Abstract
In my last contribution to this series I reflected upon the case of a young woman whose life had literally been turned around with the use of estradiol implants. To balance the picture I will now present another patient for whom we have made a considered decision to move away from implants as the chosen route of delivery.
Maggie has a highly complex medical profile. She has Lupus which is treated with hydroxychloroquine for disease modification and as she is antiphospholipid antibody positive is anticoagulated with Warfarin. She has asthma yet continues to smoke. She has neuropathic pain and takes Diconal, Amitriptyline and Duloxetine for this but is constipated as a result. She has erosive gastritis and is prescribed high-dose proton pump inhibitors.
Maggie is now 49 but had become prematurely menopausal at 38. This had presented with a significant depressive mood change in addition to flushing and sweats. She had found that conventional oral combined hormone replacement therapy eased her symptoms a little but not enough. She had therefore had a levonorgestrel intrauterine system (IUS) inserted 10 years ago and subsequently had been prescribed nasal spray, then gel and then patches but none had been effective. She had eventually been offered implants and had been given estradiol 50 mg and testosterone 100 mg regularly with serum monitoring prior to some but not every procedure. The IUS had been replaced at five years but this had not been a pleasant experience.
Maggie had moved into our area in 2011 when estradiol implants had just been withdrawn from the UK. She was told that her usual six-month replacement would not be possible and was prescribed gels as alternatives. By about four months past replenishment date, the depth of her depression with suicidal ideation concerned her new general practitioner (GP) to the extent that she was referred to the psychiatrists due to the risk to herself that she posed. Maggie wondered whether her hormones could be involved. Following the suggestion of her previous GP and with the blessing of the new one, she came to the menopause clinic.
Maggie described being tearful and irritable and explained that this had crept in over the previous 7–8 months. She had no motivation, was anxious and lacking confidence, panicky and the effort to come to clinic had been almost overwhelming. Her mood was clearly depressed and she was able to relate thoughts of self harm. Unusually with severe depression, she had good insight into her situation.
Her last implants had been about 11 months earlier and she reported that occasionally the benefit had been lost early with previous implants. She had tried but stopped testosterone gel but had continued to supplement using single-use estradiol 0.1% gel 1.0 g sachets (Sandrena® gel, Orion Corporation, Espoo, Finland) at one a day. Flushing and sleep disturbance did occur but on their own would have been tolerable and not as bad as they had been in the past. She reported difficulty with concentration and particularly verbal memory. There was urinary frequency and urgency but no leakage. Vaginal dryness was not a marked issue but intercourse was uncomfortable and this had become worse since stopping the testosterone. She was slow to respond sexually and had little interest. She had no breast symptoms and no bleeding. An ultrasound scan two years earlier (following light spotting) had been unremarkable and there had been no further episodes.
I was provided with serum levels taken nearly six months earlier showing a serum estradiol of 308 pmol/L (reflecting only the residual implant) and total testosterone of 0.3 nmol/L. Testim® gel (Ferring Pharmaceuticals Ltd, West Drayton, UK) (testosterone 0.1%) had been used at the time at low frequency. I had no sex harmone-binding globulin (SHBG) level and was therefore unable to calculate a free androgen index. It was reasonable however to infer that the testosterone level was low with the gel and likely to be lower without. The effect of the estradiol implant could only have dropped since the test and it appeared highly likely that despite the added gel, replacement remained suboptimal.
I endorse the adage that medicine in general and menopausal medicine in particular is not an exact science. Occasionally though, some science can help clinical decision-making. I work on a rule of thumb that estradiol 150 pmol/L is likely to be bone sparing but recognize that women in their 40s may need a serum level of 500–600 pmol/L to feel well. This is broadly the mean estradiol level derived from plotting serum measurement against the day in a normal menstrual cycle. A few younger women may need slightly more. Lower levels seem generally to be associated with satisfactory symptom relief as women get older. I therefore do monitor serum levels in women requesting implants but otherwise only if high standard doses of transdermals appear ineffective. This is to determine whether an increase can be justified as an acceptable risk and to maintain exposure within the normal reproductive range.
Maggie appeared to need an increase, with pharmacokinetic factors attributable to the enzyme inducing concomitant medication as the most plausible explanation. I asked her increase the Sandrena® 1.0 g sachets to two twice a day and to resume using the Testim® gel at 1/7 of a 5.0 g sachet a day (such that each would last a week). I explained that this was an off license use. This product has a screw cap which means that fractionation is practically easier than the alternative but otherwise similar product on the market.
Despite this plan, Maggie asked for the implants to be repeated. Acceding to this, we planned for this to be done four weeks later. The medium term management plan was to achieve satisfactory estradiol replacement and then replace the IUS. Vaginal estrogen was additionally prescribed as estradiol 25 μg vaginal tablets. This would both ease symptoms and prepare the cervix for IUS replacement. I wanted to ensure that the procedure would be as comfortable as possible and to ease dilation given the warfarin.
Four weeks later Maggie was much better and had become alarmed by how close she had been to committing suicide. Flushes were minimal, sleep not bad and she thought that her memory had improved. She still had some irritative bladder symptoms but she was she was interested in sex, able to respond and comfortable. She had noticed that her aches and pains (that she had been too overwhelmed to tell me about before) had eased. She was trying to exercise and lose weight.
Hoping to maintain this benefit, after full discussion and with informed consent I inserted implants of estradiol 50 mg (Bartor Pharmaceuticals, Rye, NY, USA) and testosterone 100 mg (Organon, Cambridge, UK). I did not expect to see her again for a further two months at which point we would replace the IUS. I thought that she would only need the vaginal estrogen until then.
My confidence was misplaced and responding to a call of distress I saw her two weeks later. It had transpired that rather than the international normalized ratio of 2.6 which was normal for her it had been 4.7 at the time that the implants were inserted. She had developed a small haematoma which had then become infected, her mood symptoms had returned and it seemed reasonable to infer that absorption from the fluid of the haematoma had been minimal. The oily crystalline pellet seemingly requires fat for absorption of the steroid hormones. We reverted to the previous regimen using estradiol and testosterone gels while the haematoma resolved but a further two weeks later the pellets were extruded.
This is my first experience of such a problem but the learning point is not to accept the patient's reassurance that anticoagulation is appropriate and to measure prior to even minor procedures. I was therefore very cautious prior to the IUS change, particularly since the previous time had been painful and difficult. Maggie stopped the Warfarin four days prior to the planned fit and her INR was 1.3 the day before. Intracervical anaesthesia using Citanest 3% and Octapressin® (marketed by DENTSPLY Ltd, Addlestone, UK) 2 × 2.2 mL was an informed decision and subsequently the procedure was entirely unremarkable.
Chatting during the procedure revealed that all of her symptoms had completely resolved and specifically, both mood and sex were good. Serum levels indicated however, that replacement was too high (serum estradiol 1203 pmol/L and testosterone 10.3 nmol/L though again no SHBG). Clearly the gels could achieve the desired results and if we could adjust this we had the potential to achieve stability without the wearing off effect seen with implants. I asked her to reduce progressively to three Sandrena® 1.0 g sachets daily and Testim® 5.0 g used over 14 days.
Six weeks later all remained well. The IUS was confirmed to be appropriately sited and there had been no bleeding or pain. The free androgen index was 1.2% – in the middle of the female range and optimal as far as I was concerned. The estradiol had dropped to 816 pmol/L – still a little higher than I was looking for. It transpired that the test had been done at four Sandrena sachets daily. The reduction from previous levels being presumably an effect of reduced peripheral conversion from androgens. Maggie had since reduced her estradiol 1.0 g sachets to three daily and had not lost benefit.
We have therefore ended up with physiological replace ment and satisfactory symptom relief using an alternative to the implants. In this woman, implants had failed to provide the stability that is usually seen. The best solution was once again an individual one: that which was most effective for the patient involved.