This has really caught my interest. I spent the afternoon reviewing the evidence because there are so many opinions here and I wanted to see the data for myself. After doing so, I can see why WPATH 8 was cautious after reading the papers it cites on HRT therapy per-operatively.
The papers WPATH 8 cites are, directly, Hembree, Gaither, Prince and Kozato. Indirectly, WPATH 8 cites Berli and Kailas.
Hembree is an evidence based guideline dating to 2009 and the only relevant statement it makes is: 'There is some concern that estrogen therapy may cause an increased risk for venous thrombosis during or after surgery' (p3149). They don't provide any credible evidence either way and the guideline is old.
Gaither (2017) is a retrospective review of 330 patients who had undergone a penile inversion vaginoplasty and found that, 'Age, BMI, and HRT were not associated with complications.' They tapered their patients down to 2mg of estradiol at least two weeks pre-op, but numerically theirs is the largest series.
Prince (2020) is a expert review whose only mention of the matter is: 'There is also debate as to whether transgender women should cease estrogen therapy preoperatively, with some surgeons advocating a suspension of HT 2-4 weeks before surgery (citing Berli 2017). By contrast, in 2019 at the national meeting of the Unites States Professional Association for Transgender Health in Washington, DC, a team from Mount Sinai Health System in New York City reported no observed increase in VTE events among transgender women who remained on estrogen treatment during gender affirming genital surgeries (citing Kailas 2017).'
So I found and read those two references too...
Berli summarises WPATH 7, which was published in 2012. So Berli is a summary of a guideline, yet it states 'However, oral estrogen, especially ethinyl estradiol, is associated with an increased risk of venous thromboembolism; therefore, it is common practice for the use of estrogens to be discontinued 2 to 4 weeks before GCS.' Tracking the citations, this statement is based on Hembree and Van Kesteren 1997.
Where on earth did this recommendation come from? I took a deep breath and read through Van Kesteren.
Van Kesteren (1997) Despite being referenced in Berli and by proxy in Prince, this retrospective study didn't split out the risk of per-operative DVT in AMAB patients taking oestrogen, but instead looked at adverse events overall in 816 patients taking oestrogens whether they had been operated on or not. The authors found a 20 fold increase in venous thromboembolism overall in patients on an oral dose of 100 ug of ethinyl oestradiol daily. Recent studies haven't replicated this finding. However, this paper is probably the source of Berli's statement that 'oral estrogen, especially ethinyl estradiol, is associated with an increased risk of venous thromboembolism.'
That is unhelpful because through Prince using Van Kesteren as a reference, WPATH 8 picked up on a single, then 25 year old study which was nothing to do with per-operative risk. The guidelines picked up on Van Kesteren twice, because Hembree also quotes Van Kesteren as justification for their. '...there is some concern...' statement. Despite the number of cases review, the scope of Van Kesteren doesn't address per-operative risk directly and in any case it is grade III or IV evidence at best. Better than any other example I can think of, this shows the danger of relying on a single paper, because its findings can colour guidelines written a quarter of a century later.
That leaves us with the Prince review Kailas reference...
Kailas (2017) was a retrospective review of 99 patients of whom 71 were transfemale, of whom only 27% had any kind of transgender surgery. Of those only two had a vaginoplasty and there were another two where the nature of the operation was unknown.
I would leave Prince aside, because ultimately, from our POV its recommendations in this speficic area are based either on Hembree, or on two patients in Kailas. Berli is irrelevant because it recycles Hembree and WPATH 7. With only two paitients in our group we can't learn anything from Kailas.
Retracing my steps back to the four papers quoted by WPATH 8...
Kozato (2021) is another retrospective study which looks larger than it really was, but in the group we're interested in, 190 'long' cases were performed with estrogen suspended for 1 week prior to surgery, and 212 long cases were performed with estrogen continued throughout. Only one patient presented with DVT and that was in the treatment suspended group.
Which means the WPATH panel were looking at 554 patients who had had penile inversion vaginoplasty, which is what we would call a long case in the UK and the type of op most likely to cause DVT. All of these 554 patients were on oestrogen, although close to half were on a low dose. There were no DVTs in the group.
This is great news, so why does WPATH 8 not declare a position that oestrogen medication can be given per-operatively in trans patients??? Because all the studies they have cited are grade III or grade IV evidence. They didn't have an RCT to look at, which would have boosted the quality of evidence to grade II at worst and grade I at best. They didn't even have a prospective trial.
What can we take away from this? Gaither and Kozato are absolutely the papers to show to a surgeon if you're trying to persuade them. But medicine is science based, so until we have a grade I RCT in this area the quality of the evidence we have, though persuasive, is by no means rock solid. And without grade I or II level evidence, I doubt guidelines will declare on this subject.
If you think the guideline writers are being too cautious, or patronising, or whatever on this point, imagine you are invited to play a high stakes poker game where the chips represent people's lives. With your name at the top of the paper, you aren't going to declare on a pair of kings.
We need a prospective trial to put this one to bed, once and for all
If a group of specialist centres got together and if people having penile inversion vaginoplasty were prepared to enter a prospective trial, once the design was settled and ethics approved, it should be possible to get the actual case throughput done in a year or eighteen months. I can't see any reason why a trial would need to be confined to PIV surgery in trans people taking oestrogen, because the risk should be no different to any kind of procedure where AMAB patients on oestrogen underwent a 2-3 hour anaesthetic, but identifying and consenting prospective cases would be much easier in a unit setup for GAMC care. There is no prospect of running a double blind trial because the patients would know if they were given dummy meds instead of oestrogen, but even so, a well designed trial like that would answer the question forever. Who do we encourage to do it?
If anyone wants my own view on whether I would personally consent to penile inversion vaginoplasty while taking a working dose of oestrogen I will gladly post what I think in reply. But this post isn't about what I think, it is about what the research has found, about the quality of the existing papers, why guidelines leave this question unanswered, and ultimately, why some surgeons are saying the things they do.
Gaither, Thomas, et al. "Postoperative Complications Following Primary Penile Inversion Vaginoplasty among 330 Male to Female Transgender Patients." The Journal of Urology 199 (2017).
Kozato, Aki, et al. "No Venous Thromboembolism Increase among Transgender Female Patients Remaining on Estrogen for Gender-Affirming Surgery." The Journal of Clinical Endocrinology & Metabolism 106, no. 4 (2021): 1586-90.