Advertisement

Pediatric Radiology

, Volume 48, Issue 9, pp 1291–1306 | Cite as

Magnetic resonance imaging of pediatric adnexal masses and mimics

  • Christopher Z. Lam
  • Govind B. Chavhan
Pediatric Body MRI
  • 273 Downloads

Abstract

Evaluation of adnexal masses in children and adolescents relies on imaging for appropriate diagnosis and management. Pelvic MRI is indicated and adds value for all adnexal masses when surgery is considered or when ultrasound findings are indeterminate. Specifically, features on MR imaging can help distinguish between benign and malignant lesions, which not only influences the decision between surgery and conservative treatment, but also the type of surgery to be performed, including potential use of fertility-sparing approaches with minimally invasive techniques. Larger size, younger age, presentation with precocious puberty or virilization, restricted diffusion in a solid mass, and rapid and strong enhancement of solid components are all features concerning for malignancy. In addition, distinctive MR imaging features of adnexal masses, combined with clinical and laboratory biomarkers, might suggest a specific histological diagnosis.

Keywords

Adnexal mass Adolescents Children Magnetic resonance imaging Malignancy Ovaries 

Introduction

Accurate characterization of adnexal masses in children is crucial for appropriate management [1, 2, 3, 4]. An ovarian malignancy is present in 3–8% of children and adolescents who present with an adnexal mass [3, 5] and in 10–20% of all adnexal masses treated surgically [3, 5, 6]. Most non-neoplastic or benign adnexal masses can be managed conservatively [5]. When surgery is indicated for benign lesions, ovarian preservation with a minimally invasive technique is recommended [3, 4, 5]. Malignant lesions typically require oophorectomy and a potentially different surgical approach [4]. Because the optimal approach for benign and malignant adnexal lesions differs substantially, preoperative diagnosis and risk stratification is a critical component of gynecological imaging.

Ovarian neoplasms that occur in children are the same as in adults but differ in frequency. They can be classified into four types: germ cell tumors, 60–80%; surface epithelial-stromal tumors, 15–20%; sex cord–stromal tumors, 10–20%; and miscellaneous tumors (gonadoblastoma, lymphoma/leukemia, small cell carcinoma, soft-tissue tumors), <5% [6]. In contrast, epithelial tumors are the most common ovarian neoplasm in adults. Other adnexal masses include hemorrhagic ovarian cyst, massive ovarian edema, and tubo-ovarian lesions such as tubo-ovarian torsion, tubo-ovarian abscess (TOA), hydrosalpinx and paraovarian cysts. In addition, uterine, bowel and appendiceal masses can mimic an adnexal mass on presentation.

This article reviews the magnetic resonance imaging (MRI) approach and value added in the diagnostic workup of pediatric adnexal masses, with a focus on MR imaging differential diagnoses and mimics.

Value of MRI in evaluation of adnexal masses

Ultrasound is the initial modality of choice for evaluating an adnexal mass in children. However, differentiating benign from malignant lesions by ultrasound alone can be challenging in certain cases. Advantages of MRI compared to ultrasound include MRI’s utility for global abdominopelvic assessment to better localize and stage adnexal masses, along with superior soft-tissue contrast, which can aid tumor characterization, risk stratification and diagnosis. Limiting exposure to ionizing radiation is also especially important in young female patients, which is why MRI is preferred over CT in the pediatric population.

In cases indeterminate by ultrasound, pelvic MRI can be used as a second-line problem-solving modality that can provide additional diagnostic information with potential to change management decisions. A recent study from our group showed that the addition of MRI significantly altered management in 31% of children with indeterminate ultrasound findings who were subsequently referred for pelvic MRI [4]. This was on the basis of better assessment of likely benign versus malignant nature of the mass, along with better assessment of the origin of the mass. Studies in adults have shown similar findings, with pelvic MRI providing major diagnostic information in 29% and reduction of surgery in 13% of indeterminate cases [7], featuring a diagnostic accuracy of 91–93% for malignancy [8, 9], and adding particular value with a specificity of 98% that allows for confident diagnosis of many benign adnexal lesions [10, 11]. Hence we recommend pelvic MRI for any adnexal mass considered for surgery to decide both whether surgery versus conservative management is indicated, and when surgery is required, to decide between ovarian-sparing surgery and oophorectomy [4].

MRI protocol and techniques

Our pelvic MRI protocol is the same on the 1.5-T and the 3-T scanner and includes administration of a gadolinium-based contrast agent. The MRI sequences and purpose for each are summarized in Table 1. Typically we start with a fast coronal T2-weighted sequence of the entire abdominopelvic cavity for a global anatomical overview and to screen the upper abdomen. We obtain standard T1-weighted and T2-weighted fat-saturated images for pelvic anatomy, tissue characterization and detection of pathology. In-phase and out-of-phase imaging identify microscopic fat and sebaceous lipid. A pre-contrast T1-weighted fat-saturated sequence differentiates macroscopic fat from hemorrhage and aids in evaluating for enhancement on post-contrast T1-weighted fat-saturated images. Post-contrast T1-weighted fat-saturated imaging detects solid enhancing components. In adult studies, the addition of gadolinium has been shown to significantly improve sensitivity and positive likelihood ratio for diagnosis of ovarian neoplasms, with modest improvement in specificity [11]. Some authors suggest that qualitative assessment of strong early enhancement within 60 s of injection is more likely to represent a malignant ovarian mass versus benign [9]. Gadolinium can also be useful to define non-neoplastic adnexal masses, including tubo-ovarian torsion, inflammatory gynecological disease and other mimics. Optional sequences include T2-weighted fat-saturated periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER)-variant sequences, which are useful for compensating for motion artifacts, including those from bowel peristalsis or the urinary bladder.
Table 1

Sample MRI protocol for pediatric adnexal mass evaluation

Sequence

Purpose

Pre-contrast

 Coronal T2-W single-shot breath-hold

Global overview including upper abdomen

 Axial T1-W TSE

Pelvic anatomy / tissue characterization

 Axial DWI (b=0, 100, 600–800)

Risk stratification / tissue characterization / lymph nodes

 Axial T2-W fat-saturated TSE

Fluid or other pathology / tissue characterization

 Sagittal T2-W fat-saturated TSE

Fluid or other pathology / tissue characterization

 Axial T1-W in-phase/out-of-phase

Microscopic fat

 Coronal 3-D T1-W fat-saturated GRE breath-hold (VIBE/THRIVE)

Differentiate fat from hemorrhage / comparison for images acquired after contrast administration

Post-contrast

 Coronal 3-D T1-W fat-saturated GRE breath-hold dynamic (4 runs)

Qualitative tissue perfusion dynamics

 Sagittal T1-W fat-saturated TSE

Tissue characterization

 Axial T1-W fat-saturated TSE

Tissue characterization

Optional

 Coronal STIR

In small children for better SNR than T2 single shot

 Axial T2-W fat-saturated PROPELLER (BLADE/MultiVane)

Compensate for motion artifacts (including from bowel peristalsis or urinary bladder)

3-D three-dimensional, BLADE Siemens’ proprietary variant of PROPELLER, DWI diffusion-weighted imaging, GRE gradient echo, MultiVane Philips’ proprietary variant of PROPELLER, PROPELLER periodically rotated overlapping parallel lines with enhanced reconstruction, SNR signal-to-noise ratio, STIR short tau inversion recovery, THRIVE T1-weighted high-resolution isotropic volume examination, TSE turbo spin echo, VIBE volumetric interpolated breath-hold examination

Emerging advanced MRI sequences that could add value in evaluation of adnexal masses include diffusion-weighted imaging (DWI) and semi-quantitative perfusion-weighted imaging (PWI). In terms of DWI, while some adult studies have not found a benefit to DWI [12, 13, 14], others report how DWI helps differentiate malignant from benign lesions (especially when combined with conventional imaging) [15, 16, 17, 18, 19, 20] and can be useful in monitoring peritoneal carcinomatosis [21]. A pitfall with DWI is that several benign lesions notoriously restrict diffusion — including endometriomas, keratinized substance/sebum in teratomas, and fibrothecomas — but these entities can be accurately diagnosed on conventional sequences [22]. Overall, the evolving evidence from the adult literature suggests that DWI has a complementary role in characterizing specific solid ovarian tumors, with perhaps little benefit in purely cystic types. The clearest scenario of added value in adults is in evaluation of the solid adnexal mass that is hypointense on T2-weighted images, whereby if the mass is also hypointense on high b value DWI it can be confidently considered benign [18]. These masses likely represent fibromas or Brenner tumors, which are rare in children. No studies have evaluated use of DWI specifically for adnexal masses in children and adolescents, and thus future investigation is required to determine the value of DWI in this population. Nonetheless, given the available evidence in adults we recommend inclusion of DWI sequences for adnexal mass evaluation.

PWI is a promising technique that uses semi-quantitative multiphase contrast-enhanced MRI to generate time-intensity enhancement curves of solid adnexal masses [22]. Using external myometrium as a reference, a rapid rate and high level of enhancement is associated with a very high likelihood of malignancy, whereas a slow rate and low level of enhancement confers a high likelihood of benignity [15, 23]. PWI has not been studied in the pediatric population. We do not have any experience with the technique at our institution.

General MRI approach to adnexal masses

The first step in MRI evaluation of an adnexal mass is to identify the organ of origin of the abnormality. Adnexal masses are typically of ovarian origin, but other causes of an adnexal mass include fallopian, uterine and bowel pathologies. If an ovarian origin is identified or suspected, a neoplastic process needs to be distinguished from other entities such as tubo-ovarian torsion or infectious/inflammatory disease, which have characteristic imaging appearances and usually present with acute symptomatology. If a neoplastic process is suspected, worrisome signs of malignancy include size >8 cm [1, 2], age 1–8 years [2], presenting with an abdominal mass or precocious puberty/virilization (as opposed to pain or torsion-like symptoms, which are more common in benign lesions) [2], identification of a solid mass with high signal intensity on high b value DWI (where MRI is not otherwise diagnostic of a benign entity on conventional T1- and T2-weighted sequences) [15], rapid and strong enhancement of solid components after administration of contrast agent [15], or any characteristic MRI features of specific malignant entities, as described in Table 2. Elevated levels of various hormones can serve as biomarkers for ovarian tumors, as summarized in Table 3. Correlation with these biomarkers can help narrow the differential diagnosis, with elevated beta-hCG (human chorionic gonadotropin) and alpha-fetoprotein nearly diagnostic of malignant germ cell tumors [24].
Table 2

Common MRI features of pediatric adnexal masses

Adnexal mass

MRI features

Ovarian germ cell tumors

 Mature cystic teratoma

Unilocular fat-containing cystic ovarian mass

 Immature teratoma

Scattered punctate foci of fat with microcystic areas of fluid signal intensity

 Dysgerminoma

Lobulated solid mass with fibrovascular septa

 Yolk sac tumor

Heterogeneous mixed solid–cystic mass with large dilated intratumoral vessels

 Mixed germ cell tumor

Large solid mass with extensive hemorrhage and necrosis with combination of other germ cell tumor features

Ovarian surface epithelial-stromal tumors

 Benign serous cystadenoma

Large unilocular simple cyst, indistinguishable from follicular cyst when small

 Benign mucinous cystadenoma

Large multilocular cystic mass with variable signal intensity of cysts in stained-glass pattern

 Borderline or malignant epithelial tumor

Multilocular cystic mass with papillary projections or solid enhancing mural components

Ovarian sex cord–stromal tumors

 Juvenile granulosa cell tumor

Mixed solid–cystic mass with sponge-like appearance on T2-weighted images from intermediate-signal solid tissue intermixed with hemorrhagic cysts

 Thecoma-fibroma

Solid masses that are dark on T2-weighted images with low signal on high b value diffusion-weighted imaging and weak enhancement post-gadolinium

 Sertoli-Leydig cell tumor

Solid mass with peripheral cysts; can be predominantly hypointense on T2-weighted imaging

Ovarian non-neoplastic processes

 Hemorrhagic cyst

Ovarian cyst with variable signal intensity on T1- and T2-weighted images; can have heterogeneous contents

 Polycystic ovarian syndrome

Bilateral moderately enlarged ovaries with multiple small peripheral follicles

 Massive ovarian edema

Unilateral enlarged edematous ovary with peripheral follicles and preserved vascularity

Tubo-ovarian masses

 Paraovarian cyst

Unilocular simple cyst separate from the ovary near the round ligament

 Tubo-ovarian torsion

Unilateral enlarged edematous ovary with peripheral follicles and diminished enhancement

 Isolated tubal torsion

Twisted fallopian tube with normal ipsilateral ovary

 Tubo-ovarian abscess

Complex cystic mass with thick enhancing rim and surrounding adnexal inflammatory changes

Table 3

Serologic biomarkers for ovarian tumors

Biomarker

Associated tumors

Beta-human chorionic gonadotropin (βhCG)

Choriocarcinoma

Embryonal carcinoma

Alpha-fetoprotein (αFP)

Immature (malignant) teratoma

Yolk sac tumor

Embryonal carcinoma

Cancer antigen 125 (CA125)

Borderline and malignant ovarian epithelial neoplasms

Lactate dehydrogenase (LDH)

Dysgerminoma

Inhibin

Juvenile granulosa cell tumor

Ovarian germ cell tumors

Germ cell tumors are the most common ovarian neoplasm in children and adolescents, accounting for 60–80% of pediatric ovarian neoplasms [6]. Germ cell tumor subtypes include teratomas (mature, immature and monodermal), dysgerminoma, yolk sac tumor (endodermal sinus tumor) and mixed germ cell tumors (including embryonal carcinoma) [25]. Unlike in adults where the majority of germ cell tumors are benign, 10–30% of germ cell tumors in the pediatric population are malignant [6, 26, 27].

Teratoma

Ovarian teratomas are tumors derived from more than one of the three primitive embryonic layers (ectoderm, mesoderm and endoderm). When predominantly composed of a specific type of tissue (e.g., thyroid tissue in struma ovarii, neuroectodermal tissue in carcinoid tumor, and rare neural tumors), they are classified as monodermal or specialized teratomas. Ovarian teratomas are associated with torsion (16%), rupture (1–4%), malignant transformation (1–2%), infection (1%), paraneoplastic autoimmune hemolytic anemia (<1%) and immune-mediated limbic encephalitis (<1%) [28].

Mature teratoma

Mature cystic teratomas (often called “dermoid cysts”) are benign and represent the most common ovarian neoplasm in children and adolescents, constituting 84–87% of all germ cell tumors in some pathological series [26, 29], and thus approximately 50% of all pediatric ovarian neoplasms. Pathologically they are composed of well-differentiated tissues from at least two of the three primitive germ cell layers, although ectodermal tissues (skin derivatives and neural tissue) are always present [30]. They are unilocular cysts in 88% of cases [30], which are characteristically filled with sebaceous liquid 94–96% of the time [31]. There is often a focal protuberance that projects into the cyst, known as a Rokitansky nodule, which might contain hair, teeth, bone, large coarse calcifications or cartilage. Solid mature teratomas are rare and are more likely to be at least partially immature [25]. Adipose tissue is present in 67–75% of cases [32].

Mature teratomas characteristically appear as a sebaceous/fat-containing cystic adnexal mass by MRI. Both the macroscopic intratumoral fat in adipose tissue and lipid sebaceous intracystic component are hyperintense on T1-weighted imaging, and intermediate- to hyperintense signal on T2-weighted imaging. Suppression of signal on frequency-selective fat-saturated T1-weighted images is diagnostic, while signal drop on opposed-phased imaging can help to identify microscopic lipid in lesions with a low amount of macroscopic fat. These sequences differentiate fatty components from hemorrhage, which can otherwise sometimes look similar. A reversed chemical shift artifact with a hypointense inner margin and hyperintense outer margin in the frequency encode direction at the lesion boundary has been described related to the interface between intralesional fat and teratoma wall [33]. A fat–fluid level is occasionally appreciated on MR imaging and ultrasound (Fig. 1). Rarely a floating globule composed of fat, debris, desquamated material and hair is seen in the non-dependent portions of the cyst [34]. Deeply hypointense areas reflecting large coarse calcifications or teeth can be identified in the Rokitansky nodule or along the lesion walls. Mature teratomas restrict diffusion on DWI, which has been attributed to the keratinoid ectodermal components that are invariably present [35]. Keratinoid substances show low signal on T1-weighted images and high signal on T2-weighted images, and thus otherwise resemble serous fluid [35]. Solid components in the Rokitansky nodule can enhance and do not necessarily indicate malignant transformation. So while enhancement and an obtuse margin of the Rokitansky nodule with the teratoma wall have been reported to suggest malignant transformation in the adult literature [28], malignant transformation is rare in the pediatric population [36]. However it is not unusual to find a small microscopic focus of immature/malignant tissue on histology in otherwise completely cystic mature teratoma on imaging.
Fig. 1

Mature cystic teratoma (dermoid cyst) in a 16-year-old girl with sickle cell disease. a Transverse ultrasound image shows an echogenic mass with fluid–fluid level (arrow) in the left ovary. b, c Axial T1-W (b) and T2-W fat-saturated (c) MR images demonstrate the fat–fluid level (arrow) with fat/sebaceous material on the non-dependant portion that contains an irregular Rokitansky nodule (arrowhead)

Immature (malignant) teratoma

Immature teratomas are classified as a malignant ovarian germ cell tumor. Like mature teratomas, immature teratomas are also derived from more than one germ cell layer, but these contain both primitive embryonal and well-differentiated tissues [25]. The amount of immature tissue determines the histological grade, which is typically of neuroectodermal origin [37]. The younger the patient, the more likely the teratoma will be of the immature type [28]. Immature teratomas grow rapidly, spread through the peritoneal cavity, and metastasize via lymphatics [25]. At presentation, they are usually larger than mature cystic teratomas and more frequently rupture [32]. In contrast to mature teratomas, they are predominantly solid with only scattered foci of macroscopic fat, small and irregularly shaped calcifications throughout rather than large coarse calcifications along walls, have hemorrhage, and contain several microcystic structures as opposed to a single large cyst [37]. Microcystic components are typically lined with respiratory epithelium and thus are filled with simple serous fluid rather than sebaceous material as in mature teratomas [37]. Therefore, immature teratomas appear on MRI as complex heterogeneous masses with cystic areas that exhibit simple fluid signal along with scattered foci of macroscopic fat, which are confirmed with fat suppression techniques in a similar fashion as mature teratomas. The amount of solid tissue by imaging does not correlate with histological grade [37].

Dysgerminoma

Dysgerminomas are composed of uniform cells that resemble primordial germ cells (the ovarian equivalent of testicular seminoma), and are the most common malignant germ cell tumor in pediatrics [38]. Most occur in the 2nd and 3rd decades of life [6]. They are bilateral in 10–15% and can metastasize to retroperitoneal lymph nodes [6]. On imaging, dysgerminomas are almost always purely solid and appear as lobulated, heterogeneous masses that are hypointense to muscle on T1-weighted images and isointense to slightly hyperintense on T2-weighted images, with possible areas of necrosis, hemorrhage or speckled calcifications [39, 40]. Characteristically they have prominent fibrovascular septa coursing through the mass [39, 40]. Because of the fibrous content, the fibrovascular septa on MRI are hypointense on T2-weighted images and enhance intensely with administration of contrast agent [39, 40] (Fig. 2). Appearance of a lobulated ovarian mass with fibrovascular septa might be specific enough to suggest a preoperative imaging diagnosis [41].
Fig. 2

Dysgerminoma of the left ovary in a 15-year-old girl. a Transverse Doppler ultrasound image demonstrates a solid mass with vascularity (arrow). b–d Axial T2-W (b), T1-W (c) and post-contrast T1-W fat-saturated (d) MR images demonstrate a lobulated T2-hyperintense solid heterogeneously enhancing mass (arrows) anterior to the uterus. A normal right ovary is seen (arrowheads)

Yolk sac tumor

Yolk sac tumor (formerly known as endodermal sinus tumor) is a rare and highly malignant germ cell tumor with a cellular structure that is similar to the primitive yolk sac (vitelline elements) [42]. It is the second most common malignant germ cell tumor of the ovary in pediatrics, usually occurring in the second decade of life [43, 44]. Yolk sac tumors are typically large masses on presentation and can have peritoneal, hematogenous or lymphatic metastases [45]. By MRI, they are typically heterogeneous, mixed solid and cystic, with hemorrhage, necrosis and striking enhancement that is greater than myometrium [45, 46, 47, 48] (Fig. 3). The intratumoral hemorrhage and marked enhancement reflect the rich vascularity of the tumor [45, 46, 47]. In this regard, a particularly characteristic sign for yolk sac tumors is visualization of large dilated intratumoral vessels or vascular aneurysms, which can be seen by MRI as signal voids on spin-echo imaging, or directly following contrast agent administration [45, 46, 47]. Another commonly reported feature of yolk sac tumors is linear tearing of the capsule causing rupture [49], which is present in 13–27% of cases at surgery [50]. However capsular tearing and rupture are not specific because they can be seen in other ovarian neoplasms, including the benign mature cystic teratoma [42].
Fig. 3

Yolk sac tumor in a 14-year-old girl. a Sagittal Doppler ultrasound image demonstrates a complex solid–cystic mass with vascularity (arrow). b, c Sagittal T2-W (b) and post-contrast T1-W fat-saturated (c) MR images demonstrate a large complex solid–cystic heterogeneously enhancing mass occupying the pelvis and lower abdomen (arrows)

Mixed germ cell tumor and other malignant germ cell tumors

Mixed germ cell tumors are very aggressive tumors composed of more than one malignant germ cell component, usually a combination of dysgerminoma, immature teratoma and yolk sac tumor. Other malignant germ cell tumors include embryonal carcinoma, polyembryoma and choriocarcinoma, which can occur alone but are usually seen as part of a mixed germ cell tumor [42]. Imaging characteristics of mixed germ cell tumors reflect the underlying histological components as already described (Fig. 4). All typically present as large, predominantly solid masses with extensive hemorrhage and necrosis.
Fig. 4

Mixed germ cell tumor (65% dysgerminoma and 35% yolk sac) in an 8-year-old girl with elevated lactate dehydrogenase (LDH). a Transverse ultrasound image demonstrates a solid echogenic mass with a few punctate calcifications (measured). b–d Axial T1-W (b), sagittal T2-W fat-saturated (c) and sagittal post-contrast T1-W fat-saturated (d) MR images demonstrate a large complex solid–cystic heterogeneously enhancing mass occupying the pelvis and lower abdomen. The mass has two distinct morphological components: a central solid–cystic component likely corresponding with yolk sac tumor (arrowheads) and a superior and peripheral solid component likely corresponding with dysgerminoma (arrows)

Ovarian surface epithelial-stromal tumors

Epithelial ovarian tumors represent 15–20% of all pediatric ovarian tumors [6] but occur almost exclusively post-menarche and hormonal stimulation has been proposed to trigger their development [51]. Histological subtypes of epithelial ovarian tumors include serous, mucinous, endometrioid, clear cell, transitional (Brenner) and epithelial-stromal types. The vast majority of ovarian epithelial tumors in the pediatric population are serous or mucinous type [44, 51, 52, 53]. Each subtype can be graded as benign, borderline or malignant. Borderline or malignant epithelial tumors occur in only 2–3% of women younger than 21 years with an ovarian neoplasm [52] but represent a much higher proportion of cases at tertiary referral centers, ranging from 15% to 35% [51, 52, 53].

Benign serous and mucinous cystadenomas

Benign cystadenomas are cystic lesions with septa <3 mm in thickness and normally do not have any solid components [54, 55]. Serous cystadenomas are typically unilocular and indistinguishable from follicular cysts by imaging (Fig. 5). Cyst contents have signal characteristics similar to simple fluid, unless complicated by infection or hemorrhage. Occasionally they have a few locules or one or two septa. In contrast, mucinous cystadenomas are characteristically multilocular [54] (Fig. 6). Cyst contents in mucinous cystadenomas have variable signal intensity, reflecting their mucinous composition and resulting in a stained-glass appearance of the multiple cysts on T2-weighted imaging. The stained-glass appearance on T2-weighted imaging is classic but not specific [56]. Mural calcifications can be seen in both serous and mucinous cystadenomas but are more common in mucinous types [57].
Fig. 5

Serous cystadenoma in a 17-year-old girl. a, b Transverse ultrasound image (a) and sagittal T2-W fat-saturated MR image (b) demonstrate a large unilocular cystic mass without any septa or solid components. Pathology revealed serous cystadenoma

Fig. 6

Mucinous cystadenoma in an 11-year-old girl. a, b Sagittal Doppler ultrasound image (a) and coronal single-shot T2-W fat-saturated MR image (b) demonstrate a large multicystic mass composed of multiple variable-size cysts with vascularity in some of the septa (arrow). Pathology revealed mucinous cystadenoma

Borderline and malignant ovarian epithelial tumors

Borderline epithelial tumors are more frequent in children than in adults, and are also more frequent than malignant epithelial carcinomas [6]. Features suggestive of a borderline or malignant epithelial lesion are larger size, thick irregular walls, thick septa >5 mm and large soft-tissue components with necrosis [54]. Papillary projections represent folds of proliferating epithelium growing over a stromal core and should be distinguished from mural solid tissue [58]. On T1-weighted images, papillary projections have intermediate signal intensity and avidly enhance, while on T2-weighted images they have a central low-intensity fibrous core surrounded by a high-signal-intensity periphery [58, 59]. Papillary projections are distinctive features of epithelial neoplasms and are particularly characteristic of borderline lesions [58]. Although they have been found to exist in 13–20% of benign cystadenomas, they are usually smaller, fewer in number and extent, and less vascular in benign lesions [55, 60, 61]. And while papillary projections are actually more common in malignant than borderline lesions, malignant lesions are typically dominated by the solid components [55, 58, 60]. Other reported features suggestive of a borderline or malignant mucinous ovarian epithelial tumor include a greater amount of locules relative to overall lesion size compared with benign mucinous cystadenomas, a honeycomb-like pattern of locules, and very low signal of cyst contents on T2-weighted imaging [59, 62, 63]. Ancillary findings of malignancy include direct local invasion, peritoneal implants, ascites and lymphadenopathy.

Ovarian sex cord–stromal tumors

Ovarian sex cord–stromal tumors represent 10–20% of pediatric ovarian neoplasms [6]. They are derived from two groups of embryologic precursor cells, the sex cords and stromal cells. The stromal cells (fibroblasts, theca cells and Leydig cells) are formed from the mesenchymal genital ridge [64]. The sex cords (granulosa and Sertoli cells) are formed from the coelomic epithelium (primordial peritoneum) that grows into the mesenchymal ridge [64]. Ovarian sex cord–stromal tumors often have more than one cell type within the tumor. As a group, ovarian sex cord–stromal tumors occur across all age groups, but the distribution of histological types among ages is vastly different. Granulosa cell and Sertoli-Leydig cell tumors are the most common in children and adolescents while thecoma-fibromas are rare [65]; the opposite distribution is the case for women. Most sex cord–stromal tumors are benign or of low grade at presentation [64].

Juvenile granulosa cell tumor

Granulosa cell tumors can be separated into juvenile and adult types. Adult granulosa cell tumors usually present in women older than 40. In contrast, juvenile granulosa cell tumors occur 77% of the time in females 20 years or younger [66]. They are derived from the cells that surround developing follicles and are typically estrogen-secreting tumors. Pseudoprecocious puberty (development of secondary sex characteristics without ovulation) is the presenting symptom for 75–82% of prepubertal patients [64, 66], although with increased use of imaging, more cases now might be detected before symptoms emerge [67]. Nonetheless, juvenile granulosa cell tumors account for 10% of all precocious puberty cases in girls [64]. Granulosa cell tumors are considered low-grade malignancies. While the juvenile type typically has a higher mitotic index and more aggressive tumor growth compared to the adult granulosa cell tumor, the prognosis is good and late recurrences are rare [68].

Juvenile granulosa cell tumors have heterogeneous and nonspecific imaging appearances. They are typically large and most are mixed solid–cystic, although they can be completely solid or completely cystic [69, 70, 71]. Intratumoral hemorrhage, infarct, necrosis and fibrosis are common. On T2-weighted imaging, the appearance has been described as sponge-like with alternating solid tissue of intermediate signal intermixed with innumerable cystic spaces [70, 71, 72]. Cysts often have hemorrhage within them, and so are bright on T1-weighted images [64, 72]. Auxiliary findings include uterine enlargement and endometrial thickening related to estrogen secretion [70, 71] (Fig. 7).
Fig. 7

Juvenile granulosa cell tumor of the left ovary in a 2-year-old girl. a Sagittal Doppler ultrasound image demonstrates a complex solid–cystic mass with vascularity. b, c Sagittal T2-W (b) and axial post-contrast T1-W fat-saturated (c) MR images demonstrate a complex solid–cystic heterogeneously enhancing mass (solid arrows). Note the post-pubertal appearance of the uterus (arrowhead in b) in keeping with an estrogen-secreting functional tumor. Normal right ovary is also seen (dashed arrow in c)

Thecoma-fibroma tumor

Thecoma-fibroma tumors represent a group of tumors that are typically benign and rare in the pediatric population. There are many types including fibroma, thecoma, fibrothecoma, fibrosarcoma and sclerosing stromal tumor, and all are characterized by their fibrous content. Fibromas, fibrothecomas and thecomas represent a spectrum of benign tumors composed of variable amounts of fibrosis and lipid-rich theca cells, which can be estrogenic [64]. They usually are solid masses that are characteristically hypointense to myometrium on T1- and T2-weighted sequences because of their fibrous contents [73, 74, 75, 76]. When larger, however, they demonstrate cystic degeneration. Fibrothecomas show weak enhancement, typically less than myometrium or fibroids, and often show low signal on high b value DWI [73, 74, 75, 76].

Sertoli-Leydig cell tumor

Sertoli-Leydig cell tumors are rare stromal tumors that can histologically range from well-differentiated to poorly differentiated, which dictates clinical behavior [69]. They are usually detected early, at low grade and with good prognosis, but recurrences are common. Approximately 30–40% cause virilization manifesting as amenorrhea or virilized secondary sex characteristics [67, 69, 77]. Sertoli-Leydig cell tumors can also be seen in DICER1 syndrome [78]. On imaging, Sertoli-Leydig cell tumors are typically solid with multiple peripheral or intratumoral cysts, or cystic with mural solid components [6, 64, 69, 79] (Fig. 8). Calcifications are rare. The signal intensity of the solid component reflects the degree of underlying fibrous tissue but is typically low on T1-weighted imaging and low to intermediate on T2-weighted imaging, with intense enhancement after administration of contrast agent [64, 79].
Fig. 8

Sertoli-Leydig tumor in a 16-year-old girl with DICER1 syndrome. a Sagittal ultrasound image demonstrates a large complex solid–cystic mass (measured). b–d Coronal (b) and sagittal T2-W (c) along with post-contrast T1-W fat-saturated (d) MR images demonstrate a large complex solid–cystic heterogeneously enhancing mass with cystic components showing different signal intensities

Miscellaneous ovarian tumors

Other rare miscellaneous tumors can occur in the ovary in children, such as lymphoma, small cell carcinomas, soft-tissue tumors and metastases. Unlike in adults, metastases to the ovaries in children usually occur hematogenously [24]. Ovarian metastases in children are most commonly from mucinous adenocarcinoma of the colon, Burkitt lymphoma and alveolar rhabdomyosarcoma [80]. Rare causes include Wilms tumor, neuroblastoma and retinoblastoma. On imaging, presence of bilateral adnexal masses and secondary features such as ascites, peritoneal implants, matted bowel, adenopathy and pleural effusions are common with peritoneal spread [80].

Ovarian non-neoplastic processes

Non-neoplastic processes such as hemorrhagic cysts, polycystic ovarian syndrome (PCOS) and massive ovarian edema can mimic ovarian neoplasms. Hemorrhagic ovarian cyst is a common ovarian lesion that can be mistaken for a solid echogenic ovarian mass or ovarian torsion on ultrasound [4]. Hemorrhagic cysts in the ovary are usually follicular cysts that typically show resolution on follow-up ultrasound performed after a menstrual cycle. In indeterminate cases by ultrasound, MRI accurately differentiates hemorrhagic cysts from true ovarian neoplasms (Fig. 9). Massive ovarian edema is a rare benign condition affecting women in the reproductive age group and adolescent girls [81]. It is thought to be caused by intermittent or partial ovarian torsion in most cases but can also be secondary to an underlying ovarian lesion. Clinical presentation is nonspecific. Occasionally it can be a cause of precocious puberty. Findings of an enlarged ovary with stromal edema, peripherally placed follicles, and preserved vascularity on MRI might help to differentiate massive ovarian edema from a neoplasm or a completely torsed ovary [81]. Treatment for massive ovarian edema is surgical but ovary-sparing, consisting of detorsion (if a degree of torsion is found intraoperatively) along with wedge biopsy to exclude tumor. PCOS can also show an enlarged ovary with peripherally placed follicles, but it usually does not enlarge to the extent seen in massive ovarian edema or complete ovarian torsion and is seen bilaterally. In addition, PCOS manifests with a large number of follicles less than 9 mm in diameter [82]. The diagnosis of PCOS requires correlation with specific clinical and laboratory features.
Fig. 9

Hemorrhagic cyst in a 13-year-old girl presenting with abdominal pain. a Transverse Doppler ultrasound image demonstrates an echogenic avascular solid-looking mass in the right ovary (arrow) with vascularity in peripheral normal parenchyma. b–d Axial T2-W fat-saturated (b), T1-W (c) and post-contrast T1-W fat-saturated (d) MR images demonstrate a large peripherally enhancing cyst with settled hematocrit along the dependant aspect (arrows). A crescent of peripheral normal parenchyma with small follicles shows enhancement (arrowheads)

Tubo-ovarian masses

Paraovarian cyst

Paraovarian cysts arise in the mesosalpinx of the broad ligament between the ovary and fallopian tube. They can be mesothelial, paramesonephric or mesonephric in origin. Most are unilocular and filled with simple fluid, although rarely paraovarian cysts can be multilocular, multiple or bilateral [83, 84]. On MRI cyst contents show signal intensity of simple fluid unless complicated by hemorrhage. Paraovarian cysts are commonly located near the ipsilateral round ligament, with the fallopian tube stretched over the cyst walls [84] (Fig. 10). Seeing the cyst separate from the ovary is helpful, but more often paraovarian cysts abut the ovary on MR imaging [84]. Occasionally they predispose the fallopian tube to tubal torsion.
Fig. 10

Paraovarian cyst in a 14-year-old girl. a, b Sagittal (a) and coronal (b) T2-W MR images demonstrate a large cyst with slightly thick irregular wall (arrows). c Post-contrast T1-W fat-saturated axial MR image demonstrates no significant enhancement within the cyst (solid arrow). Note the fallopian tube wrapped around the cyst on its superior aspect (arrowheads in a and b). Normal right ovary is seen (dashed arrow in c). Left ovary (not shown) was separate and normal

Tubo-ovarian torsion

Tubo-ovarian torsion accounts for 30% of pediatric ovarian surgery and occurs in a bimodal distribution, with 16% of cases occurring in infants younger than 1 year and 52% occurring in children ages 9–14 years [85]. An associated lead point is present 50% of the time, which is a malignant neoplasm in 2% [85, 86]. Ultrasound is the modality of choice for evaluating suspected ovarian torsion. MRI, which is typically performed when an enlarged torsed ovary is mistaken for an ovarian neoplasm on ultrasound, shows an enlarged ovary with peripherally displaced follicles secondary to stromal edema [87, 88] (Fig. 11). There can be reduced stromal enhancement with associated hemorrhage and necrosis. A twisted pedicle and thickened fallopian tube are uncommonly seen but specific [88]. The torsed ovary is typically displaced medially while the uterus is deviated toward the affected side [88]. While ovarian stroma normally shows restricted diffusion on DWI, early studies suggest that lower apparent diffusion coefficient values and more intense DWI signal of either the ovarian stroma or the thickened fallopian tube might be specific for hemorrhagic infarction [89, 90, 91], but more validation is needed.
Fig. 11

Ovarian torsion in a 12-year-old girl, misdiagnosed as a neoplastic mass on ultrasound (not shown). a, b Sagittal T2-W fat-saturated (a) and axial T2-W (b) MR images demonstrate a markedly enlarged right ovary with predominantly peripheral follicles (solid arrows). c Post-contrast T1-W fat-saturated MR image demonstrates no significant enhancement within the right ovary (solid arrow). A twisted pedicle is seen close to the right iliac vessels (arrowheads in a–c). Normal left ovary is also seen (dashed arrow in b and c)

Isolated tubal torsion

Isolated tubal torsion is a rare condition defined by torsion of the fallopian tube without torsion of the ipsilateral ovary. It most commonly occurs in women of reproductive age, including young adolescents [92]. Reported predisposing factors include a long mesosalpinx, premenstrual tubal congestion, tubal ligation, adnexal cysts, hydrosalpinx or hematosalpinx, pelvic inflammatory disease, hypermobility or spiral course of the fallopian tube, and trauma [92, 93]. In the pediatric population, most cases of isolated tubal torsion occur in the setting of a paraovarian or paratubal cyst [92, 93]. MR imaging findings include direct visualization of a twisted fallopian tube, sometimes termed a whirlpool sign, with a normal ipsilateral ovary [94, 95] (Fig. 12). The involved fallopian tube might also be dilated and the presence of plicae, which are epithelial folds protruding into the lumen, can help identify an adnexal cystic structure as a dilated fallopian tube [94, 95].
Fig. 12

Isolated tubal torsion secondary to a paraovarian cyst in a 16-year-old girl presenting with right lower quadrant pain. a Sagittal Doppler ultrasound image demonstrates a twisted non-dilated fallopian tube (arrow) with some preserved vascularity, closely associated with a paraovarian cyst (dashed arrow) via a globular echogenic structure presumed to be mesosalpinx (arrowhead). b–d Coronal (b) and axial (c, d) T2-W fat-saturated MR images demonstrate the same findings, with a twisted non-dilated fallopian tube (arrows) leading to a paraovarian cyst (dashed arrows) via a fusiform intermediate-signal-intensity structure presumed to be mesosalpinx (white arrowhead). The bilateral ovaries were normal (black arrowheads)

Tubo-ovarian abscess in pelvic inflammatory disease

Pelvic inflammatory disease occurs in sexually active patients. Associated pyosalpinx or tubo-ovarian abscess (TOA) can present as an adnexal mass on imaging. Pyosalpinx manifests as a fluid-filled tubular adnexal structure with thickened walls. TOA presents as a complex fluid-filled adnexal mass with thick walls and septa (Fig. 13). The MRI appearance of fluid in pyosalpinx and TOA depends on the degree of proteinaceous and hemorrhagic contents, and thus shows variable signal intensity on T1-weighted imaging and is often hyperintense on T2-weighted imaging, with possible amorphous shading in pyosalpinx or internal heterogeneity and debris in TOA [96, 97]. An internal hyperintense rim on T1-weighted imaging that enhances after administration of contrast agent can be seen in TOA, postulated to represent granulation tissue combined with hemorrhage [98]. On DWI, both show restricted diffusion [96, 97]. Thickened walls or septa show variable heterogeneous signal intensity on T1- and T2-weighted images but are typically hypointense on both and enhance intensely.
Fig. 13

Tubo-ovarian abscess in a 15-year-old girl. a Sagittal Doppler ultrasound image demonstrates a complex multicystic mass posterior to the bladder, with thick irregular septa and vascularity/hyperemia (arrowhead). b, c Sagittal T2-W fat-saturated (b) and post-contrast T1-W fat-saturated (c) MR images demonstrate an irregular complex multicystic mass with heterogeneous internal contents and thick enhancing rims (solid arrows). There are extensive surrounding adnexal inflammatory changes with hazy edema and enhancement (dashed arrows)

Differentiating pyosalpinx from hydrosalpinx can occasionally be challenging. Additional findings and a major clue to the diagnosis of pelvic inflammatory disease are adnexal edema, inflammation, diffusion restriction and possible peritoneal fibrosis/adhesions. Fibrosis and adhesions can manifest as hypointense mesh-like stranding on T1-weighted imaging that enhances after contrast administration [98].

Other lesions

Various lesions that are not of tubo-ovarian origin can present as an adnexal mass, such as nerve sheath tumors, Müllerian anomalies, uterine leiomyomas, ectopic kidneys or spleen, or appendicitis. Perforated appendicitis located in the pelvis is one of the common mimickers of an ovarian mass in clinical practice.

Conclusion

Pelvic MRI is the next appropriate step after ultrasound in evaluation of the indeterminate adnexal mass, especially when surgery is contemplated. MRI can add value by distinguishing benign from malignant masses and therefore significantly alter surgical management. Identification of a benign process allows for a fertility-sparing approach with minimally invasive technique, which is critically important in children and adolescents. We have reviewed specific MR imaging findings of various adnexal masses in the pediatric population to help aid interpretation. Often a specific entity can be prospectively suggested based on a combination of clinical, laboratory biomarker, and MR imaging features.

Notes

Compliance with ethical standards

Conflicts of interest

None

References

  1. 1.
    Rogers EM, Cubides GC, Lacy J et al (2014) Preoperative risk stratification of adnexal masses: can we predict the optimal surgical management? J Pediatr Adolesc Gynecol 27:125–128CrossRefPubMedGoogle Scholar
  2. 2.
    Oltmann SC, Garcia N, Barber R et al (2010) Can we preoperatively risk stratify ovarian masses for malignancy? J Pediatr Surg 45:130–134CrossRefPubMedGoogle Scholar
  3. 3.
    Stankovic ZB, Sedlecky K, Savic D et al (2017) Ovarian preservation from tumors and torsions in girls: prospective diagnostic study. J Pediatr Adolesc Gynecol 30:405–412CrossRefPubMedGoogle Scholar
  4. 4.
    Marro A, Allen LM, Kives SL et al (2016) Simulated impact of pelvic MRI in treatment planning for pediatric adnexal masses. Pediatr Radiol 46:1249–1257CrossRefPubMedGoogle Scholar
  5. 5.
    Kirkham YA, Lacy JA, Kives S, Allen L (2011) Characteristics and management of adnexal masses in a Canadian pediatric and adolescent population. J Obstet Gynaecol Can 33:935–943CrossRefPubMedGoogle Scholar
  6. 6.
    Heo SH, Kim JW, Shin SS et al (2014) Review of ovarian tumors in children and adolescents: radiologic-pathologic correlation. Radiographics 3:2039–2055CrossRefGoogle Scholar
  7. 7.
    Chilla B, Hauser N, Singer G et al (2011) Indeterminate adnexal masses at ultrasound: effect of MRI imaging findings on diagnostic thinking and therapeutic decisions. Eur Radiol 21:1301–1310CrossRefPubMedGoogle Scholar
  8. 8.
    Hricak H, Chen M, Coakley FV et al (2000) Complex adnexal masses: detection and characterization with MR imaging — multivariate analysis. Radiology 214:39–46CrossRefPubMedGoogle Scholar
  9. 9.
    Sohaib SAA, Sahdev A, Trappen PV et al (2003) Characterization of adnexal mass lesions on MR imaging. AJR Am J Roentgenol 180:1297–1304CrossRefPubMedGoogle Scholar
  10. 10.
    Anthoulakis C, Nikoloudis N (2014) Pelvic MRI as the “gold standard” in the subsequent evaluation of ultrasound-indeterminate adnexal lesions: a systematic review. Gynecol Oncol 132:661–668CrossRefPubMedGoogle Scholar
  11. 11.
    Kinkel K, Lu Y, Mehdizade A et al (2005) Indeterminate ovarian mass at US: incremental value of second imaging test for characterization — meta-analysis and Bayesian analysis. Radiology 236:85–94CrossRefPubMedGoogle Scholar
  12. 12.
    Fujii S, Kakite S, Nishihara K et al (2008) Diagnostic accuracy of diffusion-weighted imaging in differentiating benign from malignant ovarian lesions. J Magn Reson Imaging 28:1149–1156CrossRefPubMedGoogle Scholar
  13. 13.
    Katayama M, Masui T, Kobayashi S et al (2002) Diffusion-weighted echo planar imaging of ovarian tumors: is it useful to measure apparent diffusion coefficients? J Comput Assist Tomogr 26:250–256CrossRefPubMedGoogle Scholar
  14. 14.
    Kim HJ, Lee SY, Shin YR et al (2016) The value of diffusion-weighted imaging in the differential diagnosis of ovarian lesions: a meta-analysis. PLoS One 11:e0149465CrossRefPubMedPubMedCentralGoogle Scholar
  15. 15.
    Forstner R, Thomassin-Naggara I, Cunha TM et al (2017) ESUR recommendations for MR imaging of the songographically indeterminate adnexal mass: an update. Eur Radiol 27:2248–2257CrossRefPubMedGoogle Scholar
  16. 16.
    Meng XF, Zhu SC, Sun SJ et al (2016) Diffusion weighted imaging for the differential diagnosis of benign vs. malignant ovarian neoplasms. Oncol Lett 11:3795–3802CrossRefPubMedPubMedCentralGoogle Scholar
  17. 17.
    Yuan X, Guo L, Du W et al (2017) Diagnostic accuracy of DWI in patients with ovarian cancer: a meta-analysis. Medicine 96:e6659CrossRefPubMedPubMedCentralGoogle Scholar
  18. 18.
    Thomassin-Naggara I, Darai E, Cuenod CA et al (2009) Contribution of diffusion-weighted MR imaging for predicting benignity of complex adnexal masses. Eur Radiol 19:1544–1552CrossRefPubMedGoogle Scholar
  19. 19.
    Bakir B, Bakan S, Tunaci M et al (2011) Diffusion-weighted imaging of solid or predominantly solid gynaecological adnexal masses: is it useful in the differential diagnosis? Br J Radiol 84:600–611CrossRefPubMedPubMedCentralGoogle Scholar
  20. 20.
    Zhang P, Cui Y, Li W et al (2012) Diagnostic accuracy of diffusion-weighted imaging with conventional MR imaging for differentiating complex solid and cystic ovarian tumors at 1.5 T. World J Surg Oncol 10:237CrossRefPubMedPubMedCentralGoogle Scholar
  21. 21.
    Kyriazi S, Collins DJ, Morgan VA et al (2010) Diffusion-weighted imaging of peritoneal disease for noninvasive staging of advanced ovarian cancer. Radiographics 30:1269–1285CrossRefPubMedGoogle Scholar
  22. 22.
    Mohaghegh P, Rockall AG (2012) Imaging strategy for early ovarian cancer: characterization of adnexal masses with conventional and advanced imaging techniques. Radiographics 32:1751–1773CrossRefPubMedGoogle Scholar
  23. 23.
    Thomassin-Naggara I, Toussaint I, Perrot N et al (2011) Characterization of complex adnexal masses: value of adding perfusion- and diffusion-weighted MR imaging to conventional MR imaging. Radiology 258:793–803CrossRefPubMedGoogle Scholar
  24. 24.
    Epelman M, Chikwava KR, Chauvin N, Servaes S (2011) Imaging of pediatric ovarian neoplasms. Pediatr Radiol 41:1085–1099CrossRefPubMedGoogle Scholar
  25. 25.
    Chen VW, Ruiz B, Kileen JL et al (2003) Pathology and classification of ovarian tumors. Cancer 07:2631–2642CrossRefGoogle Scholar
  26. 26.
    Cass DL, Hawkins E, Brandt ML et al (2001) Surgery for ovarian masses in infants, children, and adolescents: 102 consecutive patients treated in a 15-year period. J Pediatr Surg 36:693–699CrossRefPubMedGoogle Scholar
  27. 27.
    Schultz KA, Sencer SF, Messinger Y et al (2005) Pediatric ovarian tumors: a review of 67 cases. Pediatr Blood Cancer 44:167–173CrossRefPubMedGoogle Scholar
  28. 28.
    Park SB, Kim JK, Kim KR, Cho KS (2008) Imaging findings of complications and unusual manifestations of ovarian teratomas. Radiographics 28:969–983CrossRefPubMedGoogle Scholar
  29. 29.
    Ehren I, Mahour G, Isaacs HJ (1984) Benign and malignant ovarian tumors in children and adolescents. Am J Surg 147:339–344CrossRefPubMedGoogle Scholar
  30. 30.
    Caruso PA, Marsh MR, Minkowitz S, Karten G (1971) An intense clinicopathologic study of 305 teratomas of the ovary. Cancer 27:343–348CrossRefPubMedGoogle Scholar
  31. 31.
    Yamashita Y, Hatanaka Y, Torashima M et al (1994) Mature cystic teratomas of the ovary without fat in the cystic cavity: MR features in 12 cases. AJR Am J Roentgenol 163:613–616CrossRefPubMedGoogle Scholar
  32. 32.
    Outwater EK, Siegelman ES, Hunt JL (2001) Ovarian teratomas: tumor types and imaging characteristics. Radiographics 21:475–490CrossRefPubMedGoogle Scholar
  33. 33.
    Togashi K, Nishimura K, Itoh K et al (1987) Ovarian cystic teratomas: MR imaging. Radiology 162:669–673CrossRefPubMedGoogle Scholar
  34. 34.
    Sahin H, Abdullazade S, Sanci M (2017) Mature cystic teratoma of the ovary: a cutting edge overview on imaging features. Insights Imaging 8:227–241CrossRefPubMedPubMedCentralGoogle Scholar
  35. 35.
    Nakayama T, Yoshimitsu K, Irie H et al (2005) Diffusion-weighted echo-planar MR imaging and ADC mapping in the differential diagnosis of ovarian cystic masses: usefulness of detecting keratinoid substances in mature cystic teratomas. J Magn Reson Imaging 22:271–278CrossRefPubMedGoogle Scholar
  36. 36.
    Biskup W, Calaminus G, Schneider DT et al (2006) Teratoma with malignant transformation: experiences of the cooperative GPOH protocols MAKEI 83/86/89/96. Klin Padiatr 218:303–308CrossRefPubMedGoogle Scholar
  37. 37.
    Yamaoka T, Togashi K, Koyama T et al (2003) Immature teratoma of the ovary: correlation of MR imaging and pathologic findings. Eur Radiol 13:313–319PubMedGoogle Scholar
  38. 38.
    Young JL, Wu XC, Roffers SD et al (2003) Ovarian cancer in children and young adults in the United States, 1992-1997. Cancer 97:2694–2700CrossRefPubMedGoogle Scholar
  39. 39.
    Tanaka YO, Kurosaki Y, Nishida M et al (1994) Ovarian dysgerminoma: MR and CT appearance. J Comput Assist Tomogr 18:443–448CrossRefPubMedGoogle Scholar
  40. 40.
    Kim SH, Kang SB (1995) Ovarian dysgerminoma: color Doppler ultrasonographic findings and comparison with CT and MR imaging findings. J Ultrasound Med 14:843–848CrossRefPubMedGoogle Scholar
  41. 41.
    Guerriero S, Testa AC, Timmerman D et al (2011) Imaging of gynecological disease (6): clinical and ultrasound characteristics of ovarian dysgerminoma. Ultrasound Obstet Gynecol 37:596–602CrossRefPubMedGoogle Scholar
  42. 42.
    Shaaban AM, Rezvani M, Elsays KM et al (2014) Ovarian malignant germ cell tumors: cellular classification and clinical and imaging features. Radiographics 34:777–801CrossRefPubMedGoogle Scholar
  43. 43.
    Lack EE, Young RH, Scully RE (1992) Pathology of ovarian neoplasms in childhood and adolescence. Pathol Annu 27:281–356PubMedGoogle Scholar
  44. 44.
    Norris HJ, Jensen RD (1972) Relative frequency of ovarian neoplasms in children and adolescents. Cancer 30:713–719CrossRefPubMedGoogle Scholar
  45. 45.
    Li YK, Zheng Y, Lin JB et al (2015) CT imaging of ovarian yolk sac tumor with emphasis on differential diagnosis. Sci Rep 5:11000CrossRefPubMedPubMedCentralGoogle Scholar
  46. 46.
    Choi HJ, Moon MH, Kim SH et al (2008) Yolk sac tumor of the ovary: CT findings. Abdom Imaging 33:736–739CrossRefPubMedGoogle Scholar
  47. 47.
    Yamaoka T, Togashi K, Koyama T et al (2000) Yolk sac tumor of the ovary: radiologic-pathologic correlation in four cases. J Comput Assist Tomogr 24:605–609CrossRefPubMedGoogle Scholar
  48. 48.
    Levitin A, Haller KD, Cohen HL et al (1996) Endodermal sinus tumor of the ovary: imaging evaluation. AJR Am J Roentgenol 167:791–793CrossRefPubMedGoogle Scholar
  49. 49.
    McCarthy S, Schwartz PE (2010) CT scan detects linear tear in ovarian yolk sac tumor. Eur J Radiol Extra 83:e73–e75CrossRefGoogle Scholar
  50. 50.
    Kurman RJ, Norris HJ (1976) Endodermal sinus tumor of the ovary: a clinical and pathologic analysis of 71 cases. Cancer 38:2404–2419CrossRefPubMedGoogle Scholar
  51. 51.
    Tsai JY, Saigo PE, Brown C, La Quaglia MP (2001) Diagnosis, pathology, staging, treatment, and outcome of epithelial ovarian neoplasia in patients age < 21 years. Cancer 91:2065–2070CrossRefPubMedGoogle Scholar
  52. 52.
    Diamond MP, Baxter JW, Peerman CG Jr, Burnett LS (1988) Occurrence of ovarian malignancy in childhood and adolescence: a community-wide evaluation. Obstet Gynecol 71:858–860PubMedGoogle Scholar
  53. 53.
    Morowitz M, Huff D, von Allmen D (2003) Epithelial ovarian tumors in children: a retrospective analysis. J Pediatr Surg 38:331–335CrossRefPubMedGoogle Scholar
  54. 54.
    Jung SE, Lee JM, Rha SE et al (2002) CT and MR imaging of ovarian tumors with emphasis on differential diagnosis. Radiographics 22:1305–1325CrossRefPubMedGoogle Scholar
  55. 55.
    Ghossain MA, Buy JN, Ligneres C et al (1991) Epithelial tumors of the ovary: comparison of MR and CT findings. Radiology 181:863–870CrossRefPubMedGoogle Scholar
  56. 56.
    Tanaka YO, Nishida M, Kurosaki Y et al (1999) Differential diagnosis of gynaecological “stained glass” tumors on MRI. Br J Radiol 72:414–420CrossRefPubMedGoogle Scholar
  57. 57.
    Okada S, Ohaki Y, Inoue K et al (2005) Calcifications in mucinous and serous cystic ovarian tumors. J Nippon Med Sch 72:29–33CrossRefPubMedGoogle Scholar
  58. 58.
    Outwater EK, Huang AB, Dunton CJ et al (1997) Papillary projections in ovarian neoplasms: appearance on MRI. J Magn Reson Imaging 7:689–695CrossRefPubMedGoogle Scholar
  59. 59.
    Ma FH, Zhao SH, Qiang JW et al (2014) MRI appearances of mucinous borderline ovarian tumors: pathological correlation. J Magn Reson Imaging 40:745–751CrossRefPubMedGoogle Scholar
  60. 60.
    Granberg S, Wikland M, Jansson I (1989) Macroscopic characterization of ovarian tumors and the relation to the histological diagnosis: criteria to be used for ultrasound evaluation. Gynecol Oncol 34:139–144CrossRefGoogle Scholar
  61. 61.
    Hassen K, Ghoussain MA, Rousset P et al (2011) Characterization of papillary projections in benign versus borderline and malignant ovarian masses on conventional and color Doppler ultrasound. AJR Am J Roentgenol 196:1444–1449CrossRefPubMedGoogle Scholar
  62. 62.
    Zhao SH, Qiang JW, Zhang GF et al (2014) MRI in differentiating ovarian borderline from benign mucinous cystadenoma: pathological correlation. J Magn Reson Imaging 39:162–166CrossRefPubMedGoogle Scholar
  63. 63.
    Okamoto Y, Tanaka YO, Tsunoda H, Minami M (2007) Malignant or borderline mucinous cystic neoplasms have a larger number of loculi than mucinous cystadenoma: a retrospective study with MR. J Magn Reson Imaging 26:94–99CrossRefPubMedGoogle Scholar
  64. 64.
    Outwater EK, Wagner BJ, Mannion C et al (1998) Sex cord-stromal and steroid cell tumors of the ovary. Radiographics 18:1523–1546CrossRefPubMedGoogle Scholar
  65. 65.
    Schneider DT, Calaminus G, Wessalowski R et al (2003) Ovarian sex cord-stromal tumors in children and adolescents. J Clin Oncol 21:2357–2363CrossRefPubMedGoogle Scholar
  66. 66.
    Young RH, Dickersin GR, Scully RE (1984) Juvenile granulosa cell tumor of the ovary. A clinicopathological analysis of 125 cases. Am J Surg Pathol 8:575–596CrossRefPubMedGoogle Scholar
  67. 67.
    Cecchetto G, Ferrari A, Bernini G et al (2011) Sex cord stromal tumors of the ovary in children: a clinicopathological report from the Italian TREP project. Pediatr Blood Cancer 56:1062–1067CrossRefPubMedGoogle Scholar
  68. 68.
    Schneider DT, Janig U, Calaminus G et al (2003) Ovarian sex cord-stromal tumors — a clinicopathological study of 72 cases from the Kiel Pediatric Tumor Registry. Vichows Arch 443:549–560CrossRefGoogle Scholar
  69. 69.
    Jung SE, Rha SE, Lee JM et al (2005) CT and MRI findings of sex cord-stromal tumor of the ovary. AJR Am J Roentgenol 185:207–215CrossRefPubMedGoogle Scholar
  70. 70.
    Morikawa K, Hatabu H, Togashi K et al (1997) Granulosa cell tumor of the ovary: MR findings. J Comput Assist Tomogr 21:1001–1004CrossRefPubMedGoogle Scholar
  71. 71.
    Kitamura Y, Kanegawa K, Muraji T, Sugimura K (2000) MR imaging of juvenile granulosa cell tumor of the ovary: a case report. Pediatr Radiol 30:360CrossRefPubMedGoogle Scholar
  72. 72.
    Rusterholz KR, MacDonald W (2016) An unusual case of juvenile granulomas cell tumor of the ovary. Radiol Case Rep 4:178CrossRefPubMedPubMedCentralGoogle Scholar
  73. 73.
    Shinagare AB, Meylaerts LJ, Laury AR, Mortele KJ (2012) MRI features of ovarian fibroma and fibrothecoma with histopathologic correlation. AJR Am J Roentgenol 198:296–303CrossRefGoogle Scholar
  74. 74.
    Chen J, Wang J, Chen X et al (2017) Computed tomography and magnetic resonance imaging features of ovarian fibrothecoma. Oncol Lett 14:1172–1178CrossRefPubMedPubMedCentralGoogle Scholar
  75. 75.
    Chung BM, Park SB, Lee JB et al (2015) Magnetic resonance imaging features of ovarian fibroma, fibrothecoma, and thecoma. Abdom Imaging 40:1263–1272CrossRefPubMedGoogle Scholar
  76. 76.
    Yin B, Li W, Cui Y et al (2016) Value of diffusion-weighted imaging combined with conventional magnetic resonance imaging in the diagnosis of thecomas/fibrothecomas and their differential diagnosis with malignant pelvic solid tumors. World J Surg Oncol 14:5CrossRefPubMedPubMedCentralGoogle Scholar
  77. 77.
    Young RH, Scully RE (1985) Ovarian Sertoli-Leydig cell tumors: a clinicopathological analysis of 207 cases. Am J Surg Pathol 9:543–569CrossRefPubMedGoogle Scholar
  78. 78.
    Bueno MT, Martínez-Ríos C, la Puente GA et al (2017) Pediatric imaging in DICER1 syndrome. Pediatr Radiol 47:1292–1301CrossRefPubMedGoogle Scholar
  79. 79.
    Cai SQ, Zhao SH, Qiang JW et al (2013) Ovarian Sertoli-Leydig cell tumors: MRI findings and pathological correlation. J Ovarian Res 6:73CrossRefPubMedPubMedCentralGoogle Scholar
  80. 80.
    McCarville MB, Hill DA, Miller BE, Pratt CB (2001) Secondary ovarian neoplasms in children: imaging features with histopathologic correlation. Pediatr Radiol 31:358–364CrossRefPubMedGoogle Scholar
  81. 81.
    Dahmoush H, Anupindi SA, Pawel BR, Chauvin NA (2017) Multimodality imaging findings of massive ovarian edema in children. Pediatr Radiol 47:576–583CrossRefPubMedGoogle Scholar
  82. 82.
    Fondin M, Rachas A, Huynh V et al (2017) Polycystic ovary syndrome in adolescents: which MR imaging-based diagnostic criteria? Radiology 285:961–970CrossRefPubMedGoogle Scholar
  83. 83.
    Kim JS, Woo SK, Suh SJ, Morettin LB (1995) Sonographic diagnosis of paraovarian cysts: value of detecting a separate ipsilateral ovary. AJR Am J Roentgenol 164:1441–1444CrossRefPubMedGoogle Scholar
  84. 84.
    Kishimoto K, Ito K, Awaya H et al (2002) Paraovarian cyst: MR imaging features. Abdom Imaging 27:685–689CrossRefPubMedGoogle Scholar
  85. 85.
    Oltmann SC, Fischer A, Barber R (2009) Cannot exclude torsion — a 15 year review. J Pediatr Surg 44:1212–1216CrossRefPubMedGoogle Scholar
  86. 86.
    Oltmann SC, Fischer A, Barber R (2010) Pediatric ovarian malignancy presenting as ovarian torsion: incidence and relevance. J Pediatr Surg 45:135–139CrossRefPubMedGoogle Scholar
  87. 87.
    Duigenan S, Oliva E, Lee SI (2012) Ovarian torsion: diagnostic features on CT and MRI with pathologic correlation. AJR Am J Roentgenol 198:W122–W131CrossRefPubMedGoogle Scholar
  88. 88.
    Rha SE, Byun JY, Jung SE et al (2002) CT and MR imaging features of adnexal torsion. Radiographics 22:283–294CrossRefPubMedGoogle Scholar
  89. 89.
    Fujii S, Kaneda S, Kakite S et al (2011) Diffusion-weighted imaging findings of adnexal torsion: initial results. Eur J Radiol 77:330–334CrossRefPubMedGoogle Scholar
  90. 90.
    Kato H, Kanematsu M, Uchiyama M et al (2014) Diffusion-weighted imaging of ovarian torsion: usefulness of apparent diffusion coefficient (ADC) values for the detection of hemorrhagic infarction. Magn Reson Med Sci 13:39–44CrossRefPubMedGoogle Scholar
  91. 91.
    Moribata Y, Kido A, Yamaoka T et al (2015) MR imaging findings of ovarian torsion correlate with pathological hemorrhagic infarction. J Obstet Gynaecol Res 41:1433–1439CrossRefPubMedGoogle Scholar
  92. 92.
    Harmon JC, Binkovitz LA, Binkovitz LE (2008) Isolated fallopian tube torsion: sonographic and CT features. Pediatr Radiol 38:175–179CrossRefPubMedGoogle Scholar
  93. 93.
    Narayanan S, Bandarkar A, Bulus DI (2013) Fallopian tube torsion in the pediatric age group: radiologic evaluation. J Ultrasound Med 33:1697–1704CrossRefGoogle Scholar
  94. 94.
    Sakuragi M, Kido A, Himoto Y et al (2017) MRI findings of isolated tubal torsions: case series of 12 patients: MRI findings suggesting isolated tubal torsions, correlating with surgical findings. Clin Imaging 41:28–32CrossRefPubMedGoogle Scholar
  95. 95.
    Park BK, Kim CK, Kim B (2008) Isolated tubal torsion: specific signs on preoperative computed tomography and magnetic resonance imaging. Acta Radiol 49:233–235CrossRefPubMedGoogle Scholar
  96. 96.
    Rezvani M, Shaaban AM (2011) Fallopian tube disease in the nonpregnant patient. Radiographics 31:527–548CrossRefPubMedGoogle Scholar
  97. 97.
    Kim M, Rha SE, Oh SN et al (2009) MR imaging findings of hydrosalpinx: a comprehensive review. Radiographics 29:495–507CrossRefPubMedGoogle Scholar
  98. 98.
    Ha HK, Lim GY, Cha ES et al (1995) MR imaging of tubo-ovarian abscess. Acta Radiol 36:510–514CrossRefPubMedGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Department of Diagnostic Imaging, The Hospital for Sick Children, Medical ImagingUniversity of TorontoTorontoCanada

Personalised recommendations