The clinical risks associated with the diagnosis and management of disorders of gender identity

Principles of treatment

In common with other disorders, treatment is based on first establishing a diagnosis. Hormonal and surgical treatment is needed only when a diagnosis of transsexualism applies. Disorders of gender identity can be particularly challenging and it is agreed by the draft United Kingdom and other guidelines that it is inappropriate for patients to be treated outside the remit of a specialized clinic with a full range of professional expertise, which will include psychiatrists, psychologists, endocrinologists, surgeons, and speech and language therapists. ¹

Treatment for transsexualism is triadic, with each stage being less reversible than the preceding stage and only undertaken if the previous stage has been satisfactorily negotiated.

First there must be a change of social gender role, then treatment with cross-sex hormones and, lastly, surgical intervention.

Each of these stages carries significant risks.

Diagnosis

Given the drastic nature of the treatments in transsexualism it is appropriate for decisions to be made by two clinicians who should both be psychiatrists or psychologists with experience in this field. This is mandatory for gender reassignment surgery ¹ and, given that hormone treatment has irreversible effects (especially androgen treatment), the same safeguards should probably apply to diagnosis. ²

It is a cardinal error to accept the patient's self-diagnosis, as it is made by someone with neither objectivity nor experience. Diagnoses by other medical practitioners (including general psychiatrists) may be more objective, but will still lack the necessary experience.

Frequently, patients seek to present their circumstances as simple and meriting early intervention with hormones. Such simplicity is uncommon and transsexualism may be claimed by patients with other diagnoses, including psychosis, mood disorders, homosexuality and a variety of endocrine disorders and types of sexual deviance. For this reason, properly expert assessment is very necessary. ²

Clearly, inappropriate treatment produces drastic, irreversible effects and can be viewed as disastrous. It has been the subject of successful negligence litigation and General Medical Council censure. ³

Hormone treatment for born-male patients

Hormone treatment should only be undertaken if there has already been a change of social gender role. Hormone treatment before such a change of role has been deemed a constituent of Serious Professional Misconduct by the General Medical Council. ³ Further, commencing hormone treatment in advance of a change of social gender role does not make such a change of role more likely. Rather, there often follows an interminable period of treatment with ever escalating doses of hormones, the effects of which may be irreversible. ²

Hormone treatment should only be commenced when the patient has been confirmed as physically fit for such treatment. Safety monitoring is summarized in Table 1.

The presence of grossly elevated androgens and unsuppressed gonadotrophins at initial screening suggests a partial androgen insensitivity syndrome. This requires endocrine follow-up, as genetic counselling may be required for the patient's family. ⁴

If all these indices are within normal limits, and provided the patient has changed social gender role, hormone treatment may commence.

Hormone treatment for born-male patients always involves the administration of estrogens and cessation of smoking. This may be augmented by a gonadotrophin-releasing hormone analogue, with cyproterone acetate employed in the first two weeks of such treatment to prevent that initial rise in androgens which precedes their precipitous fall which is sometimes called ‘androgen flare’. Cyproterone acetate treatment should not be continued for more than two weeks as cyproterone acetate causes deranged liver function tests, depression, lethargy and fatigue. These unwanted effects are seemingly more common in the context of gender identity problems than in the context of malignancy, which is why cyproterone acetate is not used as a long-term androgen antagonist. ⁴

There is no role for progesterone in the management of disorders of gender identity because it has no action to feminize breast tissue (it is not present in born-females until breast development is complete) and may serve to raise the risk of breast cancer. ⁴

Estrogen treatment characteristically employs estradiol valerate. There is the suggestion that this may be less thrombogenic than ethinyl estradiol and, unlike ethinyl estradiol it is easily measured on hormone assays. Earlier, equine conjugated estrogens were employed, but the animal origin is unacceptable to some patients and some patients develop a photosensitive rash on this treatment.

Excessive doses of estradiol valerate employed from the outset cause rapid initial breast development but early duct fusion, resulting, ultimately, in small, hard conical breasts. Maximal breast development requires the estrogen dose to be progressively increased, a typical initial dose being estradiol valerate 2 mg daily for three months, rising by 2 mg aliquots at three-monthly intervals until reaching the dose that results in a serum estradiol level of 400–600 pmol/mL – typically 2 mg TDS.

This hormone regime carries low risks. Overall all cause mortality is not increased. ⁵ Safety monitoring (Table 1) indices should remain within normal limits, with the exception of prolactin, where levels under 1000 iu are acceptable. Bone screening should be undertaken in any individual having a break from sex steroid treatment of greater than six months.

In those patients (typically aged under 40 years) where this regime does not cause suppression of native androgen production (usually clinically evident in the form of ongoing erections) there may be a role for a gonadotrophin-releasing hormone analogue. This should be given in exactly the same manner as employed in prostate cancer, with cyproterone acetate given for the first two weeks of such treatment, at a dose of 50 mg BD.

Patients should be advised that estrogen therapy will probably permanently impair their fertility. They may decide to seek gamete storage.

Estrogen treatment should cease six weeks prior to gender reassignment surgery, to minimize the risk of perioperative thromboembolic events. Gonadotrophin-releasing hormone analogue treatment can, however, be continued up to the point of surgery. ⁴

After gender reassignment surgery estrogen dosing is that required to obtain a serum estradiol level of 400–600 pmol/L. The dose is often less than the preoperative dose unless there has been adjunctive gonadotrophin-releasing hormone analogue treatment, when it tends to remain unchanged. ⁴

The commonest side-effect of estrogen treatment is increased appetite. Weight gain, even to the point of obesity, may consequently follow. This may threaten gender reassignment surgery, for which the patient must have a body mass index of less than 32 and waist measurement of less than 90 cm. ⁶

The most important danger in estrogen treatment is deep vein thrombosis. The risk is 2.6% (20 times population risk) with the majority of thromboembolic events occurring in the first two years of treatment and an ongoing risk of 0.4% per year. This risk is almost certainly increased by smoking and obesity. ⁴

The risk of breast cancer on estrogen therapy seems very low. Only four cases have ever been reported. Progesterone and estrogen therapy probably raises this risk. ⁴

Severe hyperprolectinaemia (prolactin greater than 1000) is rarely caused by estrogen treatment. It usually responds to the dose reduction allowed by adjunctive gonadotrophin-releasing hormone analogue treatment. ⁴

Deranged liver function is rarely caused by estrogen treatment. It usually responds to dose reduction or transdermal delivery. The risk of cholelithiasis, Budd-Chiari syndrome, hepatic adenoma and pancreatitis is probably raised. ⁴

Prostatic malignancy rates seemed to be lowered on estrogen treatment. ⁴

Hormone treatment for born-female patients

The mainstay of hormone treatment for born-female patients is intramuscular androgen treatment. A typical regime would be Sustanon^®, given at a dose of, initially, 250 mg monthly. The aim is to achieve testosterone levels of 25–30 nmol/L a week after dosing and 8–12 nmol/L immediately after dosing, once a steady state has been obtained (after three doses). Peak levels may be adjusted by varying the dose, trough levels by varying the frequency. This frequency dosing usually achieves menopause within three doses. ⁴

The principal side-effect of this treatment is polycythaemia, so haematocrit should be monitored. It is directly proportional to the amount of supraphysiological testosterone being administered, and is more likely in smokers. Treatment is by dose reduction or venesection. ⁴

Hepatic dysfunction occurs in a little of 4% of cases, so hepatic function should be monitored also. It usually responds to dose reduction. ⁴

Lipid profile changes to male, but this does not seem to carry through to increased cardiovascular risk. ⁴

The risk of gynaecological malignancy is probably raised because testosterone causes endometrial hyperplasia in 15% of patients. Two-yearly ultrasonic endometrial thickness monitoring is needed if hysterectomy has not occurred after three years of treatment. ⁴

Androgens treatment seems not to cause insulin resistance or increased risk of ovarian carcinoma. It does seem to worsen the symptoms of sleep apnoea and the rate of cardiac arrhythmia in that condition. ⁴

Surgical treatments in born-male patients

Surgical treatments for born-males are contraindicated unless the patient has been living in a female role, with documented success, for a period of, at the very least, one year and usually two years. ² Most clinics require patients to have been meaningfully occupied for at least half that period of time, and all require authorization from two competent professionals.

Gender reassignment surgery, in the sense of genital surgery, typically consists of:

Penectomy;

This surgery carries the risks of any prolonged ‘lithotomy’ procedure. The main risk specific to neovaginal creation is rectal wall tear with the attendant risk of fistula. The bladder is much more rarely injured. Haematoma is an early complication, as is failure of the vascular supply to the neoclitoris. Later complications include prolapse of the neovagina (in up to 15% of cases), urethral stenosis and intravaginal hair growth. ⁶

The neovagina must be dilated to remain unstenosed. Dilators or sexual intercourse can achieve this. The frequency of dilation falls from three times daily immediately postoperatively to weekly in the longer term. ⁶

Orgasm is reported, postoperatively, in 85% of those patients in whom it was possible preoperatively. ⁶

Augmentation mammoplasty for born-male patients is indicated only when adequate levels of estradiol have been achieved for a period of at least two years, and breast development has ceased. The most common early complications are haematoma, malpositioned implants and altered nipple sensation. Infection is rarer, and requires implant removal.

The most common late complication is capsular contraction (11%). Implants have a finite life, limited by rupture which may be confirmed by ultrasonography. ⁷

Born-male patients who have failed to respond to an adequate level of speech and language therapy may benefit from a cricothyroid approximation aimed at altering the vocal pitch. Tall, thin patients may require a thyroid chondroplasty in order to reduce thyroid cartilage (Adam's apple). These two procedures can be combined, or performed independently. Thyroid chondroplasty alone, from an incision high in the midline neck, carries a small risk of anterior detachment of the vocal cords, with an associated risk to the voice and airway. ^8,9

Surgery for born-female patients

Born-female patients are considered candidates for bilateral mastectomy and male chest reconstruction only if they have lived in a male role for a period of at least a year, and have been undergoing androgen treatment. ² This surgery is more complex than bilateral mastectomy in malignancy, carrying as it does the requirement for a male chest reconstruction. A surgeon who has experience of working in this way is, accordingly, preferred.

The most common complication is haematoma. If the nipple-areola complex has been moved as a free graft it may fail – particularly in smokers. Later, scars may become hypertrophic or keloid. ‘Dog ears’ may need excision at six months postoperatively. ⁷

Genital surgery for born-female patients (phalloplasty) is a complex and multistage procedure with a significant morbidity. About one-third of such patients opt for this surgery. Some patients opt for a metoidioplasty, in which the hypertrophied clitoris is freed from its hood and mobilized to simulate a small penis. This carries a lower rate of complications, but an inferior cosmetic result. There are a number of approaches in phalloplasty. All involve very complex surgery and all should only be undertaken by a surgeon with extensive experience in this challenging area. The commonest of the many possible complications are fistulae and strictures. The latter raise the risk of recurrent urinary tract infections, hairballs and stone formation. ¹⁰

Born-female patients, whether undergoing cosmetic genital surgery or not, should undergo a bilateral oophorectomy and hysterectomy since long-term androgen treatment causes endometrial hyperplasia and nulliparity (common in these patients) is associated with ovarian carcinoma. This surgery can easily be combined with cosmetic genital surgery. If it precedes such surgery the most useful surgical approach would be transvaginal since this avoids those abdominal scars which might render a subsequent phalloplasty difficult. The shortened vagina which may result does not trouble these patients. If a transvaginal approach proves impossible, the next best option would be laparoscopic. If an abdominal incision must be employed, a midline incision is least likely to compromise subsequent phalloplasty. An abdominal transverse suprapubic (Pfannenstiel) incision should be avoided, because it is pathegnomic of gynaecological surgery. ^4,10

There is evidence that once patients are assessed as suitable candidates for gender reassignment surgery their outcome will be better if such surgery is provided promptly rather than in a delayed way. ¹¹ The benefits of prompt gender reassignment surgery have only been demonstrated in patients who have been assessed as suitable for such surgery. There is no evidence that prompt gender reassignment surgery provided before this point is helpful. The clinical impression is that it can be decidedly unhelpful and premature, or referral for surgery has prompted censure. ³