Tales from the clinic: Hormone replacement therapy (HRT) in perimenopausal women with migraine

Introduction

Migraine prevalence, particularly menstrual migraine, increases during perimenopause but is under-reported, under-diagnosed and under-treated. ¹ Additionally, migraine is associated with an increased frequency of hot flushes and night sweats, particularly during the late menopausal transition. ² Given the benefits of hormone replacement therapy (HRT) on vasomotor symptoms, it is important for healthcare professionals to understand the potential effects of HRT on perimenopausal migraine and the options for effective management. In this edition of ‘Tales from the Clinic’, we consider a relatively typical case of perimenopausal migraine.

Case study

KR first attended the menopause service in October 2019, aged 51. She had a history of monthly premenstrual attacks of episodic migraine without aura, but in the preceding year or so the frequency of attacks had become weekly, and she was also having troublesome hot flashes. Her GP had started her on Evorel Conti® with some improvement in migraine, but she developed troublesome spotting and bleeding. She switched to Evorel Sequi® but had monthly heavy bleeding accompanied by migraine. She had no personal or family medical history of note. Blood pressure was normal, BMI was 26 and she did not smoke. She was not taking any regular medication.

Migraine had started during her teens but was not particularly troublesome until her late 30s, when she first sought help from a healthcare profession. She first noticed a cyclical pattern of attacks when she was 45, but lack of sleep and alcohol were also migraine triggers. Hormonal treatment was considered but was not pursued, as she was able to treat the attacks effectively with symptom treatment alone. She managed the attacks well for several years until her late 40s, when the cyclical attacks became more difficult to treat and were less responsive to medication. She was going through a difficult divorce and assumed that the increase in migraine was due to stress, despite the attacks typically occurring a few days before menstruation.

At the menopause service, she had a 52 mg LNG-IUD fitted and was prescribed Estradot® 50 mcg patches twice weekly, which was increased to Estradot® 100 mcg twice weekly in an attempt to reduce migraine attacks. For the next 3 years, from October 2019 until Autumn 2022, KR reports that she had no bleeding and only occasional migraine attacks with obvious non-hormonal triggers such as stress, tiredness, alcohol and changes in barometric pressure. In Autumn 2022, she experienced a significant escalation in migraine, which she ascribed to increased work and home stresses. Around the same time, there were supply issues with Estradot® so she was switched to Evorel® 100 mcg patches, but these did not stick as well. Despite the stresses resolving in the early months of 2023, migraine attacks continued. In April 2023, she observed that the attacks were occurring monthly, lasted several days, and were more severe and disabling than in recent months, despite her efforts to manage non-hormonal triggers. At review in November 2023, she was advised to reduce the estrogen dose on the basis that at 55, she was likely to be in the postmenopausal period and cycle suppression was no longer required. She reduced the dose initially to 75 mcg and then to 50 mcg, but migraine frequency increased. She also had some night sweats and light bleeding. At her subsequent review in December 2023, her diary showed that the attacks had occurred every 25 days on average. This combination of symptoms and pattern of attacks was suggestive of an ongoing endogenous hormone cycle triggering cyclical migraine attacks. We advised KR to increase the patches back up to 100 mcg twice weekly and have the 52 mg LNG-IUD replaced.

Discussion

Menstrual migraine is defined as regular attacks of migraine starting within a 5-day window including the 2 days before the first day of menstruation and the first 3 days of bleeding, where the association is more than by chance. ³ Menstrual attacks result from one or both of two independent mechanisms – perimenstrual estrogen ‘withdrawal’ during the late luteal phase of the natural menstrual cycle and prostaglandin release during menstruation. ¹ The increase in menstrual migraine prevalence during perimenopause could be accounted for by higher luteal phase estradiol levels during the menopause transition compared to premenopausal luteal phase levels, resulting in a greater premenstrual drop. ⁴ Following menopause, migraine gradually improves with increasing time from the last menstrual period. With respect to the effect of HRT on migraine, two large cross-sectional studies have shown that HRT use is associated with a significant increase in the risk of migraine, which was similar for estrogen-only and combined HRT.^5,6 Small studies using different routes of delivery of estrogen suggest that the increase in migraine is associated with oral estrogen, while transdermal estrogen has a neutral effect. ⁷ There are no data on the effect of different progestogens on migraine, but one randomized controlled study reported that cyclical HRT had a greater adverse effect on migraine compared to continuous combined HRT. ⁸

During the early menopausal transition, menstrual migraine can be prevented with continuous combined hormonal contraception, which effectively manages both estrogen ‘withdrawal’ and prostaglandin triggers, as well as menstrual disorders, vasomotor symptoms in addition to providing effective contraception. ⁹ In late perimenopause, the potential risks of synthetic ethinylestradiol may outweigh the potential benefits, and in the United Kingdom combined hormonal contraception is not currently recommended for women over age 50. ¹⁰ However, other contraceptive methods can also suppress ovarian activity. Of note, ovarian suppression, while highly variable, can occur in 55% of women in the first year following fitting of a 52 mg LNG-IUD, associated with higher serum levels of levonorgestrel compared to women with ovulatory cycles. ¹¹ This effect is the most likely reason for KR’s improvement in migraine. The 100 mcg estradiol patches twice weekly may have also played a role in suppressing ovarian activity given that the initial 50 mcg dose did not appear to be as effective. However, it may be that time for the LNG-IUD to take full effect was the more important parameter, as some research has shown that 200 mcg estradiol patches twice weekly are necessary to achieve suppression of ovarian activity. ¹²

While the range of age at last menstrual period is 45–55 years, with average age at menopause being 52, some research shows that late menopause (≥55 years) may occur in around 10% of women. ¹³ Ovulation is usually suppressed when serum levonorgestrel levels are above 200 pg/mL. ¹⁴ The maximum concentration is reached within 2 weeks after insertion following which levels gradually decline, associated with the declining release rate of levonorgestrel from the intrauterine device. KR’s return of a cyclical pattern of migraine and vasomotor symptoms suggests that she is not yet in the postmenopausal period and the HRT is no longer suppressing ovarian activity. Estrogen fluctuations may have been further exacerbated by the change to Evorel and the patches not sticking properly. We hypothesized that restoring cycle suppression with a new 52 mg LNG-IUD and 100 mcg patches for a further few years should re-establish effective control of KR’s hormonally triggered migraine. She had the 52 mg LNG-IUD replaced mid-January 2024, and at follow-up in mid-February she reported that she was only experiencing mild headaches and no migraine.

This case study highlights the importance of considering and managing the potential underlying mechanism(s) of migraine in perimenopausal women with vasomotor symptoms.