Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=iejc20

Download by: [ Jordan Univ. of Science & Tech] Date: 28 August 2016, At: 23:58

The European Journal of Contraception & ReproductiveHealth Care

ISSN: 1362-5187 (Print) 1473-0782 (Online) Journal homepage: http://www.tandfonline.com/loi/iejc20

Efficacy and safety of the contraceptive vaginalring (NuvaRing) compared with a combinedoral contraceptive in Chinese women: a 1-yearrandomised trial

Guang Sheng Fan, Mulan Ren, Wen Di, Ping Su, Qin Chang, Shuying Wu,Yun Qin, Tjeerd Korver, Maya Marintcheva-Petrova, Carol Yacik, ChristineMcCrary Sisk & Guoqin Wang

To cite this article: Guang Sheng Fan, Mulan Ren, Wen Di, Ping Su, Qin Chang, ShuyingWu, Yun Qin, Tjeerd Korver, Maya Marintcheva-Petrova, Carol Yacik, Christine McCrary Sisk& Guoqin Wang (2016) Efficacy and safety of the contraceptive vaginal ring (NuvaRing)compared with a combined oral contraceptive in Chinese women: a 1-year randomised trial,The European Journal of Contraception & Reproductive Health Care, 21:4, 303-309, DOI:10.1080/13625187.2016.1186269

To link to this article: http://dx.doi.org/10.1080/13625187.2016.1186269

Published online: 24 Jun 2016.

Submit your article to this journal

Article views: 80

View related articles

View Crossmark data

CLINICAL STUDY

Efficacy and safety of the contraceptive vaginal ring (NuvaRing) compared witha combined oral contraceptive in Chinese women: a 1-year randomised trial

Guang Sheng Fana, Mulan Renb, Wen Dic, Ping Sud, Qin Change, Shuying Wuf, Yun Qing, Tjeerd Korverg,Maya Marintcheva-Petrovag, Carol Yacikg, Christine McCrary Siskg and Guoqin Wangg

aPeking Union Medical College Hospital, Beijing, China; bZhongda Hospital, Southeast University, Nanjing, China; cDepartment ofObstetrics and Gynecology, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China; dFamily PlanningResearch Institute of Tongji Medical College of Huazhong University of Science and Technology, Wuhan, China; eSouthwest Hospital, ThirdMilitary Medical University, Obstetrics and Gynecology, Chongqing, China; fPeking University Third Hospital, Beijing, China; gMerck & Co.,Inc., Kenilworth, NJ, USA

ABSTRACTObjectives: The aim of the study was to assess the efficacy and tolerability of the monthly vaginalring (NuvaRing; 15 lg ethinylestradiol [EE] and 120 lg etonogestrel per day) compared with a mono-phasic (21/7) combined oral contraceptive (COC) containing 30 lg EE and 3 mg drospirenone inhealthy Chinese women aged 18–40 years.Methods: This was a phase III, open-label, randomised multicentre trial conducted in China.Participants received NuvaRing or COC for 13 cycles (3 weeks of ring/pill treatment followed by a 1-week ring-free/pill-free period). Contraceptive efficacy was assessed by in-treatment pregnanciesand expressed by the Pearl Index (PI; number of pregnancies/100 woman-years of use). Cycle con-trol was assessed by unscheduled (breakthrough) and absence of scheduled (withdrawal) bleedingevents. Safety and tolerability were assessed throughout the study.Results: Participants were randomised either to the NuvaRing (n¼ 732) or to the COC (n¼ 214); 588(82.4%) and 182 (78.4%) participants, respectively, completed the study. There were 10 in-treatmentpregnancies in the NuvaRing group (PI 1.92; 95% confidence interval [CI] 0.92, 3.53) and five in theCOC group (PI 3.12; 95% CI 1.01, 7.29). Breakthrough bleeding/spotting ranged from 18.6% (Cycle 1)to 4.2% (Cycle 11) for NuvaRing and from 21.6% (Cycle 1) to 7.9% (Cycle 11) for COC. Absence ofwithdrawal bleeding ranged from 8.6% (Cycle 1) to 3.0% (Cycle 11) for NuvaRing and from 14.6%(Cycle 1) to 6.4% (Cycle 5) for COC. For NuvaRing and COC, respectively, 26.6% and 25.0% of partici-pants had treatment-related adverse events, and 7.0% and 9.1% discontinued the study as a result.Conclusions: Once-monthly NuvaRing is efficacious and safe for use in Chinese women.

ARTICLE HISTORYReceived 31 December 2015Revised 25 April 2016Accepted 30 April 2016Published online 17 June2016

KEYWORDSCompliance; contraceptivepill; efficacy; NuvaRing;tolerability

Introduction

NuvaRing is a vaginal contraceptive ring used over a 28-dcycle (21 d with the ring in place, 7 d without a ring), whichavoids daily oral dosing and provides more stable systemicdrug levels relative to combined oral contraceptives (COCs).The ring consists of an ethylene-vinyl acetate copolymercore (28% vinyl acetate), containing 11.7 mg etonogestreland 2.7 mg ethinylestradiol (EE), covered by a skin of ethyl-ene-vinyl acetate (9% vinyl acetate). The ring continuouslyreleases an average amount of 120 lg/day etonogestrel and15 lg/day EE over the 3-week period of use. Peak serumconcentrations of EE and etonogestrel occur approximately1 week after ring insertion and are 60–70% lower than thedaily peak concentrations produced by a COC containing150 lg desogestrel and 30 lg EE.[1] NuvaRing provides con-sistent ovulation inhibition that is rapidly reversible upondiscontinuation.[2] The non-oral route with NuvaRing avoidshepatic first-pass metabolism and the potential for gastro-intestinal interference with absorption. In addition, thismethod is user-administered, unlike some other longer-acting contraceptive methods, including injectables,implants and hormone-releasing intrauterine systems.

Data from clinical trials conducted in the USA, Canada,Latin America and Europe demonstrate that NuvaRing is an

effective and well-tolerated method of contraception with awell-defined safety profile appropriate for this indicationwhen used according to the product labelling.[3–10]However, the efficacy and tolerability of NuvaRing have notbeen compared with those of a COC in Chinese women.With this in mind, the present study was conducted inChina to compare the efficacy and tolerability of the twocontraceptive regimens. The primary objective of this trialwas to compare the contraceptive efficacy, cycle control,safety and tolerability of NuvaRing with that of the COC inhealthy Chinese women aged 18–40 years. At the time ofstudy initiation, the monophasic (21/7) COC containing30 lg EE and 3 mg drospirenone (DRSP [Yasmin]; BayerHealthCare Pharmaceuticals, Whippany, NJ, USA) was theleading contraceptive drug used in China, and no similarproducts were approved for use; it was thus selected as thecomparator for this study.

Methods

Study design

This was a phase III, open-label, randomised, comparative-group multicentre trial conducted at 29 obstetrics andgynaecology centres in China (study protocol no. P06450;

CONTACT Guang Sheng Fan fgsxiehe@163.com Peking Union Medical College Hospital, No. 1 Shuai Fu Yuan, Dongcheng District, Beijing, China� 2016 The European Society of Contraception and Reproductive Health

THE EUROPEAN JOURNAL OF CONTRACEPTION & REPRODUCTIVE HEALTH CARE, 2016VOL. 21, NO. 4, 303–309http://dx.doi.org/10.1080/13625187.2016.1186269

ClinicalTrials.gov NCT01277211). The study protocol wasapproved by the institutional review boards at each studycentre, and written informed consent was obtained from allparticipants before the initiation of any study procedures.The study was conducted in accordance with the principlesof good clinical practice. The primary study objective was toassess the contraceptive efficacy of NuvaRing compared witha COC. The key secondary study objectives were to assessthe vaginal bleeding pattern (cycle control), and the generalsafety and tolerability of NuvaRing, compared with a COC.

Trial participants

The participants in this study were healthy Chinese womenof reproductive age (18–40 years), seeking contraception. Atotal of 720 women in the NuvaRing group were plannedfor enrolment based on the requirement of the ChineseFood and Drug Administration to obtain 500 women onNuvaRing who would complete 13 cycles of treatment, andbased on the 30% discontinuation rate observed in previousNuvaRing clinical trials.[7,8] In addition, 240 women wereplanned to be randomly assigned to the DRSP-EE COC, anestablished contraceptive in China, and as such this groupmight serve as an internal reference in the study, facilitatingthe interpretation of data. Key exclusion criteria includedcontraindications to contraceptive steroids; previous use ofan injectable hormonal method of contraception morerecently than the pre-specified washout period or beforespontaneous menstruation had occurred following a deliv-ery or abortion; breast-feeding within 2 months of the startof trial medication; undiagnosed vaginal bleeding; an abnor-mal cervical smear diagnosed during screening; and use ofdrugs that interfered with the metabolism of contraceptivehormones. No restrictions were set for cycle length. Therewas an additional exclusion of conditions that predisposedto hyperkalaemia (renal insufficiency, hepatic dysfunctionand adrenal insufficiency), in relation to the antimineralocor-ticoid activity of DRSP.

Interventions

Participants were randomly allocated in a 3:1 ratio toNuvaRing or COC for 13, 28-d cycles, each consisting of3 weeks of ring or COC use, followed by a 1-week ring-free/pill-free period. Participants were randomised to one of thetwo treatment groups using an interactive voice responsesystem by the study investigator (or qualified designee).

Women randomised to NuvaRing received verbal andwritten instructions upon study entry on the use of thering, including insertion and removal procedures and tim-ing. Women who had been taking no prior hormonalcontraception inserted the ring between days 1 and 5 ofthe spontaneous onset of menses, and were advised to usea barrier method of contraception during the first 7 d ofring use if not started on day 1. Women who had been tak-ing prior COCs started using the ring following their usualpill-free interval. Women in the NuvaRing group received anew ring for each cycle.

Women randomised to the COC who had been takingno prior hormonal contraception took one COC tablet dailyfor 21 d, starting on the first day of spontaneous menses.Women who had been using prior COCs started using the

study COC on the day after the last active tablet of the priorCOC, but, at the latest, on the day following the usual pill-free interval. Pills were to be taken at approximately thesame time each day.

Contraceptive efficacy and cycle control

The primary efficacy endpoint was the incidence of in-treat-ment pregnancies, expressed by the Pearl Index (PI; numberof pregnancies per 100 woman-years of use, with 1 woman-year defined as a period of 365.25 d). In-treatment pregnan-cies were pregnancies with an estimated date of conceptionbetween the first insertion/intake of trial medication up toand including 7 d (the scheduled ring-free/pill-free period)after the last day of ring use or pill intake. Post-randomisa-tion assessments were performed during the first weekafter the ring-free/pill-free period of cycles 1, 3, 6 and 9.End-of-trial assessments were performed after Cycle 13 or atpremature discontinuation.

Vaginal bleeding pattern (cycle control) was assessed asa secondary endpoint by the occurrence of unscheduled(breakthrough) bleeding/spotting and absence of scheduled(withdrawal) bleeding/spotting. Diary booklets were usedfor daily recording of vaginal bleeding events up to andincluding the ring-free/tablet-free period of Cycle 13. If vagi-nal bleeding was present, participants were asked to indi-cate whether it was considered to be bleeding (requiringtwo or more pads/tampons per day) or spotting (requiringone pad/tampon per day or no sanitary protection at all).Breakthrough bleeding/spotting was defined as any bleed-ing/spotting episode that occurred during the expectednon-bleeding period that was neither an early nor a contin-ued withdrawal bleeding. Absence of withdrawal bleedingwas defined as no bleeding/spotting episode that beganduring or continued into the expected bleeding period. Theexpected bleeding period was the 7-d period starting onday 22 of the cycle; the expected non-bleeding period wasthe 21-d period starting on day 1 of the cycle.

Dysmenorrhoea

Dysmenorrhoea was assessed by interviewing the partici-pants at screening and at all visits after randomisation,using a verbal rating scale (1¼none; 2¼mild, no analgesicsneeded; 3¼ severe, analgesics needed; 4¼ very severe,analgesics had insufficient effect).

Safety and tolerability

Safety and tolerability assessments were performedthroughout the study. In addition to the assessment ofadverse events, safety assessments included physical exam-ination, breast/gynaecological examination, vital signs(blood pressure, heart rate and body weight and height)and laboratory safety parameters. These assessments wereperformed during treatment and/or after the woman hadended pill intake/ring use (after Cycle 13 visit or at prema-ture discontinuation).

Compliance

Participants used paper diaries to record information dailyabout ring use/tablet intake and vaginal bleeding pattern

304 G. S. FAN ET AL.

throughout the trial. In addition, condom use and informa-tion about sexual intercourse during each cycle wererecorded at the end of each 28-d cycle, to assess whetherthe woman had been at risk of pregnancy. Compliance wasdetermined for the cycles at risk of pregnancy. For theNuvaRing group, compliance with the 21/7 regimen wasdefined as cycles at risk with both a ring-use periodbetween 456 and 552 h (19–23 d) and a ring-free periodbetween 144 and 192 h (6–8 d). Compliance with ring usewas defined by the ratio of the hours of actual ring useover the total number of hours in the ring-use periods ofcycles at risk. Compliance with respect to temporary ringremoval was defined as the average number of temporaryring removal hours per cycle at risk. For the COC group,compliance with cycle length was defined as cycles at riskwith a length of at least 26 d and at most 30 d. Compliancewith pill intake was defined as the ratio of the days withpill intake over the total number of days with non-missingresponses of pill intake, over the scheduled 21-d pill intakeperiods of the cycles at risk. Compliance with no pill intakeduring the scheduled pill-free periods was defined as thenumber of days in which no pill was to be taken relative tothe number of days with non-missing responses during thescheduled pill-free periods (over cycles at risk).

Statistical methods

The primary PI analysis was based on the restricted intent-to-treat (R-ITT) population, consisting of all randomised par-ticipants who used at least one ring/pill and had at leastone cycle at risk of pregnancy; cycles which were not at riskof pregnancy (i.e., cycles without intercourse or with use ofcondoms) were excluded from the analysis. Two-sided 95%confidence intervals (CIs) for the PI were based on Poisson(kT) distribution, where k denotes the PI and T is the totalexposure in 100 woman-years of exposure. Supportive PIanalyses were performed in the intent-to-treat (ITT) popula-tion, consisting of all randomised participants who used atleast one ring/pill, and the per-protocol (PP) population,consisting of all participants from the ITT population with-out any major protocol violation (cycles with minor protocolviolations and cycles not at risk of pregnancy were excludedfrom the PP analysis of contraceptive efficacy). The cyclecontrol analyses were based on the ITT population.Between-group comparisons were performed per cycleusing the unconditional asymptotic method of Miettinenand Nurminen.[11] Compliance parameters were summar-ised for the cycles at risk of pregnancy for the R-ITT popula-tion. The analyses for all safety and tolerability outcomeswere performed in the all-subjects-treated (AST) population,defined as all randomised participants who used at leastone ring/pill and were performed using descriptivestatistics.

Results

Participant disposition

The trial was initiated on 11 November 2011 and completedon 5 September 2013. A total of 1137 women werescreened, and 983 were randomised across 29 study sites inChina. The number of women randomised ranged from 11to 84 across the 29 study sites. In the NuvaRing and COC

groups, 714 and 232 participants, respectively, were treated,representing the ITT group and the AST group. The R-ITTgroup comprised 878 participants (NuvaRing n¼ 662; COCn¼ 216). Overall, 588 (82.4%) and 182 (78.4%) participants,respectively, completed the study (Figure 1). For the 37 indi-viduals who were randomised but not treated, the most fre-quent reasons for discontinuation from the study were:pretreatment pregnancy (n¼ 13), withdrawal of consent(n¼ 11) and not meeting protocol eligibility (n¼ 6).

There were no notable differences between the twotreatment groups in baseline demographic and clinical char-acteristics (Table 1). The overall age of participants rangedfrom 18 to 40 years, with a mean age of 31.8 years in theNuvaRing group and 31.2 years in the COC group.

Contraceptive efficacy

In the R-ITT population, the NuvaRing group was exposedfor an equivalent of 520.8 woman-years, and the COC groupwas exposed for an equivalent of 160.1 woman-years. In theR-ITT population, there were 10 in-treatment pregnancies inthe NuvaRing group and five in-treatment pregnancies inthe COC group (Table 2). The difference in ratio of PIsbetween NuvaRing over COC was not statistically significant.

For the supportive analyses, the estimated PIs for the ITTand PP populations for both groups were lower comparedwith those for the primary analysis based on the R-ITTpopulation. For the ITT population, all 15 pregnancies wereincluded in the analysis, and the estimated PIs were 1.59(95% CI 0.76, 2.93) for the NuvaRing group and 2.57 (95%CI 0.83, 5.99) for the COC group. For the PP population, fourpregnancies (three NuvaRing, one COC) were excluded dueto protocol violations; the estimated PIs were 1.58 (95% CI0.64, 3.26) in the NuvaRing group and 2.73 (95% CI 0.74,6.99) in the COC group. As with the R-ITT population, thePIs were not statistically significantly different between thetreatment groups in the ITT and PP populations.

Cycle control analysis

The incidence of breakthrough bleeding/spotting overcycles 1–13 decreased with time and ranged from 18.6%(Cycle 1) to 4.2% (Cycle 11) for NuvaRing and from 21.6%(Cycle 1) to 7.9% (Cycle 11) for COC. The incidence of break-through bleeding/spotting was lower in the NuvaRing groupthan in the COC group for all 13 cycles, with a statisticallysignificant (p< 0.05) between-treatment group difference incycles 2, 3, 4, 8, 10, 11, 12 and 13 (Table 3). Absence ofwithdrawal bleeding also decreased with time and rangedfrom 8.6% (Cycle 1) to 3.0% (Cycle 11) for NuvaRing andfrom 14.6% (cycle 1) to 6.6% (cycle 10) for COC. Absence ofwithdrawal bleeding also occurred less frequently withNuvaRing than with COC for all 13 cycles, with a statisticallysignificant (p< 0.05) between-treatment group difference incycles 1, 2, 3, 4, 7, 9, 11, 12 and 13 (Table 4).

Dysmenorrhoea

At baseline, 31.2% of NuvaRing users and 35.8% of COCusers reported dysmenorrhoea. At the last in-treatmentmeasurement, the proportions of women having dysmenor-rhoea had decreased to 3.9% and 8.0% in the NuvaRing

THE EUROPEAN JOURNAL OF CONTRACEPTION & REPRODUCTIVE HEALTH CARE 305

and COC groups, respectively. For the vast majority ofwomen reporting dysmenorrhoea, it was of mild intensity(i.e., no analgesic use was needed) at each time point.

Safety and tolerability

NuvaRing and COC were both generally well tolerated.Throughout the study, 39.8% of NuvaRing users and 36.2%of COC users reported an adverse event, of which 26.6%and 25.0%, respectively, were considered by the investigatorto be at least possibly related to treatment (Table 5).

Overall, 16 participants (2.2%) in the NuvaRing groupreported full device expulsion, and five of these discontin-ued the study as result of the device expulsion.Nasopharyngitis, upper respiratory tract infection and vagi-nal infections were reported more frequently by NuvaRingusers, whereas nausea and vomiting were reported morefrequently by COC users (Table 5).

Investigations included laboratory tests for haematology,blood chemistry and D-dimer; it was up to the investigator todecide whether or not an observed abnormal laboratoryvalue was clinically relevant and reportable as an adverseevent. In 20 (2.8%) women in the NuvaRing group and 7women (3.0%) in the COC group, an increase in D-dimer wasreported as an adverse event. Based on actual levels, D-dimershowed numerically similar increases from baseline in boththe NuvaRing and COC groups, with respective mean (stand-ard deviation) changes from baseline to last measurement of0.058 (0.286) and 0.086 (0.843) mg/L fibrinogen equivalentunits (FEU). The reference range in this population (female,19–40 years of age) was 0–0.499 mg/L FEU.

Two serious adverse events were reported in theNuvaRing group: one participant had sinusitis and anotherhad an intraductal proliferative breast lesion. Neitheradverse event was considered to be related to the studymedication. Four serious adverse events were reported inthe COC group: two cases of spontaneous abortion (oneconsidered probably related and the other consideredunlikely related to the study medication), one case of deepvein thrombosis and one case of invasive ductal breast

Screened N=1137

Randomised N=983

All subjects treated N=946

COC

ASTITTR-ITTPP

Discontinued N=50 (21.6%)

Completed N=182 (78.4%)

ReasonsAdverse eventsPregnancyLost to follow-upSubject did not wish to continue,not treatment relatedSubject did not wish to continue,treatment relatedSubject withdrew consentNon-compliance with protocolDid not meet protocol eligibilityAdministrativePregnancy wish

22 (9.5%)4 (1.7%)9 (3.9%)0

0

9 (3.9%)4 (1.7%)002 (0.9%)

N=232 (100%)N=232N=216N=215

NuvaRing

ASTITTR-ITTPP

Discontinued N=126 (17.6%)

Completed N=588 (82.4%)

ReasonsAdverse eventsPregnancyLost to follow-upSubject did not wish to continue,not treatment relatedSubject did not wish to continue,treatment relatedSubject withdrew consentNon-compliance with protocolDid not meet protocol eligibilityAdministrativePregnancy wish

60 (8.4%)9 (1.3%)

14 (2.0%)7 (1.0%)

10 (1.4%)

9 (1.3%)13 (1.8%)

1 (0.1%)3 (0.4%)0

N=714 (100%)N=714N=662N=662

Figure 1. Flow of participants through the study.

Table 1. Baseline characteristics of the NuvaRing and COC treatment groupsin the AST population.

NuvaRing (N¼ 714) COC (N¼ 232)

Age, years 31.8 ± 4.0 31.2 ± 3.9Weight, kg 57.1 ± 7.4 56.5 ± 7.5Height, cm 161.3 ± 4.9 160.8 ± 5.0Body mass index, kg/m2 21.9 ± 2.5 21.8 ± 2.5Last used contraceptive method

Oral contraceptive 32 (4.5) 12 (5.2)Foam, condom, suppositories

or diaphragm550 (77.1) 169 (72.8)

Hormonal IUC 3 (0.4) 1 (0.4)Non-hormonal IUC 7 (1.0) 3 (1.3)Progestogen-only pill 1 (0.1) 0Vaginal ring 1 (0.1) 0Other 30 (4.2) 15 (6.5)None 89 (12.5) 32 (13.8)Missing 1 (0.1) 0

All data are presented as mean ± SD or n (%).IUC: intrauterine contraception.

306 G. S. FAN ET AL.

carcinoma (both considered possibly related to the studymedication).

A similar proportion of participants discontinued due toan adverse event in the two treatment groups: 60 women(8.4%) in the NuvaRing group and 22 women (9.5%) in theCOC group. In the NuvaRing group, the most frequentlyreported adverse event which led to discontinuation wasdecrease in libido (1.0%). In the COC group, the most fre-quently reported adverse events which led to discontinu-ation were nausea (3.0%) and vomiting (1.7%). In bothgroups, irregular vaginal bleeding resulted in discontinu-ation in 0.8% and 0.9% of participants in the NuvaRing andCOC groups, respectively. There were no major differencesin the rates of drug-related adverse events leading to

discontinuation. The most common drug-related adverseevents that led to discontinuation were nausea (0.6% and3.0%) and vomiting (0% and 1.7%) in the NuvaRing andCOC groups, respectively.

Physical and gynaecological examinations revealed veryfew clinically relevant abnormalities. There was no overalleffect on diastolic blood pressure, systolic blood pressure orbody weight observed in either treatment group. The per-centage of participants with an increase in body weightfrom baseline was 0.2–7.0% higher in the COC group com-pared with the NuvaRing group at each assessment.Clinically relevant changes in blood chemistry or haematol-ogy occurred infrequently, and there were no significant dif-ferences between groups in these parameters at study end.

Table 3. Occurrence of breakthrough bleeding/spotting in the ITT population.

NuvaRing COC Differenced

Cycle Na n (%)b 95% CIc Na n (%)b 95% CIc Estimate (95% CI) p Value

1 598 111 (18.6) 15.5, 21.9 199 43 (21.6) 16.1, 28.0 �3.0 (�9.9, 3.1) 0.3462 604 74 (12.3) 9.7, 15.1 195 36 (18.5) 13.3, 24.6 �6.2 (�12.7, �0.6) 0.0293 578 49 (8.5) 6.3, 11.1 196 28 (14.3) 9.7, 20.0 �5.8 (�11.8, �0.9) 0.0194 587 44 (7.5) 5.5, 9.9 194 30 (15.5) 10.7, 21.3 �8.0 (�14.0, �2.9) 0.0015 592 40 (6.8) 4.9, 9.1 187 15 (8.0) 4.6, 12.9 �1.3 (�6.4, 2.6) 0.5566 590 41 (6.9) 5.0, 9.3 185 20 (10.8) 6.7, 16.2 �3.9 (�9.5, 0.5) 0.0897 583 44 (7.5) 5.5, 10.0 186 21 (11.3) 7.1, 16.7 �3.7 (�9.4, 0.8) 0.1108 585 37 (6.3) 4.5, 8.6 188 20 (10.6) 6.6, 16.0 �4.3 (�9.8, �0.0) 0.0499 576 40 (6.9) 5.0, 9.3 187 15 (8.0) 4.6, 12.9 �1.1 (�6.2, 2.9) 0.62110 572 34 (5.9) 4.2, 8.2 183 19 (10.4) 6.4, 15.7 �4.4 (�10.0, �0.2) 0.04111 578 24 (4.2) 2.7, 6.1 177 14 (7.9) 4.4, 12.9 �3.8 (�8.9, �0.1) 0.04612 569 30 (5.3) 3.6, 7.4 176 17 (9.7) 5.7, 15.0 �4.4 (�9.9, �0.2) 0.03713 544 35 (6.4) 4.5, 8.8 173 19 (11.0) 6.7, 16.6 �4.5 (�10.4, �0.0) 0.048aNumber of participants with evaluable cycle.bIncidence of breakthrough bleeding/spotting.cTwo-sided 95% CI based on the Clopper–Pearson method.dBased on the unconditional asymptotic method of Miettinen and Nurminen.[11]. Significant p values (<0.05) appear in bold font.

Table 4. Occurrence of absence of withdrawal bleeding in the ITT population.

NuvaRing COC Differenced

Cycle Na n (%)b 95% CIc Na n (%)b 95% CIc Estimate (95% CI) p Value

1 596 51 (8.6) 6.4, 11.1 199 29 (14.6) 10.0, 20.3 �6.0 (�12.0, �1.1) 0.0152 602 29 (4.8) 3.2, 6.8 195 19 (9.7) 6.0, 14.8 �4.9 (�10.1, �1.0) 0.0123 577 30 (5.2) 3.5, 7.3 196 23 (11.7) 7.6, 17.1 �6.5 (�12.0, �2.2) 0.0024 586 24 (4.1) 2.6, 6.0 194 15 (7.7) 4.4, 12.4 �3.6 (�8.5, �0.1) 0.0445 590 32 (5.4) 3.7, 7.6 187 12 (6.4) 3.4, 10.9 �1.0 (�5.7, 2.5) 0.6096 588 34 (5.8) 4.0, 8.0 185 18 (9.7) 5.9, 14.9 �3.9 (�9.3, 0.2) 0.0627 581 20 (3.4) 2.1, 5.3 186 14 (7.5) 4.2, 12.3 �4.1 (�9.0, �0.6) 0.0198 583 29 (5.0) 3.4, 7.1 188 13 (6.9) 3.7, 11.5 �1.9 (�6.7, 1.6) 0.3089 575 25 (4.3) 2.8, 6.4 187 16 (8.6) 5.0, 13.5 �4.2 (�9.3, �0.4) 0.02710 571 21 (3.7) 2.3, 5.6 183 12 (6.6) 3.4, 11.2 �2.9 (�7.6, 0.5) 0.09811 576 17 (3.0) 1.7, 4.7 177 13 (7.3) 4.0, 12.2 �4.4 (�9.4, �1.0) 0.00912 568 23 (4.0) 2.6, 6.0 176 18 (10.2) 6.2, 15.7 �6.2 (�11.7, �2.1) 0.00213 543 32 (5.9) 4.1, 8.2 173 18 (10.4) 6.3, 15.9 �4.5 (�10.2, �0.1) 0.043aNumber of participants with evaluable cycle.bIncidence of absence of withdrawal bleeding.cTwo-sided 95% CI based on the Clopper–Pearson method.dBased on the unconditional asymptotic method of Miettinen and Nurminen.[11] Significant p values (<0.05) appear in bold font.

Table 2. Contraceptive efficacy: estimated PIs for NuvaRing and COC in the R-ITT population.

Exposure

Contraceptive N Cyclesa YearsbIn-treatmentpregnancies

Estimated PI(95% CI)c

Ratio of estimated PIs, NuvaRingversus COC (95% CI)d p Value

NuvaRing 662 6793.9 520.8 10 1.92 (0.92, 3.53) 0.61 (0.19, 2.29) 0.531COC 216 2088.1 160.1 5 3.12 (1.01, 7.29)a28-d cycles of exposure.b365.25-d woman-years of exposure.cCIs are based on Poisson (kT) distribution, where k denotes the PI and T is the total extent of exposure in 100 woman-years of exposure.dEquality of the PIs is tested by conditioning on the total number of pregnancies in both treatment groups (resulting in a binomial distribution) and rejecting

for large and small relative number of pregnancies in one arm (two-sided, a¼ 0.05).

THE EUROPEAN JOURNAL OF CONTRACEPTION & REPRODUCTIVE HEALTH CARE 307

Compliance with treatment

For the NuvaRing group, 92.7% of the cycles at risk werecompliant with the 21/7 regimen. Overall, 70.1% of partici-pants used the ring without temporary ring removals for allof their cycles at risk, and 95% of participants used the ringfor at least 97.3% of the ring-use treatment time; 12.1% ofparticipants temporarily removed the ring for at most 2 hover their cycles at risk, and 2.1% removed the ring formore than 24 h.

For the COC group, 97.6% of the cycles at risk were com-pliant with respect to cycle length. Overall, 52.3% of partici-pants never missed a tablet during the scheduled 21 d ofpill intake for all of their cycles at risk, and all pills hadbeen taken in 79.4% of the scheduled pill intake periods ofthe cycles at risk. Ninety percent of participants had at least98.6% of no pill intake during their scheduled pill-freeperiods.

Discussion

Findings and interpretation

This open-label, randomised, comparative-group multicentretrial demonstrated that NuvaRing is an effective and well-tolerated contraceptive method compared with a commonlyused COC in Chinese women. In the R-ITT population, therewere 10 in-treatment pregnancies in the NuvaRing groupand 5 in-treatment pregnancies in the COC group, resultingin PIs of 1.92 and 3.12, respectively. Although the PI forNuvaRing was numerically lower than that observed forCOC, the difference between the treatment groups was notstatistically significant. Similarly, the between-treatmentgroup differences were not statistically significant for thesupportive analyses in the ITT and PP populations.

NuvaRing provided superior cycle control compared withthe COC, as demonstrated by the statistically significantlylower incidences of breakthrough bleeding/spotting in 8 of13 treatment cycles and the absence of withdrawal bleed-ing during 9 of 13 treatment cycles.

In addition to contraceptive protection and cycle control,a favourable tolerability profile is a major factor associatedwith the overall acceptability of a contraceptive method.[12]In our study, the nature, frequency and intensity of theadverse events reported, as well as the absence of clinicallymeaningful findings on physical examinations, laboratorytests and vital signs, indicate that both NuvaRing and theCOC are safe and well tolerated in Chinese women. Asexpected, there were differences in device-related com-plaints (e.g., device difficult to use, device dislocation,device expulsion) observed with NuvaRing. Differencesbetween NuvaRing and COC were also observed for localvaginal complaints (e.g., vaginal infection, vaginal discharge,vaginal inflammation); however, the incidences were gener-ally low and there was a low incidence of such events lead-ing to discontinuation.

Lower incidences for NuvaRing versus COC werereported for gastrointestinal adverse events such as nauseaand vomiting. Additionally, the percentage of women withan increase in body weight from baseline was 0.2–7.0%higher in the COC group compared with the NuvaRinggroup at each assessment, although there were no signifi-cant differences between the groups.

In patients with dysmenorrhoea at baseline, both treat-ments induced an overall decrease in dysmenorrhoea.Similar numbers of participants discontinued in both theNuvaRing (17.6%) and COC (21.6%) treatment groups: 8.4%and 9.5% of women, respectively, discontinued due to theoccurrence of an adverse event.

Differences in results and conclusions in relation toother studies

The comparable contraceptive efficacy of NuvaRing withthat of a COC is consistently observed across large-scaleclinical trials, confirming that NuvaRing provides contracep-tive protection as robust as a COC; however, in an absolutesense, the PIs observed for NuvaRing and the DRSP-EE COCin this study are higher than previously reported.[3,6,8]

Table 5. Summary of adverse events in AST population.

Summary of adverse events NuvaRing, N¼ 714 COC, N¼ 232

One or more clinical 284 (39.8) 84 (36.2)Treatment-related clinicala 190 (26.6) 58 (25.0)Serious clinical 2 (0.3) 4 (1.7)Severe/life-threatening 9 (1.3) 3 (1.3)Serious treatment-related clinical 0 (0.0) 3 (1.3)Discontinued due to adverse event 60 (8.4) 22 (9.5)Discontinued due to treatment-related adverse event 50 (7.0) 21 (9.1)Died 0 (0.0) 0 (0.0)Adverse events with an incidence �2% in either group

Gastrointestinal disorders 32 (4.5) 23 (9.9)Nausea 13 (1.8) 9 (3.9)Vomiting 3 (0.4) 5 (2.2)

General disorders/administration site conditions 34 (4.8) 4 (1.7)Device expulsion 16 (2.2) n/a

Infections and infestations 98 (13.7) 13 (5.6)Nasopharyngitis 15 (2.1) 2 (0.9)Upper respiratory tract infection 29 (4.1) 6 (2.6)Vaginal infection 20 (2.8) 4 (1.7)

Investigations 41 (5.7) 10 (4.3)Fibrin D-dimer increased 20 (2.8) 7 (3.0)

Reproductive system and breast disorders 105 (14.7) 28 (12.1)Breast hyperplasia 17 (2.4) 6 (2.6)Vaginal haemorrhage 19 (2.7) 7 (3.0)

Values are n (%).aConsidered by the investigator to be definitely, probably or possibly related to study treatment.n/a: not applicable.

308 G. S. FAN ET AL.

Most likely, comparatively poor compliance with medicationuse in this particular trial population contributed to theobserved elevated PIs.

In addition, the superior cycle control with the NuvaRingcompared with the COC, as demonstrated by the lower inci-dences of breakthrough bleeding/spotting and the absenceof withdrawal bleeding during all treatment cycles, is similarto that found in previous studies.[9]

NuvaRing has been shown to exhibit similar tolerabilityand safety to those of COCs. The overall pattern of localvaginal complaints was similar to that found in previousclinical trials comparing NuvaRing and COC methods.[3–6]Our findings are also consistent with previous results show-ing that NuvaRing has a neutral effect on body weight.[3,6]

Relevance of the findings: implications for clinicians

Based on the results of this open-label, randomised, com-parative-group multicentre trial, the contraceptive efficacyof NuvaRing is expected to be as high as that of the COC.In addition, NuvaRing provided superior cycle control com-pared with the COC, which will enhance acceptability forusers. NuvaRing also exhibited similar tolerability and safetyto those of the COC, yet it has a formulation that does notrequire daily dosing. These results indicate that once-monthly NuvaRing is an efficacious and well-toleratedcontraceptive method compared with a commonly usedCOC for use in Chinese women.

Strengths and weaknesses of the study

To the best of our knowledge, ours is the first clinical trialto directly compare the efficacy and tolerability of NuvaRingwith those of a COC containing 30 lg EE and 3 mg DRSP inhealthy Chinese women. The strengths of our study includeits multicentre, randomised design and large sample size.The limitations of the study include the open-label studydesign and reliance on self-reporting in patient diaries.

Unanswered questions and future research

This study demonstrated that NuvaRing provides safe andeffective contraception and was generally well tolerated inChinese women. Larger-scale observational studies are war-ranted to identify long-term safety and particularly toidentify rare side effects.

Conclusion

The present results indicate that once-monthly NuvaRing isan efficacious contraceptive method for use in Chinesewomen. Compared with the COC, NuvaRing resulted in anumerically lower PI, and better cycle control as indicatedby statistically significantly lower incidences of break-through bleeding/spotting and absence of withdrawalbleeding during eight and nine treatment cycles, respect-ively. The safety profile of NuvaRing was unremarkable andcomparable to that of the COC, with no unexpected find-ings in the nature, frequency or intensity of reportedadverse events.

Acknowledgements

The authors thank Kristen Lewis of Merck & Co., Inc., Kenilworth, NJ,USA, for editorial assistance with the preparation of this manuscript forjournal submission.

Disclosure statement

Y.Q., C.Y., C.M.S. and G.W. report that they are current employees ofMerck & Co., Inc. and hold stock or stock options in the company. T.K.reports that he has received personal fees from Merck & Co., Inc. duringthe conduct of the study. M.M.-P. reports that she is a former employeeof Merck & Co., Inc. G.F., M.R., W.D., P.S., Q.C. and S.W. report that theyserved as investigators on the study and received study funding fromMerck & Co., Inc. All authors are responsible for the work described inthis manuscript. All authors were involved in at least one of the follow-ing: conception, design, acquisition, analysis, statistical analysis or inter-pretation of data and drafting the manuscript and/or revising it forimportant intellectual content. All authors provided final approval ofthe version to be published.

Funding information

Funding for this study was provided by Merck & Co., Inc., Kenilworth,NJ, USA.

References

[1] Timmer CJ, Mulders TM. Pharmacokinetics of etonogestrel andethinylestradiol released from a combined contraceptive vaginalring. Clin Pharmacokinet. 2000;39:233–242.

[2] Mulders TM, Dieben TO, Bennink HJ. Ovarian function with anovel combined contraceptive vaginal ring. Hum Reprod.2002;17:2594–2599.

[3] Nov�ak A, de la Loge C, Abetz L, et al. The combined contracep-tive vaginal ring, NuvaRing: an international study of useracceptability. Contraception 2003;67:187–194.

[4] Dieben TO, Roumen FJ, Apter D. Efficacy, cycle control, and useracceptability of a novel combined contraceptive vaginal ring.Obstet Gynecol. 2002;100:585–593.

[5] Roumen FJ, Apter D, Mulders TM, et al. Efficacy, tolerability andacceptability of a novel contraceptive vaginal ring releasing eto-nogestrel and ethinyl oestradiol. Hum Reprod. 2001;16:469–475.

[6] Bjarnadottir RI, Tuppurainen M, Killick SR. Comparison of cyclecontrol with a combined contraceptive vaginal ring and orallevonorgestrel/ethinyl estradiol. Am J Obstet Gynecol.2002;186:389–395.

[7] Oddsson K, Leifels-Fischer B, de Melo NR, et al. Efficacy andsafety of a contraceptive vaginal ring (NuvaRing) compared witha combined oral contraceptive: a 1-year randomized trial.Contraception 2005;71:176–182.

[8] Ahrendt HJ, Nisand I, Bastianelli C, et al. Efficacy, acceptabilityand tolerability of the combined contraceptive ring, NuvaRing,compared with an oral contraceptive containing 30 microg ofethinyl estradiol and 3 mg of drospirenone. Contraception2006;74:451–457.

[9] Milsom I, Lete I, Bjertnaes A, et al. Effects on cycle control andbodyweight of the combined contraceptive ring, NuvaRing, ver-sus an oral contraceptive containing 30 microg ethinyl estradioland 3 mg drospirenone. Hum Reprod. 2006;21:2304–2311.

[10] Roumen FJ, Mishell DR Jr. The contraceptive vaginal ring,NuvaRing, a decade after its introduction. Eur J ContraceptReprod Health Care 2012;17:415–427.

[11] Miettinen O, Nurminen M. Comparative analysis of two rates.Stat Med. 1985;4:213–226.

[12] Newton JR. Classification and comparison of oral contraceptivescontaining new generation progestogens. Hum Reprod Update1995;1:231–263.

THE EUROPEAN JOURNAL OF CONTRACEPTION & REPRODUCTIVE HEALTH CARE 309

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具