Menopause is a universally experienced event in the natural course of every woman's life. This transition may lead not only to somatic, menstrual and psychological symptomatology but also to increased comorbidities, in particular, reduced BMD and increased cardiovascular disease, all of which may be exacerbated by HIV infection itself or ART. Managing menopause in the context of HIV requires an understanding of the overlap between menopausal and HIV-related symptomatology, knowledge of DDIs to anticipate and facilitate dose titration of MHT where necessary, and skill in counselling on known and unknown risks, particularly, for example, with regard to cardiovascular disease and venous thromboembolism (VTE).
The mean age of women attending the clinic was 49 years and, on average, symptoms were present for 31 months prior to attendance. There was a long delay between onset of symptoms and initial clinic attendance/initiation of MHT. Additionally, the prevalence of POI (18%) was higher than in the general population (1%). Although we cannot make a direct comparison, in contrast to cohorts described in other high-income countries, reported risk factors for early menopause were not evident in this group of patients. None of the women attending the clinic were regular injecting drug users and rates of hepatitis C were low. Smoking prevalence was higher than that of women in the general UK population (18% vs. 12.5%) and in keeping with other studies comparing women living with HIV and HIV-negative women. All the patients attending the service were established on ART and most had an undetectable viral load demonstrating overall good viral suppression. Univariate analysis found no factors or patient characteristics that were significantly associated with menopause before the age of 45, although it is acknowledged that this is a very small sample from a specialist service and hence highlights the need for further, more generalized data before establishing a true link between HIV infection and rates of POI/early menopause.
Although vasomotor symptoms were the most common menopausal symptom, notably over half of patients had psychological symptoms. This is an important consideration when providing integrated care for this group of women with a high prevalence of pre-existing depression (24%). Risk factors for and/or established cardiovascular disease were present in 42%, and there was reduced BMD in two-thirds, all highlighting the need for multidisciplinary teamworking.
Most referrals were from HIV clinicians rather than GPs, possibly due to lack of awareness of the existence of the dedicated menopause service. An alternative explanation could be that women are more likely to report menopausal symptoms to their HIV clinician rather than their primary care provider. Additionally, clinicians may be poorly or ill-informed about menopause and/or the risks and benefits of treatment. A small number of women were investigated for other differential diagnoses such as tuberculosis and cardiovascular disease prior to referral. On review of the case notes, these initial differential diagnoses were not unreasonable as patients presented with symptoms that were atypical for menopause (e.g. weight loss) and a clinical background warranting further investigation before assuming a diagnosis of menopause.
The long delay from onset of symptoms to referral and review may be due, in part, to failure to report symptoms, as menopause is often perceived as a taboo subject among women from sub-Saharan Africa, as well as among many in the UK general population. Other patient factors may be a lack of awareness of available effective treatment for menopausal symptoms or concerns about the risks of taking MHT. This may be overcome by improved access to information about the menopause and connections to peer support from women living with HIV transitioning through the menopause. Currently, the reasons for this long delay are not fully understood; however, patient experience suggests that being caught between HIV clinician and GP plays a significant role.
Oestrogen is metabolized by the cytochrome P450 (CYP) 3A4, CYP1A2 and glucuronidation. Progesterone is metabolized by CYP3A4. The effect of coadministration of MHT with interacting NNRTIs, which are CYP3A4 inducers, is predicted to reduce both oestradiol and progesterone levels. Consequently, higher oestrogen doses may be needed to achieve a therapeutic effect. Lower progestogen levels may potentially result in inadequate endometrial protection, with possible endometrial proliferation, and BTB. The anticipated effect of ritonavir-boosted PIs on MHT is reduced oestrogen exposure (due to induction of CYP1A2 and glucuronidation) and increased progesterone levels (due to inhibition of CYP3A4). Conversely, cobisistat-boosted ART (a selective CYP3A4 inhibitor) is predicted to increase both oestrogen and progesterone exposure.
The clinical significance of DDIs (causing increased progesterone and oestrogen exposure) on the risk of thromboembolic disease and cardiovascular events in postmenopausal women is unknown. Half the women in our small cohort experiencing BTB were on potentially interacting NNRTI regimens. It is our recommendation that women should be switched, where possible, away from interacting ART preparations to non-interacting third agents such as rilpivirine, doravirine and unboosted INSTIs.
Menopausal hormone therapy was acceptable and beneficial to most women in this hospital population of symptomatic, predominantly black, women living with HIV. Nearly 60% (17/29) of women reporting an improvement in symptoms required an increase in their oestrogen dose for better symptom management, of whom 71% were on an enzyme inducer (eight on interacting NNRTI and four on ritonavir-boosted regimen). In these individuals it is possible that known DDIs between oestradiol and enzyme-inducing ART may have resulted in sub-optimal hormone concentrations, although we cannot be certain of this as hormone levels were not measured routinely. Currently, there is no consensus on recommended oestradiol (E2 levels) required for resolution of menopause symptoms. This is because of intraindividual variation and differences related to time of day and site of application of transdermal oestrogen. In addition, E2 levels are of no value in individuals on oral MHT. Given this lack of clarity on the interpretation of E2 levels, non-clinical response to an average MHT dose within a clinically defined period (usually 3 months), would be assumed to signify sub-optimal E2 levels, based on potential and known DDIs between oestradiol and ART. This would warrant an increase in oestrogen dose with further observation of clinical response. There were only two cases of ART modification specifically due to a DDI interfering with effective MHT dosing, reflecting the reality that most drug interactions can be anticipated and managed effectively.
Our data have certain limitations; information on further follow-up or future adverse events on MHT is not available due to the retrospective nature of the study. Selection bias is a potential limiting factor as most of these patients were referred by their HIV clinician and thus more likely to be engaged in care and stable on ART. This cohort may also reflect the more symptomatic individuals rather than the whole spectrum of menopause in women with HIV and may explain the higher rates of early menopause and POI observed. We unfortunately did not have comprehensive information on socioeconomic status, final last menstrual period, mammogram history, BMI or sexual history, which would be helpful in building a richer dataset to better understand the experiences and outcomes of these women as they transition through menopause. It is important that work is done to collect such data.
HIV Medicine. 2022;23(4):426-433. © 2022 Blackwell Publishing