Clinicopathologic features and proposed grossing protocol of orchiectomy specimens performed for gender affirmation surgery

Summary

Gender affirmation surgery performed for gender dysphoria is increasing to instigate changes more closely approximating gender identity. We investigated the clinicopathologic features of gender-affirming orchiectomies performed at our institution and devised a grossing protocol for these increasingly encountered specimens. We obtained 45 orchiectomies from 23 patients and reviewed clinicopathologic features. The number of sections per case was noted and reviewed to devise an optimal grossing protocol to assess pathologic findings. Twenty-three patients had bilateral orchiectomy with 1 unilateral. The average patient age was 39.4 years (range, 21–71 years); all received hormones for a mean of 66.1 months (range, 12–348 months). The average number of slides per orchiectomy was 8 slides (range, 1–11). Aspermatogenesis occurred in 32 (71%), hypospermatogenesis in 8 (18%), and normal spermatogenesis in 5 (11%) testes. Twenty-five (56%) exhibited scattered cells with nuclear cytomegaly, concerning for germ cell neoplasia in situ (GCNIS), but OCT4 negative. Six (13%) had multinucleated stromal cells. Leydig cells were markedly reduced/absent in 38 testes (85%). Epithelial hyperplasia was identified in 15 rete testes (33%) and 24 epididymes (53%), while 18 (40%) showed periepididymal muscular hyperplasia. All findings were identified in the initial 2 slides including rete testis/epididymis, except for 3 cases, missing only focal tubular sclerosis. Despite all received treatment, only a subset showed changes of exogenous hormone therapy. The presence of nuclear cytomegaly can mimic GCNIS and may be a potential pitfall. Two sections to include rete testis/epididymis and a third of cord margin are sufficient to identify the relevant pathology and germ cell tumors overall are uncommon in orchiectomies performed for gender affirmation.

1. Introduction

Patients with gender dysphoria seeking medical and surgical treatment to instigate changes in secondary sex characteristics that more closely approximate the gender which they self-identify, has been progressively more common over the past several years [1,2]. Treatment often includes cross-sex hormonal therapy, comprised of antiandrogens with estrogen, followed by gender-affirming surgery. Gender-affirming surgeries in male-to-female (MtoF) individuals encompass a combination of mammoplasty, facial feminization with voice surgery and thyroid cartilage reduction, gluteal reconstruction, penectomy with orchiectomy, and the creation of a labia and neovagina [3]. Bilateral orchiectomy is often performed at an early stage of their transition as it reduces the production of endogenous androgens, allows for the discontinuation of anti-androgen therapy, and reduces the dosage of estrogen [3]. Previous studies have focused on the histologic changes related to hormone therapy, particularly estrogen, on spermatogenesis in the testis [4-11]. Therefore, we aimed to provide a clinicopathologic overview of the features identified within gender-affirming orchiectomies performed at our institution to summarize our experience, confirm previous findings, and devise a proposed grossing protocol to adequately assess these increasingly encountered specimens.

2. Materials and methods

2.1. Patient cohort

We reviewed the archived surgical pathology cases from the Massachusetts General Hospital from January 2019–December 2021 in patients who underwent orchiectomy for gender-affirming surgery with institutional review board approval (IRB Protocol# 2020P000199). Electronic medical records were reviewed and noted for presurgical cross-sex hormone therapy and duration. Controls included nontumoral sections with normal spermatogenesis of orchiectomies performed for germ cell tumors (n = 12), fibrous pseudotumor (n = 2), inflammation/ischemia (n = 2), and testicular lymphoma (n = 1). All hematoxylin and eosin (H&E) stained slides were counted and reviewed by at least 2 genitourinary pathologists. Twenty cross-sections of round seminiferous tubules were evaluated at 100× objective and measured the diameter using an oculometer in millimeters (mm) and the mean (mean tubular diameter [MTD]) was calculated [4]. Additional characteristics such as the incidence of spermatogenesis and if in existence, the presence of spermatogonia, spermatozoa, and spermatids were recorded. Classification of the histologic findings include: (1) normal spermatogenesis when there is complete spermatogenesis throughout the entire specimen and presence of normal interstitial parenchyma; (2) hypospermatogenesis when there is complete but reduced spermatogenesis; and (3) maturation arrest when there is complete cessation at a specific stage, most often at the spermatogonial or spermatocyte phase. The quantity of interstitial Leydig cells was stratified into absent or reduced (only rare Leydig cells present in a high-power field [HPF]) and other which included hyperplasia (interstitial growth with small nodules <5.0 mm) or normal (approximately 30–35 Leydig cells/HPF).

2.2. Immunohistochemistry

Immunohistochemical staining was performed in 6 testes for OCT4 (mouse monoclonal antibody, clone N1NK, Leica) per the manufacturer’s protocol.

3. Results

3.1. Patient clinical characteristics

We identified 23 patients who underwent 45 orchiectomies (22 bilateral and 1 unilateral), all of which were intrascrotal and none associated with a history of cryptorchidism. One patient had a prior unilateral orchiectomy for an unknown reason. The average age of the patients was 39.4 years (range, 21–71 years), and all received cross-sex hormone therapy with variable estradiol (n = 22), spironolactone (n = 16), and Lupron (n = 8) for a mean duration of 66.1 months (range, 12–348 months). The average number of slides per orchiectomy was 8 slides (range, 1–11 slides).

3.2. Histologic and immunohistochemistry findings

The histologic features are summarized in Table 1. The mean gonad size was 4.1 cm (range, 2.0–5.5 cm) and weighed 30.0 grams (range, 19–46 grams). Upon serial sectioning, none of the orchiectomies were grossly identified to harbor a mass. Seminiferous tubules were small in diameter ranging from 0.11 to 0.43 mm (mean, 0.1902 mm) with most displaying peritubular fibrosis. Aspermatogenesis occurred in 32 testes (71%), of which 22 and 10 displayed maturation arrest at the level of spermatogonia and spermatocytes, respectively (Fig. 1). Eight (18%) showed hypospermatogenesis with a reduction in quantity but complete maturation and 5 (11%) displayed normal spermatogenesis (Fig. 2). Six testes (13%) had multinucleated stromal cells and 25 (56%) exhibited scattered cells with nuclear cytomegaly (≥3× Sertoli cell nuclei), concerning for germ cell neoplasia in situ (GCNIS), but all 6 specimens tested were OCT4 negative. These cells displayed coarse nuclear chromatin with clear cytoplasm and lacked prominent nucleoli, resulting in a degenerative appearance. Leydig cells were markedly reduced/absent in 38 (85%), normal in 6 (13%), and hyperplastic in 1 (2%) testis. Epithelial hyperplasia with stratification and micropapillae and/or cribriform formation was identified in 15 rete testes (33%) and 24 epididymes (53%), and 18 (40%) showed periepididymal muscular hyperplasia.

Table 1.

Histologic characteristics of orchiectomy specimens in MtoF transgender patients (n = 45).

Characteristic	Mean Result/Frequency
	Gender Affirmation (n = 45)	Control Group (n = 17)
Testicular size (cm)	4.1 (2.0–5.5)	5.5 (3.6–8.5)
Testicular weight (gm)	30.0 (19–46)	59.4 (36–126)
Spermatogenesis:
- Aspermatogenesis with maturation arrest at spermatogonia/spermatocyte	32 (71%)	0 (0%)
- Hypospermatogenesis (complete but reduced spermatogenesis)	8 (18%)	0 (0%)
- Normal spermatogenesis	5 (11%)	17 (100%)
Mean tubular diameter (mm)	a0.1902 (0.11–0.43)	a0.2482 (0.19–0.31)
Nuclear cytomegaly (≥3× Sertoli cell nuclei)	25 (56%)	0 (0%)
Multinucleated stromal cells	6 (13%)	0 (0%)
Quantity of Leydig cells:
- Absent/markedly reduced	38 (85%)	0 (0%)
- Normal	6 (13%)	15 (88%)
- Hyperplastic	1 (2%)	2 (12%)
Rete testis hyperplasia	15 (33%)	3 (18%)
Epididymal hyperplasia:
- Epithelial hyperplasia	24 (53%)	0 (0%)
- Periepididymal muscular hyperplasia	18 (40%)	0 (0%)
GCNIS	0 (0%)	–
Germ cell tumor	0 (0%)	–
Regression	0 (0%)	–
Sex cord-stromal tumor	0 (0%)	–

Abbreviations: Cm, centimeter; GCNIS, germ cell neoplasia in situ; gm, grams; mm, millimeter; MtoF, male to female; n, number.

^a(P = .0019) by t test.

Fig. 1.

Histologic findings in seminiferous tubules. A, Seminiferous tubules with aspermatogenesis and maturation arrest at the level of spermatogonia and reduced Leydig cells (H&E stain, 100×). B–C, Scattered cells with nuclear cytomegaly (H&E stain, 400×). D, Absence of staining for OCT3/4 (200×). E–F, Another example of nuclear cytomegaly with clear cytoplasm, enlarged nuclei with coarse chromatin and lacks prominent nucleoli, imparting a degenerative appearance (H&E stain, 100× and 400×).

Fig. 2.

Histologic findings in testicular parenchyma. A, Hypospermatogenesis with reduced but complete maturation and normal number of Leydig cells (H&E stain, 200×). B, Aspermatogenesis at the level of spermatocytes with Leydig cell hyperplasia (H&E stain, 100×). C, Multinucleated stromal cells (H&E stain, 200×). D, Efferent ductule/epididymal and E, rete testis epithelial hyperplasia with stratification and micropapillary formation (H&E stain, 100×). F, Muscular hyperplasia of epididymis (H&E stain, 100×).

The histologic findings were similar in both testes in all patients who underwent bilateral orchiectomy except for one in which each testis displayed hypospermatogenesis compared to aspermatogenesis with maturation arrest, respectively. All findings were identified in the initial 2 slides reviewed per case which included rete testis and epididymis, except for 3 cases in which focal tubular sclerosis would have been missed. In addition, none of the cases displayed GCNIS or evidence of regression and all were negative for a germ cell tumor, sex cord-stromal tumor, or other malignancy. An incidental adrenal rest and cord lipoma were the only benign neoplasms identified.

We identified 17 control orchiectomy specimens that had sections of nontumoral testicular parenchyma with normal spermatogenesis from men with an average age of 33.0 years (range, 14–64 years; Table 1). None were taking hormonal therapy. The mean gonad size was 5.5 cm (range, 3.6–8.5 cm) and weighed 59.4 grams (range, 36–126 grams). The MTD was 0.2482 mm (range, 0.19–0.31 mm). Leydig cells were in normal quantities within the interstitium, except in 2 cases with hyperplasia.

4. Discussion

Gender dysphoria has been increasing in the United States with approximately 1.4 million adults (0.6% of the total population) identifying as being transgender [12]. Cross-sex hormone therapy is often initiated to achieve a better resemblance between their affirmed and biological gender, and estrogen has known but varying effects on spermatogenesis in the testis [9]. The effect of hormone therapy on spermatogenic activity can range from aspermatogenesis with maturation arrest and Sertoli cell-only syndrome to normal spermatogenesis, regardless of the duration of treatment [5-11,13]. Complete spermatogenesis has been reported to remain in 4–40% of orchiectomy specimens despite undergoing hormonal therapy [6-8,10,13].

Prior studies have also described some of the morphologic findings observed in orchiectomies performed for gender-affirmation, such as nuclear cytomegaly, marked reduction or absence of Leydig cells, and hyperplasia of the rete testis and/or epididymis, in addition to the features typically described with estrogen therapy such as germ cell hypoplasia and aspermatogenesis with maturation arrest, which is similar to what we have observed in our investigation [4,14-16]. Matoso and colleagues reported that 80% of cases showed aspermatogenesis with maturation arrest at the level of spermatogonia, while 20% displayed hypospermatogenesis with spermatid/mature spermatozoa formation [4]. Comparably, we identified in our cohort that 71% had aspermatogenesis with maturation arrest, and 18% showed hypospermatogenesis. Alike to prior studies, which have shown that the variable stages of spermatogenesis are independent upon the duration of cross-sex hormone therapy, 11% in our cohort also displayed normal spermatogenesis despite being on treatment for an average interval of 108 months in these cases [5-8,16].

A marked decrease or absence of Leydig cells has been previously described and is similar to what we observed in the majority of our cases (85%) [4,14]. Also analogous to our findings, epidydimal and rete testis hyperplasia has also been recognized and likely attributed to the presence of estrogen receptors that have been identified in both tissue types [4,14]. In our study, epithelial hyperplasia of the rete testis and epididymis showed pseudostratification as well as micropapillary and cribriform formation, while the periepididymal muscle showed thickening with periepididymal cuffing. Prominent smooth muscle hyperplasia of testicular adnexa has been previously described in a similar periductal distribution and in rare instances can be mass forming, mimicking a neoplasm [17]. Although a few cases in the control group also showed features of rete testis hyperplasia, this occurred in orchiectomies performed for germ cell tumors and is a finding that has been previously been described as a pseudoneoplastic reaction [18].

The reduced MTD of seminiferous tubules with a thickened basement membrane are findings that have previously been described in orchiectomies performed for gender-affirmation [16]. Matoso and colleagues noted an MTD of 0.137 ± 0.002 mm compared to their control group with an MTD of 0.237 ± 0.007 mm (P < .001) [4]. Similarly, in our cohort, the MTD was 0.1902 mm, which was significantly smaller than the control group with an MTD of 0.2482 mm (P = .0019; t test).

An interesting finding is the nuclear cytomegaly, classified as at least 3 times the size of neighboring Sertoli cells, previously described by Matoso and colleagues, which can be misinterpreted as GCNIS [4]. These dispersed cells contain clear cytoplasm with enlarged nuclei, coarse chromatin, and absence of prominent nucleoli, resulting in a degenerative appearance. Helpful clues are the scattered versus clustered distribution of these large cells within the tubules, the presence rather than the absence of associated spermatogenesis, and most importantly the lack of prominent nucleoli. In addition, immunohistochemistry can be performed to exclude GCNIS as they have been previously reported to be absent of staining for OCT4, CD117, and PLAP [4]. In our study, approximately one-half of the cases demonstrated nuclear cytomegaly and all the cases that were tested for OCT4 were comparably negative.

A histologic finding that has not been previously mentioned is the identification of multinucleated stromal cells which were seen in a minor subset (13%) of the cases and not seen in the control specimens. These giant cells often displayed smudged rather than hyperchromatic nuclei and are likely degenerative and reactive stromal cells, similar to what was previously described in other sites such as the vulva, vagina, bladder, and prostate [19,20]. Furthermore, these multinucleated cells have been reported to be reactive for androgen receptors, with variable expression for progesterone and estrogen, and are perhaps a consequence of hormone therapy [19-21].

Given orchiectomies performed for gender affirmation are becoming more frequent, a grossing protocol for adequate evaluation is needed. The histologic findings were similar in both testes in all patients who underwent bilateral orchiectomy except for one. In addition, all findings were identified in the initial 2 slides reviewed per specimen, which included rete testis and epididymis, except for 3 cases in which focal tubular sclerosis would have been missed but are overall inconsequential findings.

The pathologic diagnosis can include the specific histologic findings identified or may be consolidated into a general statement such as “testicular parenchyma with features of exogenous hormone therapy.” The most important finding, given the variable effects of estrogen on spermatogenesis is well-established, is to exclude a malignancy. None of the cases in our cohort displayed GCNIS or evidence of regression and all were negative for a germ cell tumor, sex cord-stromal tumor, or other malignancy. Although rare, in the likelihood that an incidental neoplasm may be identified, an additional section of spermatic cord margin should be submitted. Therefore, 3 sections per orchiectomy (spermatic cord margin, n = 1; testicular parenchyma with rete testis and epididymis, n = 2) should be adequate, unless a mass is seen or evidence of regression and/or GCNIS is microscopically identified, which would necessitate additional sampling.

The overall incidence of identifying a malignancy within the surgical specimens of MtoF transgender patients is extremely rare (0.7%), with only 6 germ cell tumors identified in approximately 920 transgender females who underwent orchiectomy for gender affirmation [4-10,14,15,22-27]. In a cohort of 94 patients who underwent vaginoplasty with orchiectomy, only 1 testis from a 29-year-old transgender female was found to harbor a mixed malignant germ cell tumor composed of yolk sac tumor and embryonal carcinoma [22]. Subsequently, only a handful of cases of testicular cancer arising in transgender women have been described [22-25,27]. A malignant germ cell tumor has been incidentally discovered in a 26-year-old (seminoma, embryonal carcinoma, and yolk sac tumor) and a 30-year-old (seminoma) transgender female undergoing simple orchiectomy and a vaginoplasty with bilateral orchiectomy, respectively [23,24]. The latter was also noted to have elevated testosterone levels and halted progression of physical feminization resulting in increased dosage and added hormone therapy [24]. Both individuals did not have a testicular mass on examination.

The persistence of elevated testosterone levels and failed progression of physical feminization are helpful clues in identifying these occult malignancies, which can trigger the initiation of physical examination, additional serum marker testing, and imaging, which subsequently led to the identification of a seminoma with syncytiotrophoblasts and embryonal carcinoma in a 31-year-old and 28-year-old transgender female, respectively, who were on cross-sex hormone therapy [25,26]. There has also been a case of a choriocarcinoma and a metastatic germ cell tumor, hypothesized to be a seminoma, with complete resolution after neoadjuvant chemotherapy [28,29]. In addition, Kobori and colleagues reported a 30-year-old MtoF individual who developed a mature testicular teratoma which was positive for estrogen receptors, while receiving progesterone and estradiol, implying a potential contributable role in its development [27].

The 2017 Endocrine Society of Clinical Practice Guidelines recommend routine hormone screening with both serum estradiol and testosterone levels measured in 3 month increments for the first year and annually or biannually thereafter to observe for appropriate signs of feminization and adverse reaction development [30,31]. In monitoring transgender women on cross-sex hormone therapy, serum testosterone levels should be less than 50 ng/dL and estradiol levels between 100 and 200 pg/mL [25,30]. Persistent elevation should initiate additional testing to rule out ectopic etiologies such as a testicular cancer or an adrenal neoplasm [25,30].

5. Conclusion

In this study, we confirm what has previously been reported and describe our institutional experience regarding the morphologic findings in orchiectomy specimens performed for gender-affirmation in MtoF transgender individuals and propose a grossing protocol to adequately assess these increasingly encountered specimens. Although all patients received hormone therapy, only a subset had changes consistent with exogenous hormone exposure such as germ cell hypoplasia and aspermatogenesis with maturation arrest, diminished/absent Leydig cells, and hyperplasia of the rete testis and/or epididymis. The presence of nuclear cytomegaly, which can mimic GCNIS in some cases, may be a potential pitfall. Lastly, given identifying a malignancy is rare, a total of 3 sections, composed of 2 sections of testicular parenchyma to include rete testis/epididymis and 1 of the spermatic cord margin, appear to be sufficient to identify the relevant pathology, unless a mass or other unusual finding such as regression or GCNIS is seen.