An Overview of the Role of Multiparametric MRI in the Investigation of Testicular Tumors

Simple Summary Although conventional ultrasonography remains the primary imaging modality for the assessment of testicular tumors, multiparametric MRI of the scrotum, which combines morphologic and functional data, serves as a powerful adjunct. Based on the recommendations issued by the Scrotal and Penile Imaging Working Group of the European Society of Urogenital Radiology, scrotal MRI is strongly recommended after equivocal US findings. In cases of testicular masses, the main clinical indications are as follows: lesion characterization when sonographic findings are non-diagnostic, discrimination between germ-cell and non-germ-cell testicular tumors, local staging of testicular tumors in patients who are candidates for testis-sparing surgery, and preoperative histological characterization of testicular germ-cell tumors in selected cases. This article aims to provide an overview of the role of multiparametric MRI in the investigation of testicular tumors. Abstract Conventional ultrasonography represents the mainstay of testis imaging. In cases in which ultrasonography is inconclusive, scrotal MRI using a multiparametric protocol may be used as a useful problem-solving tool. MRI of the scrotum is primarily recommended for differentiating between benign and malignant testicular masses when sonographic findings are ambiguous. This technique is also accurate in the preoperative local staging of testicular tumors and, therefore, is recommended in patients scheduled for testis-sparing surgery. In addition, MRI may provide valuable information regarding the histological characterization of testicular germ-cell tumors, in selected cases. Scrotal MRI may also help in the differentiation between testicular germ-cell neoplasms and non-germ-cell neoplasms. Axial T1-weighted imaging, axial and coronal T2-weighted imaging, axial diffusion-weighted imaging, and coronal subtracted dynamic contrast-enhanced imaging are the minimum requirements for scrotal MRI. A variety of MRI techniques—including diffusion tensor imaging, magnetization transfer imaging, proton MR spectroscopy, volumetric apparent diffusion coefficient histogram analysis, and MRI-based radiomics—are being investigated for testicular mass characterization, providing valuable supplementary diagnostic information. In the present review, we aim to discuss clinical indications for scrotal MRI in cases of testicular tumors, along with MRI findings of common testicular malignancies.


Introduction
Scrotal ultrasonography (US), with a combination of grayscale and color Doppler US, is the first-line imaging technique for the evaluation of testicular pathology, due to its low cost, wide availability, convenience, short examination time, lack of radiation exposure, and high diagnostic accuracy [1][2][3][4][5][6][7]. This technique represents the initial imaging diagnostic tool for the confirmation of the presence of a testicular mass and for the evaluation of the in the characterization of LCTs, allowing for the adoption of conservative surgery or active surveillance in compliant patients [14].

mpMRI Protocol of the Scrotum
Correct positioning is essential for scrotal MRI [14]. Patients are placed into the MRI unit in the supine position, with the feet first. The penis is raised and fixed to the lower abdominal wall, covered with a gauge. A towel is placed between the thighs to elevate the scrotum and to ensure that both testes have a similar distance from the coil. A circular surface coil is placed over the scrotum, on top of a second towel [14,22,29].
Thin-section, small-field-of-view T2WI in at least two orthogonal planes should be obtained, along the testes' transverse axis and length. T2WI is accurate for the detection, localization, and characterization of lesions. The coronal plane allows a comparative evaluation of both testes, along with the paratesticular spaces and the spermatic cords. Sagittal T2WI may be added, and is helpful in lesion localization, assessment of small mass lesions, and local staging of testicular tumors [14]. Axial T1WI should be included in the routine MRI protocol, and provides information on scrotal anatomy and the presence of areas of increased signal, corresponding to either fat or methemoglobin [14]. DWI should be acquired in the axial plane, with at least three b-values (0, 400-500, and 800-1000 s/mm 2 ). Apparent diffusion coefficient (ADC) maps should be created on the MRI console and analyzed both qualitatively and quantitatively. ADC calculation is performed using the highest b-value (800-1000 s/mm 2 ) [14]. Coronal DCE imaging is performed after the bolus intravenous injection of gadolinium-based contrast medium. Five to seven consecutive imaging sets, each of 60 s duration, are acquired after the injection of the contrast material. The total examination time is 8 min. Subtraction is subsequently performed using the dataset obtained before administration of the contrast agent as a mask, to subtract slice-byslice from each of the datasets obtained after contrast injection [14].
High magnetic fields (3.0 T) provide a higher signal-to-noise ratio compared to 1.5 T magnets, resulting in improvement of the diagnostic efficacy of the MRI examination. However, the biological effects of the higher field strength on spermatogenesis have not been investigated [22].

Normal Conventional MRI Findings
Normal testes are depicted as well-defined, ovoid, homogeneous structures, with signal intensity similar to that of the surrounding muscle, on T1WI. They appear hyperintense on T2WI, with thin, hypointense septa coursing radially and towards a low-signal-intensity band in the posterior aspect of the testis-the mediastinum testis. The T2 hyperintensity of the normal testicular parenchyma is ideal for the detection of intratesticular masses. The tunica albuginea is seen as a thin hypointense rim around the testis, best detected via T2WI ( Figure 1) [22,26,29]. The normal epididymis is slightly inhomogeneous, being isointense to the testis on T1WI, and mostly hypointense on T2WI (Figure 1c). The spermatic cord is better depicted on coronal T2WI, and appears hyperintense with internal flow voids, due to the presence of fat and vessels. The vas deferens demonstrates a dark wall and a bright lumen on T2WI. A small hydrocele is often seen as a hyperintense rim surrounding the testis on T2WI, and represents a normal finding [26,29,35]. testis, best detected via T2WI ( Figure 1) [22,26,29]. The normal epididymis is slightly inhomogeneous, being isointense to the testis on T1WI, and mostly hypointense on T2WI (Figure 1c). The spermatic cord is better depicted on coronal T2WI, and appears hyperintense with internal flow voids, due to the presence of fat and vessels. The vas deferens demonstrates a dark wall and a bright lumen on T2WI. A small hydrocele is often seen as a hyperintense rim surrounding the testis on T2WI, and represents a normal finding [26,29,35]. (c) Coronal T2WI shows bilateral epididymal heads and tails (arrows), which are mostly hypointense compared to the normal testicular parenchyma. Thin, hypointense septa are detected within the right testis. A small bilateral hydrocele is also seen (normal finding). Coronal (d) ADC and (e) color-coded FA maps show bilateral normal testes. The mean ADC and FA of the normal right testis are 1.07 × 10 −3 mm 2 /s and 0.06, respectively. (f) 1H-MR spectrum of the right testis. Prominent metabolic peaks include the following: Cr, creatine; Glx complex, glutamine and glutamate; mI, myo-inositol; Cho, choline; TLM, total lipids and macromolecules resonating at 2.0 ppm, 1.6 ppm, 1.3 ppm, and 0.9 ppm (parts per million).

Diffusion-Weighted Imaging
DWI is a functional imaging technique based on the measurement of random diffusion movements of water molecules. Tissue cellularity and the presence of intact cell membranes largely determine the impedance of diffusion. This is quantitatively assessed by calculating the ADC, using different b-values. ADC is displayed as a parametric map, and expressed in units of mm 2 /s. DWI has a short acquisition time, and does not require contrast agents or expensive and complex hardware [50][51][52][53][54][55][56][57]. The tunica albuginea appears as a thin hypointense rim surrounding the testes, best seen on T2WI. (c) Coronal T2WI shows bilateral epididymal heads and tails (arrows), which are mostly hypointense compared to the normal testicular parenchyma. Thin, hypointense septa are detected within the right testis. A small bilateral hydrocele is also seen (normal finding). Coronal (d) ADC and (e) color-coded FA maps show bilateral normal testes. The mean ADC and FA of the normal right testis are 1.07 × 10 −3 mm 2 /s and 0.06, respectively. (f) 1H-MR spectrum of the right testis. Prominent metabolic peaks include the following: Cr, creatine; Glx complex, glutamine and glutamate; mI, myo-inositol; Cho, choline; TLM, total lipids and macromolecules resonating at 2.0 ppm, 1.6 ppm, 1.3 ppm, and 0.9 ppm (parts per million).

Diffusion-Weighted Imaging
DWI is a functional imaging technique based on the measurement of random diffusion movements of water molecules. Tissue cellularity and the presence of intact cell membranes largely determine the impedance of diffusion. This is quantitatively assessed by calculating the ADC, using different b-values. ADC is displayed as a parametric map, and expressed in units of mm 2 /s. DWI has a short acquisition time, and does not require contrast agents or expensive and complex hardware [50][51][52][53][54][55][56][57].
The testis' structural complexity-including the presence of densely packed seminiferous tubules, which are lined with a compact fibroelastic connective tissue and interstitial stroma, containing fibroblasts, blood vessels, lymphatics, and Leydig cells-results in restricted diffusion. Normal testes appear hyperintense and slightly hypointense on high-b-value DWI and ADC maps, respectively, with an ADC ranging between 1.08 and 1.31 × 10 −3 mm 2 /s [50][51][52].
DWI improves the diagnostic efficacy of MRI in the characterization of testicular masses. Testicular carcinomas often present with restricted diffusion and, therefore, are detected as hyperintense and hypointense on DWI and ADC maps, respectively, with a decrease in ADC (Figure 2), when compared to normal testes and benign testicular lesions. Tumors' histological characteristics-such as increased cellularity, densely packed neoplastic cells, enlargement of the nuclei, and angulation of the nuclear contour-result in impedance of the motility of the water molecules [50][51][52][53][54][55][56][57].
DWI improves the diagnostic efficacy of MRI in the characterization of testicular masses. Testicular carcinomas often present with restricted diffusion and, therefore, are detected as hyperintense and hypointense on DWI and ADC maps, respectively, with a decrease in ADC (Figure 2), when compared to normal testes and benign testicular lesions. Tumors' histological characteristics-such as increased cellularity, densely packed neoplastic cells, enlargement of the nuclei, and angulation of the nuclear contour-result in impedance of the motility of the water molecules [50][51][52][53][54][55][56][57].  The overall accuracy of conventional MRI, DWI alone, and DWI combined with conventional sequences in the characterization of testicular lesions has been reported to be 91%, 87%, and 100%, respectively [50]. An ADC cutoff of 0.99 × 10 −3 mm 2 /s has been proven to be reliable for the characterization of testicular lesions, with 93.3% sensitivity Figure 2. Right testicular seminoma in a 48-year-old man: The patient had a history of left radical orchiectomy due to seminoma. The incidence of metachronous TGCT in the contralateral testis is approximately 1-5% in men affected by TGCTs. Coronal (a) T2WI, (b) ADC, (c) color-coded FA map, and (d) subtracted DCE images demonstrate a multinodular right intratesticular tumor (arrow)-mainly homogeneous-of low T2 signal. Intratumoral septa are seen as hypointense bands on T2WI, enhancing more than the remaining tumor. The carcinoma appears with low signal on the ADC map, due to restricted diffusion. The ADC and FA of testicular seminoma are 0.51 × 10 −3 mm 2 /s and 0.15, respectively. (e) 1H-MR spectrum of testicular seminoma shows a significant decrease in choline peaks (Cr, creatine; Glx complex, glutamine and glutamate; mI, myo-inositol; Cho, choline; Lac, lactate; Lip: lipids).
The overall accuracy of conventional MRI, DWI alone, and DWI combined with conventional sequences in the characterization of testicular lesions has been reported to be 91%, 87%, and 100%, respectively [50]. An ADC cutoff of 0.99 × 10 −3 mm 2 /s has been proven to be reliable for the characterization of testicular lesions, with 93.3% sensitivity and 90% specificity [52]. In a recent study, an ADC lower than 0.90 × 10 −3 mm 2 /s was considered to be suggestive of the diagnosis of a testicular tumor [56]. In the same study, the authors found that adding DWI to the conventional MRI sequences increased the specificity and accuracy of the technique in the characterization of testicular masses by 9.8% and 3.2%, respectively [56]. Table 1 summarizes data from the literature on the diagnostic efficacy of DWI in the characterization of intratesticular masses.

Dynamic Contrast-Enhanced MRI
DCE-MRI is a functional imaging tool that provides information on tissue perfusion, vessel permeability, and the structure of the extracellular-extravascular space. Semiquantitative analysis of DCE imaging data evaluates changes in signal intensity over time within a specific region of interest (ROI), and describes tissue enhancement by calculating parameters such as maximum enhancement, time to peak, wash-in rate, and washout. Early changes in the time-signal intensity (TSI) curve correlate with the blood flow, and late changes are mainly related to the extravascular accumulation of the contrast medium through capillary permeability [58].
DCE-MRI of the scrotum is strongly recommended when US results on testicular mass lesion characterization are equivocal [14]. Differences in TSI curves between normal testes, benign testicular lesions, and testicular tumors have been described [59][60][61][62][63][64]. Normal testes enhance moderately and homogeneously, with a gradual, progressive increase in enhancement throughout the dynamic scan (type I curve) ( Figure 3) [59][60][61][62][63][64]. This behavior is probably related to an intact blood-testis barrier [65]. TGCTs typically present with a type III curve, which is early and avid enhancement, followed by a gradual washout of the contrast medium ( Figure 4). Absence of contrast enhancement represents a specific sign for the characterization of the benign nature of testicular masses (type 0 curve). Alternatively, benign testicular lesions may enhance homogeneously or heterogeneously, with early, strong enhancement, followed by either a plateau or a further slight, progressive contrast enhancement (type II curve) ( Figure 3) [62,63].
In a retrospective study of 27 testicular lesions, type I curves were detected in all (100%) normal testes, type III curves were seen in all (100%) TGCTs, and type II curves were found in 63.6% of benign lesions. The maximum time to peak, which is defined as the difference between the arrival of the contrast medium and the time of peak intensity, proved to be an independent predictor of malignancy in the same study [62].
This behavior is probably related to an intact blood-testis barrier [65]. TGCTs typically present with a type III curve, which is early and avid enhancement, followed by a gradual washout of the contrast medium ( Figure 4). Absence of contrast enhancement represents a specific sign for the characterization of the benign nature of testicular masses (type 0 curve). Alternatively, benign testicular lesions may enhance homogeneously or heterogeneously, with early, strong enhancement, followed by either a plateau or a further slight, progressive contrast enhancement (type II curve) ( Figure 3) [62,63].   In a retrospective study of 27 testicular lesions, type I curves were detected in all (100%) normal testes, type III curves were seen in all (100%) TGCTs, and type II curves were found in 63.6% of benign lesions. The maximum time to peak, which is defined as the difference between the arrival of the contrast medium and the time of peak intensity, proved to be an independent predictor of malignancy in the same study [62]. Table 2 presents data from the literature on the role of DCE-MRI in the characterization of intratesticular masses. (a) Coronal T2WI shows a large, extremely heterogeneous testicular tumor, replacing the right testis. The neoplasm has hyperintense and hypointense parts (arrows), the latter corresponding to the seminomatous histological component. (b) Transverse T1WI image depicts the tumor's inhomogeneity. Coronal (c) ADC and (d) color-coded FA maps. Note: the seminomatous part (arrows) causes significant restricted diffusion (mean ADC: 0.78 × 10 −3 mm 2 /s). The neoplasm shows increased anisotropy. The mean FA of the seminomatous and nonseminomatous parts of the tumor is 0.16 and 0.12, respectively-higher than that of the contralateral normal testis (star, FA: 0.04). (e) Coronal DCE imaging and TSI curve show the neoplasm enhancing strongly and heterogeneously, with a type III curve. Table 2 presents data from the literature on the role of DCE-MRI in the characterization of intratesticular masses. DTI was developed based on DWI to depict the direction and speed of water molecules' diffusion. Diffusion is essentially a three-dimensional process; therefore, molecular mobility in structured tissues may be anisotropic [65,66]. The normal testis is a structured organ, with the seminiferous tubules, septa, and vessels orientated towards the mediastinum testis [42]. DTI calculates both ADC and fractional anisotropy (FA), providing information not only on the amount of diffusion, but also on the anisotropy of diffusion, which may occur in an anisotropic manner ( Figure 1). Tissue structural integrity and microstructural changes can be assessed by DTI [66,67].
Testicular lesions-both benign and malignant-often have increased anisotropy when compared to normal testes. Specifically, TGCTs usually present with higher FA compared to normal testicular parenchyma (Table 3; Figures 2 and 4). Increases in the number of neoplastic cells, the amount of cell membranes, and intracellular viscosity, combined with a decrease in the extracellular space, may explain the anisotropy seen in testicular tumors [42]. MTI is a technique that enables measurements beyond those in conventional MRI images, based on the interactions between free water protons, which are responsible for the conventional MRI signal and restricted protons, which are bound to proteins and macromolecules. Exchange and cross-relaxation of magnetization between these two Cancers 2022, 14, 3912 9 of 24 proton pools creates the MT effect. The MTI phenomenon is quantified by the MT ratio (MTR) [68].
The macromolecular content of normal testes is responsible for an MTR of 46%. The main contributing factors to the MT effects of the normal testicular parenchyma include the collagenous content and the presence of ample smooth endoplasmic reticulum-a membranous structure that plays an important role in the synthesis of testosterone [43].
MTI represents a supplementary MRI tool for the characterization of testicular masses. TGCTs often have high MTR (Table 4, Figure 5) when compared to normal testes and benign testicular lesions-a finding that may be related to tumoral heterogeneity and the presence of fibrovascular septa in cases of seminomas, and to the presence of hemorrhagic foci in cases of nonseminomas [43].  MTI is a technique that enables measurements beyond those in conventional MRI images, based on the interactions between free water protons, which are responsible for the conventional MRI signal and restricted protons, which are bound to proteins and macromolecules. Exchange and cross-relaxation of magnetization between these two proton pools creates the MT effect. The MTI phenomenon is quantified by the MT ratio (MTR) [68].
The macromolecular content of normal testes is responsible for an MTR of 46%. The main contributing factors to the MT effects of the normal testicular parenchyma include the collagenous content and the presence of ample smooth endoplasmic reticulum-a membranous structure that plays an important role in the synthesis of testosterone [43].
MTI represents a supplementary MRI tool for the characterization of testicular masses. TGCTs often have high MTR (Table 4, Figure 5) when compared to normal testes and benign testicular lesions-a finding that may be related to tumoral heterogeneity and the presence of fibrovascular septa in cases of seminomas, and to the presence of hemorrhagic foci in cases of nonseminomas [43].

Proton Magnetic Resonance Spectroscopy
1H-MRS provides information about the biochemical milieu of normal tissues an pathological entities [69,70]. The prominent metabolic peaks of the 1H-MR spectrum normal adult testes include choline, creatine, myo-inositol, and lipids ( Figure 1f) [71]. Base on preliminary data, a decrease in choline levels is detected in testicular malignancies (Figu 2e) [44].

Volumetric ADC Histogram Analysis
Volumetric ADC histogram analysis is a quantitative method that describes the st tistical information contained in an image. Most histogram analyses use descriptive p rameters-including mean ADC; standard deviation; mode, maximum, and minimu ADC; percentiles of ADC; kurtosis; skewness; and entropy-representing the first-ord statistical properties of the image. This technique uses a volume of interest (VOI) cove ing the entire lesion, which assesses diffusion properties and heterogeneity [72,73 Volumetric ADC histogram analysis performs better than mean ADC, which is based o an ROI placed on the most representative part of the lesion and, therefore, not reflectin histologic heterogeneity, and possibly resulting in measurement sampling errors an subjective bias [72,73]. Changes in histogram shape and asymmetry reflect microstru tural and functional differences in tumor composition, and are correlated with the hi tological grade, TNM, tumor aggressiveness, and therapeutic response [72,73].

Proton Magnetic Resonance Spectroscopy
1H-MRS provides information about the biochemical milieu of normal tissues and pathological entities [69,70]. The prominent metabolic peaks of the 1H-MR spectrum of normal adult testes include choline, creatine, myo-inositol, and lipids (Figure 1f) [71]. Based on preliminary data, a decrease in choline levels is detected in testicular malignancies (Figure 2e) [44].

Volumetric ADC Histogram Analysis
Volumetric ADC histogram analysis is a quantitative method that describes the statistical information contained in an image. Most histogram analyses use descriptive parameters-including mean ADC; standard deviation; mode, maximum, and minimum ADC; percentiles of ADC; kurtosis; skewness; and entropy-representing the first-order statistical properties of the image. This technique uses a volume of interest (VOI) covering the entire lesion, which assesses diffusion properties and heterogeneity [72,73]. Volumetric ADC histogram analysis performs better than mean ADC, which is based on an ROI placed on the most representative part of the lesion and, therefore, not reflecting histologic heterogeneity, and possibly resulting in measurement sampling errors and subjective bias [72,73]. Changes in histogram shape and asymmetry reflect microstructural and functional differences in tumor composition, and are correlated with the histological grade, TNM, tumor aggressiveness, and therapeutic response [72,73].
In a recently published retrospective study of 61 histologically proven testicular lesions, the minimum ADC and the 10th percentile of ADC were reduced in testicular tumors compared to benign testicular lesions. The minimum ADC had the highest diagnostic performance in characterizing testicular lesions, with 81.4% sensitivity and 77.78% specificity [45]. The lower percentiles of ADC better represent the highest cellular areas within tumors, explaining their diagnostic superiority in the characterization of malignancies (Table 5) [72,73].

MRI-Based Radiomics
Radiomics is an emerging quantitative approach to imaging that aims to provide information beyond that which can be perceived from human imaging interpretation alone, by means of advanced mathematical analysis [74,75]. Texture analysis represents a technique used for the quantification of tissue heterogeneity. Histogram analysis features and intra-perinodular textural transition analysis features based on manual segmentation of testicular lesions on T2WI proved helpful in the differential diagnosis between benign and malignant testicular lesions (Table 6) [47].
Conventional MRI-including T1WI, T2WI, and contrast-enhanced T1WI-had 100% sensitivity and 87.5% specificity in the differentiation between benign and malignant testicular masses, in a prospective study of 36 testicular lesions [76]. A multicenter retrospective review identified 34 patients (1.4%) examined with scrotal MRI following indeterminate clinical and US examination. MRI was accurate in the characterization of scrotal lesions in 91% of cases, and helped to improve the treatment provided by general urologists and urological oncologists in 56% and 50% of patients, respectively [36] (Table 7).

MRI Findings of TGCTs
TGCTs typical appear isointense on T1WI compared to normal testes. Areas of hemorrhage may be seen as intratumoral hyperintense foci on T1WI. These neoplasms are either detected with a low T2 signal, or are heterogeneous, with variable signal, on T2WI [7,19,24,[76][77][78]. Testicular tumors often cause restricted diffusion and enhance heterogeneously after the administration of gadolinium, with a type III curve (Figures 2  and 4-6) [52,53,62]. Areas of intratumoral necrosis are hyperintense on T2WI, with a lack of enhancement (Figure 6d). Tumor extension to the testicular tunicae, paratesticular space (Figure 7), and/or the spermatic cord confirms the diagnosis of malignancy [7,19,24,[76][77][78].   (a) (b) Figure 7. Right testicular seminoma in a 27-year-old man: (a) Coronal T2WI and (b) sagittal contrast-enhanced T1WI images depict a large, heterogeneous tumor, replacing the right testis and extending into the paratesticular space (arrow). Large seminomas may have areas of necrosis, which appear hyperintense on T2WI, not enhancing after gadolinium administration.

Epidermoid Cysts
Testicular epidermoid cysts (ECs) represent a subtype of teratoma of the prepubertal type, included in the group of TGCTs unrelated to GCNIS [39]. The ability of preoperative imaging to suggest the diagnosis of an EC may prompt a conservative treatment, including TSS, instead of radical orchiectomy [8,[79][80][81][82][83].
Histologically, EC is a squamous epithelium-lined unilocular cystic lesion, composed of concentric layers of laminated keratinous material, which are responsible for its characteristic imaging appearance [79][80][81][82][83]. MRI may provide supportive findings for the diagnosis of EC, and is recommended as a complimentary imaging examination in cases of equivocal US findings [14].

Epidermoid Cysts
Testicular epidermoid cysts (ECs) represent a subtype of teratoma of the prepubertal type, included in the group of TGCTs unrelated to GCNIS [39]. The ability of preoperative imaging to suggest the diagnosis of an EC may prompt a conservative treatment, including TSS, instead of radical orchiectomy [8,[79][80][81][82][83].
Histologically, EC is a squamous epithelium-lined unilocular cystic lesion, composed of concentric layers of laminated keratinous material, which are responsible for its characteristic imaging appearance [79][80][81][82][83]. MRI may provide supportive findings for the diagnosis of EC, and is recommended as a complimentary imaging examination in cases of equivocal US findings [14].
Typically, an EC is an oval encapsulated lesion surrounded by a hypointense halo on both T1WI and T2WI, which does not enhance after the administration of the contrast material. Alternating zones of high and low signal intensity on T2WI ("onion skin sign"), or a hyperintense central area ("target appearance")-the latter corresponding to the paradox T1 hyperintensity of calcium depositions-are imaging findings suggestive of the diagnosis [79][80][81][82][83].

Testicular Lymphoma
Although uncommon, primary testicular lymphoma is the most common bilateral testicular neoplasm. It is an aggressive malignancy, and may infiltrate the epididymis, spermatic cord, and/or scrotal skin. Most testicular lymphomas are diffuse large B-cell lymphomas [18,28,41,56,84,85].
On MRI, these tumors are often hypointense on T2WI, enhance strongly and heterogeneously, and cause significant diffusion restriction (Figure 8) [56,85]. Ipsilateral lymphomatous epididymal involvement and contralateral testis infiltration may be better depicted on MRI.
(a) (b) Figure 8. Right primary diffuse large B-cell testicular lymphoma in a 52-year-old man: Coronal (a) T2WI and (b) post-contrast T1WI images depict a multinodular right testicular mass. The tumor is mainly homogeneous, and has a low T2 signal, enhancing inhomogeneously after the administration of gadolinium. A moderate hydrocele (arrow) is also seen ipsilaterally.

Characterization of Testicular Tumors: Germ-Cell Versus Sex-Cord Stromal Testicular Tumors
In recent years, the widespread use of scrotal US has resulted in a high incidence of asymptomatic, impalpable, small, and solid testicular tumors, detected as incidental findings. The prevalence of benign histology is approximately 80% in these cases, and LCT represents the most common pathology [86]. TSS for small US-detected, non-palpable intratesticular lesions is highly recommended to obtain a histological diagnosis. When a non-germ-cell tumor is suggested by frozen section examination (FSE), orchiectomy may be avoided [8,87]. Active surveillance through clinical and radiological follow-up, once the diagnosis is strongly suggested based on imaging findings, represents another alternative option for small LCTs [87].
Although no reliable imaging features yet exist, mpMRI may help in the characterization of LCTs, and especially in their differentiation from testicular seminomas

Characterization of Testicular Tumors: Germ-Cell versus Sex-Cord Stromal Testicular Tumors
In recent years, the widespread use of scrotal US has resulted in a high incidence of asymptomatic, impalpable, small, and solid testicular tumors, detected as incidental findings. The prevalence of benign histology is approximately 80% in these cases, and LCT represents the most common pathology [86]. TSS for small US-detected, non-palpable intratesticular lesions is highly recommended to obtain a histological diagnosis. When a non-germ-cell tumor is suggested by frozen section examination (FSE), orchiectomy may be avoided [8,87]. Active surveillance through clinical and radiological follow-up, once the diagnosis is strongly suggested based on imaging findings, represents another alternative option for small LCTs [87].
Based on TSI curves, quantitative DCE data can also be calculated, including the volume transfer constant, rate constant, extravascular extracellular space volume fraction, and initial area under the curve in the first 60 s, providing valuable information on tumor angiogenesis [58]. Increases in the volume transfer constant, rate constant, and initial area under the curve have been reported in LCTs, confirming a typical pattern of enhancement for these neoplasms, which is an avid and rapid wash-in [90].
pointense on T2WI, with strong, early, homogeneous enhancement, followed by slow washout (Figure 9) [63,[88][89][90][91]. Additional T2WI characteristics include the presence of capsular high signal intensity and a hyperintense central scar [88]. Ill-defined margins, mild T2 lesion hypointensity, mild T1 hyperintensity, and weak and gradual enhancement, without de-enhancement, is more often seen in seminomas [63,[88][89][90][91]. Differences in semi-quantitative DCE-MRI parameters were found between LCTs and seminomas, with an increase in peak enhancement and wash-in rate, along with a decrease in time to peak, observed in LCTs [90]. Based on TSI curves, quantitative DCE data can also be calculated, including the volume transfer constant, rate constant, extravascular extracellular space volume fraction, and initial area under the curve in the first 60 s, providing valuable information on tumor angiogenesis [58]. Increases in the volume transfer constant, rate constant, and initial area under the curve have been reported in LCTs, confirming a typical pattern of enhancement for these neoplasms, which is an avid and rapid wash-in [90].
DCE-MRI parameters proved useful in the differentiation between benign testicular stromal tumors, malignant neoplasms, and BOTs in a study of 31 impalpable, incidentally discovered testicular tumors on sonography. Specifically, benign stromal tumors had a higher maximal relative enhancement, shorter time to peak, higher initial slope, and higher transfer constants when compared to malignancies and BOTs [63]. Multiparametric MRI findings in a series of 10 BOTs, discovered in asymptomatic infertile DCE-MRI parameters proved useful in the differentiation between benign testicular stromal tumors, malignant neoplasms, and BOTs in a study of 31 impalpable, incidentally discovered testicular tumors on sonography. Specifically, benign stromal tumors had a higher maximal relative enhancement, shorter time to peak, higher initial slope, and higher transfer constants when compared to malignancies and BOTs [63]. Multiparametric MRI findings in a series of 10 BOTs, discovered in asymptomatic infertile men, were as follows: well-defined intratesticular nodule, with low T2 signal, high ADC, and lack of enhancement [92].

First Author (Year) Study Aims MRI Scanner b Values (s/mm 2 ) Main Findings Study Limitations
Feliciani et al. [49] To assess the ability of MRI-based radiomics to differentiate between testicular germ-cell and non-germ-cell tumors

Sertoli-Cell Tumors
SCTs are less common than LCTs, accounting for about 1% of all testicular neoplasms. These neoplasms are rarely malignant [40]. TSS may be recommended, with subsequent radical orchiectomy if histology proves malignant [8].

Histological Characterization of TGCTs
Radical inguinal orchiectomy with removal of the entire testis, containing the tumor along with the spermatic cord, to the level of the internal inguinal ring, is the preferred treatment for all testicular tumors, and should be performed within a week of initial diagnosis, unless clinical indications require immediate chemotherapy [8].
In patients presenting with disseminated disease and/or life-threatening metastases, orchiectomy may be delayed. The first-line treatment for metastatic TGCTs mainly depends on the histology of the primary tumor [8]. MRI is recommended to differentiate between seminomas and NSGCTs in these cases [14]. MRI features have been found to closely correlate with the histopathological characteristics of TGCTs, with an accuracy of up to 91% [7,17,19,20,22,24,46,48,64,77,78].
Testicular seminoma is typically seen as a multinodular tumor, mainly homogeneous, with a low T2 signal. Intratumoral hypointense bands on T2WI are often detected, which enhance more than the remaining tumor, and correspond to fibrovascular septa in terms of pathology (Figures 2 and 5) [77,78]. Conversely, nonseminomas often appear markedly heterogeneous on both T1WI and T2WI, with inhomogeneous contrast enhancement (Figures 4 and 6)-findings that are mainly associated with the presence of hemorrhage and/or necrosis histologically. A hypointense T2 rim surrounding the tumor is more commonly seen on nonseminomas, corresponding to a fibrous capsule in terms of histology ( Figure 6) [77,78]. Moreover, seminomas usually cause significant diffusion restriction compared to nonseminomas (Figures 2 and 4-6)-a finding that is related to the presence of densely packed uniform cells, with abundant cytoplasm and large nuclei, surrounded by fibrous septa containing eosinophils and lymphocytes in seminomatous tumors. An ADC cutoff of 0.68 × 10 −3 mm 2 /s is reliable for the histological characterization of TGCTs [56,64].
Whole-tumor ADC histogram analysis may also help in the characterization of TGCTs. Decreases in the median, mean, minimum, maximum, and percentiles of ADC, and increases in kurtosis and skewness, have been observed in testicular seminomas compared to nonseminomas, reflecting the compact, uniform histological composition of seminomatous tumors, in contrast to the heterogeneous, cystic appearance of NSGCTs [46]. Among volumetric ADC histogram parameters, the 10th percentile of ADC yielded the highest diagnostic performance, with 100% sensitivity and 92.86% specificity [46]. Recently, a T2-based radiomics signature had 90% sensitivity and 100% specificity in the histological differentiation of TGCTs (Table 9) [48]. Table 9. Diagnostic performance of MRI in the differential diagnosis between seminomas and nonseminomas (ADC: apparent diffusion coefficient; T2WI: T2-weighted imaging; TPR: true positive rate; DCE: dynamic contrast enhancement).

Local Staging of TGCTs
In patients with TGCTs, radical orchiectomy is the treatment of choice, as pathological studies report multifocal and/or adjacent GCNIS in 20-30% of cases [8,94,95]. TSS, when feasible, can be attempted in special circumstances, including synchronous bilateral testicular tumors and tumors in a solitary testis [8]. TSS, in properly selected cases, has the advantage of better preserving endocrine function and fertility, with a reduced impact on sexual and psychological aspects, and satisfactory oncological outcomes [8,96]. TSS should be followed by FSE, which has been proven to be highly accurate and consistent with the final histological diagnosis [97,98]. However, in cases of discordance between FSE and final histopathology, delayed orchiectomy is justified [8].
Accurate preoperative knowledge of the local extent of TGCTs is important in patients who are candidates for TSS. MRI is recommended for local staging of TGCTs in these cases [14]. The technique performs well in the preoperative evaluation of the local staging of TGCTs. Accurate information regarding tumor dimensions and possible invasion of the rete testis, testicular tunicae, paratesticular space, and/or spermatic cord is assessed by MRI. Tumor pseudocapsules detected as hypointense halos surrounding lesions on T2WI have been described as a feature facilitating enucleation [14,17,22,24,25,28,32,35,76].
Regarding the local staging of the primary tumor (pT) in testicular malignancies, the 2016 TNM classification of the International Union Against Cancer includes the following stages: pT1-tumor limited to the testis and epididymis, without vascular/lymphatic invasion; tumor may invade the tunica albuginea, but not the tunica vaginalis; pT2-tumor limited to the testis and epididymis, with vascular/lymphatic invasion or tumor extending through tunica albuginea, with involvement of the tunica vaginalis; pT3-tumor invades the spermatic cord, with or without vascular/lymphatic invasion; pT4-tumor invades the scrotum, with or without vascular/lymphatic invasion [8,99]. MRI criteria for the evaluation of the local extent of testicular tumors are as follows: pT1-intratesticular tumor surrounded by a rim of normal testicular parenchyma and/or intact testicular tunicae, which are detected as a continuous hypointense rim in the periphery of the testis; pT2-intratesticular tumor invading the testicular tunicae, detected as a disruption of the T2 hypointense peripheral halo, with or without the presence of a paratesticular mass; pT3-enlargement, heterogeneity, and enhancement of the spermatic cord, due to neoplastic invasion; pT4-neoplastic invasion of the scrotal wall [76]. The rate of correspondence between MRI and histological diagnosis in local staging of TGCTs has been reported to be 92.8% (Table 10) [76].

Conclusions
Multiparametric MRI of the scrotum has emerged as a valuable, second-line diagnostic tool for imaging of testicular tumors. This technique is recommended for the characterization of testicular masses when conventional US findings are indeterminate. MRI is accurate in the differentiation between benign and malignant testicular masses, helping to avoid unnecessary radical surgical procedures. Scrotal MRI estimates the local extent of testicular carcinomas, allowing for surgical planning in candidates for testis-sparing surgery. MRI features are closely correlated with the histopathological characteristics of testicular germcell tumors. Therefore, this technique is recommended in the rare cases presenting with disseminated disease and/or life-threatening metastases, where chemotherapy represents the primary treatment, to differentiate between seminomas and nonseminomatous germcell tumors. Although still under investigation, multiparametric MRI of the scrotum may provide valuable information in differentiating between germ-cell and sex-cord stromal tumors-especially in cases of small, impalpable, incidentally detected testicular tumors.
Functional MRI techniques-including diffusion tensor imaging, magnetization transfer imaging, proton MR spectroscopy, volumetric ADC histogram analysis, and MRI-based radiomics-may provide new insights in the understanding of testicular tumors in the future.