Understanding hormone action: Does route of administration matter?

With the advent of increasing molecular characterization of individual differences, it has become ever more important to understand how we might wisely individualize our approach to hormone use. The need to individualize, in turn, has spurred an interest in determining and understanding more about how estrogens differ. From the clinical point of view, we know that not all women are the same and we know that not all estrogens are the same. But which of the differences between women or hormones are the ones that really matter when making clinical decisions about birth control or menopause management?

When I was in training, we were taught that all estrogens are roughly equivalent. Therefore, we should use the least expensive preparation, i.e. the least expensive oral contraceptive regimen or the least expensive estrogen preparation, depending on the context. For menopause, in most markets the least expensive estrogen was typically conjugated equine estrogens (CEE) and this likely explains why it became the predominant estrogen preparation used in the USA for menopause management. With the availability of transdermal estradiol formulations for menopause management in the early 1990s, the dogma that all estrogens were equivalent was called into question. It was recognized that the transdermal route of administration of estradiol had less hepatic impact than the oral route and that this might confer benefits, including a reduced risk of venous thromboembolism (VTE). This line of reasoning seemed to make sense until the recent report [1] showing that the transdermal contraceptive patch leads to higher rates of VTE than oral contraceptives. Taken together, these observations beg the question of whether route of administration of estrogens does indeed matter.

Why might route of administration matter? Let us review what we know. The oral route of estradiol administration leads to levels in the liver sinusoids that are 4–5 times higher than those found in the systemic circulation. Because the liver makes coagulation factors, it stands to reason, then, that an oral route of administration would elevate the risk of VTE more than a transdermal route of administration. Indeed, the ESTHER (Estrogen and Thromboembolism Risk) Study Group found that the odds ratio for oral vs. transdermal estrogen was 4.2 (confidence interval (CI) 1.5–11.6). Because the ESTHER study was conducted in France, of 271 cases and 610 controls, only two cases and no controls took CEE [2,3]. Most of the oral estrogen users took estradiol in doses ranging from 0.5 to 2.0 mg daily. Further, the mechanistic data on the impact of transdermal vs. oral estradiol on hemostatic variables associated with venous thrombosis [4] support the observational epidemiological data. The ESTHER study also found that women with prothrombotic mutations displayed fewer thrombotic events when given transdermal vs. oral estradiol [5]. In aggregate, the data regarding estradiol are consistent and thus it is reasonable to conclude that the transdermal route of administration of estradiol minimizes the risk of VTE compared with the oral route. The nagging question not answered by the ESTHER study is whether women using transdermal preparations were exposed to less estradiol (lower area under the curve) over time. If so, it might be that it is not route alone that matters for estradiol administration, but a combination of route and overall exposure (dose). If so, one still would expect a higher rate of VTE in oral estrogen users, because estrone sulfate, which has a half-life longer than 24 h, accumulates in the circulation of oral estrogen users but does not do so in those using transdermal estradiol [6].

Does dose (exposure) or route of administration (or both) explain the higher rates of VTE in those using the contraceptive patch? In the study by Cole and colleagues, there was a more than twofold increase in the rate of VTE (incidence rate ratio 2.2, CI 1.3–3.8) among women using transdermal ethinyl estradiol (EE) in a contraceptive patch compared with women using an oral contraceptive containing 35 μg of EE [1]. Suffice it to say that this study has generated a lot of confusion over principles, particularly the understanding that route of administration of estrogen matters. However, the authors did not discuss the mechanisms that might explain these findings in their discussion. Should we abandon the notion that transdermal estrogen reduces the risk of VTE when compared with the oral route of administration? If not, how do we reconcile the observations in menopausal women with those obtained in younger women using contraceptives? Some have suggested that the contraceptive patch delivers a higher dose of EE, but what differs dramatically is the circulating pattern of EE. There is a daily peak and trough in the circulating level of EE in women taking oral contraceptives while circulating levels of EE are steady in those using the patch. While women using the oral contraceptive patch may have slightly more EE exposure than those using oral contraceptives, this would not seem to fully explain the excess risk of VTE unless dose was far more important than route of administration for EE. For the moment, the provisional conclusion has to be that route of administration does not minimize the risk of VTE when the estrogen is EE.

Why would route of administration matter for estradiol and not EE? Part of the explanation might rest with the unique characteristics of EE. Since EE differs from estradiol only in the placement of an ethinyl group at carbon position 16, it is not immediately obvious that these two estrogens would be very different at a cellular or molecular level or that the route of administration should matter for estradiol and not for EE. However, one attribute that the addition of the ethinyl group confers is stability to degradation and metabolism. This stability is partly conferred by the triple bond in the ethinyl group. Since the main site of degradation and metabolism is the liver, it stands to reason that EE would maximally challenge the liver regardless of route of administration. Because the addition of ethinyl to the estradiol moiety also prolongs the half-life, the route of administration might not alter circulating metabolite levels for EE as much as with estradiol. Further, on a molar basis, EE has been estimated to be 400-fold more potent at the level of liver than estradiol. There may be more to it than these considerations; but to cut a long story short, the transdermal route of administration does confer a lower risk of VTE when the estrogen is estradiol but not when the estrogen is EE and this difference in clinical outcomes does not appear to be fully attributable to dose (exposure) alone.

The above discussion is meant to illustrate the concept that no estrogens are the same and to highlight our knowledge gaps. We will likely be afforded a more clinically relevant perspective about the clinical differences between estrogens that are more chemically similar as we learn more about how each differentially harnesses intracellular machinery, such as protein chaperones. It will take even longer to understand how selective estrogen-receptor modulators (SERMs) differ from one another and also from estrogens, especially when it comes to the clinical impact upon important target tissues such as brain, bone and breast. It remains important to elucidate the critical cellular and molecular differences between estrogens and between SERMs, however, so as to maximize benefit and efficacy while minimizing risk in the setting of contraception and menopause management.