テストステロンのレビュー：子宮内膜は大半が萎縮性または増殖性で、増殖症が2.3%；35%に多嚢胞性に見える卵巣

性別肯定のためのテストステロン後の子宮および卵巣の病理組織学：系統的レビュー

Authors: Maris K. Toland, Kyra Bonasia, Jessica Bentz, Regina M. DelBaugh, Elaina J. Vitale, Paige N. Scudder, Ella A. Damiano
Journal: Transgender Health
Year: 2024
Source: https://doi.org/10.1089/trgh.2022.0207

Summary

Dartmouth doctors combined 17 published studies on what the womb and ovaries look like under the microscope after transmasculine people took testosterone and then had them removed. The womb lining was roughly half thin and inactive and half still growing, and a few (3%) showed signs of recent ovulation despite testosterone. About 2.3% had precancerous thickening of the lining and one person had a tiny cancer. No ovarian cancers were reported, though about a third had ovaries that looked polycystic (full of small cysts). The authors concluded testosterone does not raise precancer risk, and routine biopsies before surgery are not needed without other risk factors. This is useful descriptive data, but it cannot measure risk: there was no comparison group, the studies were small and often biased, average testosterone use was only about 2.5 years, and the cancer comparison was indirect.

Key findings

このレビューには、658件の記録から選ばれた、患者を観察した17件の研究（試験なし）が含まれ、2021年4月5日まで検索されました。
522件の子宮内膜のうち、254件（49%）は薄く不活発で、243件（47%）はまだ増殖しており、16件（3%）は最近の排卵の兆候を示したため、テストステロンが子宮内膜を確実に静めるわけではありません。
12人の患者（2.3%）に子宮内膜の前がん状態の肥厚があり、1人（0.2%）はそれに小さながんの部分を伴っていました。これらの症例のうち9件は、出血がない人でした。
410人（卵巣765個）のうち、143人（35%）の卵巣は多嚢胞性に見え、3人（0.7%）に無害な増殖があり、卵巣がんの報告はありませんでした。
患者の5.6%の卵巣に最近の排卵の兆候が見つかり、テストステロンを使っていても排卵が続きうることを示しています。著者らは避妊の必要性を強調しています。
その他の無害な所見には、筋腫が12.5%、子宮内膜ポリープが7.3%、子宮内膜が子宮の筋肉に入り込むものが2.7%、子宮内膜症に関連する卵巣嚢胞が1.2%ありました。
テストステロンの平均使用期間は約30か月で、大半の研究は、誤解を招くリスクが中程度から高いと評価されました。
著者らは、テストステロンが子宮の前がん状態のリスクを高めることはなく、テストステロンを使っているというだけで手術前に生検が必要なわけではないと結論づけました。

Abstract

PURPOSE: The objective of this systematic review was to evaluate the effects of testosterone on uterine and ovarian pathology in transmasculine patients at the time of gender-affirming surgery.

METHODS: Searches were conducted using MEDLINE, CINAHL with Full Text, Cochrane Library, Scopus, and Europe PMC for articles published before April 5, 2021, containing concepts of testosterone use, uterine histopathology, and transgender population. The initial search yielded a total of 658 abstracts. Studies were analyzed in full text if they reported one or more uterine or ovarian histopathological findings in a noncisgender population receiving testosterone for gender affirmation. Rayyan software was used for the abstract review.

RESULTS: This review was registered with PROSPERO. Twenty-eight full-text articles were reviewed, of which 17 met the inclusion criteria. Two reviewers performed data extraction and risk of bias assessment. Thirteen studies included endometrial histopathology results, representing 522 patients. Endometrial findings demonstrated that 254 patients (48.7%) were atrophic, 243 (46.6%) were proliferative, and 16 (3.0%) were secretory. Twelve patients (2.3%) had endometrial intraepithelial neoplasia (EIN), including one (0.2%) with focal adenocarcinoma. Eleven studies reported ovarian histopathology in 410 patients and 765 ovaries. A total of 143 (34.9%) patients had polycystic-appearing ovaries. Three (0.7%) patients had benign ovarian neoplasms. No ovarian malignancies were reported.

CONCLUSION: The endometrium was predominately found to be atrophic or proliferative, with low rates of EIN and carcinoma. No ovarian malignancy was observed. Based on this review, patients on testosterone do not have an increased risk of precancerous endometrial lesions due to unopposed hormones in the setting of amenorrhea.

Conclusion

The endometrium was predominately found to be atrophic or proliferative, with low rates of EIN and carcinoma. No ovarian malignancy was observed. Based on this review, patients on testosterone do not have an increased risk of precancerous endometrial lesions due to unopposed hormones in the setting of amenorrhea. (Full-text conclusion adds: a minority had secretory endometrium, supporting evidence that ovulation can occur on testosterone; benign lesions such as fibroids and endometriosis occur and symptoms should be evaluated as in cisgender patients.)

Methodological limitations

No comparison group: None of the data come from a matched group of people not on testosterone. The claim that testosterone does "not increase" precancer risk rests on comparing with rates from separate studies of women (1–7.9% in women before menopause with any bleeding pattern, or 0.4% in one group without bleeding), not on a direct comparison.
Short use of testosterone: People had taken testosterone for about 30 months on average (roughly 1–10 years where reported), far too short to judge cancer risk from decades of use. Two studies gave only ranges.
Simple head-counts: Results were just numbers added up across studies. There was no proper statistical combining, weighting by study size, check of how much studies disagreed, or measure of uncertainty.
Small, low-quality studies: All 17 studies only observed patients, most were rated at moderate-to-high risk of being misleading, samples were small, and the pathologists often knew the patients were on testosterone.
Things defined differently: Studies defined womb-lining stages, "polycystic" ovaries (from 6% to 80% across studies depending on the definition) and absent periods differently or not at all. So the combined percentages mix things that don't match, and two different systems for naming precancer were used.
Overlapping and mixed groups: Two studies shared many of the same patients, and three included some people not on testosterone whose results could not be separated out.
Only people who had surgery: Only people who had their womb or ovaries removed were studied. The results may not apply to people who keep their womb on long-term testosterone, the group for whom check-up advice matters most.
Key details missing: Bleeding patterns were reported in only nine studies, weight and other illnesses in few, race in two, and testosterone dose and blood levels were not analysed. So risk factors cannot be studied.
Out of date and English only: The search ended in April 2021 but the paper came out in 2024, and studies not in English were left out.
Narrow focus: The review only covers what tissue looks like under the microscope. It does not cover other reported effects of testosterone on the pelvis, such as pelvic pain, symptoms from thinning tissue or managing bleeding, so it should not be read as an overall safety check.

detrans.ai

Summary

Key findings

Abstract

Conclusion