Abstract
“I would be grateful if you could see this 50 year old woman in your clinic. She has a longstanding mood disorder. She describes herself as always being feisty and prone to extremes of mood and has often drunk to excess to blot out days when she feels bad. More recently her mood seems to have been more unstable resulting in antisocial behaviour, where she has been arrested, suicidal ideation and at other times is very over stimulated. Worsening of her mood has coincided with perimenopause and for a short while was on HRT which possibly seemed to improve matters. However there is a marked family history of breast cancer, her mother developed it in her 30’s and her sister had ovarian cancer at the age of 28. She has now been discharged from the family history clinic and screening there and is being screened by the NHS breast programme 3 yearly. She had a vaginal repair earlier this year. She is awaiting assessment by the mental health services. I would however be grateful for your input on the level of risk involved with HRT and what you feel the potential benefits are in her. Many thanks”.
Along with this excellent summary, I was sent a list of significant previous medical history and prescribing such that I already knew this lady had been given both a serotonin uptake inhibitor (citalopram) and a hypnotic (zopiclone). In my view, this is exactly the place of a menopause clinic; to support primary care (and the patient) when evaluation is beyond the experience of my colleagues.
Unfortunately, due to demand pressure, it was five months before she was seen in the clinic. In that time, the dose of citalopram had been increased to 70 mg daily. Tolteridine 1 mg at night had been added. Zopiclone was only being taken occasionally, but she was also using pregaberlin 20 mg twice daily.
A structured history was taken which revealed
that having had no periods for six months prior to referral, these had returned monthly for four months. However, she had just had a delayed and heavy bleed that necessitated her to change sanitary protection hourly and to sleep on a bath towel to protect the bed; severe flushes and sweats which were both frequent and intense but which eased a little after a period. These caused her to wake at night, dripping, such that she had to change bedding; a diagnosis of depression having seen both a psychiatrist and a psychologist with consequent increase in the citalopram. In addition, she reported having been very irritable and intolerant of others before every period since her daughter had been born 31 years earlier; difficulty with memory and concentration that was consistently worse before a period; continued problems with overactive bladder which were worse at menstruation. She was self-catheterising twice daily as well as taking tolteridine; that she remained sexually active. She felt that the mesh repair had been successful. She had not been offered vaginal estrogen but reported no dryness.
This picture was interpreted as indicative of perimenopause, with the latest heavy bleed likely to be anovulatory. There appeared to be a significant hormonal component to the mood change. The severe vasomotor symptoms were likely to be due to estrogen deficiency but might have been exacerbated by the high-dose selective serotonin reuptake inhibitor (SSRI). While there was no overt vaginal dryness, the bladder symptoms could be indicative of atrophy. Estrogen replacement appeared likely to help and significantly, Lynda's husband reported that she had been much better when hormone replacement therapy (HRT) was previously prescribed. She could not remember what this was other than it was a tablet. She had, however, only had a month before another doctor had decided that with her family history this was too much of a risk.
Was it?
Unpicking a family history can be tricky. One of Lynda's sisters had had a hysterectomy at 28 with no other treatment and Lynda seemed to think that her problem was related to her uterus and was not ovarian cancer. Though details were sketchy, it seemed likely that the surgery was at least partly indicated by early cervical malignancy. As such it would represent neither a genetic risk nor a contraindication. I found out that Lynda's mother had had a breast lump at age 50, which was removed, and she had had no other treatment. She was still alive at 76. This suggests that it was either very low grade or benign. There was no other breast cancer in the family though one sister had had a benign lump removed. One first-degree affected family member diagnosed at 50 tends to suggest that a high-risk gene is unlikely to be implicated. I was encouraged that our local breast unit was of the opinion that she had no greater than a population risk and was suitable for routine NHS breast screening surveillance. Standard population risk for the use of hormone replacement could therefore be advised.
We could go further in exploring the implication of family history to the risk-benefit analysis. Lynda's mother had established osteoporosis with vertebral collapse and hence an argument could be developed that the protective effect of HRT on bone might be valuable. There was no significant family history of thrombotic disease as one aunt having had a deep-vein thrombosis (DVT) in her 50 s would be too remote to represent a significant genetic risk.
My conclusion was that impact of Lynda's family history on her cancer risk did not appear to increase it above the population baseline. Any additional risk from exogenous hormones would therefore be no greater than for anyone else. With an onerous symptom profile as well as benefit of bone protection, an informed decision to try HRT again was reasonable.
Given the high levels of concomitant medication, I suggested starting with an estradiol 50 µg/day matrix patch to be changed twice a week with medroxyprogesterone acetate 10 mg from days 15–28 for opposition and bleeding control. I also suggested estradiol 25 µg vaginal tablets to help bladder irritability.
Five months further on, she was much improved though still had some flushes and cyclical mood change. She was, however, coping much better with this. In the interim, she had been reviewed by the gynaecology team and offered a bilateral salpingo-oophorectomy (BSO) at the time of further vaginal reconstructive surgery. We discussed the implications of this and agreed that it was appropriate. The patches were increased to 75 µg/day and we discussed the possibility that further adjustment might be needed after surgery.
Two months later, she rang me to say that she had had the BSO and an anterior vaginal wall repair. She had been catheterised in the two weeks since surgery. She had had a Levonorgestrel Intra Uterine System (IUS) inserted at the same time and asked what to do about her HRT. We agreed that she should continue to use 75 µg patches for a little longer but that these might need to be increased. She could resume use of the vaginal estrogen when it was comfortable to use the inserter but that oral progestogen was not needed.
Having moved house, she did not receive her next appointment. I was however able to advise her GP regarding symptoms that would justify an increase and asked that if this was still insufficient, to check serum levels.
Two months later, I was sent a result showing that when using 100 µg/24-hour patches, this resulted in a serum estradiol of 149 pmol/L. This is likely to be just bone sparing but being at the bottom of the range, would allow an increase as she remained symptomatic. Though total testosterone was low at 0.3 nmol/L, her Sex Hormone Binding Globulin (SHBG) was also low at 28.8 nmol/L, resulting in a free androgen index that was close to the female median. An increase in estradiol patches was advised, but the message was not relayed.
This was because the IUS had been removed on grounds of persistent nuisance bleeding and because Lynda thought that it was causing pain. The GP who removed it told Lynda to stop the patches. This rapidly resulted in a desperate patient with dreadful flushes, sleep disturbance and mood swings, who was forgetful, not interested in sex and ached all over but was not bleeding.
The repeat blood test intended to monitor the increased dose showed instead a serum estradiol of <18 pmol/L.
Having discussed the options of opposed estrogen or Tibolone, we elected to return to the former. Lynda was prescribed two estradiol 75 µg/24 h patches to use together twice a week with medroxyprogesterone acetate 5 mg daily for continuous opposition (this had previously been tolerated) and estradiol 10 µg vaginal tablets twice a week (as the 25 µg were no longer available).
Three months later, this has proven to be highly successful. Flushes have resolved completely. Sleep is good, mood stable and dose reduction of the citalopram has begun. There is no urinary leakage, urgency or infection and sex is reported as good. There has been no further bleeding.
Lynda is delighted and feels that she is back in control of her life and can enjoy her family and grandchildren. At 51 and a half, she is exactly at mean menopausal age and can be advised that it would be reasonable to continue this regimen for the next five years with annual review. Reduction, slowly and progressively should be attempted from the mid 50 s. We would aim at all times at the lowest effective dose. The emphasis is placed on the word effective.
The process has been successful for this particular patient and has demonstrated the value of a referral clinic. It would have been so easy for Lynda's GP to have been overwhelmed by the perceived risk and not to have made the initial referral. This would have denied her the opportunity she needed.
Family history is important, but its implication needs to be understood and evidence-based. On this occasion, it did justify more thorough evaluation. Once the status of genetic probability is established, it can then be factored into the balancing act that determines where risk and benefit lie.