, improving image quality by mitigating respiratory motion and patient movement. In addition, CT is extremely versatile and can provide consistent and reproducible imaging for surveillance imaging for pediatric oncology patients. CT is widely available for pediatric imaging (31), and CT technology continues to define practice patterns because of its rapid and reproducible acquisition times, and high quality clinical data sets (30). Although beyond the scope of this modality review, appropriate imaging modalities and the salient use of CT for pediatric oncologic imaging have also been described in the literature (32).

Over the past two decades, comprehensive literature has detailed the potential carcinogenic risks of ionizing radiation, particularly in the context of pediatric diagnostic imaging (33–35). Risk estimates are generally derived from the linear-no-threshold (LNT) hypothesis which suggests that any dose of ionizing radiation has the potential to increase the risk of malignancy later in life. Although the LNT hypothesis is regarded as the best predictive model for risk, there are other studies that suggest that no statistically significant risk exists at ionizing radiation dose levels (36). As a result, there have been substantial efforts to optimize CT doses in an effort to maximize patient safety while maintaining diagnostic confidence. Many of these collective efforts can be traced back to the Image Gently Alliance (31, 37, 38) which led the charge of improving pediatric patient safety and ionizing radiation dose optimization.

Although it is always prudent to minimize pediatric patient exposure to ionizing radiation whenever possible, in general, in the context of staging or surveillance of pediatric malignancies, the benefits provided by CT imaging will generally outweigh the potential risks, particularly in the setting of oncologic imaging. Nonetheless, minimizing long-term ionizing radiation exposure for cancer survivors remains an important goal, and surveillance imaging with MRI or ultrasonography should at least be considered, whenever possible.

When performing CT imaging for pediatric oncologic indications, in general a single-phase CT imaging is generally sufficient; routine multiphase imaging should be avoided whenever possible (28). There are circumstances where different phases of contrast may provide additional characterization of malignance including impact on regional structures such for hepatic (39), pancreatic (40, 41) and renal cancer (23). When these circumstances are present, consideration should also be given to contrast enhanced MR imaging where multiple phases of enhancement can be performed without the attendant increased radiation dose found with multiphase CT examinations. Diagnostic reference levels, as radiation dose benchmarks for a practice’s most common pediatric CT examinations, have recently been described (42).

The risks of monitored anesthesia care (MAC) in infants and children are very small, particularly when performed by a dedicated pediatric anesthetist. Although the risk profile for pediatric sedation and anesthesia is very low, a small risk still exists, which can be divided into two time periods: immediate risks from anesthesia and long-term neurocognitive risks related to anesthesia drug exposure. Traditionally, both CT and MRI required the use of MAC for pediatric imaging, particularly in young patients and developmentally delayed individuals. Modern advances in pediatric CT imaging scan speed and iterative imaging algorithms continues to diminish the image acquisition time and further decrease patient doses. These advances in CT speed and dose optimization allow us to image the vast majority of pediatric oncology patients without MAC, and at a substantially lower dose than older generation scanners (43). Although CT is a relatively expensive imaging modality (44), it is typically less costly than MRI. This is particularly true when sedation is required for MRI as MAC adds significantly to imaging costs (43).

When imaging pediatric oncology patients with in neck, chest, and abdominopelvic CT, intravenous iodine-based contrast media (IV IBCM) is virtually always required for an accurate diagnosis. Notable exceptions would include CNS tumors and primary musculoskeletal imaging of primary or metastatic disease. Enteric contrast is not mandatory for abdominal oncologic imaging with modern multidetector CT scanners (45), although many adult and pediatric centers advocate for its use in this clinical situation (46). Contrast reactions are less common in children than adults (47) but are still a consideration when performing CT. Recently, there has been attention directed to the issue of thyroid dysregulation in young children following IBCM administration (48).

Photon-counting CT is an evolving technology which utilizes novel energy-resolving x-ray detectors which count the number of incoming photons and measure photon energy (49) as opposed to the more traditional averaging (called integrating) a collection of photons with CT. This new CT technique results in higher contrast-to noise, a superior spatial resolution, and enhanced spectral imaging. Photon-counting CT has promising possibilities for pediatric oncologic imaging, including the ability to diminish exposure to ionizing radiation, a higher reconstruction resolution, enhanced and simplified utilization of contrast media and new qualitative information for the radiologist.

Magnetic Resonance Imaging

Following its introduction in the 1980s, magnetic resonance imaging (MRI) was rapidly adopted for the evaluation of pediatric cancer. MRI is now the modality of choice for imaging brain, spinal, and musculoskeletal tumors. MRI is complementary to ultrasonography in the assessment of female genital tract and fetal tumors, and is a reasonable (and often preferred) alternative to CT for imaging abdominal and mediastinal solid tumors.

Magnetic resonance imaging generates images based on interactions between a strong external magnet field, radiofrequency waves, and the hydrogen nuclei in the body. The most common strength of external magnetic fields found on current MRI scanners are 1.5 T (Tesla) and 3 T, or about 30,000 or 60,000 times, respectively, the strength of the earth’s magnetic field. To achieve these magnetic field strengths, superconducting magnets are used. Greater field strength provides higher signal-to-noise ratio and thus better image quality but at the tradeoff of higher magnet cost and an increase in some image artifacts.

To generate a diagnostic image, contrast in appearance between normal tissues is needed to provide anatomic detail, and contrast in appearance between normal and diseased tissue is needed to detect pathology. The primary sources of tissue contrast in MRI are differences in proton densities, T1 relaxation times, and T2 relaxation times, and the weighting of these sources of tissue contrast is determined by the MRI pulse sequence design. On T1-weighted images, substances with short T1 relaxation times (e.g., fat, proteinaceous fluid, certain blood products, and gadolinium contrast) have bright signal intensity, and those with longer T1 relaxation times (e.g., edema, hemosiderin, many tumors) have intermediate to low signal intensity. On T2-weighted images, substances with short T2 relaxation times (e.g., tendon, hemosiderin, fibrosis) have low to intermediate signal intensity, and substances with longer T2 relaxation times (e.g., edema, many tumors) have bright signal intensity (50).

Pulse sequences are continuously being developed and refined for various applications, such as enhancing or suppressing signal from certain tissues, shortening scanning time, depicting blood flow and tissue perfusion, and assessing metabolic activity. Short tau inversion recovery (STIR) and fat-saturated sequences suppress the signal from fat which is very valuable in body and musculoskeletal applications. Diffusion-weighted MRI (DWI) is a technique that characterizes tissues based on Brownian molecular motion of water. The signal intensity of the resultant images depends on diffusion. Solid tumor components tend to show restricted diffusion (brighter signal) related to densely packed cells while liquefactive necrotic or cystic components tend to show increased diffusion (lower signal). Rapid imaging techniques, which include gradient echo sequences, fast (turbo) spin echo sequences, single shot sequences, echo planar imaging, non-Cartesian k-space sampling, compressed sensing, and parallel imaging using multiple receive coils, allow more coverage of the body or more sequences to be performed in less time for improved temporal resolution, or more detailed images to be obtained in the same time for improved spatial or contrast resolution (51). These are particularly valuable in children in whom small structures are commonly imaged and minimizing patient motion is imperative. Pediatric whole-body MRI scanning using rapid sequences is now feasible in 45 minutes or less of acquisition time and is well suited for screening for tumors in the setting of cancer predisposition syndromes and surveying the bone marrow for metastases or skeleton for osteonecrosis complicating therapy (8–11).

MR angiography (MRA) is a technique that noninvasively depicts blood vessels for indications such as detecting vascular occlusion or invasion by tumor or delineating vascular anatomy for surgical planning. Gadolinium-based contrast media (GBCM) produce contrast by shortening T1 relaxation time within the lumen of the vessel of interest, and the timing of injection of contrast and image acquisition allows arteries or veins to be selectively imaged. For settings in which administration of GBCM is contraindicated or not feasible, MRA can be obtained without administration of contrast using time-of-flight, phase contrast, or 3D steady-state free-precession (SSFP) techniques (52).

Analogous to iodinated contrast agents in CT, GBCM are used in MRI to make tumors more conspicuous, assess tumor vascularity, and evaluate intratumoral necrosis. Gadolinium is a paramagnetic metal that provides enhancement by shortening the T1 and T2 relaxation times and is administered as a linear or macrocyclic chelate to reduce toxicity and increase elimination, predominantly by the kidneys. An exception is the linear GBCM gadoexetate disodium, which is excreted by the kidneys and hepatobiliary system, a property that can be exploited to increase sensitivity for detection and characterization of liver lesions. Acute, allergic-like reactions to GBCMs are rare and usually mild in children (53). The administration of linear GBCMs is associated with the development of nephrogenic systemic fibrosis (NSF), primarily in adult patients with severe renal insufficiency, but this risk is mitigated with use of macrocyclic GBCM. More recently, concern has arisen over the unknown risk of gadolinium retention or deposition in the brain and other organs, particularly with repetitive use of linear GBCM in patients with renal insufficiency. Because of this potential risk, most pediatric centers have now transitioned to using macrocyclic GBCM. Given the potential risks of NSF and gadolinium retention in patients with renal insufficiency, the risk-benefit profile should be assessed and alternative imaging techniques that do not entail the use of GBCM should be considered, an approach analogous to the ALARA (as low as reasonably achievable) principle applied to imaging studies involving ionizing radiation exposure (54).

Viable tumor tends to show early and intense enhancement by GBCM, while necrotic and fibrotic areas show slower and less intense. Dynamic contrast-enhanced MRI (DEMRI) consists of rapid sequential MR imaging following the bolus administration of a GBCM to predict accessibility of the tumor to drugs before treatment or to assess the viability of residual tumor during or after treatment. DEMRI has been investigated to estimate the degree of tumor necrosis and guide risk-adapted therapy in bone sarcomas, but routine application has been limited by the need for mathematical post processing of the imaging data and the lack of a simple standardized approach across institutions with varying MRI equipment (55).

The iron oxide nanoparticle agent ferumoxytol approved by the FDA as an iron supplement is metabolized by the liver, lacks renal toxicity, and can be used off-label as an alternative MRI contrast agent to GBCAs, particularly in patients with renal insufficiency or for MR angiography applications. As a blood pool contrast agent, ferumoxytol provides long-lasting strong vascular enhancement, allowing a wider window of time for administration and the opportunity for repeated scans without compromising image quality. Although uncommon, severe allergic reactions can occur with ferumoxytol and it is contraindicated in patients with iron overload (56).

Magnetic resonance spectroscopy (MRS) is based on the principle that differences in the resonance frequencies of nuclei are determined by the chemical environment and permits noninvasive determination of the relative concentrations of various metabolites in vivo. Concentrations of metabolites such as choline (Cho), N-acetyl aspartate (NAA), and creatine (Cr) in pediatric brain tumors determined by MRS can be used to discriminate tumor types, improve biopsy yield, and predict outcome (57).

MRI offers several advantages over CT. MRI delineates normal soft tissues and discriminates abnormal from normal soft tissues better than CT and without the use of ionizing radiation. Disadvantages of MRI include its relatively higher cost, longer image acquisition time, limited ability to assess lung parenchyma and bony cortex, and limited availability and pediatric expertise at some locales or during certain periods, such as overnight or weekends. Compared to radiography, ultrasonography, and CT, the image quality of MRI is more susceptible to degradation by gross patient motion, respiratory motion, bowel peristalsis, and vascular pulsations. Young children more often require sedation or anesthesia to hold still for relatively long examination times for MRI than for other imaging modalities. Given the expanding applications and use of pediatric MRI and concerns regarding potential adverse neurodevelopmental effects of anesthetic agents in infants, it is imperative to implement strategies for streamlining pediatric MRI protocols and workflow to reduce the need for sedation or anesthesia (58, 59).

In addition to possible exposure to contrast agents, sedation or anesthesia, an MRI examination involves other potential risks. An MRI subjects a patient to static magnetic fields, gradient magnetic fields, radiofrequency electromagnetic fields, and acoustic noise. The interaction of ferromagnetic objects with the magnetic fields of MRI can result in induction of electrical current in the object, heating of the object, or movement of the object. Projectile injuries have occurred with oxygen canisters, scissors, and intravenous poles. For implanted devices that exhibit weak ferromagnetic qualities (e.g., certain intravascular stents, coils, filters, occluders), it has been recommended to wait 6–8 weeks after implantation before exposure to MRI to allow time for retentive tissue ingrowth. This waiting period may not be necessary for weakly ferromagnetic implanted devices that are rigidly fixed, such as bone screws (60).

Even if a device is not subject to significant motion from magnetic attraction, devices such as cardiac pacemakers, implantable cardiac defibrillators, cochlear implants, neurostimulators, implantable drug-infusion pumps, and programmable ventriculoperitoneal shunt valves may still be susceptible to heating or malfunctioning in the MR environment. As defined by the American Society for Testing and Materials, an MR unsafe item poses unacceptable risks to the patient, medical staff or other persons within the MRI environment, an MR conditional item is safe in the MRI environment within defined conditions of the magnetic and radiofrequency fields and other parameters that affect the safety of the item, and an MR safe item poses no known hazards resulting from exposure to any MR environment (61). Even MR safe implanted devices can compromise MRI exam quality by inducing image distortion. This is particularly pertinent for MRI exams performed for surveillance of local-regional recurrence of musculoskeletal neoplasms in patients with orthopedic implants (62).

The risks of MRI fields are incurred not only by the patient being scanned but also by accompanying family members, health care professionals, housekeeping and security personnel who may come in proximity to an MRI scanner. The 5 Gauss line of an MRI suite defines the border to an area in which the magnetic field could affect implanted devices such as pacemakers. Explicit guidelines for safe MRI practice, including the establishment of safety zones defined by distance from the MRI scanner, screening of patients and personnel, and special considerations for pediatric patients, are available from the American College of Radiology and other sources (63, 64).

Nuclear Medicine

Nuclear Medicine examinations are imaging examinations that focus on visualizing function by utilizing tracer amounts of radioactive material that are administered intravenously, inhaled or ingested. Once the radioactivity is instilled into the body, a camera placed close to the patient detects the emission of the small amount of radioactivity to create the images. The distribution of the radiopharmaceutical depends largely on the attached agent which has propensity for certain regions such as the kidneys for a renal scan or bones, for a bone scan. Imaging can occur over the half-life of the instilled radiopharmaceutical without an increase in absorbed dose to the patient for the radiopharmaceutical-based imaging. The radiation dose to the patient depends on the radiopharmaceutical but is generally below that of a typical CT scan for routine imaging, such as of the kidneys, or bones. Higher doses than a CT scan result when CT is hybridized with a PET scan.

This functional imaging allows for assessment of metabolic abnormalities that may be detected and correlated with anatomic findings. However, the value is that in oncology, nuclear medicine examinations may show an abnormality when normal anatomy is visualized or for response when anatomic imaging may show a change but does not confirm metabolic response.

Conway in the 2007 article “Quo Vadis Pediatric Nuclear Medicine” stated that only 10% of positron emission tomography/computed tomography (PET/CT) in pediatrics was for oncologic indications compared to 93% in adult practice at that time. He cited absorbed radiation dose, lower incidence of oncology pathology in children, as a significant economic impediment for a major financial investment of such new technology for the pediatric population. But he quoted: “Yet, one cannot deny the value of such a diagnostic and management capability for children with oncology disorders. The concern about absorbed radiation dose is of lesser importance when considering the risk versus benefits in children with oncology disorders.”(65). These discussions still hold true when discussing the use of scintigraphic examinations in children with cancer.

While historically bone scan and mIBG scintigraphy have been most common oncologic pediatric nuclear medicine studies, by far, the most common pediatric oncologic scintigraphic procedure is now PET scanning either PET/CT or PET/MR with 18F-fluorodeoxyglucose (18F-FDG). Lymphoma and sarcoma are the commonly imaged tumors however, all solid tumors can be evaluated for metabolic activity at time of staging, for response assessment and for follow-up assessment. Currently, there is not always validated semi-quantitative criteria available in the literature, but visual assessment correlated with the available localization technique, either CT or MR, can provide important metabolic information for patient management. It is the hope that increased standard utilization of imaging guidelines that are presented in these published monographs will enable collection of data to provide further scientific validation (66–74).

Whole-body thyroid imaging with radioiodine in children with differentiated thyroid cancer is also a common indication as is mIBG imaging for neuroblastoma (75–77). Ancillary scintigraphic studies either imaging or non- imaging are also utilized in pediatric oncologic patient management and include scintigraphic glomerular filtration rate (GFR) evaluation, gastrointestinal bleeding studies, CSF shunt studies, lung scans, myocardial perfusion and ventricular function scans, brain scans and bone densitometry. Almost any available scintigraphic study may be utilized in a child with oncologic diagnosis depending on their clinical symptomatology.

Theranostic techniques have long been utilized in treating pediatric thyroid cancer and are increasing in utility in front line treatments for children with neuroblastoma (78, 79). Use of other novel PET radiotracers for diagnosis and their companion treatment radiopharmaceuticals are being explored particularly in neuroendocrine tumors (NET). Use of radiotherapy for pain palliation with alpha-emitters is also beginning in the pediatric population. In general, if it is a scan performed in adults, with a lag time for implementation, there is spillover use in pediatrics. Due to differing regulatory bodies worldwide there has not been a uniform adoption of some of these techniques.

What has been adopted on a global basis is standardized guidelines for administration of weight-based radiopharmaceutical doses. Through a working collaboration between the Pediatric Imaging Council (PIC) of the Society of Nuclear Medicine and Molecular Imaging (SNMMI) and the Pediatric Committee of the European Association of Nuclear Medicine (EANM) guidelines for consensus administered amounts of radioactivity for nuclear medicine studies has been harmonized and most recently updated in 2016 (80–82). The next planned update is to begin in mid-2022 (80, 81, 83).

Further reduction of administered dose is possible when newer technologies such as PET/MRI are utilized. Schmall et al (84) determined that it was possible to reduce administered 18-FDG dose by 33% when performing PET/MRI studies compared to recommended weight-based dose for PET/CT degrading image quality thus reducing absorbed dose to the patient.

Imaging systems used for pediatric scintigraphic studies include gamma cameras with SPECT and SPECT/CT capabilities, PET/CT, and PET/MRI scanners, so called hybrid imaging systems. SPECT imaging is routinely used for most general scintigraphic studies and would be available in most centers that would image children. SPECT/CT systems are now widely available, but this does potentially introduce another potential radiation exposure from the CT examination. In general, the CT is usually obtained as a lower dose attenuation correction study but these machines can produce optimized CT studies and studies can be obtained with diagnostic quality images with resultant increased radiation exposure. The advantage is that these studies provide important anatomic localization information and can improve diagnostic certainty (85–87). These are studies that are acquired sequentially on the same gantry with the patient not moved between studies (86, 87). As with conventional CT studies in children, protocols should be pediatric based to ensure appropriate technique and dose modulation based on pediatric conversion should be utilized (88). These studies should also be tailored to the patient being imaged and the CT may not comprise the entire SPECT acquisition. In oncology the major utility would be in mIBG and thyroid imaging (88). While no specific diagnostic reference levels (DRLs) have been established for specific SPECT/CT studies, guidance could be obtained from recent publication citing DRLs for 10 most commonly performed pediatric CT studies in the United States (42). SPECT/CT would be routinely utilized in children with neuroblastoma for staging and response (75, 89–91).

PET imaging can be performed using PET/CT or PET/MR systems. Currently most studies are performed utilizing PET/CT systems. These scanners usually have multidetector CT systems that can adequately perform all diagnostic CT imaging as any current state of the art stand-alone CT system. The CT study combined with PET/CT can be performed for attenuation correction only as a non-contrast low dose study but some centers do perform combined optimized contrast enhanced CT for PET/CT studies. Standard guidelines for the pediatric diagnostic CT studies should be followed (92). The metabolic information obtained from the PET study can be used for staging and restaging, metastatic work-up, response assessment, identification of site for biopsy, for surveillance assessment, and for screening in cancer predisposition syndromes or suspected occult malignancy (85). Optimization of the CT portion according to standard pediatric protocols and attention to guidelines for weight-based radiopharmaceutical administered dose can ensure adherence to ALARA principle for these studies while providing important diagnostic information to the referring physician to help with patient management. One published study from Korea estimated the dose for pediatric 18F-FDG PET/CT to be 12.2 mSv but identified that newer hardware technology with image-enhancing algorithms can further help to reduce the absorbed dose from these studies in children (93). Mohammedi et al has suggested that any modelling used to assess dose from these scintigraphic studies should ensure a variety of ages and anatomies are used for accurate dose evaluation (94).

PET/MRI systems can further reduce the radiation exposure by utilizing MRI instead of CT for attenuation correction for the PET imaging. The MRI imaging will determine the length of the PET acquisition and because it is usually longer than optimal PET acquisition per area of body imaged. It has been accepted that the injected dose of radiopharmaceutical can be reduced as noted above (84, 95). There are currently a limited number of these systems in dedicated pediatric hospitals but there is early literature validating the PET imaging with either PET/CT or PET/MRI to be essentially interchangeable for clinical usage (68, 69, 72–74, 96, 97). One difference between adult and pediatric PET imaging is the recommendation to perform true whole-body imaging i.e., vertex to toes rather than “eyes to thighs” imaging, as pediatric cancer can often occur anywhere in the body including brain and distal extremities, more so than in adults.

With the addition of these hybrid imaging techniques there has been an increase in the need for sedation, either general anesthesia or conscious sedation. Most general nuclear medicine studies can be performed with standard distraction techniques and the help of Child Life Specialists, but when we add in longer SPECT, SPECT/CT, and PET studies with IV contrast administration and the need for adequate patient immobilization to optimize imaging, there is increased utilization of anesthesia (98, 99). There then must be a trade-off of the concern for potential neurologic sequelae and black-box warning for long sedations with the need for the diagnostic study (100, 101).

Conclusion

Medical imaging is an essential tool in the care of children with known or suspected cancer. The choice of a single modality or more than one modality in the pediatric oncology domain depends on the diagnostic benefit of the modality, operational considerations (such as portability), and the safety profile. While there are situations in which individual event choices and pathways may overlap between the various options, there are also situations in which a single modality may be sufficient by providing unique advantages. Arming with an understanding of the available imaging options affords an opportunity for more effective and efficient imaging management, as well as the ability for clarity in discussions with children and their caregivers. Moreover, while not specifically addressed in the previous material, continued engagement between the imaging and clinical teams regarding imaging appropriateness, protocol design (102), and recognition of and support for pediatric-focused imaging care will also serve to advance the safe and effective care of children with cancer.

Abbreviations

18F-FDG

18F-fluorodeoxyglucose

ALARA

as low as reasonably achievable

CHO

choline

CT

computed tomography

Cr

creatinine

DWI

diffusion-weighted imaging

DCEMRI

dynamic contrast-enhanced magnetic resonance imaging

GBCM

gadolinium-based contrast media

GFR

glomerular filtration rate

IBCM

iodine-based contrast media

LNT

linear-no-threshold

MRA

magnetic resonance angiography

MRI

magnetic resonance imaging

MRS

magnetic resonance spectroscopy

MAC

monitored anesthesia care

NAA

N-acetylaspartate

NSF

nephrogenic systemic fibrosis

NET

neuroendocrine tumors

NM

nuclear medicine

PET

positron emission tomography

STIR

short tau inversion recovery

SPECT

single-photon emission computed tomography

SSFP

steady-state free precession

US

ultrasonography