American Journal of Obstetrics and Gynecology (2005) 192, 2055–9

www.ajog.org

Effect of oral versus transdermal steroidalcontraceptives on androgenic markers

Terry White, MD,* John K. Jain, MD, Frank Z. Stanczyk, PhD

Division of Reproductive Endocrinology and Infertility, Department of Obstetrics and Gynecology,University of Southern California Keck School of Medicine, Los Angeles, CA

KEY WORDSSex hormone-binding

globulin

AndrogensTestosteroneOral contraceptives

Contraceptive patch

Objective: The purpose of this study was to compare biochemical androgen profiles in women

treated with the contraceptive patch versus an oral contraceptive (OC).Study design: Twenty-four healthy women were randomly assigned to receive 3 cycles of either thecontraceptive patch (ethinyl estradiol [EE] 20 mg/d and norelgestromin 150 mg/d) or OC (EE 35 mgand norgestimate 250 mg). Blood samples were taken at baseline and end of treatment. Serum levelsof sex hormone-binding globulin (SHBG), total testosterone (T), androstenedione (A),dehydroepiandrosterone sulfate (DHEAS), dihydrotestosterone (DHT), and 3a-androstanediolglucuronide (3a-diol G) were quantified by immunoassay methods; free T was calculated. The

paired t and Student t tests were used for statistical analysis.Results: Nineteen women completed the study (patch, n = 10; OC, n = 9). Despite a 1.6-foldrelative increase in SHBG levels with the patch versus OC (449% vs 274%, P = .03), free T

decreased equally in both groups (patch 60%, P ! .0001; OC 59%, P ! .0001). DHEASdecreased by 26% in the patch group (P ! .01) and 32% in the OC group (P ! .001). 3a-diol Gwas reduced by 52% in the patch group (P ! .0001) and 51% in the OC group (P ! .0001). In

addition, the OC was associated with significant decreases in A and DHT.Conclusion: The contraceptive patch had an effect comparable to the OC on several keyandrogenic markers. Given these biochemical findings, the contraceptive patch has significant

potential as a therapeutic agent for disorders of androgen excess.� 2005 Elsevier Inc. All rights reserved.

Acne and hirsutism are 2 of the most common signsof androgen excess, an endocrine disorder affectingapproximately 7% of reproductive-aged women. Thereis substantial evidence that oral contraceptives (OCs)suppress excess androgens and their clinical manifesta-tions. Randomized trials have shown that treatment ofacne with OCs can reduce lesion count and improve

Presented at the 71st Annual Meeting of the Pacific Coast Obstet-

rical and Gynecological Society, October 19-24, 2004, Phoenix, Ariz.

* Reprint requests: Terry White, MD, 1835 Arch St, #1504,

Philadelphia, PA 19103.

E-mail: terryw@wharton.upenn.edu

0002-9378/$ - see front matter � 2005 Elsevier Inc. All rights reserved.

doi:10.1016/j.ajog.2005.02.067

global assessment.1 In addition, OCs have documentedefficacy in treating hirsutism, as demonstrated by animprovement in Ferriman-Gallwey scores.2

Several potential mechanisms for the therapeutic effectof OCs on androgen excess have been suggested. Estro-gen in the OC induces hepatic sex hormone-bindingglobulin (SHBG) production, leading to reduced freetestosterone (T) levels. In addition, OCs suppress lutei-nizing hormone and follicle-stimulating hormone, lead-ing to decreased ovarian production of androstenedione(A) and T. Through a mechanism not well understood,there is decreased production of the adrenal androgen,dehydroepiandrosterone sulfate (DHEAS). Finally, the

2056 White, Jain, and Stanczyk

progestin in the OC can lead to antagonism of 5a-reductase, an enzyme that converts T to the more potentdihydrotestosterone (DHT) in the skin.

In 2001, the Food and Drug Administration (FDA)approved Ortho Evra (Ortho-McNeil Pharmaceutical,Raritan, NJ), the first transdermal system for combina-tion hormonal contraception. Each patch delivers adaily dose of 20 mg ethinyl estradiol (EE) and 150 mgnorelgestromin (NGMN) for 7 days. NGMN is theprimary active metabolite of norgestimate (NGM), theprogestin used in certain OC formulations, includingOrtho Cyclen (Ortho-McNeil Pharmaceutical). NGM-containing OCs have been demonstrated to reduceandrogen levels and clinically improve acne.3,4

Although the patch has been studied for contraceptiveefficacy and safety,5 there is no published data on itseffects on androgens. Because transdermal administra-tion of a drug avoids the hepatic first-pass effect andsubsequent peaks and troughs of the circulating drug,characteristic of oral dosing, changes in serum levels ofSHBG and androgens may be different by this route ofadministration. The objective of this randomized studywas to compare biochemical androgen profiles in womenusing the patch versus an OC, Ortho Cyclen, which hasdocumented efficacy in treating hyperandrogenism.3

Materials and methods

The study was conducted at Los Angeles County CUniversity of Southern California (LAC C USC)Medical Center from December 2003 to April 2004.Study participants were recruited from the gynecologyand family planning clinics at Women’s and Child-ren’s Hospital, LAC C USC Medical Center. Theproposal was approved by the USC InstitutionalReview Board.

Healthy women aged 18 to 40 years with regularmenstrual cycles of 25 to 35 days and a body mass index(BMI) between 18 and 30 kg/m2 were eligible for thestudy. Exclusion criteria were the following: (1) use ofsteroidal contraception during the 3 months precedingenrollment, or use of depot-medroxyprogesterone ace-tate within 12 months of enrollment; (2) history ofmedical conditions associated with abnormal androgenlevels; (3) use of medications that alter the metabolism orpharmacokinetics of steroid contraceptives; (4) historyof significant cardiovascular, hepatic, renal, or throm-boembolic disease; (5) undiagnosed, abnormal uterinebleeding; and (6) active cigarette smoking. Screening ofsubjects included a complete history and physical exam-ination, and basic laboratory studies. Informed consentwas obtained from all subjects before their participation.

After pretrial evaluation, each subject began cycle1 of treatment within 3 days of the onset of her nextmenstrual flow. Subjects were assigned randomly totreatment with either transdermal EE 20 mg/d plus

NGMN 150 mg/d (Ortho Evra) or oral EE 35 mg plusNGM 250 mg (Ortho Cyclen) according to a computer-generated random number table. A baseline bloodsample was drawn before beginning treatment, and afinal sample was taken at the end of cycle 3 (duringdays 17-21). Serum was separated and frozen for laterdetermination of biochemical markers of androgenproduction: total T, free T, DHT, A, DHEAS, 3a-androstanediol glucuronide (3a-diol G), and SHBG.Blood pressure and weight were reassessed after com-pletion of cycle 1. Adherence to the treatment protocolwas assessed at each visit.

Serum androgen levels were measured by validatedimmunoassay methods. Total T, DHT, and A werequantified by radioimmunoassay (RIA) after organicsolvent extraction and Celite column partition chroma-tography as described previously.6-9 3a-diol G wasmeasured by direct RIA with the use of commercialkits obtained from Diagnostic Systems Laboratories(Webster, Texas).10 DHEAS and SHBG were measuredby direct chemiluminescent immunoassay on the Im-mulite analyzer (Diagnostic Products Corporation, In-glewood, Calif). Intra- and interassay coefficients ofvariation range from 4% to 7%, and 9% to 12%,respectively. Baseline and final serum samples for eachsubject were assayed together to eliminate interassayvariation. Free T was calculated by using an algorithmderived from equations described by Sodergard et al11

and Vermeulen et al.12 Reliability of this algorithm hasbeen described by Rinaldi et al.13

Within each treatment group, changes from baselinein androgen and SHBG levels after 3 cycles of treatmentwere analyzed with the use of the paired 2-sided t test.Comparison of percent changes from baseline in andro-gen and SHBG levels between the 2 study groups wasbased on the 2-sided Student t test. Two-sided alpha of.05 was considered statistically significant.

Sample size calculations were determined as follows.The primary endpoint was percent change from baselinein free T levels; secondary endpoints were percentchanges from baseline in other biochemical androgenicmarkers. With the use of a 2-sided Student t test todetect a 25% difference in percent change in free T levelsbetween the 2 groups, we determined that a sample sizeof 10 per group was necessary to achieve a power of70% with an alpha of .05. The standard deviation froma previous study was used for this calculation.3 A finalsample size of 12 per group was chosen to account foran estimated 20% dropout rate.

Results

Twenty-four women, ranging in age from 20 to 33 years,were enrolled in the study. During the course of thestudy, 1 patient withdrew for personal reasons, 1 patientwas lost to follow-up, and 3 patients were excluded for

White, Jain, and Stanczyk 2057

Table I Serum androgen and SHBG levels (mean G SEM) in women treated with transdermal EE/NGMN and oral EE/NGM

Transdermal EE/NGMN (n = 10) Oral EE/NGM (n = 9)

Variable Baseline End of therapy Baseline End of therapy

SHBG (nmol/L) 55.7 G 17.8 303 G 99 58 G 14.9 201 G 63.5Total T (ng/dL) 29.2 G 9.6 35.4 G 11.5 36.4 G 10.6 34.6 G 10.9Free T (ng/dL) 0.51 G 0.16 0.19 G 0.57 0.65 G 0.19 0.25 G 0.78DHEAS (mg/mL) 1.52 G 0.51 1.12 G 0.38 2.10 G 0.71 1.37 G 0.43A (ng/mL) 1.20 G 0.39 1.04 G 0.35 1.51 G 0.40 1.01 G 0.36DHT (ng/mL) 0.17 G 0.06 0.18 G 0.05 0.22 G 0.05 0.17 G 0.053a-diol G (ng/mL) 2.41 G 0.67 1.10 G 0.31 2.84 G 0.62 1.13 G 0.36

nonadherence to the study protocol (discontinuation oftreatment before the final blood sampling). Demo-graphic data on the 19 women (patch, n = 10; OC,n = 9) who completed the study were as follows. In thepatch group, the mean (G SD) age was 27.5 G 4.3years; mean BMI was 23.9 G 3.0 kg/m2. Means for ageand BMI in the OC group were 25.1 G 2.8, and 21.8 G3.4, respectively. There were no statistically significantdifferences between the groups with respect to age orBMI.

Analysis of biochemical androgen markers was basedon the 19 women who completed the study. Table Iincludes values of biochemical markers at baseline andafter 3 cycles of treatment, and Table II shows percentchanges from baseline. Treatment with both the patchand OC resulted in significant reductions in free T,DHEAS, 3a-diol G, and significant increases in SHBG.In addition, the OC treatment was associated withsignificant reductions in A and DHT.

The Figure displays a graphic comparison betweenthe patch and OC groups with respect to percentchanges from baseline in SHBG, T, and free T. SHBGincreased significantly in both the patch and OC groupswhen compared with baseline. There was a significantlygreater increase in SHBG with the patch versus the OC(449% vs 274%, P = .03). The patch was associatedwith a small increase in total T, whereas the OC groupshowed no change. However, the difference between the2 groups with respect to total T was not statisticallysignificant. Furthermore, both the patch and OC groupsshowed significant, equal decreases in free T (patch60%, P ! .0001; OC 59%, P ! .0001).

Comment

This study compared biochemical androgen profiles inwomen treated with the contraceptive patch versus anNGM-containing OC that has documented efficacy inreducing androgens.3 Despite the use of different drugdelivery systems, both treatments led to comparableeffects on important androgens. The significant reduc-tions in free T, DHEAS, and 3a-diol G demonstrated

modulations in multiple compartments of androgenproduction.

Although there are no published studies on the effectof the contraceptive patch on androgens, there havebeen several studies comparing different combinationOCs with respect to androgenic markers. These datahave demonstrated that OCs containing different dosesof EE (20-35 mg) and different progestins have highlyvariable effects on SHBG but comparable reductions inandrogens, even when controlling for dose of EE.1,3,14-16

Coenen et al3 studied 4 different OC formulations,including Ortho Cyclen, and found an average increasein SHBG of 263% and reductions in free T by 51% to58% and DHEAS by 21% to 44%.

Thorneycroft et al1 found that OCs containinglevonorgestrel (LNG)/EE and norethindrone acetate(NETA)/EE caused SHBG to increase by 106% and234%, respectively, whereas androgens decreased com-parably. In the LNG/EE and NETA/EE groups, bio-available (non-SHBG-bound) T decreased by 31% and26%; DHEAS decreased by 19% and 22%; and 3a-diolG decreased by 39% and 36%, respectively. In contrast,total T decreased by 27% in the LNG/EE group and didnot change in the NETA/EE group. The difference inthe effects of the 2 formulations on total T levels maypartly explain the comparable reduction in bioavailableT despite a 2.2-fold relative increase in SHBG withNETA.

A similar discrepancy was observed in the currentstudy. Despite the 1.6-fold greater increase (449% vs274%) in SHBG in the patch versus OC group, thereductions in free T were almost identical in the 2 groups.As in the Thorneycroft study, the nearly equal decreasesin free T may be explained partly by the divergent effectson total T in the 2 groups: a 27% increase in the patchgroup but no change (–0.5%) in the OC group. Althoughit has been stated that the magnitude of increase inSHBG after treatment with an OC is inversely related toits androgenicity, our data and those reported byThorneycroft et al1 do not support this hypothesis.

Another inconsistency in the present data was thefinding of a significant decrease in 3a-diol G without a

2058 White, Jain, and Stanczyk

corresponding decrease in DHT in the patch group. Incontrast, the OC group showed significant decreases inboth DHT and 3a-diol G. Although logically one wouldexpect DHT to serve as the primary marker of periph-eral androgen production, serum DHT levels do notaccurately reflect androgen activity in peripheral tissues.This has been attributed primarily to the rapid cellularturnover of DHT and its very high affinity for SHBG.17

However, the distal metabolite of DHT, 3a-diol G, hasbeen shown to serve as a better marker of peripheralandrogen action. An excellent correlation is foundbetween serum 3a-diol G levels and tissue 5a-reductaseactivity, as well as clinical manifestations of hyper-androgenism, e.g., hirsutism.17

The lack of significant reductions in A and total Tafter treatment with the patch was an unexpectedfinding of the current study. By contrast, the OCgroup demonstrated a reduction in A, consistent withprior studies. It has been documented that treatmentwith Ortho Evra results in suppression of follicle-stimulating hormone and luteinizing hormone in amanner similar to combination OCs.18 Assuming thatgonadotropin suppression is related to reduced ovar-ian androgen production, one would expect the patchto result in reduced ovarian production of A and T.However, gonadotropin suppression is due primarilyto the progestin component of combination OCs.Because the progestins administered in the 2 studygroups differed in formulation, dose, and route ofadministration, treatment with the patch may havehad a less suppressive effect on A and T in compar-ison with the OC.

The marked increase in SHBG with the patch wasalso an unexpected finding. Previous studies on the effectof various OCs on SHBG have shown increases in therange of 50% to 400%.19 The 449% increase in SHBGwe observed with the patch was significantly greaterthan the 274% increase noted with the OC. We hadhypothesized that the increase in SHBG with the patchwould be less than the increase with the OC, given the

Table II Percent changes in serum androgen and SHBGlevels (mean G SEM) from baseline to end of treatment

VariableTransdermalEE/NGMN

OralEE/NGM

SHBG C449 G 64*y C274 G 36*Total T C27 G 9.4* �0.5 G 9.2Free T �60 G 7.1* �59 G 3.2*DHEAS �26 G 7.7* �32 G 4.9*A �9 G 9.5 �29 G 8.5*DHT C3 G 10.8 �22 G 5.2*3a-diol G �52 G 3.2* �51 G 4.5*

* P ! .05 vs baseline within treatment group.y P ! .05 vs oral EE/NGM group.

lack of hepatic first-pass effect with transdermal admin-istration. The increase in SHBG with the patch suggestsa biochemical mechanism similar to that of combinationOCs, specifically, induction of hepatic SHBG produc-tion. This may seem inconsistent with data from studiesin postmenopausal women on transdermal estrogentherapy that demonstrated no significant effects onSHBG.20-23 However, the comparison may not beappropriate given that the transdermal estrogen patchcontains 17ß-estradiol, whereas the contraceptive patchdelivers EE. Differences in potency and metabolism ofthe steroids and differences in the populations studied(postmenopausal vs reproductive-aged) may be under-lying factors.

The marked increase in SHBG raises the question ofwhether production of other hepatic proteins would beaffected by the transdermal contraceptive. Prior studieson OCs have demonstrated an increase in hepaticglobulin production. Wiegratz et al14 reported thatincreases in SHBG, corticosteroid-binding globulin,and thyroxine-binding globulin occurred in all 4 groupsrandomly assigned to different combination OC for-mulations. Van der Vange et al15 studied the effects of7 different OCs and found that all produced significantincreases in SHBG and corticosteroid-binding globulin.Finally, van Rooijen et al24 found that increasedSHBG after treatment with 2 different OCs correlatedwith changes in resistance to activated protein C.Further studies are needed to study the effect of thepatch on SHBG and other hepatic proteins.

We acknowledge certain limitations in the study.Given the relatively small sample size and absence ofclinical variables (e.g., acne lesions, Ferriman-Gallweyscores), the results do not warrant recommendation ofthe patch for treatment of hyperandrogenism. In addi-

Figure Percent change from baseline (mean G SEM) forserum concentrations of SHBG, total T, and free T after 3cycles of treatment with transdermal EE/NGMN (dark grey

bar) and oral EE/NGM (light grey bar). *P ! .05 vs baselinewithin treatment; yP ! .05 difference between groups.

White, Jain, and Stanczyk 2059

tion, the small sample size may have given rise to baselinedifferences between the 2 study groups, although baselinevariablesdandrogen levels, age, and BMIdwere notsignificantly different between groups. Finally, the studyallows limited comparison of transdermal versus oraldelivery systems for steroidal contraception. The pro-gestins in the 2 formulations were similar but notidentical, and comparison of different doses is technicallychallenging. Steady-state concentrations of estrogen andprogestin after transdermal administration and averageserum concentrations with oral delivery are consideredpharmacokinetically distinct variables that cannot bedirectly compared.25

In summary, our study demonstrates that treatmentwith the contraceptive patch reduced several importantandrogens comparable to the OC. Given the establishedbiochemical and clinical effects associated with OCtreatment, the patch has significant potential as a ther-apeutic agent for problems of androgen excess such asacne and hirsutism. Further studies on the patch areneeded to establish clinical efficacy for affected patients.

Acknowledgment

We would like to thank Dr Begum Ozel for assistancewith statistical analysis and review of the manuscript.

References

1. Thorneycroft IH, Stanczyk FZ, Bradshaw KD, Ballagh SA,

Nichols M, Weber ME. Effect of low-dose oral contraceptives on

androgenic markers and acne. Contraception 1999;60:255-62.

2. Breitkopf DM, Mitchell PR, Young SL, Nagamani M. Efficacy of

second versus third generation oral contraceptives in the treatment

of hirsutism. Contraception 2003;67:349-53.

3. Coenen CMH, Thomas CMG, BormGF, Rolland R. Comparative

evaluation of the androgenicity of four low-dose, fixed-combination

oral contraceptives. Int J Fertil 1995;40(Suppl 2):92-7.

4. Redmond GP, Olson WH, Lippman JS, Kafrissen ME, Jones TM,

Jorizzo JL. Norgestimate and ethinyl estradiol in the treatment of

acne vulgaris: a randomized, placebo-controlled trial. Obstet

Gynecol 1997;89:615-22.

5. Burkman RT. The transdermal contraceptive system. Am J Obstet

Gynecol 2004;190:S49-53.

6. Goebelsmann U, Bernstein GS, Gale JA, Kletzky OA, Nakamura

RM, Coulson AH, et al. Serum gonadotropin testosterone

estradiol and estrone levels prior to and following bilateral vasec-

tomy. In: Lepow IH, Crozier R, editors. Vasectomy: immunologic

and pathophysiologic effects in animals and man. New York:

Academic Press; 1979. p. 165-75.

7. Goebelsmann U, Arce JJ, Thorneycroft IH, Mishell DR Jr. Serum

testosterone concentrations in women throughout the menstrual

cycle and following hCG administration. Am J Obstet Gynecol

1974;119:445-52.

8. Goebelsmann U, Horton R, Mestman JH, Arce JJ, Nagata Y,

Nakamura RM, et al. Male pseudohermaphroditism due to testic-

ular 17-hydroxysteroid dehydrogenase deficiency. J Clin Endocrinol

Metab 1973;36:867.

9. Serafini P, Ablan F, Lobo RA. 5a-Reductase activity in the

genital skin of hirsute women. J Clin Endocrinol Metab 1985;60:

349.

10. Narang R, Rao J, Savjani G, Peterson J, Gentzschein E, Stanczyk

FZ. Radioimmunoassay kit for the quantitative measurement of

androstanediol glucuronide in unextracted serum. Poster presented

at: 17th National Meeting of the Clinical Ligand Assay Society;

April 10-13, 1991; Chicago, Ill.

11. Sodergard R, Backstrom T, Shanbhag V, Carstensen H. Calcula-

tion of free and bound fractions of testosterone and estradiol-17b

to human plasma protein at body temperature. J Steroid Biochem

1982;26:801-10.

12. Vermeulen A, Verdonck L, Kaufman JM. A critical evaluation of

simple methods for estimation of free testosterone in serum.

J Clin Endocrinol Metab 1999;84:3666-72.

13. Rinaldi S, Geay A, Dechaud H, Biessy C, Zeleniuch-Jacquotte A,

Akhmedkhanov A, et al. Validity of free testosterone and free

estradiol determinations in serum samples from postmenopausal

women by theoretical calculations. Cancer Epidemiol Biomarkers

Prev 2002;11:1065-71.

14. Wiegratz I, Kutschera E, Lee JH, Moore C, Mellinger U, Winkler

UH, et al. Effect of four different oral contraceptives on various

sex hormones and serum binding globulins. Contraception

2003;67:25-32.

15. Van der Vange N, Blankenstein MA, Kloosterboer HJ, Haspels

AA, Thijssen JHH. Effects of seven low-dose combined oral

contraceptives on sex hormone binding globulin, corticosteroid

binding globulin, total and free testerone. Contraception

1990;41:345-52.

16. Murphy A, Cropp CS, Smith BS, Burkman RT, Zacur HA. Effect

of low-dose oral contraceptive on gonadotropins, androgens, and

sex hormone binding globulin in nonhirsute women. Fertil Steril

1990;53:35-9.

17. Lobo RA. Hirsutism, alopecia, and acne. In: Becker KL, editor.

Principles and practice of endocrinology and metabolism. Phila-

delphia: Lippincott, Williams and Wilkins; 2001. p. 991-1008.

18. Pierson RA, Archer DF, Moreau M, Shangold GA, Fisher AC,

Creasy GW. Ortho Evra versus oral contraceptives: follicular

development and ovulation in normal cycles and after an inten-

tional dosing error. Fertil Steril 2003;80:34-42.

19. Odlind V, Milsom I, Persson I, Victor A. Can changes in sex

hormone binding globulin predict the risk of venous thromboem-

bolism with combined oral contraceptive pills? Acta Obstet

Gynecol Scand 2002;81:482-90.

20. Campagnoli C, Biglia N, Altare F, Lanza MG, Lesca L,

Cantamessa C, et al. Differential effects of oral conjugated estro-

gens and transdermal estradiol on insulin-like growth factor 1,

growth hormone and sex hormone binding globulin serum levels.

Gynecol Endocrinol 1993;7:251-8.

21. Basbug M, Aygen E, Tayyar M, Muhtaroglu S, Demir I, Okten S.

Twenty two weeks of transdermal estradiol increases sex hormone-

binding globulin in surgical menopausal women. Eur J Obstet

Gynecol Reprod Biol 1997;73:149-52.

22. Nachtigall LE, Raju U, Banerjee S, Wan L, Levitz M. Serum

estradiol-binding profiles in postmenopausal women undergoing

three common estrogen replacement therapies: associations with

sex hormone-binding globulin, estradiol, and estrone levels. Men-

opause 2000;7:243-50.

23. Serin IS, Ozcelik B, Basbug M, Aygen E, Kula M, Erez R.

Long-term effects of continuous oral and transdermal estrogen

replacement therapy on sex hormone binding globulin and free

testosterone levels. Eur J Obstet Gynecol Reprod Biol 2001;99:

222-5.

24. Van Rooijen M, Silveira A, Hamsten A, Bremme K. Sex hormone-

binding globulinda surrogate marker for the prothrombotic

effects of combined oral contraceptives. Am J Obstet Gynecol

2004;190:332-7.

25. Abrams LS, Skee D, Natarajan J, Wong FA. Pharmacokinetic

overview of Ortho Evra�/Evra�. Fertil Steril 2002;77(Suppl 2):

S3-12.