Risk of cardiovascular disease according to type of HRT: analysis based on Swedish registry data

For the majority of women, the menopausal transition is accompanied by an incapacitating climacteric syndrome, with its sometimes highly embarrassing flushing. Hormonal menopause treatments (HRT) are the most effective way of alleviating these symptoms. In non-hysterectomized women, they combine an estrogen with a progestin (or progesterone). The estrogen used may be conjugated equine estrogen or estradiol, administered orally or transdermally. Finally, Tibolone has been proven effective in reducing flushing, and can therefore also be used.

However, very few menopausal women (probably symptomatic) use HRT (currently only 6% of menopausal women in France). This is because the largest randomized trial, the WHI study, showed an unfavorable benefit-risk balance for the HRT used (conjugated equine estrogens combined with medroxyprogesterone acetate or conjugated equine estrogens alone in hysterectomized women). This trial has been the subject of much discussion, particularly in view of the fact that the women recruited were elderly (mean age 63) and well beyond the onset of menopause (timing hypothesis).



The aim of this epidemiological study, based on Swedish registers, is to analyze the arterial and venous cardiovascular risks associated with the use of different types of HRT in early menopause.

The study population is drawn from various Swedish registers (basic national statistics, vascular pathologies, cancers, causes of death, prescribed treatments). The methodology was based on the simulation of an emulated target trial recruiting women entering the menopausal period over 2-year periods. The study analyzed 919,614 women aged 50 to 58 between July 2007 and December 2018 who had not used HRT in the two years prior to enrolment in the study, and who were followed up for a period of 2 years. Thus, 138 nested "trials" compared women who initiated one of the six HRT analyzed during this period (HRT: continuous oral-combined, sequential oral-combined, oral estrogen alone, oral estrogen combined with levonorgestrel IUD, tibolone, transdermal combined, transdermal estrogen alone) versus women not initiating HRT during the same period.

Relative risks and their confidence intervals were estimated for venous events (VTE), ischemic heart disease, ischemic stroke, myocardial infarction (MI) and all the above vascular events.

As the methodology simulates a trial, two types of risk are calculated: intention-to-treat and per-protocol.



Confirmation of the neutrality of the transdermal route, and even the cardiovascular benefits, in immediate post-menopausal women.

During the study period, 77,512 women started HRT (8.4%), while 842,102 were non-users. 24,089 women had a vascular event of interest during the follow-up period: 10,360 (43.0%) a cardiac ischemic event, 4,098 (17.0%) a stroke, 4,312 (17.9%) a MI and 9,196 (38.2%) a venous event.

Two points are important in analyzing the results of this study.

- Between 2007 and 2020, the authors observed a significant reduction of over 50% in cardiovascular disease in women aged 50 to 58. This reduction is mainly due to a drop in ischemic heart disease, from 22.2 per 10,000 women in 2007 to 11.8 in 2020. On the other hand, the incidence of venous thromboembolic disease remains stable, at 14.5 per 10,000 women in 2007 and 13.9 in 2020.

- The most frequently initiated HRT is combined oral estrogen therapy, which accounts for 1/3 of all treatment initiations. Transdermal estradiol was the second most frequently initiated HRT (1/5 of all initiations). During the follow-up period, the authors observed a 50% increase in the use of transdermal HRT, with a parallel 50% decrease in the use of oral estrogen therapy and Tibolone.

Compared with women not initiating HRT, women initiating HRT were slightly younger (mean age 53.0 versus 53.9), had a higher level of education, were more likely to live in an urban area (39.8% versus 32.3%), purchased more treatments for heart disease (13.5% versus 10.7%) or hypertension (29.7% versus 27.0%), but had fewer prescriptions for diabetes (5.8% versus 7.8%).

When all data were pooled (all "trials" lasting 2 years), resulting in a total of 154,433 person-years, 664 events were observed among women initiating treatment, leading to an incidence of 4.37 per 1000 person-years versus 3.56 /1000 person-years among non-initiating women.

On an intention-to-treat basis, Tibolone was associated with a significant increase in the risk of cardiovascular disease (hazard ratio 1.52 95% CI; 1.11 - 2.08) and ischemic heart disease (1.46; 1.00 - 2.14) compared with non-initiating women. Combined oral HRT was associated with an increased risk of ischemic heart disease (1.21; 1.00 - 1.46).

With regard to the risk of VTE, continuous oral-combined HRT was associated with an increased risk (1.61; 1.35 - 1.92), as was sequential oral-combined HRT (2.00; 1.61 - 2.49) and oral estrogen-only HRT (1.57; 1.02 - 2.44), but not Tibolone.

Per-protocol analysis showed a significant increase in the risk of stroke (1.97; 1.02 - 3.78) and MI (1.94; 1.01 - 3.73) in Tibolone users.

Combined transdermal estrogen use is associated with a reduced risk of cardiovascular disease and MI (per protocol) and neutrality for other events. Only the risk of VTE appeared to be increased with combined HRT. This latter result could potentially be explained by the type of progestin used in Sweden, essentially norethisterone or medroxyprogesterone acetate and very little natural progesterone. Meta-analyses published in this field show that the combination of transdermal estradiol and progesterone or dydrogesterone is neutral.

Like all registry crossover studies, this study has the strength of the number of women analyzed and a good knowledge of the treatments prescribed and purchased.

But it does have its limitations, notably: - the failure to take account of certain major risk factors, - the lack of precision concerning the date of onset of menopause, - the notion of actual intake of prescribed HRT.



In conclusion

This study confirms the poor cardiovascular tolerance of Tibolone (already suspected in the LIFT trial), and the (short-term) increased cardiovascular risk of oral estrogens, probably due to rapid thrombotic phenomena linked to coagulation changes. It confirms the neutrality of the transdermal route for all cardiovascular events. However, the venous risk of the combined transdermal route needs to be better determined, taking into account the progestin used in this study.

Published in gynéco-online