Testicular mass frozen section examination: Pathological insights and diagnostic accuracy

Abstract

Background and Objectives

Testicular frozen section examination on excisional biopsy (FSEB) is an underused pathological and surgical approach, considering the increasing number of small benign testicular lesions found on radical orchidectomy specimens. This study aims to determine the diagnostic accuracy of FSEB and to provide a pathological summary of the most frequent diagnoses and pitfalls.

Methods

We report the pathological findings and definitive outcome of 135 FSEB for small testicular masses performed between 2005 and 2024 in a single institute.

Results

The median tumour size was 0.9 cm (Interquartile Range [IQR] 0.5–1.3 cm). The most common FSEB diagnoses were Leydig cell hyperplasia/tumour (n = 37; 28%) and seminoma (n = 36; 27%). On FSEB, benign diagnoses represented 58% of cases which allowed us to avoid 81 unnecessary radical orchidectomies. The sensitivity and specificity of FSEB for malignancy were 100% and 96.3%, respectively. Excluding three indeterminate cases on FSEB, the concordance rate was 97.7% (129/132). On definitive assessment, the majority of cases were benign (84/135, 62%) and 51 (38%) cases were malignant. The three indeterminate cases were ultimately confirmed as benign. There were three false‐positive diagnoses of (favoured) malignancy and no false negatives.

Conclusions

FSEB is accurate for patient management of small testicular lesions, allowing us to save young men from unnecessary radical orchidectomy. We provide an in‐depth overview of the most prevalent pathological diagnoses encountered.

Frozen section examination on testicular excisional biopsy is accurate and can help avoid unnecessary radical orchidectomy. A very high concordance rate between frozen section diagnosis and permanent diagnosis was obtained with no false negatives.

Abbreviations

FISH

fluorescent in‐situ hybridization

FSEB

frozen section examination on excisional biopsy

GCT

Germ cell tumours

H&E

haematoxylin and eosin

IQR

Interquartile range

Introduction

Germ cell tumours (GCT) account for approximately 95% of testicular neoplasms and are typically managed with radical orchidectomy, yielding excellent survival outcomes. ¹ , ² , ³ In contrast, sex cord stromal tumours, such as Leydig and Sertoli cell tumours, are relatively rare and generally exhibit indolent clinical behaviour. ⁴ Due to the lack of definitive preoperative criteria to distinguish benign from malignant testicular lesions, radical orchidectomy remains the standard treatment for most testicular masses. ⁵ This approach risks overtreatment of indolent lesions, especially in a young male patient population. ⁶

The increasing detection of incidental, predominantly benign, small testicular masses has been leading to an additional burden for urologists and pathologists. ⁷ , ⁸ In general, preoperative factors, such as medical and family history, physical exam, serum tumour markers and scrotal ultrasound, have high predictive value for the distinction of malignant and benign lesions. ⁹ , ¹⁰ However, histopathology is required for definitive diagnosis and preoperative percutaneous testicular biopsy does not represent a safe option. ¹¹ As such, intraoperative frozen section examination may provide a helpful tool in determining surgical management. This method is most useful in testis‐sparing approaches for small (<2 cm), non‐palpable testicular masses with distinctive ultrasound features including hypoechogenicity, vascularization and microlithiasis and is recommended by the European Association of Urology. ² , ¹² , ¹³ Despite its high sensitivity and specificity, however, testicular frozen section examination on excisional biopsy (FSEB) remains seldom used in most centres. ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷

One reason for its scarce application may be the lack of major post‐orchidectomy negative effects for men with otherwise normal testicular anatomy and function. However, now there is a growing concern regarding quality of life and long‐term effects following radical orchidectomy for these young patients. ¹⁸ , ¹⁹ As a result, further attention has been given to testis‐sparing options to preserve hormonal function, fertility but also self‐esteem and patient satisfaction. ²⁰ , ²¹ , ²² On the other hand, intraoperative FSEB assessment requires the availability of pathologists with knowledge on the spectrum of benign and malignant testicular tumours. ²³ The aim of this study was to determine the accuracy of FSEB for small testicular masses and to provide guidance for reliable assessment and recognition of pitfalls for practicing pathologists.

Methods

Study Cohort

This retrospective observational study included consecutive patients older than 16 years who underwent FSEB for a small testicular mass between 2005 and 2024 in the Erasmus Medical Centre of Expertise for rare urinary tract and testicular tumours in Rotterdam, the Netherlands. All patients had a small testicular lesion on ultrasound. Preoperative semen cryopreservation and testicular prosthesis implantation in case of radical orchiectomy were offered and discussed with all patients. In case of azoospermia, simultaneous testicular sperm extraction (oncoTESE) was performed. This project was approved by the Erasmus MC Medical Ethics committee (MEC‐2025‐0090).

Surgical Technique and Intraoperative Decision‐Making

Open testicular excisional biopsy was carried out by delivery of the testicle through an inguinal incision. Leaving the gubernaculum intact, a tunica vaginalis incision was performed to locate the lesion through palpation and per‐operative ultrasound. After clamping the spermatic cord, the lesion was then enucleated in toto through a small albugineal incision. Per‐operative ultrasound confirmed complete excision of the lesion which was submitted to the Department of Pathology for immediate FSEB.

The clinical algorithm of the therapeutic approach following FSEB assessment is illustrated in Figure 1. Radical orchidectomy was carried out for malignant GCT, while the testis was left in situ in case of a benign diagnosis. If the FSEB diagnosis was indeterminate, the decision whether to proceed with radical orchidectomy was dependent on the level of suspicion of the urologist and pathologist.

Figure 1.

Algorithm of testicular frozen section examination on excisional biopsy. *Non‐tumoral lesions include fibrosis, granulomas, sclerosis, haemorrhage and inflammation, atrophy and calcifications.

Pathological Assessment

After reception and gross description, the FSEB was submitted in toto either after bi‐sectioning or as is, followed by freezing in liquid nitrogen, cutting 5 μm cryostat sections and staining with haematoxylin and eosin (H&E). When available, a uropathologist performed the FSEB diagnosis, preferably with prior notice by the surgeon; if not available, a non‐uropathologist handled the FSEB. After intraoperative diagnosis, tissue was formalin‐fixed, paraffin‐embedded and processed for routine pathological assessment. In case a radical orchidectomy was prompted by the FSEB diagnosis, the specimen was processed and reported in accordance with international consensus recommendations. ²⁴ , ²⁵ Definitive diagnostic evaluation included the use of appropriate immunohistochemistry and chromosome 12p fluorescent in‐situ hybridization (FISH) in selected cases.

Statistical Analysis

Clinical and pathological data were retrospectively collected, and descriptive statistics are provided. For study purposes, intraoperative diagnostic terminology ‘favour malignant’ and ‘malignant’ were grouped as malignant, and ‘favour benign’ and ‘benign’ as benign, since clinical consequences were the same for both categories. Fisher's exact test was used to compare the proportion of indeterminate and false‐positive diagnoses (SPSS version 28). A P‐value of <0.05 was considered statistically significant.

Results

Patient Characteristics

Over a 19‐year study period, 135 FSEB in 127 patients were performed. The median age at the time of FSEB was 32.0 years (Interquartile Range [IQR] 26.5–37.5 years). The grossly measured median specimen size was 1.2 cm (IQR 0.9–1.6 cm) and the median tumour size was 0.9 cm (IQR 0.5–1.3 cm). Sixty‐two (46%) FSEB were derived from the right testis and 73 (54%) from the left. Most patients underwent a single FSEB (119/127, 94%), and two FSEB were performed on eight (6%) patients; once on both sides for six patients and twice on one side for two patients. Seventeen (13.4%) patients presented with a mass in a solitary testis following previous orchidectomy of the contralateral testis and seven (5.5%) men were undergoing simultaneous contralateral radical orchidectomy for GCT within the same operation.

FSEB Outcomes

FSEB diagnoses were benign/favour benign in 78 (57.8%) cases, malignant/favour malignant in 54 (40.0%) and indeterminate in 3 (2.2%) cases. The most common benign diagnoses (Table 1; Figure 2) were sex cord tumours (N = 47; 34.8%), of which Leydig cell tumour/hyperplasia (N = 37; 27.4%) was most prevalent, followed by non‐tumoral reactive changes (N = 16; 11.9%). Invasive GCT (N = 48; 35.6%), in particular seminoma (N = 36; 26.7%), was the most common malignant diagnosis. Sclerosis suggestive of regressed GCT was diagnosed in three cases (2.2%). Fourteen out of 17 (82.4%) men with a previous radical orchidectomy were diagnosed with malignancy on FSEB; the remaining three men had Leydig cell tumours. FSEB was handled by a uropathologist in 122/135 (90.4%) and by a general pathologist in 13 (9.6%) cases.

Table 1.

FSEB diagnoses and clinical characteristics

FSEB diagnosis	Number (n, %)	Age (median, range)	Tumour size (median, range)	Reported by uropathologist (n, %)	Orchidectomy (n, %)
Benign/favour benign
Leydig cell tumour/hyperplasia	37 (27)	34 (21–59)	0.6 (0.3–2)	34 (92)	0 (0)
Non‐tumoral findings a	16 (12)	35 (18–79)	1.1 (0.4–4)	15 (94)	0 (0)
Epidermoid cyst	9 (7)	21 (19–39)	1.6 (0.6–2)	8 (88)	0 (0)
Sertoli cell nodule/hyperplasia	7 (5)	27 (21–42)	0.9 (0.5–1.5)	6 (86)	1 (14) b
Adenomatoid tumour	3 (2)	26 (22–33)	1.2 (0.7–2)	2 (67)	0 (0)
Sex cord stromal tumour NOS	3 (2)	35 (20–39)	0.9 (0.8–3)	3 (100)	0 (0)
Spindle cell tumour	1 (1)	53	0.9	1 (100)	0 (0)
Unclassified benign tumour	2 (1)	28 (27–29)	1.4 (1.3–1.5)	2 (100)	0 (0)
Indeterminate/non‐conclusive
Unclassified tumour, cannot exclude GCT c	3 (2)	44 (37–45)	0.6 (0.3–0.9)	1 (33)	0 (0)
Malignant/favour malignant
Seminoma	36 (27)	32 (17–54)	1.0 (0.3–3.0)	34 (94)	36 (100)
Mature teratoma	3 (2)	25 (22–48)	1.4 (1.2–1.7)	3 (100)	2 (67) d
Embryonal carcinoma	2 (1)	36 (35–37)	1.0 (0.5–1.4)	1 (50)	2 (100)
GCT NOS	7 (5)	32 (28–41)	1.4 (0.8–2.5)	6 (86)	7 (100)
Suggestive of regressed GCT	3 (2)	36 (28–53)	1.6 (0.9–1.6)	3 (100)	3 (100)
GCNIS e	2 (1)	29 (27–31)	0.4 (0.3–0.5)	2 (100)	1 (50) f
Necrosis, cannot exclude GCNIS	1 (1)	43	0.6	1 (100)	1 (100)

Abbreviations: GCNIS, germ cell neoplasia in situ; GCT, germ cell tumour.

^a Non‐tumoral changes include fibrosis, granulomas, sclerosis, haemorrhage and inflammation, atrophy and calcifications.

^b Case where the lesion had a stellate and irregular shape which was clinically considered to be at risk for malignancy despite the Sertoli cell tumour FSEB diagnosis.

^c Permanent diagnosis was of Leydig cell hyperplasia/tumour for all 3 cases.

^d One case of mature teratoma was followed up with yearly imaging for 5 years with no evidence of recurrence or metastasis.

^e Both cases of GCNIS presented with a mass: GCNIS adjacent to a Leydig cell tumour and GCNIS in an ill‐defined mass on ultrasound with a differential of hematoma or GCT.

^f One case of GCNIS was treated by radiotherapy.

Figure 2.

Most common diagnoses on FSEB. (A, B) Leydig cell tumour with sheets of uniform round tumour cells with abundant eosinophilic cytoplasm and prominent nucleoli. Inhibin was strongly positive post‐freezing (A: FSEB H&E 10×; B: permanent H&E and inhibin, 20×). (C–F) Seminoma with dense lymphocytic infiltrate with enlarged nuclei and abundant clear cytoplasm (C: FSEB H&E, 10×; D: permanent H&E, 10×). (E) OCT3/4 was mildly weaker following freezing and showed some cytoplasmic smudging (OCT3/4, 20×). (F) CD117 exhibited a strong membranous staining post‐freezing (CD117, 20×).

Following FSEB, 81 (60.9%) testes remained in situ and 52 (39.1%) orchidectomies were performed. Orchidectomy was carried out in one benign case of clinically suspicious Sertoli cell tumour (Table 1). Two cases diagnosed as malignant on FSEB did not undergo radical orchidectomy; one case of germ cell neoplasia in situ (GCNIS) underwent radiotherapy and one case of mature teratoma was followed up with imaging according to the patient's preferences.

FSEB Concordance with Definitive Pathology

Excluding the three indeterminate FSEB cases, diagnostic concordance for malignancy between FSEB and definitive diagnosis was achieved in 129/132 cases (97.7%) (Table 2). The sensitivity and specificity of FSEB for malignancy were 100% and 96.3%, respectively. The positive predictive value was 94.4% and the negative predictive value was 100%. There were no false‐negative cases. The three false‐positive cases showed (i) seminiferous tubules with necrosis suspicious for GCNIS or intratubular embryonal carcinoma later revealed to be associated with recent testicular torsion, (ii) mature teratoma in an adult that later revealed to lack chromosome 12p alterations finally diagnosed as a pre‐pubertal mature teratoma (Material S1) and (iii) FSEB suspicious for GCT handled by a non‐uropathologist that revealed to be a Leydig cell tumour (Figure 3, Material S2). Two of three FSEB indeterminate diagnoses deferred to permanent section were diagnosed by non‐uropathologists and included a differential of benign and malignant tumours; all later revealed to be benign Leydig cell tumour/hyperplasia (Figure 3), none of which underwent radical orchidectomy. The sensitivity and specificity of FSEB for malignancy for the uropathologist cases was 100% and 97.3%, respectively. The sensitivity and specificity of FSEB for malignancy of the cases handled by non‐uropathologists were 100% and 87.5%, respectively (Table 2). The proportion of indeterminate and false‐positive diagnoses for uropathologists (2.5%) and non‐uropathologists (23.1%) was statistically significant (Fisher's exact test, P = 0.012).

Table 2.

Concordance between FSEB and definitive diagnosis for total cohort, and separated between uropathologist and non‐uropathologist

	Total	Definitive diagnosis
		Benign	Malignant
FSEB diagnosis	Benign/favour benign	78	0
	Indeterminate	3	0
	Malignant/favour malignant	3	51

	Uropathologist	Definitive diagnosis
		Benign	Malignant
FSEB diagnosis	Benign/favour benign	71	0
	Indeterminate	1	0
	Malignant/favour malignant	2	48

	Non‐uropathologist	Definitive diagnosis
		Benign	Malignant
FSEB diagnosis	Benign/favour benign	7	0
	Indeterminate	2	0
	Malignant/favour malignant	1	3

Figure 3.

Examples of discordant and indeterminate cases. (A, B) Discordant case showing an alternating pattern of necrotic and ghost hyalinized atrophic tubules signed out as suggestive of intra‐tubular embryonal carcinoma on FSEB, but was non‐tumoral ischemic change without evidence of GCNIS or GCT (A: FSEB; B: permanent H&E, 2.5×). (C, D) Discordant case showing a diffuse pattern of monotonous cells interpreted as possibly consistent with GCT by a non‐uropathologist finally diagnosed as Leydig cell tumour (Inhibin positive, OCT3/4 negative, not shown) (C: FSEB; D: permanent H&E, 20×). (E, F) Indeterminate case showing moderately pleomorphic eosinophilic cells diagnosed as an unclassifiable tumour on FSEB finally diagnosed as a benign Leydig cell tumour (Inhibin and Calretinin positive, not shown) (E: FSEB; F: permanent H&E, 10×).

Clinicopathology of Challenging FSEB Diagnosis

While FSEB diagnosis and clinical handling was straightforward in the vast majority of patients, some diagnoses led to uncertainty inherent to their specific clinicopathological features: (i) reactive changes and regressed GCT and (ii) benign and malignant mature teratoma.

Reactive changes and regressed GCT

Non‐tumoral FSEB diagnoses were made in 16 cases including fibrosis (n = 4), granulomas (n = 4), sclerosis (n = 3), haemorrhage and inflammation (n = 2), testicular atrophy (n = 2) and calcifications (n = 1). Of these, a single case underwent orchidectomy for severe granulomatous and purulent orchiditis following antibiotic therapy failure. Three FSEB were classified by a uropathologist as ‘suggestive of regressed GCT’ (Figure 4). Microscopically, these cases all consisted of a poorly cellular sclerotic mass radiating between pre‐existent seminiferous tubules without inflammation or haemorrhage and adjacent sclerotic tubules. One case underwent a second immediate FSEB showing seminoma followed by radical orchidectomy without residual tumour. The second case underwent immediate radical orchidectomy showing seminoma. The third case did not undergo immediate radical orchidectomy. One month after operation, however, a paraaortic lymph node seminoma metastasis was diagnosed by biopsy and treated by chemotherapy, followed by radical orchidectomy 1 year later with sclerosis consistent with therapy effect without residual tumour.

Figure 4.

Examples of non‐neoplastic (A–D) and neoplastic (E–H) fibrotic changes. In reactive cases, fibrosis and scarring were multifocal and ill‐defined (A) and can present with well‐formed granulomas (B), mixed lympho‐plasmocytic inflammation (C) and Sertoli‐cell only seminiferous tubules mimicking GCNIS (D). In regressed GCT cases, fibrosis was vascularized (E) and scarring was stellate with atrophic seminiferous tubules (F). Coarse stromal calcifications could be found (G) and the presence of GCNIS confirmed the diagnosis of regressed GCT (H). (A: permanent H&E, 10×; B: FSEB H&E, 10×; C1&D: FSEB H&E, 20×; C2: permanent H&E, 20×; E–G: FSEB H&E, 5×; H: FSEB H&E, 20×).

Benign and malignant mature teratoma

There were nine cases of benign epidermoid cysts on FSEB, all of which did not undergo radical orchidectomy. In addition, the diagnosis of ‘mature teratoma’ was made in three cases of FSEB. Two of these patients immediately underwent radical orchidectomy which showed GCNIS with signs of tumour regression in one patient; the other was a 22‐year‐old man, whose permanent section did not show residual tumour or GCNIS, and FISH testing of the FSEB sample did not reveal chromosome 12p alterations, which led to the final diagnosis of a benign pre‐pubertal mature teratoma (Material S1). The third patient was a 48‐year‐old male, who did not undergo radical orchidectomy but was followed up with annual abdominal ultrasound for 5 years with no recurrence or metastasis.

Discussion

FSEB is an underused technique despite its high accuracy and potential in preventing unnecessary radical orchidectomy for benign testicular lesions in 61% of cases. In our study, FSEB was performed on 127 patients and revealed a benign diagnosis in 78 (58%) cases. The three most common diagnoses were Leydig cell tumour/hyperplasia (27%), seminoma (27%) and reactive changes (12%). Sensitivity and specificity for malignancy were 100% and 96.3%, whereas the positive and negative predictive values were 94.4% and 100%, respectively. This is concordant with previous series which reported excellent accuracy of FSEB in testicular lesions with sensitivity and specificity for malignancy ranging from 95%–100% to 86%–100%, respectively. ⁸ , ¹⁵ , ¹⁶ , ²³ , ²⁶ , ²⁷ Our study shows that FSEB is an accurate diagnostic procedure saving young men with small testicular masses from unnecessary orchidectomy and provides practical guidance for its clinicopathological assessment.

While the vast majority of intraoperative assessments were straightforward, some pose histologic challenges and/or rely on ancillary testing. First, interpretation of fibrotic changes in FSEB can be challenging due to subtle histologic differences of a potentially regressed (burned‐out) tumour. ²⁸ , ²⁹ Indeed, it was not possible to formally distinguish regression from non‐neoplastic fibrosis in the absence of residual tumour and/or GCNIS; however, one should be aware of suggestive histologic signs and not hesitate to mention the possibility of regression when making the diagnosis (Figure 4). ²⁸ , ³⁰ In our series, regressed GCT was suggested three times, of which all were diagnosed with seminoma at immediate radical orchidectomy or during follow‐up. In contrast, fibrotic changes may occur in the scope of infective orchitis or trauma, whereas one should be aware that some seminomas could be masked by an inflammatory background. ¹⁴ , ³¹ None of the FSEB with reactive changes were diagnosed with GCT at definitive FSEB assessment or during follow‐up, indicating these can be distinguished from regressed GCT based on morphologic features.

As for the second challenge, the lack of immediate ancillary technique was rarely an issue for diagnosing malignancy on FSEB with the exception of morphologically mature teratomas without GCNIS. Those cases may require FISH studies to determine malignancy and this situation remains, albeit rare, a limitation of the FSEB technique and should be communicated to the urologist (Material S1). Immunohistochemistry on permanent section helped refine the diagnosis in certain benign cases from Leydig cell to Sertoli cell tumour, or from Sertoli cell tumour to adenomatoid tumour where frozen morphology was challenging (Material S3). This had minor implications as none of these cases underwent radical orchidectomy and both the FSEB and permanent diagnoses were benign. One should, however, be aware that OCT3/4 on previously frozen tissue can appear weaker and smudge into the cytoplasm of GCNIS and seminoma cells (Figure 2). The interpretation of the stain was still readily made for our cases; however, one might also consider adding CD30, CD117, D2‐40, PLAP, SOX2 and SOX17 to the panel if necessary (Figure 2, Material S4).

Our and other studies show that FSEB can be a valuable diagnostic tool to assist the surgeon in intraoperative management, thereby avoiding unnecessary surgery and preserving endocrine function, fertility and psychological benefits. ³² , ³³ , ³⁴ Some authors have suggested avoiding performing FSEB for various reasons; most importantly, because it would involve the availability of an experienced uropathologist. ³⁵ In the 13 cases of our study where a uropathologist was not available, 10 cases were correctly diagnosed; however, two cases were diagnosed as indeterminate, and one case was falsely diagnosed as malignant leading to an avoidable radical orchidectomy. Nonetheless, many centres will have the availability of specialized pathologists and notification in advance further maximizes their immediate availability. On the other hand, education regarding the most important diagnoses and pitfalls, as in this current report, may improve diagnostic accuracy and certainty (Material S5).

The impact of removing testis, especially in young men, may be significant. Patients who were able to avoid radical orchidectomy often mention being satisfied with not having to go through prosthesis implantation or having an empty scrotum, helping the preservation of the male body image. ³⁶ , ³⁷ In published reports, post‐operative testosterone levels and paternity are also encouraging even in malignant testis‐sparing cases, a subgroup with known lower testicular function. ²⁰ , ²² , ³² Therefore, offering FSEB is a safe procedure, especially if a uropathologist is available, and contributes to preserving endocrine function, fertility and self‐esteem.

Our study is one of the largest single‐institution cohorts on testicular FSEB, specifically providing a comprehensive pathological overview of the most frequent diagnoses and important pitfalls. Our retrospective study had several limitations, such as a lack of long‐term follow‐up. Testosterone profiles, postoperative paternity status and patient quality‐of‐life surveys were not systematically collected, limiting our evaluation of the functional and psychological success of our approach. Benign lesions excised in toto on FSEB were not subjected to standardized clinical follow‐up. Finally, the clinical indications for performing FSEB were not consistently documented, and no standardized size criteria defined the eligible small lesions. The decision to proceed with FSEB then reflected a shared‐decision process between the patient and urologist.

Conclusion

The FSEB approach for small testicular lesions represents a valid alternative to first‐line radical orchidectomy considering the significant rate of benign lesions (58%). FSEB allowed for the preservation of 81 testes during the study period with high accuracy, rare false‐positive results and no false‐negative outcomes. We conclude that, in experienced centres, FSEB should be offered as a way to avoid unnecessary radical surgery and preserve organ function.

Funding statement

No funding was received for this study.

Conflict of interest

The authors declare no conflicts of interest.

Patient consent statement

Not applicable.

Permission to reproduce material from other sources

Not applicable.

Supporting information

Data S1. Supporting Information.

Material S1. False‐positive mature teratoma classification. (A, B) On FSEB, mature intestinal epithelium adjacent to mildly atrophic tubules lacking GCNIS was interpreted as a post‐pubertal mature teratoma (A, B: FSEB H&E, 10×, 5×). (C, D) On permanent section (C, D: permanent H&E, 10×, 5×) and orchidectomy, no GNCIS nor other GCT components were found. No chromosome 12p alterations were found by FISH. The final diagnosis was of a pre‐pubertal type teratoma.

Material S2. Summary table of discordant and indeterminate cases.

Material S3. Adenomatoid tumour. (A, B) FSEB diagnosis was benign unclassified tumour, favour Sertoli cell tumour, based on the cords of pale monotonous cells in hyalinised stroma. (C, D) On permanent section, the intracytoplasmic vacuoles and the positive D2‐20 (E) and WT1 (F) confirmed the diagnosis of adenomatoid tumour. (A, B: FSEB H&E, 20×; C, D: permanent H&E, 20×; E: D2‐40, 10×; F: WT1, 10×).

Material S4. Embryonal carcinoma. A‐D: On FSEB, necrosis (A) was present adjacent to solid nests (B), papillary architecture (C) and GCNIS (D). (E–H) On permanent section, complex glandular architecture, basophilic pleomorphic cells (E) and GCNIS (F) were identified. Strong positivity for CD30 (G) and OCT3/4 (H) confirmed the diagnosis of embryonal carcinoma. Note OCT3/4 in embryonal carcinoma and GCNIS with some cytoplasmic smudging but retained strong nuclear staining (H). (A: FSEB H&E, 1.25×; B: FSEB H&E, 5×; C, D: FSEB H&E, 10×; E: permanent H&E, 20×; F: permanent H&E, 5×; G: CD30, 20×; H: OCT3/4, 5×).

Material S5. Freezing pitfalls in seminiferous tubules. (A, B) Due to freezing and formalin fixation, larger cells filled lumens of the seminiferous tubules. (C, D) Sertoli cell‐only seminiferous tubules mimic GCNIS on FSEB due to mild enlargement of the nuclei from freezing. (E, F) Due to freezing, GCNIS cells did not display their usual sharp cellular membranes and later, on permanent section, their cytoplasm was more eosinophilic. (A,C,D: FSEB H&E, 20×;B,D,F: permanent H&E, 20×).

Data availability statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.