Histologic Findings in Surgical Pathology Specimens From Individuals Taking Feminizing Hormone Therapy for the Purpose of Gender Transition: A Systematic Scoping Review

This review was conducted based on the guidelines of the Joanna Briggs Institute on conducting systematic scoping reviews. The review protocol was registered with the Open Science Framework on March 22, 2020.

A search strategy was developed in consultation with a medical reference librarian. An electronic search was conducted in March 2020 and covered the PubMed and Medline databases. The search strategy included both controlled vocabulary (National Library of Medicine's Medical Subject Headings: Transgender persons; Transsexual; Pathology, surgical; histology) and key words (transgender; transsexual; Pathology, anatom*), and encompassed articles published from January 1, 1946, to December 31, 2019 (see Search Strategy in the Supplemental Digital Content, also containing one table, at https://meridian.allenpress.com/aplm in the February 2022 table of contents). Briefly, the search strategy was designed to identify all studies that included descriptions of histopathologic findings from surgical specimens collected from transgender patients. No filters were applied to limit the retrieval by study design or language.

We conducted an additional search to identify studies that may not have been captured in the PubMed and Medline searches. For this search, we used Scopus to identify all studies referenced by studies identified in the initial electronic search (ie, backward search).

References were stored and reviewed using Covidence software for systematic reviews (Covidence, Melbourne, Australia). Title and abstracts were screened, and the full-text review of the potentially relevant articles was performed independently by 2 authors (A.A. and A.K.). Discrepancies were resolved by discussion. Papers were included if: (1) the study population was transgender patients/persons on feminizing hormone therapy, (2) the study protocol included biopsy or resection of tissue, and (3) the study objective was specific/relevant to the workup of pathology specimens for transgender patients. Papers were excluded if they did not include analysis of pathology specimens. At the level of the full-text review, papers that did not specifically discuss patients on feminizing hormone therapy were excluded. Papers were also excluded if the analysis of the specimens from transgender patients was pooled with other populations. Additional duplicate studies not removed by the Covidence software were also excluded. Interrater reliability was calculated using a Cohen κ score in the Covidence software.

Data extraction was performed by a single author (A.A.) using a pilot-tested collection form. We collected the following items: author, date of publication, country of study, organ system, aim or scope, number and characteristics of study participants (including age, gender, history of hormone treatment, indication for procedure), methods (including imaging modalities used, type of biopsy or excision performed, staining techniques used, other ancillary tests performed), main results or findings, and implications for practice or research.

The database searches identified 2748 references, including 233 duplicates. One additional study was identified through manual screening of the references of a related review article.⁶  There were 2516 unique references after duplicates were removed (Figure 1). We excluded 2270 references in the title and abstract screen. Full-text review was performed on 245 articles. Of these, 166 were excluded (reasons for exclusion are provided in Figure 1). One Russian-language study was excluded because the full text was not available online and attempts to contact the corresponding author received no response.⁷  Seventy-nine studies were included in the review (see Table 1 in the Supplemental Digital Content).

The Cohen κ score for the inclusion/exclusion of studies by the 2 authors involved in the screening process was 0.62 for the title and abstract screening and 0.66 for the full-text screening. There were 169 discrepancies out of 2516 titles and abstracts screened (6.7%), and 35 discrepancies out of 245 full-text articles screened (14.3%). These discrepancies were all resolved through discussion between the authors.

The basic characteristics of the 79 included studies are summarized in Table 1. The number of studies addressing surgical pathology specimens from patients taking feminizing hormones for gender affirmation has increased markedly over time, with the most marked increase in the last decade: 1968–1979 (n = 6), 1980–1989 (n = 7), 1990–1999 (n = 12), 2000–2009 (n = 13), and 2010–2020 (n = 41). More than half of the published literature (n = 60) is in the form of descriptive studies, including case reports and case series. There are more studies addressing neoplastic than nonneoplastic findings (n = 49 and n = 33, respectively). The most prevalent specimen types described in the identified studies are breast tissue (n = 24) and urologic excisions (n = 36). Reports of benign or neoplastic histologic changes in the cardiovascular system (n = 1), central nervous system (n = 9), anal canal (n = 3), skin (n = 1), musculoskeletal system (n = 2), kidney (n = 2), and lung (n = 1) were also identified. The following sections will summarize the key findings and entities that have been described. These findings are summarized in Table 2.

One comparative study and 3 case reports that described the benign effects of feminizing hormones on the breast were identified. Kanhai et al⁸  compared 6 transgender women taking combined progestative antiandrogen and estrogen therapy (group 1) with 7 transgender women after orchiectomy with estrogen-only treatment (group 2) and 2 cisgender men, after prostate carcinoma, taking nonprogestative antiandrogen therapy who developed gynecomastia (group 3). The breast tissue of patients in group 1 showed well-developed acini and lobules as well as proliferative and pseudolactational change.⁸  In group 1, one patient who did not take hormones regularly had less developed lobules with apoptosis and fibrosis.⁸  In group 2, breast tissue showed some lobule formation, but less than that seen in group 1.⁸  The breast tissue from patients in group 3 showed duct hypertrophy in both cases and a minor element of lobular development in 1 case, but less than that seen in breast tissue from persons in group 1.⁸  Pritchard et al⁹  and Richards et al¹⁰  reported background breast tissue with well-formed lobules in cases of adenocarcinoma and benign phyllodes tumors, respectively, in 2 transgender women with histories of longstanding (more than 10 years) estrogen therapy. Sam et al¹¹  reported a case of granulomatous mastitis in a 36-year-old transgender woman with a history of 6 years of estrogen therapy and 2 years of antiandrogen therapy. There was no mention of augmentation of the breast with exogenous material, and no foreign body response was identified in the biopsy material of the initially self-detected mass.¹¹ 

Twenty-one studies describing cases of neoplasia in the breasts of persons taking feminizing hormones were identified, of which 19 were case reports or case series and 2 were studies of cohorts of transgender persons assessing the incidence of breast carcinoma. The characteristics of these cases are summarized in Table 3. One case report that discussed a transgender woman with a family history of breast cancer gene (BRCA) mutation who underwent a prophylactic mastectomy was also identified.¹² 

Of the 38 cases of adenocarcinoma described, most (n = 23) were diagnosed as invasive ductal carcinoma.^9,13–24  Four cases of carcinoma in situ with no invasive lesions were reported.^17,24  In addition, single cases of secretory carcinoma²⁵  and poorly differentiated carcinoma of unknown primary cause were reported.¹⁶  The remaining 8 cases did not include a description of the type of carcinoma present.^21,22,24,26  Based on the 2 cancer cases in their cohort of 2307 transgender women, Gooren et al¹⁶  calculated an incidence of 4.1 per 100 000 person-years (95% CI, 0.8–13.0), which is comparable to the incidence in cisgender men.¹⁶  However, de Blok et al²⁴  found 18 cases of breast adenocarcinoma (15 invasive and 3 in situ) in their cohort of 2260 transgender women. From this, they calculated a standardized incidence ratio of 46.7 (95% CI, 27.2–75.4) when comparing the incidence of breast carcinoma in transgender women treated with feminizing hormones to that in Dutch hormone-naive cisgender men. The risk remained lower than that for cisgender women (standardized incidence ratio, 0.3; 95% CI: 0.2–0.4).²⁴ 

Reports of 5 other noncarcinomatous neoplasms were also identified, including 4 cases of fibroepithelial lesions and an angiolipoma. Two cases of fibroadenoma in young transgender women (aged 35 and 41 years) on hormone therapy have been reported, with features no different than those seen in the hormone-naive breast.^27,28  In addition, 1 benign phyllodes tumor has been reported,¹⁰  as well as 1 malignant phyllodes tumor with liposarcomatous differentiation.²⁹  One case of angiolipoma in the breast of a 61-year-old transgender woman who had been on feminizing hormone therapy for 18 months has also been reported.³⁰ 

Eighteen studies that described the histologic changes in the testes of persons taking feminizing hormones were identified: 3 case reports or case series, 1 cross-sectional study, 2 other descriptive studies, 1 case control study, and 11 other comparative studies. Grossly, atrophy was the most common finding.^31–35  The degree of atrophy correlated with serum hormone concentration, as opposed to duration of therapy.^33,36  Significant correlations (reported as P < .005) between testicular size and both testosterone and free testosterone were identified in 1 study of 108 patients treated with feminizing hormones and undergoing gender-affirming surgery.³³ 

The most common histologic finding was a reduced diameter of the seminiferous tubules, with increased basement membrane thickness, coupled with maturation arrest.^32–44  A compensatory increase in Sertoli cells within tubules accompanied by cytoplasmic vacuolation was also observed.^41,45  The seminiferous tubules in 1 study of 99 testes from transgender women had significantly smaller diameters than controls (0.137 ± 0.002 versus 0.237 ± 0.007 mm; P < .001).³²  The level of maturation arrest was independent of duration of therapy.^42,46  In the stroma, the most common findings were increased fibrosis and reduced Leydig cell numbers, as well as Leydig cell vacuolization.^* In some studies (5 of the 18), a subgroup of treated cases showed normal spermatogenesis and no histologic changes.^{33,35,39,44,46}  One study also reported 2 patients with ectopic adrenal tissue within the testes.⁴⁶ 

Immunohistochemical staining for estrogen receptor (ER) was strongly and diffusely positive in the head of the epididymis and was focally positive in the rete testis in all cases and controls in 2 studies where immunohistochemical staining was performed.^31,32 

Four case reports of neoplasms in the testes of transgender women on feminizing hormones were identified. There have been 2 cases of choriocarcinoma of the testes reported, both of which showed complete resolution on final resection after neoadjuvant chemotherapy.^49,50  The remaining 2 cases were germ cell tumors. The first was a pure, mature teratoma, showing skin, skin appendages, hair, cartilage, bone, bone marrow, respiratory, and gastrointestinal epithelial elements.⁴⁰  Immunohistochemical stains showed that ER-β was present, whereas ER-α and progesterone receptor (PR) were not detected.⁴⁰  The second was a mixed germ cell tumor with predominantly embryonal carcinoma (75%) as well as immature teratoma (15%), seminoma (9%), and yolk sac tumor (<1%).⁵¹ 

One comparative study examining the prostates of 9 transgender women was identified.⁵²  No difference in epithelium-to-stromal ratio was observed compared with estrogen-naive cisgender controls.⁵²  Prostatic acini showed an increased number of layers of epithelial cells (versus usual monolayer) with squamous metaplasia.⁵² 

Immunohistochemical staining was performed in 5 of the 9 cases described above for prostate-specific antigen, prostate acid phosphatase, and androgen receptor in the epithelium; staining was variable but less than in the controls.⁵²  The epithelium of 4 of the 5 cases showed ER and PR positivity compared with none of the controls.⁵²  Stromal staining for hormone receptors was variable in both estrogen-treated and estrogen-naive groups.⁵² 

Nine case reports of prostatic adenocarcinoma in transgender women with a history of estrogen therapy were identified. Patients' ages ranged from 54 to 78 years.^53–60  Gleason scores were available for 6 cases and ranged from 7 to 9 (4 + 5).^{54–57,60,61}  Four cases were diagnosed as poorly differentiated prostatic adenocarcinoma with no Gleason score provided.^53,58,59  One of these cases was diagnosed on the basis of increased serum prostate-specific antigen and biopsy of metastatic disease only.⁵⁹  This case was also reported as the only identified case of prostate adenocarcinoma in a cohort study of 2306 transgender women who were post orchiectomy and taking feminizing hormones in the Netherlands.⁶²  The incidence of prostate adenocarcinoma in this cohort was 0.04%. The incidence in a subgroup analysis of 779 transgender women who initiated hormonal treatment at age 40 years or later was relatively higher (0.13%), but still low in absolute terms.⁶² 

One case report and 1 comparative study that described changes in the urethras of persons taking feminizing hormones were identified. The case report described a transgender woman who presented twice with urethral obstruction.⁶³  At the first presentation in 1979, a biopsy of the verumontanum showed marked squamous metaplasia as well as prostatic hypertrophy.⁶³  The patient presented a second time in 1983, at which time the verumontanum was biopsied, showing a smooth muscle with focal myomatous changes and sparse glandular elements, some with transitional cell metaplasia.⁶³  The urothelium was edematous and inflamed but did not show the same degree of squamous metaplasia as seen in the previous biopsy.⁶³  Squamous metaplasia of the urethra affecting more than 50% of the lining and involving the periurethral glands was seen in 16 of 21 transgender women taking estrogen and 1 taking progesterone in the comparative study.⁶⁴  Immunohistochemical staining for ER was negative.⁶⁴ 

One comparative study that used immunohistochemistry to assess the corpora cavernosa of the penises of hormone-naive cisgender men (n = 8) compared with transgender women who had been taking feminizing hormones (n = 12) was identified.⁶⁵  Both androgen receptor and ER-α were detected in both study populations, and PR was not detected in the corpora cavernosa of either group.⁶⁵  There was no significant difference in the percentage of cells expressing androgen receptor (P = .17). However, the expression of ER-α was significantly increased in the hormone-treated tissue (P < .05).⁶⁵  The distribution and amount of androgen receptor and ER-α were not influenced by age.⁶⁵ 

One comparative study of the acquired smooth muscle hamartoma of the scrotum used the scrotal tissue from 7 transgender women undergoing gender-affirming surgery as control material. Histologic descriptions of the tissue describe the usual amount of dartos smooth-muscle tissue, lymph vessels, and collagen matrix for the site.⁶⁶  The distribution of CD34⁺ cells, as highlighted by immunohistochemistry, was also described as normal for the site.⁶⁶ 

Three case reports that described a total of 4 cases of pituitary lesions in patients taking feminizing hormones were identified. The patients' ages ranged from 32 to 71 years.^67–69  One case was diagnosed as lactotroph adenoma,⁶⁷  one as a gonadotroph adenoma, and one as a mammosomatotroph adenoma.⁶⁹  The fourth case showed retained reticulin networks and was diagnosed as lactotroph hyperplasia rather than an adenoma.⁶⁸  The lactotroph adenoma and the mammosomatotroph adenoma both showed immunohistochemical staining for prolactin but with the former showing a Golgi pattern and the latter showing diffuse cytoplasmic positivity.⁶⁹ 

Six case reports or case series describing a total of 8 cases of transgender women on feminizing hormone therapy with meningiomas were identified. The patients ranged in age from 28 to 65 years.^70–75  Three cases were diagnosed as meningothelial meningiomas,^70,72,74  1 as a transitional meningioma,⁷³  and 4 were not subtyped.^71,75  All of the lesions were positive for PR, whereas ER positivity was seen in half of the cases.^70–75 

One case report of carcinoma of the lung in a 49-year-old transgender woman was identified.⁷⁶  The patient presented with coughing, dyspnea, and purulent sputum production and was initially treated for pneumonia.⁷⁶  Biopsy of a nonresolving consolidation in the right lower lobe showed papillary adenocarcinoma of the lung with intermediate differentiation.⁷⁶ 

Two studies describing musculoskeletal biopsies were identified. Lips et al⁷⁷  performed a comparative study to assess the histomorphologic changes in bone associated with feminizing hormone therapy. Statistically significant decreases in osteoid volume, osteoid surface, and osteoid thickness were observed, as were statistically significant decreases in mineralizing surface and bone formation rate compared with hormone-naive cisgender male controls. No difference in bone volume, bone surface, trabecular thickness, osteoclast number, mineral apposition rate, or adjusted apposition rate was identified.⁷⁷  In their case report, Vague et al⁷⁸  assessed the effects of androgens and estrogens on adipocytes in deltoid and trochanteric adipose tissue. They found a moderate increase in deltoid and trochanteric adipocyte numbers in a transgender woman who had been taking hormone therapy for 34 months.⁷⁸ 

One study describing pathologic findings in the heart of a transgender woman on hormone therapy was identified. This 22-year-old woman died suddenly, and an autopsy was performed, which revealed right ventricular arrhythmogenic cardiomyopathy.⁷⁹  Grossly, the right ventricle was enlarged and flaccid, with diffusely thin, lipomatous, and yellowish walls and marked dilation of the tricuspid valve.⁷⁹ 

Histologically, the right ventricle showed diffuse infiltration of fibroadipose tissue with residual bridges and islets of myocytes.⁷⁹ 

Two case reports of renal changes in a patient taking feminizing hormones were identified. Both cases described young (ages 23 and 27 years) transgender women with a history of hormone therapy who presented with photosensitive malar rashes.^80,81  Clinical and serologic findings in both cases were in keeping with lupus nephritis, and a renal biopsy showed diffuse proliferative glomerulonephritis and extensive subendothelial “wire loop” deposits.^80,81  Direct immunofluorescence was performed in 1 case and showed a “full-house” pattern with positivity for C1q, C3, immunoglobulin (Ig) A, IgG, and IgM.⁸¹ 

One case report of a new presentation of lupus erythematosus was identified in a transgender woman.⁸²  Skin biopsy from the breast showed interface dermatitis and pools of mucin consistent with lupus dermatitis.⁸² 

One comparative study and 2 case reports describing gastrointestinal pathology were identified in transgender women. Cytologic studies were not included in this review. The comparative study evaluated the histologic changes in anal biopsies of 22 patients taking feminizing hormones for the purpose of gender transition and 1448 men who have sex with men.⁸³  No statistical significance was present between the rates of squamous intraepithelial lesions identified at anoscopy between the 2 groups (91% of transgender women and 84% of men who have sex with men, P = .20).⁸³  The first case report was a locally advanced squamous carcinoma of the anus in a 64-year-old, 24 years after her gender-affirming surgery.⁸⁴  The second case was a Kaposi sarcoma in the small bowel of a 34-year-old transgender woman who was also HIV positive.⁸⁵ 

As a result of a history of discrimination and failure to understand their needs, transgender and gender nonconforming patients on feminizing hormone therapy may not self-identify to their health care providers.⁸⁶  Even when gender nonconformity is apparent in the patient/provider relationship, this information may not be available to the pathologist. There is a great deal of variability in how gender history data are collected in current electronic medical record systems. Universal recording of both sex and gender data would facilitate greater access to gender history for the pathologist and clinicians. Varied approaches, such as that proposed by Deutsch and Buchholz,⁸⁷  to include fields for “gender ID,” “birth sex,” “legal sex,” and “pronouns” provide possible solutions to this lack of available information, but there must be institutional uptake for these processes to be effective. Thus, the dilemma for pathologists is compounded by the lack of identification of specimens on requisitions. Knowledge of a patient's gender and history is invaluable for the pathologist to frame diagnoses. Accurate gender histories on all specimens, not just those from gender-affirming surgeries, lead to a greater ability to track associations with neoplasia as well as benign changes associated with feminizing hormone therapy.

However, even without a history of hormone therapy, a pathologist may be able to identify specimens as coming from patients on feminizing hormone therapy, just as they may be able to identify patients taking specific drugs based on histologic findings. For example, predictable changes in the testes, prostate, urethra, and breast can be observed in patients taking feminizing hormones (Figures 2 through 5). Additional images of the changes in the testes can be found in the comparative study by Matoso et al,³²  and images of the breast changes can be found in the study by Kanhai et al.⁸  Unfortunately, these changes may be erroneously interpreted as pathologic if the pathologist is unfamiliar with the spectrum of histologic changes that may occur in patients on feminizing hormone therapy.

The changes described above are not universal. For example, the testes of some patients on feminizing hormones show no significant histologic changes. There are several theories as to why this may occur. Thiagaraj et al⁴⁴  suggest that the testes of some patients may be refractory to hormone treatment. Schneider et al³³  postulate it is due to individual responses to anti-androgen treatment regimens as well as variability in treatment between patients. Furthermore, Jindarak et al³⁵  note that in many cases, hormone therapy is discontinued before gender-affirming surgery takes place, and they theorize that this may allow for resolution of changes associated with exogenous hormone administration.

In addition to histologic changes, some papers noted differential immunohistochemical staining patterns in the tissues of transgender patients treated with feminizing hormones. One such example is found in the prostate and described by van Kesteren et al.⁵² 

This review highlights that much of the currently published data on histologic findings in persons taking feminizing hormones for gender transition comes from gender-affirming surgeries. Cases of malignancies other than breast and testes have rarely been reported, and as such it is difficult to determine if these tumor types present differently in this population. Other findings, such as the case of gastrointestinal Kaposi sarcoma reported by Schmalzle et al,⁸⁵  are likely due to factors other than feminizing hormone therapy—in this case, the patient's HIV infection.

A detailed assessment of the quality of the included studies was not performed, but no significant flaws in study design were noted. Many of the included studies are case reports and other descriptive studies, which makes pooling data for detailed analysis more difficult. Variability in terminology used when referring to transfeminine persons was noted, especially among older publications. In particular, transfeminine persons were often misgendered as male.

Over time, interest in the effects of feminizing hormone therapy on tissues seems to be increasing, as illustrated by the marked increase in publications in the last 10 years compared with the previous 10 years (n = 41 from 2010–2019 versus n = 13 from 2000–2009).

The gaps in the literature are numerous and encompass several key areas. Firstly, there is a paucity of reported findings from many organ systems, namely those outside genitourinary and breast, which makes assessing trends and incidence of changes in these other areas difficult. It is well established that many tumors have a sex predilection, and additional data are needed to determine if the incidence of these tumors is modified in transfeminine persons after the initiation of hormone treatment as opposed to baseline. Information on the prevalence of various tumors in transfeminine persons is further limited by the lack of population-based studies that include pathologic information, with only 3 such studies being identified in our search.^16,24,62  In light of these issues, and the increase in publication in this area in the last decade, a follow-up systematic review in approximately 5 years may yield additional useful information about the effect of feminizing hormone therapy on histologic findings. Furthermore, additional studies of the incidence and prognosis of tumors related to feminizing hormone therapy would be of great use to both pathologists and clinicians.

Limitations of this review are largely centered around key word searching. The initial search we performed was broad, looking for any study title including the terms “transgender” or “transsexual” as well as histology or surgical/anatomic pathology. Given the variability in terminology in this area, this may have limited our results. This was compensated for in some respect by the addition of the reference search for those initial results.

Strengths of this review include that it was performed in a systematic fashion, which is a strategy employed less frequently in anatomic pathology than many other disciplines.^88,89  In addition, our search spanned a long time period (1946–2019) and thus captured the evolution of the published data pertaining to histologic findings in persons taking feminizing hormone therapy for the purpose of gender transition. Our review is the first study to consolidate the literature pertaining to the effects of feminizing hormone therapy on all tissue types. As such, it highlights the gaps and potential for future research in this area. We hope this review will be used as a general reference for practicing surgical pathologists, pathology residents, and others as an introduction to the complex relationship between hormone treatment and histologic findings that helps to illuminate this evolving area of practice and research.