Ultrasound features for pediatric testicular lesions: 20 years of ultrasound practice interpreted through the 2022 WHO Classification Authors: Meng Zhu, Huan Xu, Yidong Huang, Hualin Yan, Xiangfeng Zeng, Buyun Ma, Juxian Liu Published: 2026 ABSTRACT AbstractBackgroundTesticular lesions are rare in pediatric populations. The World Health Organization standardized the classification of testicular tumors for urological and male genital tumors in 2022. Ultrasound is highly sensitive in detecting testicular lesions in children; however, the ultrasound features that distinguish malignant from benign lesions remain unclear. We aimed to identify ultrasound features that differentiate benign and malignant testicular lesions in pediatric patients. In addition, we evaluated the role of ultrasound in pediatric testicular tumors based on the new guidelines.MethodsWe retrospectively reviewed the clinical data and ultrasound images of 84 testicular lesions in 79 pediatric patients aged 0–18 years (≤216 months), who underwent testicular sonography between April 2004 and October 2023 at West China Hospital. Prepuberty was defined as 0–12 years of age. Univariate and multivariate logistic regression analyses were conducted to compare ultrasound features between benign and malignant tumors across all 84 lesions, 66 primary tumors, and 65 prepubertal lesions. Continuous variables were compared between groups using the Mann-Whitney U test, and Fisher’s exact test was used for categorical variables.ResultsUnivariate analysis showed that among the 84 total lesions, the malignant group exhibited a more diffuse lesion range compared with the benign group (P<0.001). Among 84 total lesions or 66 primary tumor lesions, the malignant group exhibited more hypoechoic and isoechoic echoes, unclear margins, absence of calcification, and increased vascularity (P<0.001). Furthermore, in 66 primary tumor lesions, the malignant group also exhibited non-complex echo and the small hypoechoic nodule structures (P<0.001). There were no significant differences in microlithiasis among all groups (all P>0.05). In the multivariate analysis, among the total lesions group or primary tumor group, hypoechoic and isoechoic patterns, as well as absence of calcification, remained important features of malignant manifestations (all P<0.001). Margin circumscription was statistically significant only in the total lesions group (P=0.001), whereas lesion size was statistically significant only in the primary tumor group (P=0.006). In 65 prepubertal lesions, malignant lesions presented more hypoechoic and isoechoic echoes (P<0.01), and clear margins (P=0.034) were associated with a reduced risk of malignancy. Among 84 total lesions, the predictive power of echo pattern was the highest [area under the curve (AUC) =0.815], followed by calcification (AUC =0.795) and lesion size (AUC =0.712). The predictive power of echo patterns and calcification remained good in 66 primary tumors (AUC =0.818 and 0.838, respectively) and 65 prepubertal lesions (AUC =0.842 and 0.820, respectively).ConclusionsThe echo pattern shows the highest predictive value in total lesions and prepubertal lesions, while calcification is more important for primary tumors. Different imaging features may be required to optimize the prediction of benign and malignant testicular lesions.Keywords: Testicular neoplasms, pediatric, ultrasound, diagnosis Introduction Testicular lesions are rare in pediatric populations. In Asia, the incidence of testicular tumors is approximately 4.2 per million in children (1). Testicular lesions include primary testicular tumors, non-tumorous lesions, and secondary tumors. The most common malignant testicular tumor in prepubertal children is the yolk sac tumor, whereas teratoma is the most common benign type (2-4). In postpubertal adolescents, 75% of testicular tumors are malignant, with mixed non-seminomatous germ cell tumors (GCT) being the predominant type (5). The 2016 World Health Organization (WHO) classification of Tumors of the Urinary System and Male Genital Organs introduced significant revisions to the categorization of testicular tumors (6). Specifically, the term “germ cell neoplasia in situ” (GCNIS) was introduced to describe precursor lesions for most GCTs. Additionally, prepubertal yolk sac tumors and teratomas were reclassified as non-GCT. The 2022 WHO classification (5th edition), or “Blue Book”, further standardized terminology (7,8), providing a consensus for pathologists and clinicians. Clinically, most pediatric testicular tumors often manifest as scrotal swelling. Testicular lesions may be accompanied by other manifestations, such as hydrocele, pain, or cryptorchidism (4). However, the preoperative differentiation between benign and malignant lesions remains challenging. Improved understanding of imaging features may enhance diagnostic confidence. Ultrasound offers high sensitivity for detecting painless testicular lesions in children (9). Computed tomography involves ionizing radiation, and magnetic resonance imaging often requires sedation, limiting their use in pediatric patients (10,11). Currently, no large-sample study has applied the 2022 WHO classification to assess the ultrasound features of pediatric testicular lesions in distinguishing between primary benign and malignant tumors. Thus, the ability of ultrasound to differentiate between benign and malignant testicular tumors still requires clarification. We aimed to identify the ultrasound characteristics of malignant pediatric testicular tumors based on cases and experience from our institution and establish the key ultrasound features of prepubertal testicular lesions. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2725/rc). Methods Patients This retrospective study was approved by the Biomedical Ethics Review Committee of West China Hospital of Sichuan University (No. 2022-1887), which waived the informed consent requirement. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. We retrospectively analyzed the consecutive medical records of pediatric patients who underwent testicular ultrasound examination between April 2004 and October 2023. These patients presented to our hospital either with clinical symptoms or following external ultrasound findings suggestive of testicular lesions, for further clarification of diagnosis. The inclusion criteria included preoperative ultrasound examination and surgical pathology confirmation of testicular lesions in our hospital. In total, 126 patients were enrolled, of whom 47 were excluded. Reasons for exclusion included lesions located outside the testicular parenchyma (n=24), negative ultrasound findings (n=11; 10 atrophic testes and 1 cryptorchidism), unclear pathological diagnosis (n=6), and hermaphroditism (n=2). Additional exclusions were cases without identifiable lesions on ultrasound, including underdeveloped testes (n=2), and cryptorchidism with interstitial Leydig cell focal hyperplasia (n=1). One previously published case was also excluded (12). Finally, a total of 79 patients were included, of whom 74 patients had unilateral lesions and 5 had bilateral lesions. Therefore, 84 lesions met the study criteria. From medical records, we reviewed age, clinical symptoms, cryptorchidism condition, sonographic images and features, histopathological diagnosis, and biochemical indicators. Biochemical indicators included alpha-fetoprotein (AFP) and human chorionic gonadotropin (HCG). Blood samples were collected within 1 week before surgery. Classification of testicular tumors We classified the lesions based on the postoperative pathological report. Primary testicular tumors were categorized according to the 2022 WHO classification (7,8). We adopted a similar age cutoff to a previous study, dividing the patients into a prepubertal cohort and a postpubertal cohort using the study’s classification of pediatric testicular tumors (ages 0–12 years were defined as prepubertal, and ages >12 to <19 years as postpubertal) (13). Lesions in this study were grouped as follows: total lesion group, tumor group (including primary and secondary tumors), primary tumor group, and non-tumorous lesion group (14). Malignant testicular tumors included the following: “mixed teratoma and yolk sac tumor, prepubertal type”, “yolk sac tumor, prepubertal type”; all lesions from the GCNIS-derived GCTs group (7); and secondary tumors. All other lesions were classified as benign. Ultrasonic evaluation Ultrasound examinations were performed using GE LOGIQ 7 and E9, Philips (HDI 5000, HD 11, iU 22, etc.), Mindray Resona 7, BK 5000, and Esaote MYLAB 90. A high-frequency linear array probe (frequency 5–18 MHz) was used, while abdominal cryptorchidism was assessed with a convex array probe (3–5 MHz) with abdominal preset. Longitudinal and cross-sectional sections of the testes were routinely collected. Ultrasound evaluation was jointly discussed and completed by two senior ultrasound experts with extensive experience (M.Z. and B.M.). In case of disagreement, the final decision was made by a senior pediatric ultrasound specialist (J.L.). All images were reviewed in a blinded manner. Sonographic features assessed included echo and posterior echo patterns, mass margin, diffuse degree of lesion, lesion size, and presence of small hypoechoic nodule (SHN), microlithiasis, calcification, and hydrocele. The method for assessing the degree of lesion diffusion is shown in Figure S1. SHN was defined as a small hypoechoic focus (1–4 mm) within the lesion. Echo patterns were divided into six categories: anechoic, hypoechoic, isoechoic, hyperechoic, heterogeneous cystic and solid echo, and heterogeneous solid echo. Posterior echo patterns were divided into four categories: shadowing, not changed, enhancement, and not evaluated. In addition, we evaluated the complex echo and the margin to determine whether circumscribed. Complex echo is defined as consisting of at least three different echo types. Diffuse degree was calculated as the ratio of lesion diameter to testicular diameter. SHN, echogenicity, and calcification were judged according to normal testicular parenchyma. Microlithiasis and hydrocele were evaluated according to bilateral testicular contrast examination. Based on Adler’s blood flow grading (15), the vascularity was divided into absent, minimal, moderate, and marked levels from low to high. In our study, a marked vascularity (≥4 vessels) is defined as increased blood flow. Statistical analysis Continuous variables (age, lesion size, biochemical indicators) were represented by mean ± standard deviation (SD), median, and interquartile range, and the Mann-Whitney U test was used for comparison between groups. The categorical variables were represented as number (%), and Fisher’s exact test was used for comparison between groups. Furthermore, multiple logistic regression analysis was used to evaluate the relative risk of malignant tumors under different sonographic features. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the diagnostic efficacy of important ultrasound parameters and report the corresponding area under the ROC curve (AUC). The optimal cut-off values were determined by maximizing the Youden index (sensitivity + specificity − 1). Statistical analysis was performed using R Version 4.2.1 (R Development Core Team, Vienna, Austria). P<0.05 was considered statistically significant. Results Descriptive analysis of clinical and pathological characteristics Among the 79 cases, 79.7% (63/79) presented with a scrotal mass. Additionally, 13 patients had uncertain testicular masses. In addition to scrotal mass, 13 patients exhibited other clinical symptoms, as shown in Table 1, with leukemia (n=4) and neuroblastoma (n=2) being the most common. Approximately 6.3% (5/79) of patients experienced testicular pain. Among the three patients with ipsilateral cryptorchidism, one was a 13-month-old “Teratoma, prepubertal type” patient, one was a 156-month-old “Cystic dysplasia of the rete testis” patient, and one was a 26-month-old epidermoid cyst patient. Table 1Clinical symptoms of patientsClinical symptomsNScrotal mass without other clinical symptoms50Ipsilateral cryptorchidism3Indeterminate scrotal mass13Scrotal mass with other symptoms13 Leukemia4 Prior history of perianal abscess drainage1 History of prior scrotal trauma surgery1 Contralateral oblique hernia1 Contralateral cryptorchidism1 Ipsilateral hydronephrosis1 Congenital heart disease1 Lymphoma1 Neuroblastoma2 Clinical symptoms of patients Table 2 summarizes the characteristics of the 79 patients, detailing 16 pathological patterns. Of these, 65 had primary tumors, including 59 non-GCNIS-derived GCTs, three GCNIS-derived GCTs, two sex cord-stromal tumors, and one tumor of the collecting ducts and rete testis. Among the 65 patients (mean age: 48.1 months), 56 (86.2%) were prepubertal, and nine (13.8%) were postpubertal, with 34 (52.3%) patients having tumors on the right side, 30 (46.2%) on the left, and one (1.5%) bilaterally. In total, eight patients had secondary tumors: four with leukemia, two with lymphoma, and two with neuroblastoma. Of these, five patients were prepubertal, and three were postpubertal, with one patient having a lesion on the right side, four on the left, and three with bilateral lesions. Additionally, six patients had non-neoplastic lesions, including three prepubertal and three postpubertal patients. Table 2Baseline characteristics of 79 patients with testicular lesionsCharacteristicsNPuberty†Age (months)Location†PrepubertyPostpubertyMean (SD)Median [Min, Max]RightLeftBilateralPrimary tumors6556 (86.2%)9 (13.8%)48.1 (64.9)17.0 [2.40, 216]34 (52.3%)30 (46.2%)1 (1.5%) Non-GCNIS-derived GCTs5956 (94.9%)3 (5.1%)33.1 (46.0)17.0 [2.40, 209]32 (54.2%)27 (45.8%)0 Epidermoid cyst43178.8 (88.3)42.0 [22.0, 209]220 Mixed teratoma and yolk sac tumor prepubertal type33010.5 (7.37)6.60 [5.90, 19.0]120 Teratoma prepubertal type3331 (93.9%)2 (6.1%)38.3 (49.2)17.0 [2.40, 157]17 (51.5%)16 (48.5%)0 Yolk sac tumor prepubertal type1919017.9 (19.4)12.0 [5.00, 95.0]12 (63.2%)7 (36.8%)0 GCNIS-derived GCTs303206 (9.45)209 [195, 213]120 Mixed germ cell tumors202211 (2.83)211 [209, 213]020 Teratoma postpubertal type101195 (NA)195 [195, 195]100 Sex cord-stromal tumors‡202189 (38.2)189 [162, 216]011 Large cell calcifying Sertoli cell tumor101216 (NA)216 [216, 216]001 Leydig cell tumor101162 (NA)162 [162, 162]010 Tumors of the collecting ducts and rete testis101180 (NA)180 [180, 180]100 Adenoma101180 (NA)180 [180, 180]100Secondary tumors85 (62.5%)3 (37.5%)117 (54.0)128 [10.0, 171]1 (12.5%)4 (50.0%)3 (37.5%) Leukemia413157 (11.5)157 [143, 171]112 Lymphoma220112 (0.707)112 [111, 112]011 Neuroblastoma22040.5 (43.1)40.5 [10.0, 71.0]020Nonneoplastic lesions63 (50.0%)3 (50.0%)108 (70.9)144 [17.0, 169]1 4 1 Cystic dysplasia of the rete testis101156 (NA)156 [156, 156]010 Inflammatory pseudotumor21187.5 (96.9)87.5 [19.0, 156]101 Leydig cell hyperplasia101169 (NA)169 [169, 169]010 Vascular malformation22074.5 (81.3)74.5 [17.0, 132]020Overall7964 (81.0%)15 (19.0%)59.6 (68.3)19.0 [2.40, 216]36 (45.6%)38 (48.1%)5 (6.3%)†, data are presented as n (%) or n; ‡, the terminology for this category was revised in the 2022 World Health Organization classification. GCNIS, germ cell neoplasia in situ; GCT, germ cell tumor; NA, not available; SD, standard deviation. Baseline characteristics of 79 patients with testicular lesions †, data are presented as n (%) or n; ‡, the terminology for this category was revised in the 2022 World Health Organization classification. GCNIS, germ cell neoplasia in situ; GCT, germ cell tumor; NA, not available; SD, standard deviation. Univariate analysis of biochemical indicator levels Table 3 details the findings regarding available biochemical indicators in patients with malignant and benign tumors. In total, the AFP levels of 50 patients were available. Thirteen patients with yolk sac tumor had AFP level >1,210 ng/mL, whereas the remaining 37 patients had AFP level <1,210 ng/mL, with a mean AFP of 77.0±246 ng/mL. The mean HCG level of patients (n=27) was 6.64±32.9 mIU/mL. The mean age of patients with teratoma prepubertal type (n=22) was 37.3±46.7 months; all had AFP levels <1,210 ng/mL, and the mean of AFP was 7.30±15.3 ng/mL. However, the mean of AFP of two patients with yolk sac tumor prepubertal type (AFP <1,210 ng/mL) was 241±224 ng/mL. Table 3Available biochemical indicators levels of 50 patients with testicular lesionsPathological patternNAge (months)AFP (N)AFP <1,210 ng/mLHCG (mIU/mL)Mean (SD)Median [Min, Max]<1,210 ng/mL≥1,210 ng/mLMean (SD)Median [Min, Max]Mean (SD)Median [Min, Max]Epidermoid cyst478.8 (88.3)42.0 [22.0, 209]401.96 (0.607)2.03 [1.28, 2.49]0.197 (0.207)0.160 [0.010, 0.42]Large cell calcifying Sertoli cell tumour1Mean: 216216 [216, 216]10Mean: 2.07 2.07 [2.07, 2.07]Mean: 0.030 0.030 [0.030, 0.030]Leukemia2150 (9.19)150 [143, 156]201.63 (1.39)1.63 [0.650, 2.61]0.115 (0.0354)0.115 [0.090, 0.140]Mixed germ cell tumours2211 (2.83)211 [209, 213]20554 (696)554 [61.5, 1,050]85.6 (121)85.6 [0.130, 171]Mixed teratoma and yolk sac tumour, prepubertal type1Mean: 6.606.60 [6.60, 6.60]10Mean: 1,060 1,060 [1,060, 1,060]––Neuroblastoma240.5 (43.1)40.5 [10.0, 71.0]2013.7 (16.5)13.7 [2.03, 25.4]Mean: 0.280 0.280 [0.280, 0.280]Teratoma, prepubertal type2237.3 (46.7)18.0 [5.60, 157]2207.30 (15.3)1.88 [0.610, 69.1]0.187 (0.148)0.150 [0.010, 0.600]Vascular malformation1Mean: 17.017.0 [17.0, 17.0]10Mean: 1.611.61 [1.61, 1.61]––Yolk sac tumour, prepubertal type1512.8 (6.24)11.3 [5.00, 24.0]213241 (224)241 [82.1, 399]0.890 (1.63)0.180 [0.110, 3.81]Overall5047.4 (64.1)18.0 [5.00, 216]37 (74.0%)13 (26.0%)77.0 (246)2.07 [0.610, 1,060]6.64 (32.9)0.150 [0.010, 171]AFP, alpha-fetoprotein; HCG, human chorionic gonadotropin; SD, standard deviation. Available biochemical indicators levels of 50 patients with testicular lesions AFP, alpha-fetoprotein; HCG, human chorionic gonadotropin; SD, standard deviation. Sonographic features of testicular lesions Table 4 presents the conventional ultrasound characteristics of all the testicular lesions. Among total lesions, benign lesions more often showed a smaller diffuse degree (P<0.001) (Figure 1), heterogeneous cystic and solid echo and heterogeneous solid echo (P<0.001), posterior acoustic features (P<0.01), mass formation (P<0.01) and complex echo (P=0.019), and absence of SHN structures (P<0.01). Furthermore, benign lesions more often showed a circumscribed margin (Figure 2) and calcification (P<0.001) (Figure 3). All malignant lesions presented as diffuse lesions >1/2. Malignant tumors more commonly showed hypoechoic (15/36) (P<0.001), isoechoic (15/36) (P<0.001), and no change in posterior echo (24/36) (P<0.01) (Figures 4,5). Meanwhile, more malignant lesions (34/36) lacked a circumscribed margin (P<0.001) and had a larger diameter (mean size 35.7 mm) than benign lesions (P<0.01) (Figure 6). Besides, more malignant lesions presented increased vascularity (94.4% vs. 33.3%, P<0.001). Table 4Univariate analysis of ultrasound features by benignity/malignancy in overall lesions and primary tumorsSonographic featuresTotalPrimary tumorsNonbenign (N=36)Benign (N=48)Overall (N=84)PNonbenign (N=25)Benign (N=41)Overall (N=66)PLesion, n (%)†<0.001<0.01 >1/2 and <110 (27.8)26 (54.2)36 (42.9)7 (28.0)21 (51.2)28 (42.4) ≈126 (72.2)14 (29.2)40 (47.6)18 (72.0)12 (29.3)30 (45.5) ≤1/20 (0)8 (16.7)8 (9.5)0 (0)8 (19.5)8 (12.1)Echo pattern, n (%)<0.001<0.001 Heterogeneous cystic and solid echo4 (11.1)26 (54.2)30 (35.7)4 (16.0)26 (63.4)30 (45.5) Heterogeneous solid echo2 (5.6)14 (29.2)16 (19.0)1 (4.0)11 (26.8)12 (18.2) Isoechoic15 (41.7)3 (6.3)18 (21.4)15 (60.0)1 (2.4)16 (24.2) Hypoechoic15 (41.7)3 (6.3)18 (21.4)5 (20.0)1 (2.4)6 (9.1) Hyperechoic0 (0)1 (2.1)1 (1.2)0 (0)1 (2.4)1 (1.5) Anechoic0 (0)1 (2.1)1 (1.2)0 (0)1 (2.4)1 (1.5)Posterior echo types, n (%)<0.01<0.01 Not evaluated6 (16.7)1 (2.1)7 (8.3)6 (24.0)1 (2.4)7 (10.6) Not changed24 (66.7)26 (54.2)50 (59.5)13 (52.0)20 (48.8)33 (50.0) Enhancement6 (16.7)11 (22.9)17 (20.2)6 (24.0)10 (24.4)16 (24.2) Shadowing0 (0)10 (20.8)10 (11.9)0 (0)10 (24.4)10 (15.2)Ultrasonic phenotype, n (%)<0.010.379 Non mass10 (27.8)2 (4.2)12 (14.3)1 (4.0)0 (0)1 (1.5) Mass26 (72.2)46 (95.8)72 (85.7)24 (96.0)41 (100)65 (98.5)Complex echo, n (%)0.019<0.001 No18 (50.0)11 (22.9)29 (34.5)16 (64.0)8 (19.5)24 (36.4) Yes18 (50.0)37 (77.1)55 (65.5)9(36.0)33 (80.5)42 (63.6)Margin, n (%)<0.001<0.001 Not circumscribed34 (94.4)26 (54.2)60 (71.4)24 (96.0)24 (58.5)48 (72.7) Circumscribed2 (5.6)22 (45.8)24 (28.6)1 (4.0)17 (41.5)18 (27.3)Small hypoechoic nodule, n (%)<0.01<0.001 No20 (55.6)42 (87.5)62 (73.8)9 (36.0)35 (85.4)44 (66.7) Yes16 (44.4)6 (12.5)22 (26.2)16 (64.0)6 (14.6)22 (33.3)Calcification, n (%)<0.001<0.001 No34 (94.4)17 (35.4)51 (60.7)23 (92.0)10 (24.4)33 (50.0) Yes2 (5.6)31 (64.6)33 (39.3)2 (8.0)31 (75.6)33 (50.0)Microlithiasis, n (%)0.7380.701 No31 (86.1)43 (89.6)74 (88.1)23 (92.0)36 (87.8)59 (89.4) Yes5 (13.9)5 (10.4)10 (11.9)2 (8.0)5 (12.2)7 (10.6)Hydrocele, n (%)0.0560.186 No25 (69.4)42 (87.5)67 (79.8)18 (72.0)36 (87.8)54 (81.8) Yes11 (30.6)6 (12.5)17 (20.2)7 (28.0)5 (12.2)12 (18.2)Lesion size<0.010.025 Mean (SD)35.7 (18.8)24.8 (15.1)29.4 (17.6)34.2 (21.2)24.1 (14.6)27.9 (18.0) Median [Min, Max]31.5 [14.0, 112]19.0 [5.00, 63.0]26.5 [5.00, 112]28.0 [14.0, 112]18.0 [5.00, 63.0]23.0 [5.00, 112]Increased vascularity, n (%)<0.001<0.001 Yes34 (94.4)16 (33.3)50 (59.5)24 (96.0)12 (29.3)36 (54.5) No2 (5.6)32 (66.7)34 (40.5)1 (4.0)29 (70.7)30 (45.5)†, longitudinal section: lesion diameter/testis diameter. SD, standard deviation. Univariate analysis of ultrasound features by benignity/malignancy in overall lesions and primary tumors †, longitudinal section: lesion diameter/testis diameter. SD, standard deviation. A 58-month-old child with an epidermoid cyst. Ultrasound longitudinal section shows a clearly defined hypoechoic nodule with a size of approximately 5 mm at the lower pole of the left testicle, showing an onion ring-like feature. A 5-month-old child with “teratoma, prepubertal type”. An anechoic nodule measuring approximately 21 mm × 15 mm is observed during the longitudinal section scan of the right testicle. A 14-month-old child with “teratoma, prepubertal type”. (A) Longitudinal sectional ultrasound shows a defined heterogeneous solid echogenic mass in the right testicle, presenting posterior acoustic shadow and calcification. (B) Several small-to-moderate blood vessels can be seen within the mass. A 143-month-old child with acute lymphoblastic leukemia affecting the right testis. (A) Initial ultrasound shows a hypoechoic area, presenting as a non-mass lesion, with abundant blood flow signals (B). A 10-month-old child with “yolk sac tumor, prepubertal type”. (A) Longitudinal section of the right testis showing a solid isoechoic mass approximately 23 mm in size occupying the entire testicular parenchyma, with indistinct borders and multiple small hypoechoic nodules. (B) Color Doppler suggests rich, branching blood flow. A 213-month-old adolescent with mixed germ cell tumors. (A) Longitudinal section of the left testis showing a cystic-solid mixed mass approximately 67 mm in size, occupying the entire testicular parenchyma. The main part of the tumor presents as isoechoic and contains multiple large anechoic nodules. (B) Multiple short rod-shaped blood flow signals were observed within the tumor. In addition, for 66 primary tumors, most ultrasound variables showed similar distributions and differences between benign and malignant tumors. The appearance of microlithiasis and hydrocele did not significantly differ between groups (all P>0.05). Multiple logistic regression analysis of sonographic features Table 5 presents the results of multiple logistic regression analyses of sonographic features associated with malignancy. Among the 84 lesions, isoechoic echo (P<0.001) and hypoechoic echo (P<0.001) were significantly associated with an increased risk of malignancy, whereas circumscribed margin (P=0.001) and calcification (P<0.001) were significant independent protective factors against malignancy. Among the 66 primary tumors, both isoechoic echo and hypoechoic echogenicity remained significantly associated with an increased risk of malignancy (P<0.001). Calcification (P<0.001) and small lesion size (P=0.006) were strongly associated with a lower risk of malignant tumors. Table 5Multivariate analyses of sonographic features affecting malignant tumorsSonographic featuresTotalPrimary tumorOR (95% CI)POR (95% CI)PEcho pattern Heterogeneous echo† Ref Ref Anechoic0.97(0.57–1.63)0.9010.86(0.58–1.3)0.475 Hypoechoic1.54(1.29–1.85)<0.0011.54(1.26–1.87)<0.001 Isoechoic1.5(1.24–1.82)<0.0011.64(1.39–1.95)<0.001 Hyperechoic0.67(0.39–1.15)0.1470.67(0.44–1.02)0.061Margin circumscribed0.76(0.64–0.88)0.0010.89(0.77–1.03)0.105Small hypoechoic nodule1.19(1.02–1.39)0.0291.08(0.94–1.23)0.264Calcification0.74(0.63–0.87)<0.0010.72(0.63–0.83)<0.001Lesion size1.00(1.00–1.01)0.1021.00(1.00–1.01)0.006Decreased vascularity0.9(0.77–1.07)0.2280.85(0.73–0.98)0.028†, include heterogeneous cystic and solid and heterogeneous solid echo. CI, confidence interval; OR, odds ratio. Multivariate analyses of sonographic features affecting malignant tumors †, include heterogeneous cystic and solid and heterogeneous solid echo. CI, confidence interval; OR, odds ratio. Multivariate analyses of prepubertal testicular lesions Among 65 prepubertal lesions (Figure 7), hypoechoic echo (P<0.01) and isoechoic echo (P<0.01) were significantly associated with an increased risk of malignancy, while circumscribed margin (P=0.034) was associated with reduced risk of malignancy. Multivariate analyses of sonographic features for 65 prepubertal lesions. *, longitudinal section lesion diameter/testis diameter. CI, confidence interval; OR, odds ratio. ROC curve analysis results The ROC curve analysis results for distinguishing benign and malignant testicular tumors are shown in Figure 8. Among all lesions (n=84), echo pattern had the highest predictive power (AUC =0.815), followed by calcifications (AUC =0.795) and lesion size (AUC =0.712), while the predictive power of the margin was relatively low (AUC =0.652). The cut-off value of lesion size was 20.50, and the corresponding specificity and sensitivity were 0.562 and 0.889. In primary tumor lesions (n=66) and prepubertal lesions (n=65), the predictive power of calcification (AUC 0.838 and 0.820, respectively) and echo patterns (AUC =0.818 and 0.842, respectively) remained good. The cut-off values for lesion size remained unchanged in the prepubertal cases, falling to 18.50 in primary tumor lesions. The predictive power of the margin was increased in primary tumor lesions (AUC =0.739). ROC curve for predicting malignant tumors of testicular lesions. (A) ROC curves of 84 total lesions. (B) ROC curves of 66 primary tumors. (C) ROC curves of 65 prepubertal lesions. AUC, area under the curve; ROC, receiver operating characteristic. Discussion We conducted a comprehensive investigation of primary testicular tumors in patients aged 0–18 years, based on the WHO classification, 5th edition (7,8). In our study, non-GCNIS-derived GCTs were the most prevalent pediatric testicular tumors, which is similar to the previous research result (13). GCNIS-derived GCTs are typically malignant. Primary testicular tumors accounted for 78.6% (66/84) of testicular lesions, with 86.2% (56/65) cases occurring in prepubertal children. AFP and HCG are key biomarkers used in clinical diagnosis (16). In our study, AFP levels were markedly higher in children with testicular yolk tumors. Thirteen patients with “yolk sac tumor, prepubertal type” had AFP levels exceeding 1,210 ng/mL. One case of mixed teratoma and yolk sac tumor, prepubertal type, showed AFP levels of 1,060 ng/mL. A prolonged AFP half-life is considered an independent risk factor for negative outcomes in patients with yolk sac tumors following radical orchiectomy (17). Alternatively, for postpubertal patients, HCG and lactate dehydrogenase are more relevant biomarkers. Elevated HCG levels are typically seen in choriocarcinoma, whereas increased lactate dehydrogenase levels can indicate inflammation as well as suggest a high tumor burden (5). Most GCTs not derived from GCNIS represent the most common prepubertal testicular lesions. The Children’s Oncology Group staging system and its management principles are commonly used (5,18). For prepubertal testicular tumors, localized tumors with negative markers are considered treated with partial testicular resection (19). If imaging suggests a likely benign diagnosis, combined with negative markers, local resection that preserves residual testicular tissue can be considered, which may be beneficial for the child’s growth and development. The results of this study highlight the high value of ultrasound in diagnosing pediatric testicular lesions. Benign testicular tumors tend to present with clear, well-defined borders (20) and minimal vascularization (20-22). In our study, in addition to the above features, we also found that benign lesions showed a smaller diffuse degree, heterogeneous echo pattern, posterior shadowing, mass formation, complex echoes, and absence of SHN structures. A systematic review by Ager et al. further supports these findings, showing that heterogeneous echotexture, hyperechoic lesions, and peripheral Doppler flow correlate with a lower malignancy risk (22). In the total lesion group, the echo patterns of benign lesions were predominantly heterogeneous, including heterogeneous cystic-solid echoes (54.2%) and heterogeneous solid echoes (29.2%), consistent with Ager et al., although their review only included three pediatric studies. Additionally, we observed that calcification is a more important distinguishing feature. In a previous study, Yu et al. compared ultrasound features of yolk sac tumors with other benign testicular lesions and found that yolk sac tumors had a higher proportion of hypoechogenicity, stronger blood flow signals, and a larger median diameter (25.0 vs. 14.2 mm) (23). Unlike their study, our analysis included other pathological types of lesions, and interestingly, malignant lesions also had similar ultrasound findings. Notably, a significant proportion of malignant masses were isoechoic, with unclear boundaries, more diffusely extensive, and appeared non-mass-like. For total lesions or primary tumors, the median size of malignant lesions was larger than that of benign lesions (31.5 vs. 19.0 mm, P<0.01; 28.0 vs. 18.0 mm, P=0.025) in this study. A lesion size smaller than 0.5 cm was associated with a significantly reduced risk of malignancy (22). It is worth noting that among total lesions, 44.4% (16/36) malignant tumors exhibited SHN on ultrasound, and for primary testicular tumors, this proportion increased to 64.0% (16/25). Multivariate logistic regression analysis identified several key ultrasound features that distinguish between benign and malignant lesions. Echo patterns showed significant statistical significance in the total lesion group, primary tumor group, and prepubertal lesion group (all P<0.01). Calcification and lesion size showed statistical significance in the primary tumor group (P<0.001; P=0.006). Circumscribed margin showed statistical significance in the total lesion group and prepubertal lesion group (all P<0.05). The predictive ability of these factors alone or in combination across diverse populations requires further validation. ROC curve analysis showed that the echo pattern had the highest predictive value in the total lesion group and the prepubertal lesion group (AUC =0.815; AUC =0.842). Calcification demonstrated the highest predictive value in primary tumors (AUC =0.838). Notably, in primary tumors, the predictive ability of lesion size decreased. These results suggest that populations with different clinical characteristics may require different imaging indicators to optimize the prediction of testicular tumor malignancy. In our study, 19 cases of yolk sac tumors occurred exclusively in prepubertal children. On ultrasound, these tumors typically presented as localized or diffuse solid lesions that were hypoechoic or isoechoic, occasionally containing small cystic areas resulting from necrosis (24,25). Some cases demonstrated higher peak systolic velocities and excessive hypervascularity (24). Teratomas appeared as either solid or cystic-solid masses, with echogenicity varying based on the histological components (26,27). Conversely, testicular epidermoid cysts are typically seen as heterogeneous solid lesions, often described with an “onion skin” appearance on imaging. Testicular lymphoma and leukemia generally manifested as uniformly hypoechoic or multifocal hypoechoic lesions on ultrasound (28,29), with diffuse enlargement of the affected testis, either unilateral or bilateral, along with increased blood flow (29). Sex cord-stromal tumors are rare in pediatric patients. On ultrasound, typical Leydig cell tumors appear as well-circumscribed, solid, hypoechoic masses less than 1 cm in diameter, with increased internal and peripheral vascularity (30). Sertoli cell tumors may present with gynecomastia in prepubertal or peripubertal males (31). Large-cell calcifying Sertoli cell tumor typically feature multiple coarse, amorphous calcifications and increased vascularity on imaging (32). Postoperative follow-up with annual ultrasound after partial testicular resection seems feasible (32,33). Other non-tumorous testicular lesions include inflammatory pseudotumor, cystic dysplasia of the rete testis, and vascular malformations. Inflammatory pseudotumor can mimic malignant tumors on imaging. Cystic dysplasia of the rete testis appears as variably sized anechoic nodules, with some lesions resolving during follow-up (34). Capillary malformations are characterized by moderately echogenic areas with increased internal blood flow (35). In our study, microlithiasis was observed in ten lesions, with no significant difference in incidence between groups. A multicenter study reported that boys with testicular microlithiasis were approximately 22 times more likely to develop malignant GCTs compared to those without (36). Multiparametric ultrasound has been successfully applied in adult testicular tumor diagnosis (37) and assessment of spermatogenesis (38,39). Contrast-enhanced ultrasound can reveal enhancement patterns and perfusion characteristics within lesions (37), but its application in prepubertal children remains limited. A previous study detected an increase in testicular hardness in children with hematological malignancies through shear wave elastography, which may indicate testicular involvement (29). This study had certain limitations. First, it was a single-center retrospective study, which may not encompass the full spectrum of pediatric testicular tumors. Second, patients with testicular torsion and acute inflammation were excluded, as these conditions typically present with a clinical history and pain. Third, the generalizability of our conclusions to other Asian populations remains uncertain, warranting future multicenter studies for validation. Fourth, ultrasound elasticity and contrast examination of children’s testicles were not included in this study and require further research. Fifth, we used a consistent reading method when analyzing ultrasound images, which may limit the ability to quantify interobserver variability. Conclusions Following the 5th edition of the WHO classification guidelines, this study provides a comprehensive assessment of primary testicular tumors in patients aged 0–18 years, showing that non-GCNIS-derived GCTs were the most prevalent. Analysis of ultrasound characteristics revealed that the echo pattern had the highest predictive value for malignant lesions in total and prepubertal cases, while calcification was more predictive in primary tumors. Notably, in primary tumors, the predictive power of lesion size decreased, while the predictive power of margin increased. These results suggest that tailored ultrasound features may be required to optimize malignancy prediction in different pediatric populations. Supplementary The article’s supplementary files as Acknowledgments We would like to express our gratitude to Zhishuang Li from Fudan University for support in statistical methods. Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Biomedical Ethics Review Committee of West China Hospital of Sichuan University (No. 2022-1887), and individual consent for this retrospective analysis was waived. Footnotes Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2725/rc Funding: None. Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2725/coif). The authors have no conflicts of interest to declare. Data Sharing Statement Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2725/dss References Associated Data Supplementary Materials The article’s supplementary files as Data Availability Statement Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2725/dss