Contrast-enhanced ultrasound of the small organs in children

No specific patient preparation is required for CEUS imaging of the thyroid gland, lymph nodes and testes. Optimal transabdominal CEUS examination of the uterus and ovaries requires a well distended urinary bladder, which serves as an acoustic window for adequate visualization of these organs, similar to gray-scale US. Therefore, oral fluid intake is recommended prior to the examination. In an urgent setting, such as the evaluation of ovarian torsion, it might not be possible to wait for physiological bladder filling. In such cases, some centers place a catheter into the urinary bladder and fill it with normal saline prior to the US examination.

The recommended intravenous (IV) catheters for the ultrasound contrast agent (UCA) injection are peripherally inserted 18- to 22-gauge (G) IV catheters; however, for newborns 24- to 26-G catheters can be used as well.

The United States Food and Drug Administration (FDA) approved Lumason (Bracco Diagnostics Inc., Monroe Township, NJ) for pediatric liver and cardiac IV CEUS examinations. The approved dosing is 0.03 mL/kg, up to a maximum of 2.4 mL and 2 mL per injection for liver and cardiac applications, respectively. Lumason is marketed outside the United States as SonoVue (Bracco Imaging SpA, Milan, Italy). Other UCAs that have been used for IV CEUS applications in children and adults in the United States are Optison (GE Healthcare, Princeton, NJ) and Definity (Lantheus Medical Imaging, North Billerica, MA). Currently, use of all UCAs to examine small organs in adults and children are off-label indications, and therefore a range of doses have been reported in the literature.

Prior to FDA approval of Lumason, various age-based or weight-based dosing regimens were used for IV CEUS examinations of solid abdominal organs for characterizing indeterminate focal lesions and detecting traumatic injuries. Some of the most commonly reported dosing regimens are: 1.2–2.4 mL SonoVue for children ≥1 year old and 0.6 mL SonoVue for children <1 year old; 0.1 mL/kg SonoVue for children weighing up to 24 kg and fixed dose of 2.4 mL for those weighing more than 24 kg; or age-based dose according to the formula: volume of UCA (mL) = age (in years)/20 for organs other than liver [4‐6].

Determining optimal UCA dosing in small organs, particularly in children, requires special consideration. In our experience, a UCA dose of 0.03–0.10 mL/kg of SonoVue/Lumason is likely to be sufficient for pediatric small organs CEUS imaging, though larger doses can be used when higher-frequency transducers are used. Two case reports of IV CEUS in children, reported use of 0.6 mL of SonoVue for imaging of the ovaries in a 6-year-old girl and 2 mL of SonoVue for imaging of the testis in an adolescent boy [2, 3].

Bolus injection doses in adults are also higher when evaluating these small organs in comparison to those used for the adult liver. When examining the adult thyroid, CEUS doses of 1.0–4.8 mL per bolus injection of SonoVue have been used [7‐13]. Lymph node CEUS has been performed with bolus injections ranging from 1.5 to 4.8 mL [14‐16]. For adult testicular CEUS examination, a bolus injection of up to 4.8 mL of SonoVue/Lumason is described [17‐19]. Published data for CEUS examination of the ovaries and uterus in women using transvaginal transducers reported bolus injections from 2.4 mL to 4.8 mL [20, 21].

For small superficial organs (thyroid, lymph nodes, testis) high-resolution, high-frequency (i.e. >10 MHz) linear transducers should be used. The ovaries and uterus, especially in children and adolescents, can be scanned with convex transducers with frequencies appropriate for the child’s body habitus (frequency range 1–9 MHz). In this age group, intracavitary transducers do not play a significant role.

By palpation, the incidence of thyroid nodules in the adult population is approximately 4–7% [25]. However, that number increases up to 30% with the use of high-resolution US of the thyroid gland in clinical practice [11, 25]. The prevalence of thyroid nodules in children ranges between 1% to 2% and is significantly lower compared to adults [27]. Most nodules in adults are benign, with fewer than 5% classified as malignant [28]. However, in children thyroid nodules carry a far greater risk of harboring malignancy than in adults, at a rate of approximately 26.4% [27].

In the CEUS literature, thyroid nodules can be characterized and differentiated by qualitative and quantitative CEUS imaging parameters. Qualitative parameters describe nodule perfusion compared to the surrounding thyroid tissue and include observations of enhancement such as intensity, homogeneity, vascular distribution pattern and rate of contrast uptake [24, 25, 29]. The main CEUS features that are highly suspicious for malignancy are hypoenhancement of the nodule throughout the enhancement phases and heterogeneous enhancement [9, 30]. In one adult study, six qualitative CEUS features were considered significant for predicting malignant thyroid nodules: slow entry of contrast agents during the enhancement phase (odds ratio [OR]=15.6, P=0 .001), slow time-to-peak (OR=7.4, P=0.002), non-uniform enhancement (OR=10.1, P=0.023), irregular enhancement pattern (OR=36.2, P=0.002), unclear enhancement boundary (OR=25.3, P=0.012) and absence of ring-like enhancement (OR=25.3, P=0.004) [25]. On the contrary, uniform enhancement and ring enhancement are typical findings for predicting benignity [31]. However, the CEUS enhancement patterns of both malignant and benign nodules can overlap [29]. Because of the subjectivity in this assessment, a quantitative approach has been the subject of much research. Quantitative CEUS with time-intensity curve analysis can be performed following image acquisition using special software programs embedded on the US unit or on separate computers. This analysis requires that regions of interest be placed in the nodule and the adjacent normal tissue. Imaging parameters obtained from time-intensity curves including wash-in-slope, peak intensity, time-to-peak, area under the curve (AUC), mean transit time, and washout time have great potential to provide objective information for differentiating benign and malignant thyroid nodules [9]. In quantitative CEUS evaluations, later arrival time of contrast, later time-to-peak with earlier start of the washout phase, lower peak intensity and overall smaller AUC were strongly associated with malignancy [9]. Some adult studies, however, showed differences in CEUS quantitative parameters between small (<1 cm) and large (>1 cm) malignant thyroid nodules [9, 30]. Specifically, larger malignant nodules had a steeper slope of the wash-in curve, faster time-to-peak and slightly higher peak intensity compared with smaller malignant nodules. This could be because in smaller malignant nodules neoangiogenesis is incomplete, whereas in larger malignant nodules aberrant neo vessels result in more blood flow. Larger malignant nodules tend to have areas of necrosis, calcifications and fibrosis that diminish blood flow compared to benign nodules [9]. Still, as further validation of these CEUS parameters for thyroid nodules continues, they should be interpreted only with the use of clinical and laboratory findings, as well as with elastography.

At the moment, no data exist about the use and importance of CEUS in thyroid applications in children. The majority of adult criteria can likely be applied to children when selecting nodules for fine-needle aspiration, with the following provisions in pediatric patients: gray-scale US pattern and clinical context including history should be considered more than size, intrathyroidal thymic tissue might mimic a suspicious nodule, and diffuse sclerosing variant of papillary thyroid cancer presents with non-nodular diffuse infiltration associated with microcalcifications (“snowstorm”) [32]. More experience with pediatric thyroid CEUS is needed to understand its role in this population.

Hashimoto thyroiditis is an increasingly prevalent autoimmune disease. Recently, it has been shown that Hashimoto thyroiditis is associated with a greater risk of thyroid cancer [33, 34]. Diagnosis of Hashimoto thyroiditis often relies on clinical presentation and laboratory findings, as well as results from gray-scale and color Doppler US. However, at times fine-needle aspiration is indicated, especially when thyroiditis co-exists with nodules [35]. Using CEUS to assess thyroid nodules might help to decrease the number of fine-needle aspirations in these patients [24, 26]. In one study, malignant papillary thyroid carcinomas in patients with Hashimoto thyroiditis tended to exhibit hypoenhancement compared to adjacent parenchyma [24]. Using time-intensity curve analysis, the authors generated ratios of parameters from regions of interest within the nodule compared to the thyroid parenchyma and showed that the peak intensity ratio (peak intensity nodule/peak intensity parenchyma) and AUC ratio (AUC nodule/AUC parenchyma) tended to be <1 in malignant compared to benign lesions [24]. In the future, CEUS evaluation of tissue microvascularity might allow quantification of Hashimoto thyroiditis activity and differentiation from other diffuse thyroid disorders such as Graves disease.

When evaluating pediatric patients with acute-onset testicular pain, testicular torsion should always be considered, especially in adolescents [49]. A reported series for the assessment of acute testicular torsion did not demonstrate a clear advantage of CEUS over conventional color Doppler US [50]. However, CEUS might be helpful for problem-solving when it is clinically suspected that torsion has been missed in young children with small testes. In this case, CEUS might be used to confirm global testicular infarction by showing the absence of parenchymal enhancement of the entire affected testis (Fig. 5; Online Supplementary Material 1). In addition, CEUS might have a role in evaluating segmental infarction, which can be barely detectable on conventional gray-scale and color Doppler US in the early phase and has variable appearance as it evolves, making it a challenging diagnosis to differentiate from a tumor [51]. A segmental infarct appears as a focal, usually wedge-shape non-enhancing abnormality on CEUS (Fig. 6) [3, 52].

In the setting of inflammation, CEUS can be valuable for evaluating more severe cases of epididymo-orchitis, especially when assessing for abscess or venous infarction. On CEUS, an abscess might exhibit a rim of increased peripheral enhancement with an absence of internal enhancement. Similar to other modalities, the lesion does not follow lobar distribution, which might help to differentiate this from a segmental infarction [46]. Extra-testicular epididymal abscesses demonstrate similar appearances [53].

The task of differentiating between benign and malignant vascular abnormalities remains a challenge for all imaging techniques. CEUS offers the potential to improve diagnostic confidence when characterizing blood flow to intratesticular focal lesions, especially when the lesion is small [46, 54]. The lack of internal enhancement on CEUS is common for benign lesions such as a simple or an epidermoid cyst. This finding can increase examiner confidence that a lesion is benign (Fig. 7) [46]. If CEUS shows the presence of enhancement within a focal testicular lesion, this should be interpreted as concern for malignancy and requires further clinical and imaging interrogation [46, 48]. Several investigations attempted applying quantitative CEUS analysis of time-intensity curves to differentiate between malignant and benign intratesticular lesions [17‐19]. In conjunction with gray-scale US findings, preliminary studies describing parameters such as wash-in rate, peak enhancement, time-to-peak and washout time are promising to distinguish benign from malignant lesions; however, others have reported similar time-intensity curves between these groups and therefore further studies are required in this field (Fig. 8).

Notwithstanding the current scarcity of published research on testicular applications, CEUS adds to the established sonographic evaluation of testes with its ability to provide a definitive assessment of vascularization. Thus, CEUS could increase examiner confidence and improve therapeutic strategies and timely management for the testicular pathology common in the children and adolescents.

Ovarian torsion can occur at any age. This is when the ovary twists on its vascular and ligamentous supports, resulting in compromised blood supply. The presence of large cysts or other ovarian lesions, and large ovarian size predispose the ovary to twist; at least one of these factors is present in almost 75% of torsion cases [55]. In children, a possible predisposing factor is when the child has relatively long fallopian tubes and suspensory ligaments with resulting increased mobility of the ovary [55]. The most common US finding in ovarian torsion in pediatric patients is enlargement of the twisted ovary [56]. Other findings include peripheralization of ovarian follicles, a mass lesion within the twisted ovary, and twisting of the vascular pedicle or “whirlpool sign.” While using combinations of sonographic signs, gray-scale US has been shown to have specificity up to 87.5% for diagnosing ovarian torsion, whereas color Doppler US has been shown to have relatively limited sensitivity and specificity in confirming the diagnosis [57‐59].

Using CEUS to evaluate ovarian torsion is promising. However, the normal enhancement characteristics of the ovary are not well described. In adults, CEUS has been shown to identify the “whirlpool sign” of the twisted vascular pedicle [60]. Absent or decreased perfusion is also seen at CEUS, with areas of necrosis and hemorrhagic infarction clearly demarcated [60]. A single pediatric case described the CEUS appearance of ovarian torsion and follow-up in an 6-year-old girl after laparoscopic untwisting and hyperbaric oxygen therapy [2]. This report included qualitatively decreased ovarian perfusion immediately following de-torsion, followed by hyperperfusion and a subsequent return to normal enhancement. Further experience might be needed before widespread clinical application.

Contrast-enhanced US was shown to slightly improve the detection of ovarian cancer in adults, typically using a transvaginal approach. A meta-analysis published in 2016 reported sensitivity (0.97) and specificity (0.92) of CEUS for ovarian malignancy as higher than either gray-scale US (0.92 and 0.86) or Doppler US (0.93 and 0.85) assessment, respectively [61]. The heterogeneous enhancement of solid tumors is more visible at CEUS than Doppler US [62]. Quantitative analysis using time-intensity curve and the vascular enhancement describe malignant ovarian lesions as having greater peak enhancement intensity, delayed washout, and overall more enhancement than benign lesions (Fig. 9) [62]. Benign lesions such as hemorrhagic cysts and endometriomas do not enhance because these lesions have no internal perfusion and therefore can be distinguished with CEUS (Fig. 10) [63]. However, among these studies, some authors reported overlap of patterns between benign and malignant lesions such that a general threshold to differentiate one from the other cannot be established. Because of this, there are no current clinical recommendations for gynecological indications [1]. Extrapolating this evidence and applying it to the pediatric population is challenging because ovarian cancer comprises only 1–2% of all childhood malignancies, and ovarian tumors in adults differ from those in children, with the most common tumors in children being of germ cell origin [64]. No studies have evaluated CEUS in pediatric ovarian tumors.

When using CEUS to image the uterus, typically the uterine artery enhances first, followed by the outer myometrium, inner myometrium and endometrium [62]. The endometrial peak enhancement is less than that of the myometrium on CEUS and washout time is longer than what occurs in the myometrium.

In the setting of endometrial hyperplasia, the endometrium enhances similar to the normal endometrium. On the other hand, endometrial polyps have more rapid arterial enhancement and delayed washout when compared to the endometrium and endometrial hyperplasia [62]. Adult studies have evaluated the use of CEUS in endometrial carcinoma; however, this disease does not occur in the pediatric population [65].

Although uncommon in the pediatric setting, abnormal uterine bleeding can be caused by fibroids, adenomyosis or rare malignancies like rhabdomyosarcoma (Fig. 11). At US, fibroids are typically hypoechoic compared to the myometrium and have a pseudocapsule as well as possible areas of calcification. Unlike fibroids, adenomyosis usually has ill-defined margins [66]. However, distinguishing between the two at US can be challenging and CEUS might help to make a more accurate distinction. Fibroids tend to have peripheral capsular enhancement with gradual centripetal filling and diffuse enhancement in later phases. A feeding capsular vessel can often be seen. Necrotic fibroids show no central enhancement [67]. Conversely, adenomyosis has more rapid, diffuse, heterogeneous enhancement, with a “moth-eaten” appearance. No studies have specifically evaluated either of these two entities in the adolescent population.