There is an estimated two- to threefold increase in the risk of venous thromboembolism with the use of oral hormonal contraceptives.1 Absolute rates of thromboembolism with contraceptives are low, at 2–3 per 10,000 women annually. Although the absolute risk is low, because thromboembolism can be fatal, it is important for women to understand the potential risk associated with contraceptives. There is also increased risk of atherosclerotic events, with a recent meta-analysis reporting that use of low-dose oral contraceptives is associated with twofold increased risk of myocardial infarction and stroke (odds ratio [OR]=2.01, 95% confidence interval [CI]=1.63–2.48).2
Unlike oral contraceptives, there is limited data on the risk of adverse vascular outcomes with the use of non-oral hormonal contraceptives. In studies of the biochemical effects of hormone therapy in postmenopausal women, transdermal treatment had smaller or no effects on coagulation and inflammation parameters compared with oral hormone treatments, presumably due to lack of first-pass effects on the liver with transdermal medications.3 It could be theorized that non-oral hormonal medication in postmenopausal women would present a lower risk of venous thromboembolism and cardiovascular events, and observational studies support this view.5
Two epidemiologic studies have examined the effect of transdermal contraceptives on venous and arterial thrombosis with varied results.6,7 The United States Prescribing Information was recently updated on this transdermal hormonal contraceptive. The new labeling states that women using this medication are exposed to higher steady state estradiol concentrations compared with an oral hormonal contraceptive containing the same estrogen and progestin. Despite this warning, it remains unknown whether the transdermal contraceptive has the same risk of thrombosis as oral contraceptives. It is possible that a higher steady state of estradiol, despite a lower peak level, may lead to more thromboembolic complications, similar to the increased risk of venous thromboembolism observed in users of contraceptives containing high dose of estrogens as compared with users of low-dose formulations.8
Several coagulation factor abnormalities are associated with increased risk of venous thrombosis, and characteristic changes in coagulation induced by oral contraceptives include lowering of protein S and antithrombin and increase in D-dimer.9–13 Oral contraceptives containing third-generation progestins cause a more pronounced increase in resistance to activated protein C and are associated with a higher risk of venous thrombosis than other contraceptives.14,15 Hormonal medications also increase C-reactive protein, a risk factor for arterial thrombosis.16
Comparing the vascular risk of transdermal contraceptives with that of oral contraceptives, using coagulation parameters, offers insight and support to the preliminary epidemiologic studies of these medications. This study investigated the effect of oral and transdermal contraceptives containing the same hormonal agents on biomarkers that are associated with increased risk of venous and arterial thrombosis.
PARTICIPANTS AND METHODS
Twenty-four nonpregnant women, 18–35 years of age, who had not been on hormonal contraception for at least 2 months before the study or who were least 3 months postpartum and nonlactating, were enrolled in this study between October 2003 and February 2005. The study protocol was approved by the University of Vermont Institutional Review Board, and all participants gave written informed consent. Exclusion criteria for participation were personal or family history of venous thromboembolism or coagulation disorders, uncontrolled hypertension, cardiovascular disease, complicated migraine headaches, breast cancer, diabetes, abnormal uterine bleeding, liver disease, pregnancy within 3 months, or desire for pregnancy in less than 6 months. Subjects were not screened for coagulation disorders before the study (Fig. 1).
This study was a randomized investigator-blind, crossover clinical trial. The daily oral contraceptive contained 35 mcg ethinyl estradiol and 250 mcg norgestimate (Ortho Cyclen, Ortho-McNeil Pharmaceutical, Raritan, NJ). The weekly transdermal hormonal contraceptive contained 0.75 mg ethinyl estradiol and 6.0 mg norelgestromin (Ortho Evra, Ortho-McNeil Pharmaceutical). Norelgestromin is the active progestin metabolite of orally administered norgestimate. For the oral form, the average steady state plasma concentration (Css) of ethinyl estradiol was 49.3 pg/mL and of norelgestromin 0.73 ng/mL, and for the transdermal form, these concentrations were 80.0 pg/mL and 0.888 ng/mL, respectively, at the end of the second month of use (Ortho Evra [package insert]. Raritan, NJ: Ortho-McNeil Pharmaceutical, Inc; Revised September 2006). The average weekly exposure, calculated as area under the curve (AUC0–168 pg·h–1·mL–1) for ethinyl estradiol was 55% higher with transdermal contraceptive than with the oral contraceptive. The maximum ethinyl estradiol and norelgestromin levels (Cmax) were 133 pg/mL and 2.16 ng/mL for the oral form and 97.4 pg/mL and 1.12 ng/mL for the transdermal form (Ortho Evra [package insert]). The application location of the patch does not alter the Css or Cmax.17 Medications were supplied to the patients by a research nurse and the completed packets were returned to document compliance. The contraceptives and the funding for the study were supplied by research grants from the University of Vermont Department of Obstetrics and Gynecology.
Before enrollment, participants underwent a physical examination including gynecologic examination. Women on hormonal contraceptives at recruitment were given barrier contraceptives for two months before starting the study. Participants were assigned a random identification number which indicated the sequence in which transdermal or oral contraceptives would be given. The daily oral contraceptive or weekly transdermal contraceptive was given with the typical dosing of 3 weeks of active treatment followed by 1 week without hormone use. After the first 2 months of hormonal contraceptive, each participant returned to barrier contraceptive for a 2-month washout period, then received 2 months of the alternative hormonal contraceptive.
Sample size was determined based on the number of subjects required to document a significant difference in coagulation parameters in studies examining the effect of oral contraceptives. Before the study, it was established that 20 subjects were required to achieve adequate power (1–β=0.8) to find relevant differences in the coagulation parameters tested.
Participants were randomized by a computer-generated list of random numbers, using a block randomization procedure, and this list was maintained by the research coordinator. Medications were supplied to subjects in the order indicated by their assignment. Medication packaging was returned to assure completion of each treatment arm. The investigators and laboratory personnel were blind to the order of contraceptive assigned for each participant.
There were four phlebotomies performed for each participant. Blood was drawn on menstrual cycle days 18–21 before the study, within 4 days of either the last pill or removal of the last patch, and on cycle days 18–21 in the second month of the washout period. Blood was collected with minimal stasis in Vacutainer tubes using standardized methods, immediately placed on ice, and centrifuged at 4°C. Plasma and serum was stored at –70°C until completion of the study.
Laboratory assays were performed in batch, with each participant's serial samples analyzed in the same run. D-dimer, von Willebrand factor, and antithrombin were measured using immunoturbidimetric assays on the STA-R analyzer (Liatest D-Di, Liatest vWF, Liatest ATIII; Diagnostica Stago, Parsippany, NJ) with coefficients of variation (CVs) of 3.0%, 3.85%, and 4.0–8.0% respectively. Factor VIII was determined by measuring the clotting time of the sample in factor VIII-deficient plasma (STA-Deficient VIII; Diagnostica Stago) with a CV of 3.5%. Free and total protein S were measured by immunoassay (Asserachrom Free and Total Protein S, Diagnostica Stago) with CVs of 14.0% and 4.0%, respectively. Fibrinogen and C-reactive protein were measured by immunonephelometry using the BNII instrument (N Antiserum to Human Fibrinogen, N High Sensitivity CRP; Dade-Behring, Deerfield, IL) with CVs of 2.3–4% and 2.6%, respectively.
The normalized activated protein C sensitivity ratio (nAPCsr) was measured by using two thrombin generation assays. In the first assay, coagulation was triggered in defibrinated plasma with tissue factor in the presence and absence of activated protein C, and the amount of thrombin captured in complex with α2-macroglobulin over 30 minutes was taken as a measure for thrombin generation and used to calculate the nAPCsr (nAPCsr α2M-IIa).14 In the second assay, the nAPCsr (nAPCsr calibrated automated thrombinography) was determined by measuring thrombin generation in the presence and absence of activated protein C in full plasma in real time with a fluorogenic thrombin substrate using calibrated automated thrombinography.18 The CVs of the nAPCsr α2M-IIa and nAPCsr calibrated automated thrombinography were 3.5% and 7%, respectively.
Repeated measures analyses of variance corresponding to a crossover design incorporating both baseline and washout periods were used to determine the significance associated with differences between the hormonal contraceptive treatments for each of the biomarkers tested and to evaluate potential carryover effects.19 Preplanned contrasts were used to test for differences between the pretreatment periods corresponding to oral and transdermal active treatments. Contrasts were also used to test for changes from pretreatment to active treatment within each treatment and to determine whether these changes were parallel across the two active treatments. Biomarkers that had nonnormal distributions based on residual plots were log transformed before analysis. Analyses were performed using SAS 8.2 (SAS Institute, Cary, NC). Statistical significance was determined based on α=0.05.
The mean (standard deviation) age of the 24 participants was 26.0 (4.9) years, weight was 144 (26) pounds, and body mass index was 23.8 (4.2). Six women had previous pregnancies (range 1–3); most recent pregnancy was 7 months before initiation of the study. None of the subjects were current smokers.
Thirty-three women where screened for the study: two failed screening, three were randomized and dropped out before taking medication, three dropped out after initiation of the study medication, one became pregnant during the washout period, and the remaining 24 women completed the trial. Assays were performed only on subjects who completed the trial.
Table 1 shows the effects of each contraceptive on the studied biomarkers. Figure 2 demonstrates all parameters with significant changes from baseline during the use of both hormonal contraceptives. Comparison of the baseline and washout values did not demonstrate any carryover effect of the treatments (P>.10 for all outcomes). Total and free protein S and antithrombin significantly declined from pretreatment levels with both hormonal contraceptives (Fig. 2A). These declines represented decreases in the three outcome measures of 19% and 13% (total protein S), 19% and 11% (free protein S), and 13% and 10% (antithrombin) for transdermal and oral treatment, respectively. Significant increases from pretreatment levels were observed for fibrinogen and C-reactive protein for both contraceptives (8% and 10% and 220% and 292% for transdermal and oral treatment, respectively, Fig. 2B). For both contraceptives, there was also a substantial increase of the nAPCsr for both activated protein C (APC) resistance assays, with an increase of 81% and 61% in nAPCsr α2M-IIa and 102% and 68% in nAPCsr calibrated automated thrombinography for transdermal and oral treatment, respectively (Fig. 2C). D-dimer significantly increased with the oral contraceptive but not with the transdermal contraceptive. However, mean levels for the active phases of the two contraceptives were similar, with pretreatment levels being 23% lower before the oral compared with the transdermal (P=.04). D-dimer was the only biomarker for which pretreatment levels for the two contraceptives were significantly different. There was no effect of either contraceptive on levels of factor VIII or von Willebrand factor.
When comparing the effects of oral and transdermal treatments, there were no significant differences in the declines observed in total protein S or antithrombin from their corresponding pretreatment periods. There was also no significant difference in the observed increases in fibrinogen or C-reactive protein for the two contraceptives (Table 1). There was evidence of a greater effect of transdermal compared with oral treatment in the decline in free protein S (P=.07) and for the increase in nAPCsr for both APC resistance assays (nAPCsr α2M-IIa, P=.06 and nAPCsr calibrated automated thrombinography, P=.03).
The major finding of this study is that transdermal and oral contraceptives containing ethinyl estradiol and norgestimate (oral) or norelgestromin (transdermal) have similar effects on biomarkers of vascular disease risk after 2 months of treatment. The coagulation parameters studied are known to be affected by exogenous estrogen and progestin use, and they are all risk factors for venous thromboembolism or myocardial infarction. Both medications were associated with decreases in total and free protein S and antithrombin, and increases in fibrinogen, C-reactive protein, and APC resistance. Neither treatment had an effect on factor VIII or von Willebrand factor. Only oral treatment raised D-dimer, but the D-dimer level before oral treatment was significantly lower than that before transdermal treatment, and the levels during the treatment phases with the two contraceptives were similar. The largest observed effects were for C-reactive protein and APC resistance. Compared with oral treatment, transdermal treatment had more pronounced effects on free protein S and on the nAPCsr determined with the α-2 macroglobulin-thrombin end point as well as the calibrated automated thrombinography assay.
A recent preliminary study from the Boston Collaborative Drug Surveillance Program reported a similar risk of venous thrombosis with the transdermal contraceptive studied here compared with oral contraceptives.6 Of more concern was a recent epidemiologic study conducted by i3 Drug Safety, showing a twofold increased risk of venous thrombosis in the users of the transdermal contraceptive compared with oral contraceptives. In this study there were not sufficient numbers to evaluate the risk of arterial events, but the authors expressed concern regarding the thrombotic risk with the transdermal contraceptive.7 These reports contrast with data on hormonal therapy after menopause, where the transdermal route may not be associated with a lower risk of deep vein thrombosis.3
Our findings, using vascular risk biomarkers, support the hypothesis that the risks of venous thrombosis and cardiovascular disease are increased with the use of both oral and transdermal hormonal contraceptives containing these hormonal formulations. Based on the most sensitive assay, APC resistance, this study suggests a potentially higher risk of venous thrombosis with the transdermal contraceptive. We recognize that further epidemiologic studies are required to confirm this hypothesis.
The mechanism of increased venous thrombosis risk with oral contraceptives is thought to relate partly to the first pass of estrogens and progestins through the liver with resultant changes in coagulation factor levels.8 Transdermal postmenopausal hormonal medications may not have the same risk of thrombosis as oral hormone therapy,3 presumably because they do not have first-pass liver effects on coagulation factors. It is unknown whether similar mechanisms might play a role in modulating risk with transdermal contraceptives, but our findings suggest that this route of administration would not attenuate the risk compared with oral agents. Our findings of similar or larger effects on the biomarkers we evaluated may be due to the known higher sustained endogenous estrogen and progestin levels observed with transdermal compared with oral treatment, even though peak hormone levels are higher with oral than transdermal treatment (Ortho Evra [package insert]). These higher hormone levels over time may cause changes in coagulation function independent of first-pass liver effects. Alternatively, the effects of oral hormones on coagulation may not be due to first-pass liver effects.
The current study examines the effect of transdermal and oral contraceptives containing the same hormone formulations on a parsimonious group of biomarkers known to be associated with vascular risk and to be affected by estrogen therapies. Previous studies have demonstrated that changes in these coagulation factors are associated with an increased risk of venous thromboembolism.9–13 Oral contraceptives adversely affect coagulation factors, including those evaluated in this study: antithrombin, total and free protein S, D-dimer, and fibrinogen. In addition, elevated C-reactive protein is associated with an increased risk of myocardial infarction and stroke, and oral contraceptives are known to increase C-reactive protein.16,20 Induction of activated protein C resistance seems to be a key factor distinguishing differential risks of hormonal medications.21,22 In this study we observed greater effects of transdermal compared with oral contraceptives on APCsr determined with two different assays. These findings suggest that this transdermal contraceptive has at least a thrombosis risk that is equivalent to oral contraceptives, with the possibility of even greater thrombosis risk with this non-oral administration. Further clinical studies are required to evaluate this risk.
The strengths of this study include the randomized crossover design. There were no effects of the order of treatment. Standardized methods were used for the timing and methods of blood collection and for analysis of all samples, which was performed in batch. Importantly, this study also used transdermal and oral contraceptives containing similar estrogen and progestins, allowing study of the difference by route of administration. The washout period of 2 months was sufficient to allow all parameters to return to the prestudy levels, assuring that there was no effect of the first medication on the results with the alternative contraceptive.
Interpretation of our findings is limited by the characteristics of the subjects enrolled. All were 35 years of age or less and were nonsmokers. Future studies should consider subjects that more closely mirror the general population because there may be different effects of these medications in other groups of women. For two of the assays, free protein S and nAPCsr α2M-IIa, the differences in effect by route of administration approached, but did not reach, statistical significance. This suggests insufficient power to demonstrate the true effect of these hormonal contraceptives on these biomarkers. A larger sample size might improve the confidence of some of the findings comparing the two contraceptives. Future studies examining ethinyl estradiol, progestin, and sex hormone binding globulin levels may offer insight into the mechanisms of our observed biomarker changes with these medications.
This study examined vascular risk biomarkers as an intermediate outcome measure of venous thromboembolism risk. In previous studies, these biomarkers have identified mechanisms of the thrombotic risk of oral contraceptives, such as the third generation progestins,14 which are associated with higher risk of thrombotic events than other contraceptives. Indeed, the findings in this study support preliminary research demonstrating no decrease in risk6 and a potential increase in risk7 of thrombosis with this transdermal contraceptive.
The risk of hormonal contraceptives is low and the benefits many. An increased risk of venous thromboembolism from 1 in 10,000 to 2–3 in 10,000 women yearly on contraceptives must be compared with the risk of venous thromboembolism of 6 in 10,000 women yearly during pregnancy. The current study shows a similar effect of the transdermal and oral hormonal contraceptives investigated on biomarkers of vascular risk, with a suggestion of a greater adverse effect of transdermal medication. Results suggest that there is no safety advantage of the currently available transdermal hormonal contraceptive. Women should be counseled regarding the vascular risks of all hormonal contraceptive medications despite the route of administration.
1. Vandenbroucke JP, Rosing J, Bloemenkamp KW, Middeldorp S, Helmerhorst FM, Bouma BN, et al. Oral contraceptives and the risk of venous thrombosis. N Engl J Med 2001;344:1527–35.
2. Baillargeon JP, McClish DK, Essah PA, Nestler JE. Association between the current use of low-dose oral contraceptives and cardiovascular arterial disease: a meta-analysis. J Clin Endocrin Metab 2005;90:3863–70.
3. Scarabin PY, Oger E, Plu-Bureau G; Estrogen and THromboEmbolism Risk Study Group. Differential association of oral and transdermal oestrogen-replacement therapy with venous thromboembolism risk. Lancet 2003;362:428–32.
4. Perera M, Sattar N, Petrie JR, Hillier C, Small M, Connell JM, et al. The effects of transdermal estradiol in combination with oral norethisterone on lipoproteins, coagulation, and endothelial markers in postmenopausal women with type 2 diabetes: a randomized, placebo-controlled study. J Clin Endocrin Metab 2001;86:1140–3.
5. Lowe GD, Upton MN, Rumley A, McConnachie A, O'Reilly DS, Watt GC. Different effects of oral and transdermal hormone replacement therapies on factor IX, APC resistance, t-PA, PAI and C-reactive protein—a cross-sectional population survey. Thromb Haemost 2001;86:550–6.
6. Jick SS, Kaye JA, Russmann S, Jick H. Risk of nonfatal venous thromboembolism in women using a contraceptive transdermal patch and oral contraceptives containing norgestimate and 35 mcg of ethinyl estradiol. Contraception 2006;73:223–8.
7. Cole JA, Norman H, Doherty M, Walker AM. Venous thromboembolism, myocardial infarction, and stroke among transdermal contraceptive system users. Obstet Gynecol 2007;109:339–46.
8. Humpel M, Wendt H, Pommerenke G, Weiss C, Speck U. Investigations of pharmacokinetics of levonorgestrel to specific consideration of a possible first-pass effect in women. Contraception 1978;17:207–20.
9. Tans G, Curvers J, Middeldorp S, Thomassen MC, Meijers JC, Prins MH, et al. A randomized cross-over study on the effects of levonorgestrel- and desogestrel-containing oral contraceptives on the anticoagulant pathways. Thromb Haemost 2000;84:15–21.
10. Middeldorp S, Meijers JC, van den Ende AE, van Enk A, Bouma BN, Tans G, et al. Effects on coagulation of levonorgestrel- and desogestrel-containing low dose oral contraceptives: a cross-over study. Thromb Haemost 2000;84:4–8.
11. The Oral Contraceptive and Hemostasis Study Group. An open label, randomized study to evaluate the effects of seven monophasic oral contraceptive regimens on hemostatic variables. Contraception 1999;59:345–55.
12. Archer DF, Mammen EF, Grubb GS. The effects of a low-dose monophasic preparation of levonorgestrel and ethinyl estradiol on coagulation and other hemostatic factors. Am J Obstet Gynecol 1999;181:S63–6.
13. Wu O, Roberston L, Langhorne P, Twaddle S, Lowe GD, Clark P, et al. Oral contraceptives, hormone replacement therapy, thrombophilias, and risk of venous thromboembolism: a systematic review. The Thrombosis, Risk, and Economic Assessment of Thrombophilia Screening (TREATS) Study. Thromb Haemost 2005;94:17–25.
14. Rosing J, Tans G, Nicolaes GA, Thomassen MC, van Oerle R, van der Ploeg PM, et al. Oral contraceptives and venous thrombosis: different sensitivities to activated protein C in women using second- and third-generation oral contraceptives. Br J Haematol 1997;97:233–8.
15. Vandenbroucke JP, Helmerhorst FM, Bloemenkamp KW, Rosendaal FR. Third-generation oral contraceptive and deep venous thrombosis: from epidemiologic controversy to new insight in coagulation. Am J Obstet Gynecol 1997;177:887–91.
16. van Rooijen M, Hansson LO, Frostegard J, Silveira A, Hamsten A, Bremme K. Treatment with combined oral contraceptives induces a rise in serum C-reactive protein in the absence of a general inflammatory response. J Thrombos Haemost 2006;4:77–82.
17. Abrams LS, Skee DM, Natarajan J, Wong FA, Anderson GD. Pharmacokinetics of a contraceptive patch (Evra and Ortho Evra) containing norelgestromin and ethinyloestradiol at four application sites. Br J Clin Pharmacol 2002;53:141–6.
18. Hemker HC, Giesen P, Al Dieri R, Regnault V, de Smedt E, Wagenvoord R, et al. Calibrated automated thrombin generation measurement in clotting plasma. Pathophysiol Haemost Thromb 2003;33:4–15.
19. Jones B, Kenward MG. Design and analysis of cross-over trials. 2nd ed. Boca Raton (FL): Chapman and Hall/CRC; 2003.
20. Cushman M, Arnold AM, Psaty BM, Manolio TA, Kuller LH, Burke GL, et al. C-reactive protein and the 10-year incidence of coronary heart disease in older men and women: the cardiovascular health study. Circulation 2005;112:25–31.
21. Rosing J, Middeldorp S, Curvers J, Christella M, Thomassen LG, Nicolaes GA, et al. Low-dose oral contraceptives and acquired resistance to activated protein C: a randomized cross-over study. Lancet 1999;354:2036–40.
22. Kemmeren JM, Algra A, Meijers JC, Tans G, Bouma BN, Curvers J, et al. Effect of second- and third-generation oral contraceptives on the protein C system in the absence or presence of the factor V Leiden mutation: a randomized trial. Blood 2004;103:927–33.