Risk factors

There are a number of risk factors for venous thromboembolism. They fit into a more contemporary and extended version of Virchow’s triad (stasis, hypercoagulability, endothelial damage) and they change the hemostatic balance toward clot formation (thrombophilia). This can be achieved by decreasing blood flow and lowering oxygen tension, by activating the endothelium, by activating inborn or acquired immune responses, by activating blood platelets, or by increasing the number of platelets and red blood cells, or modifying the concentrations of pro- and anticoagulant proteins in the blood. Numerous risk factors are known, which can be divided into genetic and acquired (Table 23.1).

An extended multiple-hit hypothesis implies that more than one risk factor is present at any one time. Superficial vein thrombosis combined with an acquired thrombotic risk factor increases the risk of venous thrombosis 10- to 100-fold [6], but one of the most well-known acquired risk factors is the use of exogenous sex steroids such as COCs and HRT.

Gender differences

A gender difference has been observed. The incidence rates of DVT are higher in younger women during the child-bearing years compared with men at similar age. With advancing age, however, rates are much higher in men than women. A relative risk (RR) of 1.3 in men with an age between 60 and 69 years versus women in the same age group is reported. Also the risk of recurrent venous thrombosis is 2-fold higher in men than in women [7].

Age

Venous thromboembolism has an annual incidence of 1–3 per 10,000 individuals per year (global background rate). It is uncommon in young individuals and becomes more frequent with advancing age. The risk is doubled between the ages of 20 and 40. The incidence is estimated to be approximately 5 per 10,000 per year in women in their 40s, increasing to 6–12 per 10,000 per year in women in their 50s, and is approximately 30–40 per 10,000 per year in women aged 70–80 years. For women in their 80s, the estimated annual risk is approximately 70 per 10,000 [8, 9]. Similarily in the Women’s Health Initiative Study (WHI), users of combined HRT in the sixth, seventh and eighth decades of life had a 2-fold, 4-fold and 7-fold VTE risk respectively [10]. An analogous though less pronounced effect of age was seen in the estrogen-only arm of the WHI [11].

Obesity

The world is experiencing an obesity pandemic, with rates of obesity (body mass index (BMI) > 25) rising for more than two decades. Although obesity has been suggested to be a risk factor for fatal pulmonary embolism for a long time, whether obesity is an independent risk factor for pulmonary embolism or VTE has not been fully determined until recently and presented in a systematic review [12]. Investigations that reported an increased risk because of obesity have been criticized because they failed to control for hospital confinement or other risk factors and although high proportions of patients with venous thromboembolic disease have been found to be obese, the importance of the association is diminished because of the high proportion of obesity in the general population. To date, however, there is notable and consistent evidence of an association of obesity with VTE, more so in women compared with men. The risk appears to be at least double that for normal weight subjects (BMI 20–24.9) [13]. In the WHI, VTE risk increased in both overweight and obese (BMI > 30) HRT users when compared with placebo users in the normal weight group. The risk of venous thromboembolism may be additive when using HRT, and with the continuous and global rise in obesity, this interaction will become a more prevalent risk issue.

Specific risk of HRT

There are a number of different estrogen and progestin preparations in both sequential and continuous combined products. These differ according to the chemical structure of both the estrogen and the progestin component, daily dose and route of administration. The estrogen compound of combined HRT consists of either natural 17β-estradiol (E2), including estradiol valerate (E2V), or the conjugated equine estrogens (CEE). E2 can be administered orally or transdermally. CEE is synthesized from the urine of gravid mares and contains more than 30 known biologically active estrogen compounds, including mainly estrone sulfate and equilin sulfate. CEE is always orally administered. A recent study demonstrated that CEE and E2, administered alone, were associated with different risks of major thrombotic events. Compared with current use of oral E2, current use of oral CEE was associated with a doubling of the risk of DVT [14]. The association of VTE risk with the different pharmacological classes of progestogens has not been elucidated in detail, despite many clinical and epidemiologic investigations. Progestogens include both progesterone, the physiologic molecule synthesized and secreted by the ovary, and synthetic compounds named progestins, which are derived from either progesterone (pregnanes and 19-norpregnanes) or testosterone (19-nortestosterone). To date the only desired effect in relation to HRT is endometrial protection [15]. An indirect comparison within the WHI clinical trials showed that the association of estrogens plus progestogens was more thrombogenic than unopposed estrogens. Indeed, compared with placebo, the VTE risk was doubled in the estrogen plus progestogen clinical trial and not significantly elevated among HRT users in the estrogen-alone clinical trial. Similarily, the Women’s International Study of long-Duration Oestrogen after Menopause (WISDOM) trial has provided a direct comparison of the effect of CEE alone and CEE plus medroxyprogesterone (MPA) on VTE risk among postmenopausal hysterectomized women [3]. Results showed that users of CEE plus MPA had a doubling in thrombotic risk compared with women treated by CEE alone [16]. While MPA may have a thrombogenic effect among postmenopausal women using estrogens, micronized progesterone could be safe with respect to thrombotic risk [3]. The effects of the different types of progestogens on VTE risk has to be further investigated.

Tibolone, a testosterone-derived progestin, has estrogenic, progestogenic and androgenic properties, and can be used alone for treatment of climacteric symptoms. All available studies, although limited in types and number, show that tibolone is not associated with an increased risk of VTE [3,17].

Route of administration

In most investigations of the relationship of venous thromboembolism and HRT, the route of hormone administration has been primarily oral. It has been proposed that orally administered estrogen may exert a prothrombotic effect through the hepatic induction of hemostatic imbalance, as has been observed with COCs. The prothrombotic effect is possibly related to high concentrations of estrogen in the liver due to the “first-pass” effect. Studies that compared oral and transdermal estrogen administration have demonstrated that transdermally administered estrogen has little or no effect in elevating prothrombotic substances and may have beneficial effects on proinflammatory markers, including C-reactive protein, prothrombin activation peptide and antithrombin activity. Also, in contrast to oral estrogen, transdermal estrogen may also have a suppressive effect on tissue plasminogen activator antigen and plasminogen activator inhibitor activity [2]. Most of the clinical data correlating HRT route of administration and VTE risk has come from the EStrogen and THromboEmbolism Risk (ESTHER) Study Group. They have demonstrated that oral but not transdermal estrogen is associated with a 4-fold increased VTE risk. In addition, they noted that norpregnane progestogens may be thrombogenic, whereas micronized progesterone and pregnane derivatives appear safe with respect to thrombotic risk [18]. Similar results were reported elsewhere and of particular importance, in women who were stratified for weight and the presence of prothrombotic mutations [2].

Smoking

No interaction between smoking and HRT was found in the WHI clinical trial [10], although smoking itself is a weak risk factor for incident VTE. Smoking further increases the risk of VTE in women using CHC [19].

Length of use

It is now generally accepted that the risk of VTE in users of HRT decreases with length of use. There is a marked increased risk in thrombosis during the first year of use with the odds ratio as high as 8 (“the timing effect”) [20]. This is likely to be caused by the presence of prothrombotic/thrombophilic abnormalities in these women, as has been shown for COCs. In several studies, a high risk was observed for users of HRT who had coagulation abnormalities, such as APC resistance, increased levels of D-Dimer or high F IX levels. In women who carry the Factor V Leiden mutation there is a 15-fold increased risk of venous thrombosis, analogous to what has been observed for COCs. No such synergy has been demonstrated for carriers of prothrombin 20210A who use HRT. Apart from type and route of administration, comparisons of the thrombosis risks between HRT regimens must therefore also account for length of use and ensure that new users of one agent are truly being compared with new users of another agent. With time, the risk of thrombosis decreases, but always remains higher than non-users. This risk disappears during the first months after stopping treatment.

Thrombophilia

Thrombophilia is a term used to describe a group of conditions in which there is an increased tendency, often repeated and over an extended period of time, for excessive clotting. These include inherited conditions, based on a demonstrated genetic mutation such as factor V Leiden, protein C and S deficiencies, antithrombin deficiency and prothrombin 20210A mutations (which may be suspected on family history); or an acquired condition such as lupus anticoagulant/antiphospholipid or anticardiolipin antibody syndrome, which can occur alone as a manifestation of an autoimmune disorder, or as part of a syndrome such as systemic lupus erythematosus. The presence of an inherited thrombophilia is a major modifier of thrombosis risk in users of both CHC and HRT. However, to qualify, all hereditary thrombophilic defects as similarly strong risk factors might be questioned. The absolute risk of venous thrombosis in factor V Leiden heterozygous carriers is estimated as being 0.15 per 100 person-years, whereas in antithrombin-, protein C-, or protein S-deficient persons, these estimates range from 0.7 to 1.7 per 100 person-years, indicating a considerably higher degree of risk. Risk estimates for thrombophilic women using HRT are less precise compared with estimates performed in CHC users, because fewer women on HRT have been included in the types of retrospective studies that are informative for this situation. In principle the known relative risks for the various thrombophilias should be multiplied by the HRT baseline risk in the relevant age category. The route of administration as well as the progestin component has been clearly shown to modulate/decrease the risk. In general, women known to be carriers of a thrombophilia, or with a positive first-degree family history of venous thrombosis, should be advised not to take oral HRT. However, as part of the shared decision-making process, the gynecologist should weigh the risks against the benefits when prescribing HRT and counsel the patient accordingly, taking into consideration the possible thrombosis-sparing properties of transdermal administration and tibolone.