Extended Use of Transdermal Norelgestromin/Ethinyl

Estradiol: A Randomized Trial

Felicia H. Stewart, MD, Andrew M. Kaunitz, MD, Katherine D. LaGuardia, MD, MPH,
Debra L. Karvois, Alan C. Fisher, DrPH, and Andrew J. Friedman, MD, for the ORTHO EVRA
Extended Regimen Study Group*

OBJECTIVE: To compare bleeding profiles and satisfaction fewer episodes of withdrawal bleeding. (Obstet Gynecol
among women using a norelgestromin/ethinyl estradiol 2005;105:1389 –96. © 2005 by The American College of
(E2) transdermal contraceptive patch in an extended regi- Obstetricians and Gynecologists.)
men to those among women using a traditional 28-day LEVEL OF EVIDENCE: I
patch regimen.
METHODS: Healthy, regularly menstruating women (N ⴝ Combined estrogen plus progestin contraceptives have
239) were randomly assigned (2:1 ratio) to receive the traditionally been administered in cycles of 21 days of
norelgestromin/ethinyl E2 transdermal patch in an ex- active hormone followed by a 7-day hormone-free inter-
tended regimen (weekly application for 12 consecutive val that results in withdrawal bleeding in most women.
weeks, 1 patch-free week, and 3 more consecutive weekly Use of extended regimens (ie, administration of active
applications, n ⴝ 158) or a cyclic regimen (4 consecutive hormones for an interval ⬎ 21 days) of combined oral
cycles of 3 weekly applications and 1 patch-free week, n ⴝ contraceptives is common among women wishing to
81). Subjects recorded bleeding data daily and completed delay or prevent withdrawal bleeding for reasons such as
satisfaction questionnaires. Subjects and investigators pro-
athletic participation or vacation.1– 4 In addition to the
vided overall assessments of the regimens.
convenience of reducing the frequency of withdrawal
RESULTS: Extended use of the norelgestromin/ethinyl E2 bleeds, elimination of the hormone-free interval report-
transdermal patch resulted in fewer median bleeding days edly reduces many menstrual-related symptoms (eg,
(6 compared with 14, P < .001), bleeding episodes (1
headaches, dysmenorrhea) that occur at a greater fre-
compared with 3, P < .001), and bleeding or spotting
quency during the hormone-free interval than during the
episodes (2 compared with 3, P < .001) compared with
cyclic use during days 1– 84; median numbers of bleeding
rest of the cycle.1– 6
or spotting days were similar between regimens (14 com- Many women prefer extended hormonal contracep-
pared with 16, P ⴝ .407) during this time. Extended use tive regimens over traditional cyclic regimens.7 Ex-
delayed median time to first bleeding to 54 days compared tended regimens are well tolerated and efficacious for the
with 25 days with cyclic (P < .001). Subjects were highly prevention of pregnancy and reduction in menstrual-
satisfied with both regimens. Although not statistically related symptoms.3,5,6,8 The traditional oral contracep-
significant, slightly more adverse events were reported tive cycle of 21 days of active pills and 7 days of inactive
with the extended than with the 28-day regimen. pills is not based on evidence of superiority and may
CONCLUSION: Compared with cyclic use, extended use of have important disadvantages for some women. Many
the norelgestromin/ethinyl E2 transdermal patch delayed of the cyclic symptoms that women complain of occur
menses and resulted in fewer bleeding days. This regimen during the hormone-free period.6 Hence, for some
may represent a useful alternative for women who prefer women, reducing the number of days off active hor-
mones may provide a significant benefit. In addition, it
* For a list of other investigators for the ORTHO EVRA Extended Regimen Study
Group, see the Appendix.

From the University of California San Francisco, Center for Reproductive Health
Research & Policy, San Francisco, California; University of Florida Health Science Financial Disclosure
Center, Jacksonville, Florida; Ortho-McNeil Pharmaceutical, Inc., Raritan, New Drs. LaGuardia and Fisher and Ms. Karvois are employed by
Jersey; and Johnson & Johnson Pharmaceutical Research and Development, Ortho-McNeil Pharmaceutical, Inc., which funded this study. Dr.
Raritan, New Jersey. Friedman is employed by Johnson & Johnson Pharmaceutical Re-
search and Development.
This study was supported by Ortho-McNeil Pharmaceutical, Inc.

VOL. 105, NO. 6, JUNE 2005

© 2005 by The American College of Obstetricians and Gynecologists. 0029-7844/05/$30.00 1389
Published by Lippincott Williams & Wilkins. doi:10.1097/01.AOG.0000160430.61799.f6
may be easier for some women to implement and adhere
to extended regimens.
However, one of the side effects reported with ex-
tended regimens is breakthrough bleeding or spot-
ting.3,8 –11 Compared with cyclic regimens, extended
oral contraceptive use may be associated with an in-
creased incidence of breakthrough bleeding and spot-
ting, especially during the first few months.3,5,12–14 Ex-
perience with a 91-day oral contraceptive extended
regimen product indicates that bleeding and spotting
decrease with repeated 91-day cycles. However, during
the first 84 days, 65% of the extended users, compared Fig. 1. Diagram of study design. Arrows indicate study
visits.
with 38% of the cyclic users, experienced seven or more
Stewart. Extended Transdermal Contraceptive Therapy. Obstet Gynecol 2005.
days of unscheduled bleeding and spotting (Seasonale
package insert, available at www.seasonale.com, re-
trieved March 2, 2005). to Good Clinical Practice, the Declaration of Helsinki of
ORTHO EVRA transdermal system (Ortho-McNeil 1975 as revised in 1983, and current International Con-
Pharmaceutical, Inc., Raritan, NJ), is a combined hor- ference on Harmonization guidelines. Signed written
monal contraceptive system containing 6.0 mg of the informed consent was obtained from all subjects before
progestin norelgestromin (the primary active metabolite any protocol-specific procedures were completed.
of norgestimate) and 0.75 mg of the estrogen ethinyl The last term pregnancy had to have been completed
estradiol (E2). The 20-cm2 patch delivers 150 ␮g/day at least 42 days before screening, and there had to have
norelgestromin and 20 ␮g/day ethinyl E2 to the systemic been at least 1 normal menstrual period since pregnancy.
circulation, and is approved as a cyclic regimen of 3 consec- Women agreed not to use any other steroid hormone
utive 7-day patches followed by 1 patch-free week.15–17 The therapy except topical corticosteroids, if needed, during
current trial compared the bleeding profiles of extended the trial.
and cyclic regimens of the norelgestromin/ethinyl E2 trans- Exclusion criteria included history or presence of any
dermal patch and evaluated investigator and subject overall disorder contraindicating steroid hormonal therapy, his-
assessments and subject satisfaction with each regimen. tory of dermal hypersensitivity, and treatment with an
extended regimen of an oral contraceptive within the 3
months before screening. Although weight heavier than
MATERIALS AND METHODS 198 pounds is a precaution in the prescribing information
Healthy, regularly menstruating females (N ⫽ 239) aged for the norelgestromin/ethinyl E2 transdermal system (due
18 to 45 years (nonsmokers if aged 35 years or older) at to decreased contraceptive efficacy compared with that in
9 clinical research sites were randomly assigned in a 2:1 women ⬍ 198 pounds), it is not a contraindication, and it
ratio to receive an extended or cyclic regimen of the was not a criterion for exclusion from the study, because
norelgestromin/ethinyl E2 transdermal system. The increased risk of pregnancy is not a deterrent to use.
computer-generated randomization list was prepared by Enrollment of women heavier than 198 pounds was left
Ortho-McNeil Pharmaceutical, Inc. before the study to to the discretion of the physician and subject.
assign subjects to treatment within centers and by per- Subjects were instructed to record bleeding and head-
muted blocks of size 6. Extended regimen subjects ap- ache data in a diary booklet at home every day starting
plied 1 patch weekly for 12 consecutive weeks (84 days) on day 1 (defined as the first day of patch application),
followed by 1 week (7 days) without patch application including days of patch use, and the number of tampons,
(patch-free) and then 3 final weeks (21 days) of weekly pads, or panty liners used. They were also to record the
patch applications, for a total of 112 days. The last presence or absence of headache along with headache
treatment cycle was included to determine the duration characteristics, such as nausea or sensitivity to light.
of menses in the extended regimen group and to allow a Diaries were collected by clinic personnel and reviewed
transition back to cyclic contraception (Fig. 1). Cyclic with the subject at each visit. Detailed results of analyses
regimen subjects used patches for 4 consecutive 28-day of data regarding headaches, collected as described
cycles (112 days), applying 1 patch weekly for 3 weeks above, are reported elsewhere (LaGuardia K, Fisher AC,
(21 days) followed by 1 patch-free week (7 days). Bainbridge JD, LoCoco JM, Friedman AJ. Suppression
The study was approved by the institutional review of estrogen-withdrawal headache with extended trans-
board at each site. Procedures were conducted according dermal hormonal contraception. Fertil Steril, in press).

1390 Stewart et al Extended Transdermal Contraceptive Therapy OBSTETRICS & GYNECOLOGY

 Study participants underwent medical history review, The initial bleeding episode for each subject was ex-
physical examination, including breast and pelvic exam- cluded from all bleeding profile summaries, because the
ination, and hemoglobin and hematocrit testing at enroll- first day of menses was the first day of study medication.
ment and study completion. A cervical Pap test was At the 4 visits during treatment, each subject completed
performed at the screening visit. Vital signs and body a questionnaire containing items evaluating overall sat-
weight were measured, and a urine pregnancy test was isfaction (very satisfied, somewhat satisfied, neither sat-
performed at each visit. Adverse events reported by isfied nor dissatisfied, dissatisfied, very dissatisfied) and
subjects were recorded at each visit. bleeding (heavier than before, about the same as before,
The primary endpoint of the study was total bleeding lighter than before, haven’t bled). At the final visit,
or spotting days during the primary reference period, investigators and subjects provided an “overall assess-
ment of the study drug regimen” (excellent, good, fair, or
days 1– 84. Secondary endpoints included headache fre-
poor).
quency and overall subject and investigator assessment.
Statistical analyses of the bleeding profile were con-
The number of bleeding–spotting days, bleeding–spot-
ducted for both an intent-to-treat and a perfectly compli-
ting episodes, bleeding days, bleeding episodes, percent
ant completer population. The intent-to-treat population
of subjects with amenorrhea, time to first bleeding or included all randomly assigned subjects who applied
spotting, time to first bleeding during the 84-day interval, study drug and for whom postrandomization data were
and the duration of menses at the conclusion of the available. The perfectly compliant completer population,
84-day interval were evaluated. Secondary reference pe- a subgroup of the intent-to-treat population, was defined
riods of 56 days and 28 days were also analyzed, because for extended regimen subjects as a patch on every day
results at each of these 28-day intervals allow compari- (through day 84), with no patch on for more than 7 days,
sons of bleeding characteristics of extended use with and for cyclic regimen subjects as a patch on the first 21
those of cyclic use, which is typically administered every days of each cycle and no patch on the subsequent 7 days
28 days. Bleeding profiles were summarized using the of each cycle (through day 84), with no patch on for more
World Health Organization (WHO) reference period than 7 days. Except for the variables of duration of menses
methodology.18,19 Definitions, adapted from the WHO, and overall satisfaction at day 84, which required that
for the bleeding data analyses are shown in the box. subjects complete the treatment period, the intent-to-treat
results are presented. The intent-to-treat and the perfectly
compliant completer groups yielded bleeding profiles that
were comparable. The safety evaluable population in-
DEFINITIONS USED FOR BLEEDING DIARY ANALYSIS cluded all subjects who applied a patch and had at least 1
Bleeding Vaginal bleeding that required sanitary safety assessment after baseline (n ⫽ 235). The incidence
protection of at least one pad or of adverse events was evaluated in this population.
tampon The Mann–Whitney U test was used to compare
Spotting Vaginal bleeding that did not require possible differences between the treatment regimens in
sanitary protection (panty liner use was
acceptable)
Bleeding day Day on which bleeding occurred
Spotting day Day on which spotting alone occurred;
if spotting and bleeding occurred on the
same day, it was recorded as a bleeding
day
Bleeding-free Day on which neither bleeding nor
day spotting occurred, bounded by
bleeding/spotting days
Bleeding/ Any set of one or more consecutive
spotting bleeding or spotting days, bounded by
episode bleeding-free days
Bleeding Any set of one or more consecutive
episode bleeding days, bounded by bleeding-
free days
Amenorrhea No bleeding or spotting occurring
during the entire reference period Fig. 2. Flow of patients through the trial.
Stewart. Extended Transdermal Contraceptive Therapy. Obstet Gynecol 2005.

VOL. 105, NO. 6, JUNE 2005 Stewart et al Extended Transdermal Contraceptive Therapy 1391
Table 1. Reasons Cited for Withdrawal Prior to Study Com- ting, time to first bleeding, and time to discontinuation
pletion for Subjects Assigned to Extended or Cyclic due to bleeding were summarized by Kaplan-Meier esti-
Regimens of Norelgestromin/Ethinyl Estradiol mates, with the statistical significance of treatment group
Norelgestromin/Ethinyl differences assessed by the log-rank test. The perfect
Estradiol Contraceptive compliance rate per 28-day reference period for the 2
Regimen groups was compared with generalized estimating equa-
Extended Cyclic tion methodology, using Proc GENMOD from SAS 9.1
Reason for withdrawal (n ⫽ 155) (n ⫽ 80) (SAS Institute Inc., Cary, NC).
Limiting adverse event* 16 (10) 4 (5) The sample size was chosen to detect, at the 5%
Lost to follow-up 3 (2) 3 (4) significance level with 80% power, a 5-day difference
Subject choice 4 (3) 3 (4)
Protocol violation 1 (1) 1 (1)
during the 84-day reference period in the mean total
Bleeding 6 (4) 0 (0) number of bleeding or spotting days, assuming a stan-
Other 2 (1) 1 (1) dard deviation of 12 days. All statistical tests were per-
Values are n (%). formed at a 2-tailed, 5% level of significance.
* Limiting adverse events (those that led to study discontinuation)
for the extended regimen group were application site reaction, breast
discomfort (defined as “exacerbated breast tenderness, breast tender-
RESULTS
ness, and exacerbation of breast tenderness”), depression, emotional This study was initiated on May 30, 2002, and com-
lability, esophageal stricture, headache, hypertension, migraine, nau-
sea, urticaria, and vomiting. For the cyclic group, they were breast pain
pleted on March 13, 2003. Two hundred thirty-nine
(defined as “sharp breast pains and mastalgia”), dizziness, emotional subjects were randomly assigned to extended regimen
lability, hypertension, nausea, sweating, and vomiting. Subjects could norelgestromin/ethinyl E2 (n ⫽ 158) or cyclic regimen
have discontinued due to 1 or more adverse events.
norelgestromin/ethinyl E2 (n ⫽ 81) (Fig. 2). Four sub-
jects (3 extended, 1 cyclic) either never received the drug
the bleeding profile variables (including duration of men- or were lost to follow-up with no information after
ses), the overall satisfaction and bleeding questionnaire, randomization. Of those 235 women with information
and the overall assessments by the subject and investiga- after randomization, 191 (81%) completed the study: 123
tor. Fisher exact test was used to compare the propor- of 155 (79%) assigned to the extended regimen and 68 of
tions of subjects in the treatment regimens for categorical 80 (85%) assigned to the cyclic regimen. Reasons cited
demographic variables, adverse events, and the variable for withdrawal before study completion and adverse
“no bleeding days.” The t test was used for continuous events that led to discontinuation are outlined in Table 1.
demographic variables. Time to first bleeding or spot- The demographic and baseline characteristics of the

Table 2. Demographics and Baseline Characteristics of Subjects Assigned to Extended or Cyclic Regimens of
Norelgestromin/Ethinyl Estradiol
Norelgestromin/Ethinyl Estradiol Contraceptive Regimen
Extended (n ⫽ 155) Cyclic (n ⫽ 80)
Mean age 关y (⫾ SD)兴 29.1 (⫾ 6.2) 28.0 (⫾ 5.9)
Race (%)
White 54 61
Hispanic or Latina 24 20
African American 18 10
Asian 3 5
Other 1 4
Smoker (%) 12 13
Mean height 关cm (⫾ SD)兴 163.8 (⫾ 6.8) 164.0 (⫾ 7.1)
Mean weight 关kg (⫾ SD)兴 70.1 (⫾ 16.0) 67.2 (⫾ 14.0)
Mean body mass index 关kg/m2 (⫾ SD)兴 26.2 (⫾ 6.0) 25.0 (⫾ 5.1)
Previous hormonal contraceptive use* (%)
Fresh start 43 45
Direct switch 45 49
Indirect switch 11 6
Unknown 1 0
SD, standard deviation.
* “Fresh start” means subject took no hormonal contraceptive for more than 60 days before screening, “Direct switch”means subject switched
from another form of hormonal contraceptive to study contraceptive within 10 days before screening, and “Indirect switch” means subject switched
from another form of hormonal contraceptive to study contraceptive between 11 and 60 days before screening.

1392 Stewart et al Extended Transdermal Contraceptive Therapy OBSTETRICS & GYNECOLOGY

Table 3. Bleeding Profile Information for Intent-to-Treat Subjects (Excluding Initial Bleeding Days) Assigned to Extended
or Cyclic Regimens of Norelgestromin/Ethinyl Estradiol
Norelgestromin/Ethinyl Estradiol Contraceptive Regimen
Extended (n ⫽ 155) Cyclic (n ⫽ 80) P
Day 1 to day 84
Median bleeding-spotting days 14 16 .407
Median bleeding-spotting episodes 2 3 ⬍ .001
Median bleeding days 6 14 ⬍ .001
Median bleeding episodes 1 3 ⬍ .001
Amenorrhea (%) 12 1 ⬍ .003
Day 1 to day 56
Median bleeding-spotting days 6 10 .009
Median bleeding-spotting episodes 1 2 ⬍ .001
Median bleeding days 1 9 ⬍ .001
Median bleeding episodes 0.5 2 ⬍ .001
Amenorrhea (%) 28 1 ⬍ .001

study groups were statistically comparable (Table 2). measured using the first menstrual cycle for the extended
Subjects were generally compliant with patch applica- regimen group and the third menstrual cycle for the
tion. Perfect compliance by 28-day reference periods cyclic regimen group, was significantly longer in the
over the initial 84 days was numerically higher in the extended regimen group than in the cyclic regimen
extended regimen group compared with the cyclic regi- group (6.9 compared with 5.2 days, P ⬍ .001).
men group (85% compared with 78%, P ⫽ .322). Rela- Overall satisfaction, evaluated on day 84, was similar
tively few patches were reported as having fallen com- between groups; 86.3% (107/124) of subjects in the ex-
pletely off (approximately 3% of patches per 28-day tended regimen group and 88.6% (62/70) of subjects in
reference period). the cyclic group were very or somewhat satisfied. Sub-
For the primary reference period, days 1– 84, there jects in the extended regimen group indicated a percep-
was no significant difference between the extended and tion of lighter or no bleeding flow compared with their
cyclic regimens in the median number of bleeding or pretreatment menstrual flow. On day 84, 35% (43/124)
spotting days (Table 3). However, there were significant of the subjects in the extended regimen group reported
differences in bleeding or spotting episodes, bleeding that they had not bled, compared with 3% (2/70) of cyclic
days, bleeding episodes, and incidence of amenorrhea subjects.
during this time. During days 1–56, all bleeding vari- For the overall global assessment of the study drug
ables were significantly lower in the extended compared regimen, ratings of “excellent” or “good” were given
with the cyclic regimen group. For those subjects who by 87% (70/80) and 79% (122/155) of subjects in the
received the extended regimen, the incidence of amenor- cyclic and extended treatment regimens, respectively,
rhea was over twice as great during days 1–56 (28%) as
it was during days 1– 84 (12%).
The median time to first bleeding was 54 days for the
extended regimen group and 25 days for the cyclic
regimen group (P ⬍ .001; Fig. 3). The extended regimen
group had significantly (P ⬍ .001) fewer bleeding days
per 28-day interval than the cyclic group through day 84.
The median number of bleeding days in the extended
regimen group was 0, 0, and 2 for the 28-day intervals of
days 1–28, 29 –56, and 57– 84, respectively, whereas the
median number of bleeding days in the cyclic regimen
group was 4, 6, and 5 for the same intervals (Fig. 4). Fig. 3. Kaplan-Meier analysis of time to first bleeding for
Overall, in each 28-day interval studied, the extended the extended and cyclic regimens. Data collected during
regimen group had fewer bleeding days but more spot- the first week (during menses) were excluded; hence, the
ting days than the cyclic regimen group. In general, the study started with 100% of subjects with no bleeding.
differences between the 2 groups were most apparent Median day: extended, 54; cyclic, 25 (P ⬍ .001).
through 56 days of use. The mean duration of menses, Stewart. Extended Transdermal Contraceptive Therapy. Obstet Gynecol 2005.

VOL. 105, NO. 6, JUNE 2005 Stewart et al Extended Transdermal Contraceptive Therapy 1393
 Fig. 4. Median bleeding or spotting
days (A), median bleeding days (B),
and incidence of amenorrhea (C) for
extended (black line with diamonds)
(n ⫽ 155) and cyclic (black line with
squares) (n ⫽ 80) regimens, by cycle.
*P ⬍ .001.
Stewart. Extended Transdermal Contraceptive
Therapy. Obstet Gynecol 2005.

(P ⫽ .111). The investigators cited the treatment regimen in the cyclic regimen group discontinued due to bleed-
as “excellent” or “good” in 86% (69/80) of subjects in the ing. In the extended regimen group, the mean time to
cyclic regimen group, compared with 79% (123/155) of discontinuation due to bleeding was 107 days, and no
subjects in the extended regimen group (P ⫽ .217). subjects discontinued due to bleeding before day 62.
Both regimens of norelgestromin/ethinyl E2 were well One pregnancy occurred during the study in a subject
tolerated. The adverse event profiles were similar with assigned to the extended regimen. The pregnancy was
respect to the types of adverse events. Adverse events discovered at the final clinic visit (day 112), and the
that occurred with an incidence of at least 5% for either subject subsequently had a spontaneous pregnancy loss
regimen included application site reaction, upper respi- later, at approximately 8 weeks gestation. An ultrasound
ratory tract infection, breast discomfort, headache, nau- indicated a blighted ovum. This was the only serious
sea, vaginitis, and emotional lability (Table 4). Although adverse event that occurred during the study. Changes
not statistically significant, there were more reports of in hemoglobin and hematocrit levels from baseline to the
headache, nausea, and breast discomfort in the extended end of the study were small and clinically insignificant in
group compared with the cyclic group. A total of 20 both groups. Changes in vital signs from baseline to the
subjects, 16 (10%) in the extended group and 4 (5%) in end of the study were also unremarkable. Mean body
the cyclic group, discontinued the study due to adverse weight changes in both treatment groups were small
events (Table 1). Six (4%) subjects in the extended (.35 ⫾ 2.31 kg in the extended regimen group and .47 ⫾
regimen group discontinued due to bleeding; no subject 2.22 kg in the cyclic regimen group).

Table 4. Adverse Events That Occurred With an Incidence of at Least 5% in Either Extended or Cyclic Regimen
Norelgestromin/Ethinyl Estradiol Subjects
Norelgestromin/Ethinyl Estradiol Contraceptive Regimen
Adverse Event Extended (n ⫽ 155) Cyclic (n ⫽ 80) P*
Application site reaction 28 (18) 12 (15) .589
Upper respiratory tract infection 21 (14) 9 (11) .684
Breast discomfort 23 (15) 5 (6) .058
Headache 18 (12) 3 (4) .054
Nausea 16 (10) 4 (5) .220
Vaginitis 10 (7) 5 (6) 1.000
Emotional lability 8 (5) 3 (4) .754
Values are n (%).
* P values were calculated using the Fisher exact test.

1394 Stewart et al Extended Transdermal Contraceptive Therapy OBSTETRICS & GYNECOLOGY

DISCUSSION approval of an extended oral contraceptive, the scientific
Based on results of a search of publications from the community will likely follow safety and efficacy data
years 1995 through 2005 included in the National Li- closely. The higher incidence of adverse event reports of
brary of Medicine’s PubMed database (search terms: breast discomfort, nausea, and headaches seen in the
continuous use OR extended use AND contraceptive extended group may represent an estrogenic effect in
NOT replacement NOT postmenopausal; limitation: some women. It is also possible that women assigned to
clinical trials; no language limitation), this randomized use the “new” extended regimen were more aware of
clinical trial comparing extended use of the and more likely to report common side effects than were
norelgestromin/ethinyl E2 transdermal patch to cyclic women using the more common monthly regimen.
use is one of the largest on extended regimen use of A limitation of this study is that its relatively short
approved cyclic regimens published to date. In addition, treatment duration did not allow a determination of the
it is the first such trial using a transdermal mode of long-term efficacy, tolerability, and safety of the ex-
hormonal delivery. The purpose of this trial was to tended regimen. Any pregnancies or serious adverse
examine the efficacy and safety of an extended regimen events that may occur beyond 84 days of continuous
of the norelgestromin/ethinyl E2 transdermal system transdermal norelgestromin/ethinyl E2 use require fur-
over an 84-day period. Extended use resulted in fewer ther study in a long-term usage trial. In addition, the
median bleeding days, bleeding episodes, and bleeding relatively small size of the population studied makes any
or spotting episodes, a greater incidence of amenorrhea generalization about efficacy or safety difficult.
during this period, and subject satisfaction that was In summary, extended-regimen dosing with the
similar to that seen with cyclic use. norelgestromin/ethinyl E2 transdermal patch was effec-
The Kaplan-Meier analysis of the bleeding data shows tive for delaying menses and was associated with reports
a significantly longer median time to first bleeding (54 of high satisfaction, low adverse events, and few discon-
days compared with 25 days) among the extended users tinuations due to bleeding.
as compared with the cyclic users (Fig. 3). The advantage
of the time to first bleeding analysis is that it provides an
REFERENCES
understandable interpretation of regimen performance.
If the objective of an extended regimen is to delay 1. Clarke AK, Miller SJ. The debate regarding continuous
use of oral contraceptives. Ann Pharmacother 2001;35:
menses, knowing how long a patient can reliably delay
1480 – 4.
bleeding is a key determinant of regimen acceptability
2. Kaunitz AM. Menstruation: choosing whether. . .and
and provides a meaningful definition to the concept of
when. Contraception 2000;62:277– 84.
“delaying menses.” Time to first bleeding analyses have
3. Loudon NB, Foxwell M, Potts DM, Guild AL, Short RV.
been done for a few oral contraceptives.20,21
Acceptability of an oral contraceptive that reduces the
The norelgestromin/ethinyl E2 transdermal system
frequency of menstruation: the tri-cycle pill regimen. BMJ
was assessed over fixed periods of time in this study. 1977;2:487–90.
However, it should also be possible to apply this contra-
4. Thomas SL, Ellertson C. Nuisance or natural and healthy:
ceptive system in a flexible manner to accommodate should monthly menstruation be optional for women?
changes in a woman’s social or personal activities. Al- Lancet 2000;355:922– 4.
though the present study demonstrated a median de- 5. Cachrimanidou AC, Hellberg D, Nilsson S, Waldenström
layed time to first bleeding of 54 days, some women may U, Olsson SE, Sikström B. Long-interval treatment regi-
choose to delay menses through 8 continuous patch- men with a desogestrel-containing oral contraceptive.
weeks, whereas others may choose a different duration, Contraception 1993;48:205–16.
such as 6 continuous patch-weeks. This ability to accom- 6. Sulak PJ, Scow RD, Preece C, Riggs MW, Kuehl TJ.
modate variable regimens provides a flexible approach Hormone withdrawal symptoms in oral contraceptive
to extended contraception. users. Obstet Gynecol 2000;95:261– 6.
Although the extended regimen users reported more 7. den Tonkelaar I, Oddens BJ. Preferred frequency and
breast discomfort, headache, and nausea than the cyclic characteristics of menstrual bleeding in relation to repro-
users, the overall adverse event profile was favorable. ductive status, oral contraceptive use, and hormone
Only 1 serious adverse event, a spontaneous pregnancy replacement therapy use. Contraception 1999;59:357– 62.
loss, occurred during this study. The discontinuation 8. Sulak PJ, Cressman BE, Waldrop E, Holleman S, Kuehl
rate due to adverse events was low, and most discontinu- TJ. Extending the duration of active oral contraceptive
ations occurred late in the study, indicating that both pills to manage hormone withdrawal symptoms. Obstet
regimens were well tolerated overall. Given the recent Gynecol 1997;89:179 – 83.

VOL. 105, NO. 6, JUNE 2005 Stewart et al Extended Transdermal Contraceptive Therapy 1395
 9. Kovacs GT, Rusden J, Evans A. A tri-monthly regimen for 21. Hamerlynck JV, Vollebregt JA, Doornebos CM, Munten-
oral contraceptives. Br J Fam Plann 1994;14:274 –5. dam P. Postponement of withdrawal bleeding in women
10. Miller L, Notter KM. Menstrual reduction with extended using low-dose combined oral contraceptives. Contracep-
use of combination oral contraceptive pills: randomized tion 1987;35:199 –205.
controlled trial. Obstet Gynecol 2001;98:771– 8.
11. Anderson FD, Hait H. A multicenter, randomized study of
an extended cycle oral contraceptive. Contraception 2003; Address reprint requests to: Katherine D. LaGuardia, MD,
68:89 –96. MPH, Director, Medical Affairs, Ortho Women’s Health,
12. Kornaat H, Geerdink MH, Klitsie JW. The acceptance of Ortho-McNeil Pharmaceutical, Inc., 1000 Route 202, Raritan,
a 7-week cycle with a modern low-dose oral contraceptive NJ 08869; e-mail: klaguard@ompus.jnj.com.
(Minulet). Contraception 1992;45:119 –27.
13. Miller L, Hughes JP. Continuous administration of 100 ␮g
levonorgestrel and 20 ␮g ethinyl estradiol: a randomized Received October 21, 2004. Received in revised form February 8, 2005.
controlled trial. Obstet Gynecol 2002;99:25S. Accepted February 10, 2005.
14. Miller L, Hughes JP. Continuous combination oral contra-
ceptive pills to eliminate withdrawal bleeding: a random-
ized trial. Obstet Gynecol 2003;101:653– 61. APPENDIX
15. Audet MC, Moreau M, Koltun WD, Waldbaum AS,
Shangold G, Fisher AC, et al. Evaluation of contraceptive The following investigators comprised the ORTHO
efficacy and cycle control of a transdermal contraceptive EVRA Extended Regimen Study Group:
patch vs. an oral contraceptive: a randomized controlled Ronald T. Burkman, MD (Baystate Medical Center,
trial. JAMA 2001;285:2347–54. Springfield, MA); Janet K. Gersten, MD, PA (New Age
16. Smallwood GH, Meador M, Lenihan JP, Shangold GA, Medical Research Corporation, Miami, FL); Andrew M.
Fisher AC, Creasy GW, et al. Efficacy and safety of a Kaunitz, MD (University of Florida Health Science Cen-
transdermal contraceptive system. Obstet Gynecol 2001; ter/Jacksonville Department of Obstetrics and Gynecol-
98:799 – 805. ogy, Jacksonville, FL); William D. Koltun, MD (Medical
17. Abrams LS, Skee D, Natarajan J, Wong FA, Anderson Center for Clinical Research, San Diego, CA); Larry S.
GD. Pharmacokinetic overview of Ortho Evra/Evra. Fertil
Seidman, DO (Philadelphia Women’s Research, Phila-
Steril 2002;77:S3–S12.
delphia, PA); Lee P. Shulman, MD (University of Illinois
18. Belsey EM, Farley TM. The analysis of menstrual bleed-
ing patterns: a review. Contraception 1988;38:129 –56.
at Chicago, Center for Women’s Health, Chicago, IL);
Felicia H. Stewart, MD (University of California San
19. Belsey EM, Carlson N. The description of menstrual
bleeding patterns: towards fewer measures. Stat Med Francisco Center for Reproductive Health Research &
1991;10:267– 84. Policy, San Francisco, CA); Arthur S. Waldbaum
20. de Voogd WS. Postponement of withdrawal bleeding with (Downtown Women’s Health Care, Denver, CO); and
a monophasic oral contraceptive containing desogestrel Carolyn L. Westhoff, MD (Columbia University Col-
and ethinyl estradiol. Contraception 1991;44:107–12. lege of Physicians and Surgeons, New York, NY).

1396 Stewart et al Extended Transdermal Contraceptive Therapy OBSTETRICS & GYNECOLOGY